sign in sign up
<<
Home , Amos Dolbear, Charles Bourseul, John Ambrose Fleming, Jonathan Zenneck, Lee De Forest, Robert von Lieben

THE AWA REVIEW

Volume 20 2007

Published by THE ANTIQUE ASSOCIATION, INC. PO Box 108, Stafford, NY 14143

http://www.antiquewireless.org

i Devoted to research and documentation of the history of wireless communications

Member: Antique Wireless Association, Inc. Founded 1952, Chartered American Association PO Box 108 as a non-profit corporation of Museums Stafford, New York 14143 by the State of New York Regional Conference of http://www.antiquewireless.org Historical Agencies

THE A.W.A. REVIEW

EDITOR

Robert P. Murray, Ph.D. Vancouver, BC, Canada

OFFICERS OF THE ANTIQUE WIRELESS ASSOCIATION

PRESIDENT: Geoffrey Bourne VICE PRESIDENTS: Brian C. Belanger Ron Frisbe SECRETARY: Christian R. Fandt TREASURER: Christian R. Fandt AWA MUSEUM CURATOR: Edward Gable, K2MP

2007 by the Antique Wireless Association, Inc. ISBN 0-9741994-5-1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, elec- tronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright owner.

Printed in the of America by Carr Printing, Inc., Endwell, NY ii Table of Contents

Volume 20, 2007

“One Hundred Years of Electronic Communications”

Foreword ...... iv One Hundred Years of Electronic Communications Ludwell A. Sibley ...... 1 Henry J. Nolte and ’s High Power, Metal Envelope Tubes John M. Anderson ...... 19 Marconi vs. De Forest Infringement Litigation Revisited Eric P. Wenaas ...... 31 First World War Aircraft Larry Babcock ...... 75 The 1902 Wireless Connection - Santa Catalina Island to San Pedro, Norwood Teague and Joe A. Knight ...... 97 The Early History of the Electric Loudspeaker John D. Jenkins ...... 125 Emil J. Simon: A Busy Life Indeed Russ Kleinmann, Karen Blisard, A.J. Link and Warren Berbit ...... 145 A Portal into Radio’s Past: Francis A. Hart and his Radio Log James E. O’Neal ...... 169 Exploring the Origins of the Loud Speaker David and Julia Bart...... 191 The Supreme Model 45 Tube Tester and the 1933 Tube Pin Standard Charles C. Kirsten ...... 213

iii Foreword The AWA Review, Volume 20, is dedicated to, “100 years of elec- tronic communication.” There are articles directly related to this theme, as well as others. We are pleased to bring you a volume of outstanding articles by some of the finest writers on the history of wireless today. • We invited an article by noted AWA tube expert Ludwell Sibley, on “one hundred years of electronic communication.” His account fo- cuses on the industry and their use of amplification for long distance connections, which he describes as the “killer applica- tion”. His account includes other major applications, and through these examples Lud describes the major achievements along the 100 year course of development. And yes, he includes the implications for wireless technology. • Eric Wenaas, also closely linked to this year’s theme, describes in detail the Marconi vs. De Forest audion infringement litigation. In explanations not seen before in detail, he explores the demonstra- tions that were used in court to support the various arguments brought forward. • Larry Babcock, also in an invited paper, describes the artifacts and literature he presented in an award-winning exhibit at last year’s AWA Conference. His theme is First World War aircraft radio, and he had much of it on display last year. • Reaching back more than a century, Norwood Teague and Joe Knight describe a regular commercial wireless link established in 1902! It connected the residents and visitors on Catalina Island to the community of San Pedro, both in California. This commercial service is as much a social and political achievement as a technical one, but it nonetheless has important implications for the develop- ment of radio. • John Anderson describes the contributions of Henry J. Nolte to the development of General Electric’s high power, metal envelope tubes. John’s background comes as close as anyone’s today to that of someone who “was there”. He describes the challenges of glass-on- metal seals, among others, that led ultimately to the ubiquitous metal receiving tube. • We have two perspectives on the history of electrical sound repro- duction. One is by John Jenkins, and the other by David and Julia Bart. After some hesitation, we decided to publish both of these ar- ticles in the same issue. The articles have much in common, and it is interesting to see how different authors put their own spin on what is essentially the same history. After reading them, you can decide for yourself which approach you prefer. • The life of Emil J. Simon, an early wireless equipment producer, is described by Russ Kleinman, Karen Blisard, A.J. Link and Warren Berbit. Simon led an unusual personal life, but in retro- spect it makes for interesting reading. Although his wireless appara- tus was not known for its high quality, he was very effective in sup- plying military wireless equipment contracts during World War One. Simon was an important historical figure, but received mixed reviews from his contemporaries as an engineer and a businessman. iv • While searching for material on , James O’Neal uncovered the radio log of amateur Francis Hart at the Smithsonian Institution. It covers the interval from 1906 to 1909. The log pro- vides insight into the operating environment of one of the very first radio amateurs, and offers sometimes scathing assessments of his peers. It includes comments on the technical innovations available in the period. • Finally, Charles Kirsten uses the design of the Supreme Model 45 tube tester to illustrate the engineering virtue of the short-lived 1933 RMA tube pin standard. Its replacement by a new standard in 1934 made 1933 testers obsolete. Charles concludes that the 1934 stan- dard was introduced not for engineering reasons but from the self interest of RCA Radiotron Co. Ltd. This volume contains both articles addressing this year’s AWA theme, and a diversity of other topics by noted authors. It should appeal to the wide interests of AWA members and others. Our thanks are due to these authors for providing the AWA Review with the results of their fine and thorough work. Last year we introduced a practice of anonymous peer review of sub- mitted articles. This practice has been generally well accepted by au- thors and has the result of raising the credibility of our articles as his- torical documentation. This year again we have drawn on a pool of re- viewers with obvious credentials. These are:

David Bart, Buford Chidester, Alan Douglas, William Fizette, Donna Halper, Bartholomew Lee, Robert Lozier and Ludwell Sibley.

The AWA Review is very grateful for their work, an effort that has no direct benefit to them apart from the improvement of the quality of this volume. AWA members and others with an interest in wireless communica- tion history are encouraged to submit manuscripts to the AWA Review. A section titled Tips for Authors follows. We try to make the publication effort more collaborative than challenging. The single most important message in this regard is to contact us early if you are considering writ- ing an article. It is my distinct pleasure to bring you this volume of the AWA Re- view. I hope you enjoy it, and I look forward to receiving your manu- scripts for next year’s volume.

Robert P. (Bob) Murray, Ph.D. Editor

v Tips for Authors The AWA Review welcomes any submitted article on aspects of wire- less communications history. In general, shorter articles can be directed to the AWA Journal and longer manuscripts to the AWA Review. If you are in any doubt about where your article should best appear, please contact either or both editors. For first time authors, articles can be prepared with the help of a more experienced co-author, or the editor can help with the text in the editing process. Members with an interesting story to tell should not be discouraged by a lack of writing experience. The AWA Review will ac- cept manuscripts in any clearly prepared writing style. A short style manual produced by the American Radio Relay League is available on request. The Elements of Style by William Strunk Jr. and E.B. White is available in most public libraries. Reference material should be cited within the text of the article in any of the accepted reference styles. Reference lists should include all of the sources mentioned in the text. Writers should look at the articles in this volume or in previous volumes for examples. Unlike writing for a regional club newsletter, articles submitted to the AWA will be laid out on the pages in a style made consistent within the entire publication. Therefore, please do not arrange your illustra- tions on each page but rather send the text in a file separately from the files for each illustration. This requirement applies equally to the Jour- nal and the Review. (see, for example, “From the Editor” in the AWA Journal, April 2006, pages 4 & 5.) Text files can be prepared on any word processing software, but preferably on Microsoft Word. Illustra- tions are best sent as .JPG or .TIF files with a resolution of around 300 dpi. JPG files should be Standard (not Progressive) and Greyscale (not RBG). Files can be submitted as e-mail attachments directed to the edi- tor. Articles submitted to the AWA Review should be developed in con- cept not later than early January of the publication year. A first draft should be submitted around March. The editor’s deadline for submis- sion of the completed volume to the printer is July 1, so that a final draft is expected around May 1. Articles not submitted on this schedule will be rescheduled for the next year’s volume. For more information con- tact: Robert Murray, Ph.D., Editor The AWA Review #605 – 1000 Beach Avenue Vancouver, BC V6E 4M2 Canada [email protected]

vi Sibley

AWA Review One Hundred Years of Elec- tronic Communications

©2007 Ludwell A. Sibley ABSTRACT Electronic communications is entering its The essential ele- second century, with good prospects for ad- ments of electronic vancements comparable to those of the first communications hundred years. It is appropriate to take a look (mainly the LRS Re- back and consider what marvels the of lay and de Forest made possible. At the same time, audion) arrived on there is useful insight in considering how far the scene in mid- and what we view as “communications” progressed late 1906. They de- without electronic assistance. veloped into practical A bit of definition is necessary here. form during the next Strictly stated, any device that operates by the five or ten years. use of is “electronic.” That’s not very Over the rest of the satisfying for our purposes: under this view, a Twentieth Century, flashlight is “electronic.” A more relevant ap- the tube (né proach would have it that “electronics” requires audion) and transis- a device that can amplify through use of elec- tor slowly displaced a trons. Once such a unit becomes available, one wide variety of me- can “obtain gain,” build signal detectors that chanical devices for deliver more signal than they’re given, make os- signal generation, cillators, and design the digital circuit elements signal amplification, of computers. Incidentally, it matters little and computation. whether the electrons in our “gain device” flow This article provides in a solid-state material, a vacuum, or a gas. It a perspective of the is also dangerous to ignore fields like or arrival and triumph computation, which are not strictly “commu- of electronic commu- nications” but are intimately related in terms nication and related of technology. arts. Before the electronic inventions that are listed below, physicists and electricians yearned for some practical means of obtaining gain. The earliest demand came from the wireline tele- phone industry, which could have used elec- tronic assistance almost from the invention of the telephone. There were large service im- provements possible, and huge amounts of capi- tal to be saved, by talking more clearly over longer distances with less copper. The emer- gence of wireless communication 20 years later opened another need that begged to be filled, in covering longer distances with less transmit- ter power and smaller antennas.

Volume 20, 2007 1 100 Years of Electronics

PRE-ELECTRONIC AMPLIFI- carried high hopes. The ERS following is a contemporary ac- Huge efforts went into de- count of the results of adding one velopment of devices to generate to a moderately long telephone high-frequency signals, amplify connection: signals of all frequencies, and de- tect radio emissions. Some of these At the Long Distance Office lasted quite a while into the elec- in Pittsburg, of the A. T. & T. Com- tronic age. The invention of non- pany, which is called the Brushton electronic, usually electromechani- Office, at present is installed a “tele- cal, was the centerpiece phone ” which has been of this work. These devices often the result of incessant and studious carried the name “telephone relay,” work of several experts employed in analogy with the relays used in by and under the direction of Mr. . Transmission engi- Hayes of the Engineering Depart- neers anticipated such a device al- ment. Mr. Grace has personally most from the start - Thomas D. tested this repeater and explained Lockwood, later the chief patent to me about as follows: (On a St. attorney for , pub- Louis talk, which Mr. Sherwood lished a three-part article on re- says is a never a talk - operator re- peaters in Electrical World in 1896. peats.) These typically relied on an electro- “Hello Smith!!!! Hello, why magnet fed by the incoming signal, I hear you well tonight.” “Say driving a diaphragm that vibrated that sounds close, doesn’t it?” a mass of carbon granules - effec- “Well, I never talked so clearly tively a telephone receiver driving to you before.” “Are you really a . The Western Elec- in St. Louis?” - whereupon the tric (WE) production versions be- repeater was cut out - “Hello, gan with a design by H. E. Shreeve Smith!!!” “Hello!!” “Hello!!” “I don’t get that.” “Where have from 1903 (Fig. 1). It evolved into 1 the 1A “Repeater” of 1904 (Fig. 2), you gone?” “Hello!!” “Hello!!” the 2A “Repeater Cartridge” of 1910 (Fig. 3), and the 3A “Repeater Car- Despite the enthusiastic re- tridge” of 1912 (Fig. 4). The early view, the poor frequency response and waveform distortion of these devices made them essentially unus- able in tandem connec- tion, a requirement for really long telephone cir- cuits. They received only halting commercial use, reaching a mere 41 units in North America as of 1910. This didn’t pre- vent some use of the re- ceiver-against-transmit- ter principle. The WE 111A was one Fig. 1. Developmental Shreeve mechanical re- such device, used in a peater, 1904. Source: , courtesy Jerry Vanicek. telephone set to boost the

2 AWA Review Sibley

Fig. 2. 1-A developmental “repeater,” 1906. Source: Bell Labs, courtesy Jerry Vanicek. received volume. These were be- similar, but included a 65A Ampli- ing installed as late as the early fier giving gain approximating that ’60s, but solid-state amplifiers soon of one . replaced them. During the ’30s WE In the U. S. amateur-wire- made two versions of its less market, there was a “Multi- “Audiphone” hearing aid. The 36A Audi-Fone” amplifier, offered be- was simply a tween 1915 and 1917 and available driving a headphone. Such an ar- in one- and two-stage versions able rangement can provide 20 dB or so to drive a horn loudspeaker. The of acoustic-to-electrical gain. The Brown Relay from S. G. Brown in 37A and 38A Audiphones were the , and a similar version from Siemens & Halske in , were other electromechanical am- plifiers. One failed attempt to develop an electronic ampli- fier is notable. Peter Cooper Hewitt, developer of full- wave mercury-arc rectifiers, obtained three U. S. patents (682,695-97) in 1901 for a negative-resistance or “booster” telephone-re- peater element for connec- tion in parallel with a tele- phone circuit. The inherent negative-resistance charac- teristic of the arc (increasing Fig. 3. 2-A production “repeater cartridge,” voltage gives less current 1910. Source: Bell Labs, courtesy Jerry Vanicek. flow) might transfer energy

Volume 20, 2007 3 100 Years of Electronics quickly became appar- ent that more effective designs were desir- able. The needs were to produce more out- put for the power fed in, and to occupy less on the air. The result was the high-frequency con- tinuous-wave (CW) al- ternator, as developed by Reginald Fessenden, the Gen- eral Electric Company (GE), and the Fig. 4. 3-A production “repeater cartridge,” 1912. company Source: Bell Labs, courtesy Jerry Vanicek. in Germany (Fig. 6). It was central to into the circuit and provide gain. Fessenden’s development efforts Greenleaf W. Pickard, better for the National Electric Signaling known for the Perikon crystal de- Company, ca. 1906, and later tector, was working in 1904 for the formed the core technology of the WE engineering department, then world-wide of RCA in Boston. He tried using a Coo- per Hewitt full-wave mercury-arc rectifier as an amplifier. The idea was to strike an arc between the and one of the , then to connect the cathode and the other across a telephone circuit. He described the results as “noisy, insensitive”.2 Another electronic amplifier that failed to work out was a mer- cury-arc device developed by H. D. Arnold of the WE laboratories about 1912. It used an electromag- net, energized by the incoming sig- nal, to deflect a stream of mercury ions between two auxiliary cath- odes. It proved to be noisy, un- stable, and short-lived; so the de Forest audion - with some heroic development effort - superseded it before any commercial use (Fig. 5).

PRE-ELECTRONIC RF GEN- Fig. 5. H. D. Arnold’s mercury repeater ERATORS element in its mounting, with deflec- While spark-based radio tion magnets. The active portion is the were the foundation left arm. Source: Bell Labs, courtesy of wireless communication, it Jerry Vanicek.

4 AWA Review Sibley

Fig. 6. Turbine-driven alternator built ca. 1906 for National Electric Signaling Co. for marine use (note steam fittings on right). Output was 3 kW at 100 kHz. Source: Smithsonian Institution annual report, 1908. Communications beginning in the with the devices that are early ’20s. Alternator stations known today. Designed in Den- stayed in regular service into the mark by ca. mid-’40s, and even now the last 1910, it was developed in the U. S. surviving installation, station SAQ in kilowatt sizes by Kilbourne and in Sweden, periodically reaches Clark and the Independent Wire- North America with commemora- less Telegraph Co. Units up to tive transmissions. It seems hardly megawatt level came from the Fed- possible to obtain a radio-fre- eral Telegraph Company.4 Euro- quency signal from a rotating ma- pean makers included the C. chine, but by equipping the rotor Lorenz Company, Berliner- with enough poles and rotating it Poulsen, and Telefunken.5 Modu- fast enough, frequencies above 20 lated by absorbing some of the out- kHz were obtainable in the GE put in a carbon microphone, the machines and up to 60 kHz by the arc transmitter made possible the Goldschmidt units from Germany, regularly scheduled broadcast sta- which doubled the generated fre- tion SJN in San Jose, California of quency by using saturable reactors. 1909. The arc effectuated commu- RCA’s machines produced a power nication between European coun- of 200 kilowatts with reasonable tries and their colonial posses- efficiency.3 sions: between France and The other CW signal-genera- Indochina, and between the Neth- tor technology of early times was erlands and Sumatra. the Poulsen arc transmitter. This is a quasi-electronic device, since PRE-ELECTRONIC AUDIO it operates with a plasma of gas GENERATORS ions immersed in a magnetic field. Besides the dramatic high- However, it certainly doesn’t fit Volume 20, 2007 5 100 Years of Electronics power rotating signal sources for later 3145 Hz, in 170-Hz steps; Fig. RF, there were humbler generators 7.). This practice held from the for audio. The Bell telephone in- mid-’20s to the mid-’40s, when dustry used audio-frequency in- tube oscillators became domi- ductor tone generators in at least nant.6 Finally, the A3 Privacy Sys- three ways. Nearly every tele- tem, a medium-security scrambler phone central office contained a device used on transoceanic high- source of 1000-Hz tone for testing frequency circuits, purposes, “the milliwatt,” which got its five carrier tones from a gen- appeared as a dial-up number. Its erator that produced harmonics of original source was a multi-pole 550 Hz (3250, 3800 . . . 4900, 5450 generator turned by a motor run- Hz).7 The carriers were used for ning from the 48-volt battery sup- frequency-shifting slices of the ply. A solid-state replacement be- voice spectrum in apparently ran- came available in the early ’60s, dom order and thus reducing in- but many of these generators con- telligibility to an eavesdropper. tinued in service until much later. In each of the above cases, an In the era of voice-frequency car- equipment designer made a con- rier telegraph systems of the am- scious choice between rotating plitude-modulated type (using on- machinery and tubes, and chose off modulation), it was the usual the former. Indeed, in the case of practice to obtain the carrier tones the carrier telegraph system, an for the 12 (later 17) channels from earlier, higher-frequency version a generator that supplied harmon- had used tubes to generate the ics of 85 Hz (425 through 2295, channel tones, but the new design

Fig.7. Equipment for six channels (of 12) of a voice-frequency carrier telegraph system, ca. 1925. The rotary tone supply and its backup unit are visible in the extreme lower right. Source: Western Electric Co. 6 AWA Review Sibley went “backward” in its technology. Extending the “telephone re- lay” principle, the General Radio Company made a “microphone hummer” consisting of an electro- magnet driving a tuning fork. A carbon microphone coupled to the fork picked up its vibrations and fed the magnet, giving an oscilla- tor having good frequency stabil- ity (Fig. 8). Versions giving 400 Hz (213-B) or 1000 Hz (213-C) were available beginning about 1930. There was also a unit based on a resonant reed (241-A) for 1000 Hz. Fig. 8. General Radio “microphone Similar units stayed in the GR hummer,” ca. 1930. product line until at least 1951. vices with potential for giving gain 8 ELECTRONIC DEVICES AR- came onto the scene in 1906. In March, , RIVE an Austrian inventor, applied for a The , a patent on the Lieben-Reisz- , appeared in 1905. It cer- Strauss Relay. This originated as tainly operated electronically but a magnetically controlled high- couldn’t amplify, contrary to the vacuum tube, roughly analogous to claims of later patent litigators who the General Electric UV-212 of had tried to make it oscillate. How- 1920 and the GE 2B23 of 1946. ever, it was utterly basic to inven- Converted to a grid-control and tions that followed. Then, four de- changed to mercury-vapor fill, the LRS Relay enjoyed considerable success in the 1913-1918 period in audio amplifiers (the Telefunken EV72), RF amplifiers (Telefunken EV75), and oscillators for receiving CW radio signals (Telefunken ER75).9,10 (Figs. 9 - 11). The German Post Office made sizable use of it in repeaters in its long-distance network. It enabled -Istanbul telephone service in 1915. It served as the nucleus of a 20-watt radiotelephone transmitter,11 and could os- cillate at “wave lengths as short as five or ten meters.” In October, the Ger- mans Max Dieckmann and Fig. 9. Two LRS relays in modern times. Gustav Glage made appli- Source: Jerry Vanicek. cation for a patent covering Volume 20, 2007 7 100 Years of Electronics tube having a filament, a control anode, and an output anode. As proposed then, it probably would have repeated the input signal but without giving useful gain. How- ever, Heintz & Kaufman Ltd. com- mercialized this format in “Gammatron” transmitting (and even receiving) tubes in the early ‘30s. In 1943-44 the Eitel- McCullough Co. experimented with four developmental power tubes of Gammatron form, up to one-kilowatt level.12 Also in November, de Forest ordered the first samples of the gridded Audion from his supplier H. W. McCandless, and had them working in the laboratory in late December (Fig. 12). He filed for a Fig. 10. Telefunken EV75 high-fre- patent in early 1907 and received quency amplifier, 1913. The EV72 audio amplifier had the same appear- it in 1908. He originally saw it only ance, less the slide rheostat at the right as a detector-with-gain, but 1912 and the circular assembly at the base he and his associates found that of the tube. Source: Telefunken cata- the same device was capable of am- log. plifying. So, by the end of 1906 the pos- sibility of “electronics” had become real.

TELEPHONY: THE “KILLER APP” It is possible to argue that the telephone industry - taken in any large country in the world - was a more fruitful application for elec- tronics than early wireless commu- Fig. 11. Basic receiver using an LRS nication. Particularly in North audio stage. (Source: adapted from Pichler.9) America, the telephone business involved far more capital invest- a vacuum magnetic-deflection am- ment than the fledgling wireless plifier. It was a cathode-ray tube trade and, even in the pre-1910 era like those found today, but having when were relatively output anodes on the inside of the scarce, had more daily impact on “screen.” The incoming signal de- the life of the citizenry. flected the beam magnetically be- To make this clear, it is nec- tween the anodes, yielding a push- essary to go into a bit of long-dis- pull output. While probably am- tance telephone technology, par- plifying successfully, it apparently ticularly as practiced in the AT&T did not receive commercial use. /Bell System environment in the In November, Dr. Lee de For- U. S. and Canada. It became clear est sought a patent on a gridless early on that the transmission 8 AWA Review Sibley A sizable advance was the in- vention of the “phantom” circuit. Here two pairs of wire could yield three circuits via the introduction of one-to-one (re- peating coils). Two of the circuits were the “physical” or “side” cir- cuits. The third or “derived” cir- cuit involved using the longitudi- nal path of each of the side circuits via center taps on the side-circuit repeating coils. The four wires were termed a “phantom group.” (The very name “phantom” is an index of the wonderment with which early telephone people viewed the idea.) The phantom cir- cuit enjoyed twice as much copper as the side circuits, and so had somewhat less transmission loss (Fig. 13). The next advance in the pre- electronic era was the invention of “loading,” almost simultaneously by George A. Campbell of the AT&T laboratories and Prof. Michael Pupin of Columbia Uni- versity. Pupin obtained patent protection, and AT&T quickly ac- quired rights to this invention. The principle of loading was to insert additional inductance in series with the wire, thereby offsetting the undesirable effects of the shunt Fig. 12. First form of the gridded capacitance. On overland lines, Audion, possibly the version that was in test in Dec. 1906. Source: ulti- loading was a matter of inserting mately unknown, but this retouched discrete inductors at fixed intervals print is from De Forest Radio Co. files, along the line. In the case of a ca. 1931. “phantomed” line, it became pos- losses of iron wire were unaccept- sible to make a triple loading-coil able for long distances, and that unit that added inductance to both copper wire of large size (up to 165 the side circuits and the phantom mils’ diameter) was indispensable. path. The extra inductance tripled The result was New York - Chicago the characteristic impedance of the telephone service as of 1893. It was line, making it more sensitive to unaffordable to go to thicker wire, leakage across wet insulators, but at a time when the current com- that was unavoidable. modity price of copper was so im- Loading worked on both open portant that weekly industry wire and cables. Phantom opera- magazines like featured tion in cables required twisting two copper quotations on their front pairs of wire together into a “quad” covers. to avoid crosstalk with other pairs

Volume 20, 2007 9 100 Years of Electronics

Fig. 13. Basic “phantom group.” In practice, four ground-return telegraph chan- nels were derived in addition by adding low-pass filters and blocking capaci- tors. Ringing between switchboards used a 135-Hz AC signal on each circuit so as not to interfere with telegraph. in the cable. Loading made pos- In 1912 de Forest, having dis- sible the construction of a covered that his audion was good “stormproof” Boston - New York - for amplifying as well as detecting, Philadelphia - Washington cable sold telephone-repeater rights in by 1913. With extraordinarily his patents to AT&T. This was the coarse cable (10- and 13-gauge) “inertialess,” relatively and phantom operation, Washing- distortionless device that had ton -New York service was possible eluded earlier researchers. There without a repeater. Extension to was talk at the time to the effect Boston required one repeater (Fig. that the Bell System was prepared 14). to pay a million dollars for such a The same technology let device. De Forest realized $50,000 AT&T extend its open wire west- at the time, plus $340,000 a few ward, reaching Denver by 1911. years later, for audion rights. He (The official story is that this was was somewhat ungracious about the farthest expansion, but there the $50,000 in his autobiography, are signs that the company terming it “however small . . . a stretched a bit more and reached morsel of food for a starving man.” Salt Lake City, at least on a dry Yet that figure was, in terms of day.) But that was the absolute today’s money, about a million limit of practical service without dollars - not a bad return for rights multiple gain devices: studies in- to a device as immature as the dicated that it would be possible to audion - gassy, unstable, short- reach the West Coast without re- lived, and a severe distortion gen- peaters, but only by using 220-mil erator before discovery of the qual- wire (or actually copper rod), an ity-improving properties of nega- unaffordable choice at 774 pounds tive grid bias. per mile per wire.

10 AWA Review Sibley cago. The circuits initially had re- peaters at Pittsburgh (Brushton), Omaha, and Salt Lake City (Fig. 15). All three types of repeater - mechanical, mercury-arc, and tube - were tried, with the best results coming from the tube units. After several weeks, repeaters were added at Chicago (Morrell Park), Denver, and Winnemucca (Ne- vada). In 1920 the loading coils were removed to improve trans- mission performance, the added attenuation being made up by add- ing six new repeater points. This was not an inexpensive project: even including the cost reduction from phantom opera- tion, a New York - San Francisco connection occupied 940 tons (125 cubic yards) of copper and about 168,000 glass insulators. The success of tube-based re- peaters led to their wide installa- Fig. 14. A slice of the New York - New tion: by 1919 the total in service in Haven “B” cable, placed in 1914 to the U. S. was about 1000. Many of reinforce the New York - Boston route. these, perhaps 40%, were not as- Wires are insulated with paper strip. sociated with a specific telephone Composition is seven quads of No. 10, 18 quads of No. 13, six pairs of No. circuit. Instead, they were “cord 13, and 18 pairs of No. 16. Circuit circuit” units, associated with a capacity with phantoming is 99 cir- pair of cords on a telephone cuits. Sample courtesy Dan Harley. switchboard at an intermediate point. The operator, in following In any event, it was now pos- routing instructions to “build up” sible to plan transcontinental tele- an overall connection, would use phone service. Construction of a ordinary cords or a cord-circuit new phantom group to close the repeater as directed. transcontinental gap began in 1913 The audion needn’t have been between Denver and Oakland. the device chosen for transconti- This required a new pole line nental service. The Lieben-Reisz- across Nevada and Utah, on which Straus group had tried to interest the four wires were the sole facil- AT&T in the LRS Relay in early ity for most of the distance. The 1913, without success. However, new circuits were ready for test in if de Forest hadn’t preceded them, July, 1914, and commercial service the LRS Relay could have become began in January, 1915. There was the standard repeater device under no New York - San Francisco cir- Western Electric manufacture. It cuit as such; a connection between could have also taken a place in the those places would be “built up” by North American “wireless” world. switching at two or three interme- diate points like Pittsburgh or Chi- Volume 20, 2007 11 100 Years of Electronics continued to proliferate. An example would be direction of artillery fire from an observa- tion aircraft, using a BC-15 spark transmitter and BC-14 ground receiver. The entire class of military “pack sets” of the 1910-1919 period appear to have been spark-and-crys- tal designs.13 Tube receivers, however, were essential in any case where the antenna had to be small (particularly in direc- tion finders) or where ambient acoustical noise was large (es- pecially aircraft installations). Tube transmitters won out, in large part because of their small signal bandwidth. Arc transmitters, although suc- cessful in the 1915-20 era, had Fig. 15. The interior of one of the repeater a specific problem: while they cabinets of the type used at Brushton in generated a “CW” signal (with 1915. See Tyne8, p. 93, for a front view. Source: Bell Labs photo, courtesy Jerry large amounts of phase noise), Vanicek. they offered little suppression of harmonic frequencies. Worse, WIRELESS since keying of the larger units re- While electronic amplifiers quired keeping the arc running and were a make-or-break proposition shifting the frequency, they put out for long-distance telephony, they two sets of interfering harmonics. were “merely very helpful” for ra- Crystal receivers retained a dio communication. It is surpris- niche use into the ’40s, as emer- ing how successful radiotelegraphy gency receivers on ships and life- was in the absence of electronics. boats. Examples include the RCA Spark transmitters and passive (e. Radiomarine AR-8503 (a tube set g., crystal) detectors made for a with crystal option), the RCA workable system, in the company Radiomarine Type B, and the Fed- of very large antennas and young eral 123B. operators who had good hearing. It was helpful that most radio com- NON-ELECTRONIC RADAR munication before perhaps 1915 The common view of radar is was over salt water, which affords that it originated in the ’30s, when ground-wave propagation far bet- the tube art was advancing rapidly. ter than over land, and that the fre- However, a radar of sorts existed quencies were low (typically 20 to far earlier. In 1904, Christian 1000 kHz). Coastal stations ap- Hülsmeyer of Düsseldorf demon- peared in great numbers, making strated, and received patents on, the distance that had to be covered his Telemobiloskop, a device for often 200 miles or less. detecting ships in fog.14 His sys- Even when tube amplifiers tem used the wireless technology began to become common during of the time: a quenched-gap trans- World War One, passive detectors mitter, a rotating horn antenna, a

12 AWA Review Sibley separate receiving antenna moving The making of developmental with the transmitter, and a tubes of pre-World War One vin- detector. It operated at VHF, at tage used wire anodes, with both frequencies variously estimated as ends of the wire brought outside 30 to 600 MHz. The device report- the bulb to allow heating by pass- edly detected ships at distances up ing a current through the anode. to 3000 meters. A later develop- Manufacture also used a direct- ment allowed rocking the anten- current “bombardment” technique nas, thereby measuring the verti- (heat the filament, apply a vigor- cal angle of the received echo and ous plate voltage). The goal was to allowing a crude estimate of dis- heat the internal elements of the tance. His system therefore satis- tube to drive out unwanted gases fied the basic definition of RAdio during vacuum pumping and seal- Detection And Ranging. Sadly, he ing-off. Mass production of tubes was unable to interest investors in as of World War One relied instead backing further development. on passing the tube through a se- ries of coils with high-frequency INDUSTRIAL ELECTRONICS current flowing, thereby inducing Electronic principles greatly large heating currents in the tube extended the span of application of elements. The source of the cur- electrical methods in manufacture rent in the early ’20s was usually a and transportation, slowly domi- quenched-gap spark oscillator, nating the field as development typically in the 40-50 kHz range. progressed. Tube-based oscillators came later One example is that of selec- (Fig. 16). tive heating of materials in manu- Perhaps a comparison of the facturing. The first important user operation of tubes in a broadcast of radio-frequency heating was station vs. that in industrial heat- probably the tube industry itself. ing will show why tubes came into

Fig. 16. An RF-heating “bombarder” in the exhaust-and-sealing area of the De Forest Radio Co. tube plant, ca. 1930. It uses six De Forest Type 500 for perhaps 3 kW output. The immense copper-tubing tank coil visible at left suggests operation at a frequency of several MHz. Note use of table fans for air-cooling. Source: Robert Lozier.

Volume 20, 2007 13 100 Years of Electronics use only slowly. In typical broad- pure tungsten. cast practice, the transmitter was There are numerous other turned on once a day. It was in a uses of high-frequency heating: relatively hospitable location as skin-hardening of gear teeth, au- regards heat, vibration, and swings tomatic soldering and brazing, in supply voltage. It was a matter welding of plastics, etc. The latter of engineering pride (and good re- requires multi-megahertz frequen- lations with upper management) cies that quenched-gap oscillators to extract long life from the tubes. (or motor-generators or mercury- Transmitters located outside cities, arc converters) cannot reach. The before the advent of remote con- “universal” use of high-frequency trol, were tended by a skilled op- heating today is the magnetron- erator whose main job was to ob- powered oven. serve and make adjustments for Another slow incursion of favorable operation. electronics was in the area of the- By contrast, in typical indus- ater light dimming. The simple trial-heating service the heater rheostat dimmers of the earliest would start up many times an days yielded to saturable reactors hour. The environment might be (magnetic amplifiers), in which an hot, dusty, and vapor-filled. Por- inductor wired in series with the table heaters received the shock lights received a variable magnetic and vibration of being moved field from a direct current flowing around the plant. The load on the in a second winding on the induc- oscillator varied widely, depending tor core. From the ’30s on (e. g, on the placement of the heating Radio City ), these non- coil(s) and the work, and on the electronic systems served reliably. point in the heating cycle. The Only with the arrival of silicon con- heater(s) would receive minimal trolled rectifiers (SCRs) in the early attention from the plant electri- ’60s did electronics really take cian, who had to attend all the over. other equipment in the place. A Magnetic amplifiers had quenched-gap oscillator, although other applications. They enjoyed requiring periodic cleaning of the a sort of golden age in the ’50s, in gap, was suitably simple and rug- which their simplicity, ruggedness, ged. One major maker, Lepel High and small size - and the arrival of Frequency Laboratories, adver- high-efficiency core materials - tised a line of quenched-gap oscil- made them attractive for control- lators as late as 1938 (Fig. 17). system applications in aircraft and The situation evolved in later military systems. The eventual years, especially after the wide- maturing of and SCR spread adoption of high-frequency technology largely displaced them. heating to speed production in There are numerous other World War Two. Not long after, applications of industrial electron- tubes optimized for heating use ics - photoelectric control, vari- became available, with extra-thick able-speed drive of motors, weld- copper anodes to store heat dur- ing controllers, etc. - which became ing the unfavorable part of the possible as electronics came of age. heating cycle, grids rated to dissi- pate unusual amounts of power, COMPUTATION rugged construction, and The science of computation thoriated-tungsten filaments to was relatively late in receiving the improve efficiency over that of benefits of electronics. Punched-

14 AWA Review Sibley

Fig. 17. The Lepel product line of quenched-gap RF “bombarders.” The small- est unit operated with input power of five kW. Source: Ad in Electronics, June 1930. card accounting systems capable of Eventually tubes became attractive sorting, calculating, summarizing, for computer systems, such as the and printing became highly so- Electronic Numerical Integrator phisticated from their introduction and Calculator (ENIAC). It used ca. 1890 to their highest develop- 6,550 6SN7GTs, 4,200 6L6s, ment in the ’60s. Analog comput- 2,600 6SA7s, and so on. Comput- ers for direction of naval gunfire, ers were a major impetus for de- using input/output shafts, preci- velopment of long-life tubes and, sion gears, and finely machined in military use, subminiature cams, attained great precision at types. the time of World War Two and stayed in use throughout the big- CRYPTOGRAPHY guns era. These grew in capability to give ever more precise firing so- The enciphering and deci- lutions by including subtle vari- phering of messages was a rela- ables like the temperature of the tively late use of electronics. Pa- gun and its powder, the degree of per-and-pencil methods, mechani- wear of the barrel, and the strength cal “cipher machines” (e. g., the U. and direction of winds aloft. In the S. M309 Converter) and electro- beginnings of digital computation, mechanical encipherment sys- between 1937 and 1946, Bell Labo- tems, e. g., Enigma) largely held ratories built six models of com- the field until the ’50s. One of the puter that used telephone relays high-level Japanese encipherment and crossbar switches, with systems of World War Two relied, punched-tape and teletypewriter quite sensibly in terms of manu- input-output. The largest used facture, on a group of 26-position 9000 relays. Some of these instal- rotary telephone stepping lations stayed in use into the ’60s. switches. Even automatic on-line

Volume 20, 2007 15 100 Years of Electronics cryptography could be done with- Radio Corporation - A Historical out electronics. One such case was Review, 1919-1946,” Electrical the Vernam (AT&T) cipher system Communication, Vol. 23 No. 4 of World War One, in which the (Dec. 1946), pp. 376-398. plain text is punched into paper 5. A. N. Goldsmith, Radio Telephony (New York: The Wireless Press, teletypewriter tape. Another tape 1918), pp. 21-43. contains the non-repeating 6. Long Lines Dept., A. T. & T. Co., encipherment key. An “exclusive- Principles of Electricity Applied to or” logic function using ordinary Telephone and Telegraph Work, telegraph relays combines the 1928 ed., p. 216; also ibid., 1953 ed., message and key into the en- p. 279. crypted signal to go out on the line 7. R. S. Blackshear, “Trans-Pacific or radio channel. A similar process Radio-Telephone Circuits and the at the receiving end extracts the A-3 Privacy Device,” AWA Review message from the incoming signal. 11 (1998), p. 153. 8. G. F. J. Tyne, Saga of the Vacuum The British “bombe” of World Tube (Indianapolis, H. W. Sams & War Two for decrypting inter- Co, 1977), pp. 75-83 and 59-62. cepted German messages 9. F. Pichler, Robert von Lieben – 100 enciphered with the Enigma ma- Jahre Patent chine (using the 4313A trigger Kathodenstrahlenrelais (Linz: tube) was the first important use Trauner Verlag, 2006). of electronics in cryptography. 10. F. Pichler, “The LRS-Relay – Postwar vacuum-tube computers Research and Development in became important in the early and in Berlin, 1905-1912,” ’50s, ultimately followed by the Tube Collector, Aug. 2006, pp. 2- 8. universalization of cryptography in 11. A. N. Goldsmith, Radio Telephony personal computers. (New York: The Wireless Press, 1918), pp. 88-89. CONCLUSION 12. L. A. Sibley, “Eimac and It has been a long path from Gammatrons,” Tube Collector, Vol. nonelectronic telephony and wire- 6 No. 4 (Aug. 2004), p. 22. less to digital signal processors, fi- 13. R. Kleinman et al., “Early Wireless Pack Sets: Spark Hits the Beach,” ber-optic transmission, and simi- AWA Review 16 (2003), pp. 96- lar marvels. We are now in the sec- 131. ond century of that journey. 14. V. J. Phillips, “The Telemobiloscope - an Edwardian REFERENCES Radar,” Wireless World, July 1978, 1. W. F. Cozad (Colorado Tel. Co.), pp. 68-70. Notes on a Trip East - 1904, originally in AT&T Historical Museum, later reproduced by the Telephone Pioneers of America (1950), pp. 45-46. 2. Personal file on 1890-1915 repeater elements prepared by the late Lloyd Espenshied of Bell Labs. Courtesy of Jerry Vanicek. 3. Glen Fuller, “The Alexanderson System for Electro-Mechanical Production of Radio-Frequency Energy,” AWA Review 3, pp. 120- 136. 4. F. J. Mann, “Federal Telephone and

16 AWA Review Sibley ABOUT THE AUTHOR Ludwell Sibley is an antique- electronics writer and editor with 200 or so articles in the AWA Journal, Radio Age, the AWA Review, and local-club publications. Currently emphasizing electron devices, he is the tube columnist for the Journal and edits Tube Collector for the Tube Collectors Association. Sibley wrote the book Tube Lore and served as technical editor of the three-volume series Transmission Engineering produced by Bell Communications Research. Holding BSEE and MBA degrees, he is a registered professional engineer. He obtained an FCC First Class license in 1962 and became an amateur operator in 1987, lately upgrading to Extra Class. He holds the AWA Houck Documentation, Tyne, and President’s Awards. Besides tubes, he collects literature, (civilian and military), and telegraph equipment.

Volume 20, 2007 17 100 Years of Electronics

18 AWA Review Anderson

AWA Review Henry J. Nolte and General Electric’s High Power, Metal Envelope Tubes ©2007 John M. Anderson ABSTRACT INTRODUCTION After World War Early vacuum tubes for radio applications One it became appar- came into this world dressed in glass envelopes. ent that a better Technology for the manufacture of incandes- means of cooling the cent lamps had been well developed and was plate of transmitting immediately applied to their construction. Dur- tubes above about ing World War One many radio tubes were pro- one kW was neces- duced by Western Electric, Westinghouse, sary. Direct water deForest, and General Electric – all with glass cooling was the logi- envelopes. A line of pliotrons (GE’s name for cal choice, and Henry an amplifying tube) with plate dissipations up J. Nolte, a technician to 250 watts was ready for sale by RCA at the in the General Elec- time it was formed in 1919. tric Research Labora- It wasn’t long before there was a cry for tory was given the higher plate dissipation as transmitters of many task in 1918. A tortu- kilowatts were requested. Making existing ous period of develop- tubes larger had little appeal, because the plates ment followed until had to run at still higher temperatures to radi- the successful UV- ate the rejected heat. Fan cooling to keep the 207 evolved about glass from sucking in was necessary for tubes 1925. This tube, more from about one kilowatt upward. Increasing the than any other factor, size came to a halt, after 5 kilowatts, with the led to replacement of UV-208, available for sale by 1922. This tube Alexanderson alter- was stretching the limits and, even before its nators by tube trans- introduction, there was a general realization mitters. Knowledge that a new approach was needed. gained from the de- velopment led to the WATER COOLED TUBE TECHNOLOGY , ignitron One rather obvious remedy was to use wa- and ultimately, in ter cooling, wherein the water flowed in con- 1935, to the highly- tact with the plate as a part of the tube’s enve- successful octal- lope, without the intervening glass. Still implied based metal receiving in this concept, however, was enough of a glass tube, introduced by envelope to allow insulation of the filament and RCA and GE. Henry grid lead-in wires. Large diameter metal-to- Nolte was granted glass seals were needed, an art little developed more than 20 patents at that time. and received the Cof- General Electric was prodded into working fin Award (GE), be- on this problem by the US military.1 It hap- fore his retirement in pened at a meeting held at the Navy’s Bureau 1958. of Steam Engineering Department in Washing- ton, DC, on November 8, 1918. The agenda was

Volume 20, 2007 19 Henry J. Nolte the standardization of transmitting It was found that Nolte had already tubes. Present were GH Clark (ci- patented most of the possibilities.2 vilian consultant for the Navy), T Nolte’s first attempt to build a Johnson, Lt. GH Lewis, Lt. Eaton, “metal” transmitting tube was Mr. Lyng (Western Electric) and somewhat less than spectacular.3 WC White (Vacuum Tube Develop- By December 27, 1918 it was ready ment in the General Electric Re- for testing. He had taken copper search Laboratory). Following the tubing, flattened to an elliptical meeting, Lt. Lewis (Radio Division, cross section so that its internal di- Navy Department) approached mensions were 0.5 x 3.5 inches, and White and asked him to develop a into this he had placed the filament metal tube (water cooled) of large and grid from a standard 250 watt size. When he got back to power tube. Air-to-vacuum seals Schenectady, White assigned were made with sealing wax. The Henry Nolte to the job on Novem- whole assembly was placed in a ber 18, 1918. This task was to oc- crock with tubes bringing in tap cupy a large fraction of Nolte’s time water to fill the crock. After some- for the next five years or more. what more than an hour of outgas- sing with current from filament to HENRY J. NOLTE grid, he did manage to pass 300 mA Henry J. Nolte (1893-1978) had at a plate current at 1200 volts, but joined the General Electric Com- then bubbles in the water started pany in 1913, and moved to the to come from point A on the at- Research Laboratory on April 5, tached drawing. This was not a vio- 1915. He was assigned to research- lation of the basic vacuum in the ers as needed to act as a technician. tube but was enough to scrap the Prior to and during World War One sealing wax approach (see Fig. 1). he worked with x-rays to examine steel welds, high voltage rectifiers, GLASS TO METAL SEALS and submarine detection along Nolte then went to glass to metal with at Boston. seals, shown on the attached sketch And to add some diversity, he of tube C.C.P #1 (copper cylinder worked on the production of NO pliotron).4 Test results were rather gas by electric arcs in air. By 1918 modest, but encouraging. The cop- he was principally working for per cylinder, bored from a solid in- White on problems concerning got, had an internal diameter of transmitting tubes and their test- 0.75 inch, length of 5 inches, and ing. His accumulated knowledge of the diameter of the grid was 0.5 assembly techniques by this time inch. Tests showed an amplifica- and in following years made him a tion factor of about 82. Plate input very valuable member of the team. power was 2.4 kilowatt (1600 volts By the time of his retirement in at 1.5 amperes). Nolte referred to 1958, he had accumulated more the main metal-to-glass seal as a than 20 patents and had received Fuller seal to French glass. Exactly many recognitions, including the what he meant is not certain, but Coffin Award (internal to GE) in usually a similar seal in this time 1936 for contributions to the metal period was made of tungsten and receiving tube that was introduced hard glass (Corning 7052). Silver simultaneously by GE and RCA in solder (alloy of silver and copper) 1935. At the time of this tube’s de- completed the seal from tungsten velopment a survey of the work was to copper. The tube was operated made to see what was patentable. in an oscillating circuit but no men-

20 AWA Review Anderson

Fig. 1. Henry Nolte’s first attempt at a water-cooled transmitting tube. Sketch taken from his Laboratory Notebook #804, pp 142/3. Special Collections, Folsom Library, Renssselaer Polytechnic Institute, Troy, NY. tion is made of the power output. Houston, TX, to listen for signals The bored copper anode was (2700 meters) from his transmit- shown to leak air, so following ter using a metal power tube. tubes, after August, 1919, used a Wallace read every word (925 drawn copper cylinder (Fig. 2). miles) and Flitch reported a good With construction of more signal but too much interference tubes, incremental progress was (1600 miles) to be intelligible. made, until by C.C.P #50 an rf Progress came slowly but surely, power output of 912 watts at 900 and on June 9, 1921, Nolte could meters wavelength was obtained. report that one metal pliotron The date was October 19, 1919, handled 12.4 kilowatts plate input, about one year after the project and 7.95 kilowatts rf output, for an began. Comparison was made to a efficiency of 64 %. DC plate volt- transmitter currently being manu- age was 11.3 kV. But seals re- factured for the Navy. Its six 250 mained a problem until he heard watt pliotrons gave very nearly the of a copper to glass seal devised by same output into a dummy load as WG Houskeeper at Western Elec- the one metal pliotron. tric. Houskeeper had feathered the An important milestone test copper as it entered the glass so took place on January 28, 1920. that the thin copper could yield, Nolte arranged with DC Wallace at and follow the contraction of the Hamline University radio station in glass as the seal cooled. The con- St. Paul, MN, and with TG Flitch at straint, equal coefficients of expan- the Army radio station at Fort Sam sion, was hereby relaxed. In addi-

Volume 20, 2007 21 Henry J. Nolte laboratory had gone through the same thought processes. While WC White in his writings seems to imply that details of the Houskeeper seal were exchanged during 1922, Nolte had already writ- ten in his Laboratory Notebook No. 804, un- der the date of Novem- ber 8, 1920, that WE was using such a seal. The glass that WE used was No. 702P (current designation for this glass is Corning 7720, a borosilicate hard glass). Later in the 1920s GE developed a rugged seal between 705FN glass (currently 7052) and an alloy of iron, nickel, and cobalt named Fernico. In this case the alloy was formulated to give a coefficient of expan- sion which matched that of the glass. It should be noted that Westinghouse Labora- Fig. 2. Henry Nolte’s attempt at a water-cooled tories developed an tube using high temperature seals. Sketch from equivalent alloy which his Laboratory Notebook #804, page 156. Spe- cial Collections, Folsom Library, Rensselaer Poly- they called Kovar. technic Institute, Troy, NY. Making good seals was not a trivial prob- tion, Nolte observed that the West- lem. It caused GE much grief and ern Electric Laboratory had re- was not solved satisfactorily until ported in the Bell System Techni- the mid 1920s. Also, it was found cal Journal for July, 1922, that that the silver solder between tung- they had constructed a metal sten and copper alloyed with the pliotron which gave 10 kW with an copper, causing a change in metal anode dissipation of 26 kW. GE structure which gave leaks. Yet an- benefited from knowledge of the other major problem was the seal but they in turn could help puncture of the glass by high volt- Western Electric with filament age rf fields acting on the gaseous emission problems, so it turned plasma in the tube, the gas com- into a two-way street. The tubes ing from the leak. These many from WE and GE were enough problems led to poor performance similar in appearance and perfor- and early failures. Never the less, mance that it was obvious each this tube was transitioned to

22 AWA Review Anderson

problem, had to be sup- pressed. Frequency was the same as that of the alternator, about 20 ki- lohertz. In fact the al- ternator was used to drive the six tubes. The task was declared com- pleted when 16 hours of traffic were conducted on one occasion, and 10 hours on another, with operators at Nauen, Germany and with Caernavon, Wales, re- spectively.5 visited the GE Research Laboratory in June of 1923. But even before the visit, Marconi was so anxious to see the UV-207 that Irving Weir was sent, with a tube in hand, to his 3. Relaxation at Rocky Point (LI) in 1922. Left to yacht Electra in the right: WC White, HJ Nolte, and WRG Baker, while New York City harbor. installing a transmitter using UV-207s. The object When Sir JJ Thompson was to equal the performance of the Alexanderson (discoverer of the elec- alternator in communications with Europe. Cour- tron) visited the United tesy WM White. States in April of 1923, manufacturing during 1922 as the he made a point of stopping at the UV-207. Research Laboratory, and was EVENTUAL SUCCESS photographed holding the 207 The summer and fall of 1922, (Fig. 5). He also stopped at West- from July 4 to November 1, was a ern Electric Laboratory to have his time of hard work but eventual picture taken with their equivalent success. Walter Baker, William tube. Langmuir fully disclosed the White, Henry Nolte, Irving Weir, tube and its applications in an ar- and Henry Oakley, along with ticle in the October, 1922, issue of people from RCA, spent many Electrical World. It became rather hours at Rocky Point, LI, install- obvious that GE was willing to tell ing a transmitter with which they the world about the development, hoped to equal the Alexanderson even if many of the tubes were alternator for overseas message shipped with attached getter bulbs handling. Six of the metal to counteract leakage of gas pliotrons were operated in paral- through the seals. Elmer Bucher lel to deliver somewhat over 100 at RCA assigned a number to the kW when fed into the station’s tube, UV-207 (Navy number was Alexanderson antenna (see Fig. 4). GG-1971), and it was standardized Parasitic oscillations, a largely new into the RCA line. By no means

Volume 20, 2007 23 Henry J. Nolte

Fig. 4. Rack of twelve UV-207 tubes used at Rocky Point (LI) in the fall of 1922 to replace the alternator in communications with Europe. In actual tests only 6 of the tubes were operative, note disconnected wires to top of some tubes. Schenectady Museum, Schenectady, NY. were all the seal problems solved, tube in the spring of 1923. It re- however. WGY, General Electric’s placed six UV-206s. Further, pioneering radio station at White has commented that the Schenectady, NY, became a test UV-207 was first used commer- bed for many newly-developed cially in a shore to ship transmit- tubes, and in its transmitter the ter at Chatham, MA.6 In the fall of UV-207 was used as a modulator 1923, business was brisk for sup-

24 AWA Review Anderson

Marian, (MA), WGY, KGO,(San Fran- cisco), Fort Douglas, (Utah), to list a few (see Fig. 7). Produc- tion was in some cases months be- hind orders. The four major prob- lems, according to White were: cracked seals, finding the correct exhaust schedule, filament to grid leakage across the glass, and sec- ondary emission Fig. 5. Sir JJ Thomson, discoverer of the electron, from the grid. visiting the GE Research Laboratory in 1923 to see The situation got the UV-207. Left to right: Dr. I Langmuir, Dr. RB to the point where Owens, Dr WCL Elgin, Sir Thomson, and Dr. WD , Presi- Coolidge. Schenectady Museum. dent of RCA, called a plying 207s in transmitters, generally installed by RCA. LEAKAGE PROBLEMS The United Fruit Com- pany had requested con- tinuous wave transmitters (2000 to 4500 meters) from GE for its network of stations in the Caribbean area. Reception was better than spark transmitters in the face of the tropical static. Headquarters was in New Orleans, and coordi- nation of traffic was by a station on Swan Island, centrally-located in the Caribbean Sea (Fig. 6). Sig- nals were received from this station in a number of countries where the Com- pany had banana planta- tions, such as Panama, Nicaragua, and others. This was in the time period Fig. 6 . The final stage, using a water cooled when GE had trouble meet- UV-207 (toward the camera), of a 20 kW ing demands for the UV- transmitter (on the factory floor) used in the 207, principally because of United Fruit Station. It was driven by a UV- leakage problems. Installa- 206. Hall of Electrical History Foundation, tion had also been made at Schenectady, NY.

Volume 20, 2007 25 Henry J. Nolte

Fig. 7. A 10 kW transmitter using one UV-207 at Fort Douglas, Utah, during 1924. The water-cooled tube is on the lower part of the furthermost structure in the photo. Photo is from the Schenectady Museum. meeting in his office in NYC on been delivered, and still the shuttle Feb. 21, 1924.7 Present were continued for another month un- Sarnoff, Koger, Taylor, Prince, til 24 replacement tubes had been Phillips, all from RCA, and Davis delivered.8 United Fruit stations and Beokes from United Fruit. had been converted from spark at From GE there was EP Edwards New Orleans, Tegucigalpa (Hon- (Head of the Radio Department), duras), and Almirante(Panama), Stein (Managing Engineer), GW by that time. When one realizes Henyan (Radio Dept.), WRG Baker that each transmitter takes but one (Designer of 207, the extent to which tubes were Transmitters), and WC White. failing becomes obvious. Venting Sarnoff, who was known to shed his displeasure at how things were his normally-gracious manner on going, GS Davis of United Fruit vis- occasion, lent gravity to the situa- ited Schenectady on February 4, tion. It was decided the UV-207s 1925, and declared they would buy would be hand carried to New Or- no more transmitters unless the leans until tube problems were cost for replacement tubes was re- corrected. duced and, as if this was not dispir- On February 27 Nolte left by iting enough, RCA began to make train for New Orleans with four comments that the Westinghouse tubes. On March 3, RW Larson UV-207 tubes were better than carried four more. Airplanes then GE’s. Ultimately, nearly all the carried tubes to the respective manufacture of 207s moved to countries. The last few miles, in Westinghouse, about 1927, as GE’s some cases, were on the back of a production slowly reduced. Ironi- mule. By March 24, 16 tubes had cally, at the same time GE’s prob-

26 AWA Review Anderson

lems were progressively being solved. Use of the 207 fell until, on February 1, 1940, RCA declared that they would make their own tubes, which was, in fact, in agree- ment with the Consent Decree of 1932 that broke up RCA. It had largely been replaced by other tubes anyway, such as the UV-848 and UV-803.

THE END OF UV-207 PRO- DUCTION AT GE The loss of UV-207 business was not as much a shock as one might imagine. During November of 1922, White asked Nolte to be- gin work on a 100 kW version of the water cooled tube; ultimately it would be called the UV-862.9 It had its share of troubles during development, leaking seals and all, but it came to fruition at a time, early 1926, when problems with the UV-207 were being solved, and Fig. 8. Later version of the UV-207, the solutions were applied to the 1927. Schenectady Museum. higher-power tube as well. This

Fig. 9. A shelf of ten UV-207s in the power amplifier of WEAF’s new 50 kW transmitter at Bellmore, LI, July, 1928. Sixteen 207s were used in the modula- tor. Schenectady Museum. Volume 20, 2007 27 Henry J. Nolte tube, being more expensive than Elmer McArthur (under White, the UV-207, tended to make up for and inventor of the lighthouse revenue loss. Experiments at WGY tube) took over the work and pur- with two of the large tubes gave sued the tube as a solid anode mag- 200 kW output on November 22, netron with a separate axial mag- 1928, and paved the way for the netic field. It was tested at WGY remarkable feat of 500 kW output with the thought that an output of at WLW during 1933, using UV- 500 KW (expected WLW Crosley 862s. The 862 was offered for order) could be obtained. Tests open sale on July 7, 1928, and were not very promising and an shortly afterward, on February 6, “economy program has halted this 1929, assigned an RCA number program.” (see Ref. 11) The (UV-862). axiotron is a good example of a While Langmuir, White and parallel development, in competi- others were working on the 20 kW tion with the UV-207 , that triode metal tube, AW Hull and JH lost out, not from a lack of good Payne were pursuing a “metal” theoretical promise, but from an tube that Hull had named the inability to overcome a host of little Axiotron. It followed the theme problems. of the day, a water-cooled anode, but used the geometry of a mag- ALL-METAL TUBES netron, another tube that Hull had Flushed with success in the de- invented. It was claimed that it velopment of the UV-207 and the would have an output of one mil- 862, the question arose – why not lion watts, 40 amperes at 25,000 make an “all-metal” tube? White volts, at a frequency of 20 kHz, again turned to Nolte in Novem- when fully developed.10 ber of 1931 to kick off the program. The tube consisted of a cylindri- The goal was a low cost phanatron cal metal anode, 30 inches long (low pressure, recti- and 1 and 3/4 inches in diameter fier) having specifications 35 A @ with a coaxial filament of tungsten 6000 V. The concept was a sealed 0.4 inches in diameter. Electron metal can (anode) with a hot fila- action resembled that of the mag- ment in the center of the can’s in- netron with filament-to-anode terior. Refrigerator seal lead-ins current cut off by the magnetic (electrical leads to the sealed com- field from the large filament heat- pressor motor) carried the heater ing current, 1800 amperes. Work current into the can. The first “tin was started in the summer of 1922. can” tube used mercury vapor and The set up for testing was in the was put together with soft solder. local car (trolley) barn where such This was during the first part of high current was available. First 1932. It was a complete failure; the there was the ever-present prob- mercury attacked the solder. lem of leaking through seals (tung- Improvements began with a sten/AJ hard glass), to say noth- shift to argon gas, welded seams ing of outgassing such a large and better feed-thrus from the structure. Development did not go Lamp Department, until one tube well.11 On June 27, 1924 White ran for 500 hours in August of wrote in his progress summary, 1932. After considerable work on “Many detailed troubles, not very the part of Nolte and others, a suc- active.” cessful product emerged, one Developmental progress was which could use either mercury or enough of a disappointment that argon. The idea blossomed into

28 AWA Review Anderson the thyratron and ignitron, both Space Vehicle Division in Philadel- products helping to keep White’s phia. group alive during the Depression. It was the time of the Great De- There were many avenues to ex- pression, and the allowable days of amine. Nolte was pushing for alu- the week for which one was paid minum heat fins, cast directly on was cut from 6 to 5 (Saturday was the can anode. This led to work on still a working day in the 30s) and a 5 kW air cooled transmitter tube, eventually to 4 days. Most still the beginning of a line. Then it was came in and worked days for which discovered that a separate anode, they were not paid. White had left usually carbon, in the vacuum was, the Research Laboratory in 1930 “…better than the can as an anode. with the thought that he could They seem to run cooler [than in a better protect his staff against lay- glass envelope] and stand higher offs. There was bantering between voltages.” White, JD Harndon, Plant Man- At this point the concept of a ager, and Wilson which resulted in “metal” radio receiving tube White pretty much getting his emerged. GE happened to be look- way.13 There were no more petty ing for something novel as they pay reductions, and productivity reentered the consumer receiver (as well as introduction of new market in 1935. George Metcalf products) was high, making it dif- managed the group, composed ficult to launch an argument of principally of Ralph Bondley and damage to the Company. In part, James Beggs, but with help from White’s rebuttal was, “It is obvious, Nolte and others. What came out of course, that a clock card is not a of the project was the immensely- measure of an employee. It is successful metal receiving tube, equally true that in certain types introduced jointly by GE and RCA of work a full time attendance on in 1935. schedule is important. This is a problem which each department EPILOG has to handle, taking into account It is often argued that no per- the facts in each case, applying dis- son is perfect, even if they success- ciplinary measures when neces- fully launched highly visible pro- sary or justified.” In other words, grams to develop high power let me worry about it. The author, transmitting tubes. On April 18, in discussions with previous em- 1932, White, now head of the ployees of White, has found no one Vacuum Tube Engineering Depart- who speaks badly of him. ment, received a letter12 from Pay- Henry Nolte joined William master BW Wilson for the White in 1918 and the two worked Schenectady Plant which listed, together until White retired from “…the number of times employees active management of tube re- in your department have come in search in 1944 and went back to late or gone out early for the week the General Electric Research ending April 9, 1932.” Sixteen Laboratory as a consultant. Nolte people were on the list, and stayed with the Power Tube De- squarely in the middle was HJ partment until his retirement in Nolte. It seems that he came late l958. four times and left early four times. Nolte’s contributions were a He was in good company. On the good knowledge of material prop- list also was George F. Metcalf, erties and the ability to assemble who later headed GE’s Missile and and test new experimental tube

Volume 20, 2007 29 Henry J. Nolte types. He had a strong hand in the ABOUT THE AUTHOR development of water-cooled John M. Anderson was born in Kansas transmitting tubes, metal thyra- City, Missouri and attended local trons, ignitrons, metal receiving schools until entering the U. S. Army tubes, and disc-sealed tubes in l943. During World War II he (lighthouse) to name a few. He was headed a radio-transmitter sometimes referred to as a general maintenance team attached to the practitioner, making contributions Army Airways Communication to the solution of many problems System. After the war he attended the rather than specializing in one University of Illinois, receiving a PhD in in l955. In field. addition to regular studies, he worked as a research associate in the field of NOTES electromagnetic-wave propagation in 1 WC White, GE Research Laboratory ionized gases. In l955 he joined the Notebook #799, p.179. White’s Technical Staff of the GE Research notebooks are at the Antique Wireless Laboratory and investigated in the Association Museum, Holcomb, NY. areas of gaseous electronics, power 2 Conversation with WH Teare. circuit breakers, and electric- 3 HJ Nolte, GE Research Laboratory discharge lamps. For the years l964- Notebook #804, p.142. Nolte’s 70 he taught plasma measurements at notebooks are on microfilm at the the Rensselaer Polytechnic Institute General Electric Research and as member of the adjunct staff. In Development Center, Niskayua, NY, l966 he was Chairman of the and the original notebooks are in the Schenectady Section of the IEEE. Special Collections, Folsom Library, John retired from GE in l987 and is Rensselaer Polytechnic Institute, presently engaged in researching early Troy, NY. radio history. He is a Fellow of the 4 HJ Nolte, Laboratory Notebook American Physical Society and the #804, p. 156, see Ref. 3. IEEE. He also is a volunteer at the 5 WC White, Early History of Schenectady Museum. Electronics in the General Electric Company, page VII-6. Copy at the Schenectady Museum, Schenectady, NY. 6 WC White, see Ref. 5, page VII-7. 7 WC White, Work diary dated Feb. 21, 1924, Special Collections, Schaffer Library, Union College, Schenectady, NY. 8 WC White, Progress Summary Book, Vacuum Tube Development, Metal Pliotron UV-207, AWA Museum, Holcomb, NY. 9 WC White, see Ref. 8, page 142. 10 General Electric Schenectady Works News, November 3, 1922, p 26. 11 WC White, Progress Book No 1,AWA Museum, p91. 12 WC White Personal Papers, Letter BW Wilson to WC White, April 18, 1932, courtesy of WM White. 13 Letter, WC White to BW Wilson, October 13, 1933, White Personal Papers, courtesy of WM White. This article was peer-reviewed.

30 AWA Review Wenaas

AWA Review Marconi vs. De Forest Audion Infringement Litigation Revisited ©2007 Eric P. Wenaas ABSTRACT INTRODUCTION The legal action initiated by the Marconi The infringement Wireless Telegraph Company of America in litigation by American 1914 against both the De Forest Radio Tele- Marconi over the De phone & Telegraph Company and Lee De For- Forest audion was one est for infringement of the Fleming patent by of the most protracted the De Forest audion vacuum tube changed the legal disputes in U.S. course of wireless history in the United States.1 history, resulting in five Oddly, very little has been written about the different lawsuits which details of this litigation, which is all the more spanned a 29 year pe- puzzling given the controversy over the court’s riod. The history of this decisions, the subsequent litigation that ensued, dispute is recounted. Of and the stifling effect of the judgments on the particular interest are development of vacuum tubes and radio in the numerous experi- America. Even the classic book Saga of the ments that the Marconi Vacuum Tube by Gerald Tyne devotes only a side presented to the few scant paragraphs to the litigation.2 Much court during the course of what is generally known about this fascinat- of the first trial to sup- ing piece of wireless history is found in the writ- port their contention ten opinions of several judges contained in three that the audion in- volumes of The Federal Reporter3 , which col- fringed on the Fleming lects and publishes the opinions of cases argued valve when used as a in the federal appellate and district courts, as detector, amplifier and well as the U.S. Court of Claims. Most of the oscillator. Comparisons other writings about this case either paraphrase of the detection effi- the opinions reported there, and/or comment ciency of numerous on them at length—some for and some against— audion detector con- but almost all with a great deal of passion. figurations with both The objective of this article is to go behind the Fleming valve and a the scenes to chronicle not only what happened, pre- but how and why it happened. The author has sented to the court are pieced together a fascinating story based on in- recounted. Also, the formation found in letters and documents from methods by which De Forest and files, a por- American Marconi tion of the trial transcript, the results of dem- demonstrated to the onstration experiments performed both before court that the two-elec- and during the trial, notes made by one Carl A. trode Fleming valve was Richmond (Patents Section, U.S. Navy) in con- able to amplify and os- nection with proposed settlement talks between cillate are explained. Fi- the two adversaries and the government, De nally, the surprise end- Forest’s autobiography, and contemporaneous ing at the U.S. Supreme articles in radio magazines and newspapers. The Court in 1943 is re- focus of this article is on the events leading up counted. to the litigation initiated by American Marconi

Volume 20, 2007 31 Marconi vs. De Forest in late 1914, the resulting trials in- years seeking a better detection cluding the experiments presented technology not already covered by, to the court comparing the sensi- or overtly infringing on, patents tivity of various audion detectors owned by other companies, but in with Fleming valve and Perikon the end neither De Forest nor his detectors, and the judgments ren- many companies ever succeeded in dered in the federal circuit (trial) achieving this goal. The competi- and appellate courts sited in New tion and conflicts between De For- York. A summary of later events est and Marconi, made public both leading up to an historic U.S. Su- in court and in the press, created preme Court decision on June 21, an atmosphere where the two com- 1943 regarding the validity of the panies could not compromise on Fleming valve patent is related in the audion infringement issue— the Epilog. much to their mutual detriment. American Marconi Detector THE COMPETITION FOR Patents. From the outset of Ameri- can Marconi in 1898, the Company NON-INFRINGING DETEC- used the coherer as a detector in its TORS patented receiving system (see Fig. The Marconi v. De Forest 1). Guglielmo Marconi had applied audion litigation was the culmina- for and received his first patent in tion of a fifteen-year competition America covering the transmission between the two companies to find and reception of radio waves— superior non-infringing methods of which included the coherer—dated detecting radio waves for their re- July 13, 1897.4 While Marconi did spective wireless apparatus—the not invent the coherer, he did make detection of “feeble” radio waves and patent the improvements being the key technological prob- needed to make it practical as a lem limiting the range and word- detector in a commercial wireless rate of early wireless systems. system. Even so, the coherer was American Marconi relied on detec- not very sensitive, and it was also tion technologies covered by their temperamental and painfully slow, own patents, which in the end were so in 1902 Marconi began the found not to infringe on those of search for a more reliable and sen- other companies. De Forest, on the sitive detector. In mid-1902, fol- other hand, spent a number of lowing its invention by Ernest Ru-

Fug. 1. From the outset of American Marconi in 1898, the Company used this coherer as a detector in its patented receiving system. (Archives Center, Na- tional Museum of American History, Behring Center, Smithsonian Institution)

32 AWA Review Wenaas therford in 1896,5 Marconi reduced who used it as the basis of a patent to practice the — entitled “Electrical Indicator” is- affectionately called the “maggie” sued on October 24, 1884.7 It was (see Fig. 2). His original patent ap- Fleming who first recognized that plication was given a filing date of this previously observed phenom- November 28, 1902, but it was sub- enon could be used in a new appli- sequently divided, and so the first cation, namely as an oscillation patent to issue (on April 14, 1908) detector in wireless telegraph sys- carries a filing date of Feb. 2, 1903.6 tems, as he described in Claim 37 Marconi used this highly-reliable of his patent: and more-sensitive detector for a “At a receiving-station in a sys- decade or so until it was replaced tem of em- by the crystal detector and vacuum ploying electrical oscillations of tube. high frequency a detector com- While the magnetic detector prising a vacuous vessel, two con- was a superior detector at the time, ductors adjacent to but not touch- , a consult- ing each other in the vessel, a ant for the British Marconi com- means for heating one of the con- pany, invented the vacuum tube ductors, a circuit outside of the detector in 1904 shortly after the vessel connecting the two conduc- maggie went into service (see Fig. tors, means for detecting a con- 3). His U.S. patent No. 803684— tinuous current in the circuit, and assigned to American Marconi— means for impressing upon the was filed on April 19, 1905, but was circuit the received oscillations.” accorded an effective date of No- The patent contained a sche- vember 16, 1904, the filing date of matic diagram of the valve with an his original patent in Britain. This external circuit con- detector used the Edison effect dis- sisting of an antenna and ground covered in 1883 by ,

Fig. 2. The magnetic detector was invented by Ernest Rutherford in 1896. In mid-1902 Marconi reduced it to practice—affectionately called the “maggie”— and it went into service circa 1903. (John Jenkins, Sparkmuseum)

Volume 20, 2007 33 Marconi vs. De Forest new scientific principle or phe- nomenon—it often involves the ap- plication of a known phenomenon or process to solve a new problem. In this case, while the Edison ef- fect was known, nowhere in the prior art had anyone suggested us- ing the Edison effect as a detector of radio waves. Fleming’s detector patent would become one of the most valuable patents in the Marconi portfolio, not because the Fleming valve as it was named was widely used,8 but because more than a decade later the Fleming patent would block the manufac- turing and distribution of De Forest’s audion triode. De Forest’s Responder Detec- tor Patents: De Forest first became interested in wireless telegraphy while a student at Yale where he built a Branley coherer, a detector which did not appeal to him be- cause it was slow and complicated, requiring a mechanical “tap” to re- store it for each succeeding pulse Fig. 3. John Ambrose Fleming, a consultant for the British Marconi com- (see Fig. 5). He immediately rec- pany, invented the two- ognized the need for a self-restor- vacuum valve in 1904 shortly after the ing detector, and carried his inter- maggie went into service. (John est in developing such a detector Jenkins, Sparkmuseum) over to his first job at the Western circuit coupled through an induc- tion coil to a secondary circuit con- nected to the plate and filament containing a galvanometer by which the detected signals could be read (see Fig. 4). The patent stated that the galvanometer could be re- placed by other means for detect- ing the continuous current in the circuit, and while not specifically stated, the indicating device could have been a telephone receiver or earpiece for spark radio telegraphy (or later radio telephony), as Marconi used in Newfoundland in 1901. Fig. 4. The Fleming patent made There are those who have said claims for a simple receiving circuit that Fleming did not really invent which played a key role in the subse- anything new. However, invention quent litigation with De Forest. (U.S. is not limited to the discovery of a Patent No. 803,684) 34 AWA Review Wenaas

between as shown in Fig. 6. Pre- sumably these two patents de- scribed the responder in the re- ceiver he used at the yacht races in 1901 and shown in Fig. 7. Unfor- tunately, the demonstrations at the yacht races were not a success, pri- marily as a result of mean-spirited jamming by a competitor without a sponsor seeking publicity (American Wireless Telephone & Telegraph Co.). Nevertheless, De Forest’s responder was prone to clogging, and so he continued his search for a more sensitive and re- liable detector.

Fig.5. De Forest’s distaste for a Branley coherer such as the one shown here prompted him to search for a superior self-restoring detector. (John Jenkins, Sparkmuseum) Electric Company in Chicago in 1899 after he graduated from Yale. In 1900 he and co-worker Ed Smythe began to experiment with a new “responder,” as he called it. Fig 6. De Forest’s first attempt at de- De Forest left Western Electric in veloping a superior non-infringing de- late March or early April of 1900 tector was the “responder” consisting and continued to develop his re- of two closely-spaced adjustable elec- sponder for another five months trodes with an intervening electrolyte. before he filed his first patent ap- (U.S. Patent No. 716,334) plication dated September 1, 1900, In searching for a better re- which listed Ed Smythe as co-in- 9 sponder, De Forest stumbled upon ventor. one consisting of a mixture of lith- In 1901, De Forest formed a arge, glycerine [sic], alcohol and tin partnership with Ed Smythe and filings placed in the gap between Clarence Freeman to develop a two conductors. Assistant Frank wireless communication system, Butler described the discovery as which was subsequently used in a follows:10 demonstration at the International “One day, while working on Yacht Races off Newport, RI dur- receivers, it was discovered that ing September and October of a salvy mixture of various ingre- 1901. In fact, it was only shortly dients reproduced the signals in before the yacht races that he filed the headphone. The ‘discovery’ a second responder patent which was thoroughly tried out but bears a filing date of July 5, 1901. found lacking in any definite This patent application, subse- merit, although it did get as far as quently divided into two patents to receive a name. It was called issuing on December 16, 1902, de- the ‘goo’ receiver, and I believe scribed a responder consisting of that somewhere in the archives of two with an electrolyte Volume 20, 2007 35 Marconi vs. De Forest cathode. One or two volts of po- tential were applied across these two electrodes, in series with a telephone receiver. During that visit Vreeland confidentially in- formed me that he, and not Pro- fessor Fessenden, was the inven- tor of this novel type of detector.” “Thereupon we resolved to use a Wollaston-wire rectifier detec- tor or its equivalent.” De Forest unabashedly admit- ted that he copied Fessenden’s Fig. 7. De Forest claimed that the work, adding some embellish- device pictured here was the re- ments, to be sure, to improve its sponder he used at the yacht races in performance. However, he never 1901. (Father of Radio, Wilcox and gave Fessenden any credit whatso- Follett Co.) ever, justifying his use without credit in his autobiography by the Patent Office may be found a claiming that the basic concept had formal application for a patent been disclosed by Dr. Michael I. made for it by Dr. De Forest.” Pupin in an article in the Electri- This “invention” appeared as cal World in 1899, and therefore one of several inventions in a pam- was in the public domain—a de- phlet published by the U.S. patent fense, however, that he did not at- office and displayed at the 1904 tempt to use in subsequent litiga- Saint Louis Exposition, which tion with Fessenden.12 opened on April 30, 1904. How- De Forest assistant Frank But- ever, by that time, De Forest had ler later recalled the introduction switched to the electrolytic detec- of their new electrolytic detector: tor described next. “Finally the electrolytic re- De Forest’s Electrolytic Detec- ceiver was introduced. This was tor Patent: De Forest recounts in such an advance over anything his autobiography how he came previously introduced that it upon the electrolytic detector—his seemed to be the height of perfec- next attempt at finding a more sen- tion. It consisted of a small glass sitive and reliable detector:11 cell containing a dilute solution of “Early in the spring of 1903 I caustic potash and water which had visited Prof. Reginald formed one anode of the circuit. Fessenden at this home labora- Into this solution was immersed tory at Fortress Monroe, Virginia. a cathode point, and the incoming There I first met Dr. Frederick wave was rectified by electrolytic Vreeland, his assistant. They were action.”13 using at that time a form of elec- Not surprisingly, Fessenden trolytic detector for wireless sig- filed a lawsuit for patent infringe- nals, which, while resembling ment soon after De Forest de- mine in the fact that both were ployed his version of Fessenden’s anti-cohering devices, neverthe- electrolytic detector. The lawsuit less differed in that it used a dragged on for several years until Wollaston wire (named after the it was finally heard before Judge inventor), dipping into a solution Wheeler in the New York Circuit of dilute acid in which also was Court, Southern District in mid- immersed a carbon or platinum 1905. De Forest’s defenses were 36 AWA Review Wenaas rather pathetic, and were quickly ally lose because in 1904 he fo- dispatched by Judge Wheeler in a cused on developing a “flame de- relatively short opinion dated Oc- tector” consisting of electrodes in tober 16, 1905.14 As a result, De the ionized gas produced by the Forest was enjoined from further flame of a Bunsen burner. This use or distribution of this detector. idea dated back to 1900 when his Apparently undeterred, De For- assistant observed that his labora- est modified the electrode by re- tory flame flickered when it was placing the small-diameter wire exposed to an electrical dis- with a narrow flat plate in the charge.17 He filed a patent applica- shape of a spade shovel as shown tion on February 2, 1905, which in Fig. 8—whence the name “spade was subsequently divided into four electrode.” He filed a patent on the separate patent applications—all new configuration circa December accorded the same effective date as 21, 1905, the filing date appearing the original patent filing. These on his Patent No. 894,317. He con- patents, which became known as tinued to use the electrolytic detec- the “Bunsen burner” or “flame” tor in this form, and so, not sur- patents, are worthy of note be- prisingly, Fessenden took De For- cause they were introduced in the est back to court, this time seek- Marconi v. De Forest litigation and ing “attachment for contempt” for played a major role in the trial and violating the injunction on the its outcome (see Table 1). Apart original configuration. In an opin- from these patents, De Forest filed ion dated April 7, 1906, the judge one additional Bunsen burner ruled that the new configuration patent on January 20, 1906 (No. was also an infringement of 824,638) claiming four electrodes Fessenden’s patents and reaf- (in contrast to the two claimed in firmed the injunction. Apparently, the other patents), which was not even this did not deter De Forest a part of the litigation and there- because he continued to make and fore is not listed in the table. sell the electrolytic detector with the spade electrode at the Navy’s request. Yet once again, Fessenden took De Forest back to court, and this time De Forest and Company President Abram White were found guilty of contempt and fined $15,922 on May 22, 1907.15 While this fine may not seem to be large by today’s standard, the fine was not insignificant—equivalent to about ten years of De Forest’s sal- ary. It was said that De Forest’s attorney advised him to leave the country for a visit to Canada to avoid incarceration until White raised the money to post a bond.16 Fig.8. In an attempt to avoid infring- As a result, White asked for and ing Fessenden’s patent on the elec- received De Forest’s resignation. trolytic detector, De Forest modified De Forest’s Flame Detector the electrode of his detector which Patents: De Forest must have rec- took the shape of a garden spade— ognized early in the litigation with whence the name “spade electrode.” Fessenden that he would eventu- (U.S. Patent No. 894,317) Volume 20, 2007 37 Marconi vs. De Forest The first patent listed in Table 1 was the “parent” patent which, in

dates) the original filing, made so many disparate claims that it was subse- quently divided into four separate patent applications. While the specifications (i.e., description) and all flame configurations (see Fig. 9) were retained in the parent patent, the claims were reduced in number, addressing only a two- electrode version of a detector in which the electrodes which ex- tended into the flame were con- nected to an external electrical cir- cuit. The second patent listed in the table made claims about a device enclosed in a receptacle and a means for heating the gas to make it conductive and to feeble oscilla- tions. The third patent made claims for methods of establishing a gas with suitable molecular and ionic activity, but without claims for the external circuits. The fourth patent made claims about a con- figuration in which the burner bar-

electrodes in an enclosed receptacle containing a heated conducting rel formed one electrode and a

Major Distinguishing Claims as an Oscillation-Responsive Device Major Distinguishing Claims as an Oscillation-Responsive Device

Major Distinguishing Claims as an Oscillation-Responsive Device

circuits including a battery and receiver gas claims for external circuits

Major Distinguishing Claims as an Oscillation-Responsive Device Major Distinguishing Claims as an Oscillation-Responsive Device

1907 Methods of establishing a gas with suitable molecular & ionic activity; no

8, 1907 Bunsen burner barrel used as one of two electrodes

Date Issued Date Issued

Date Issued Dec. 20, 191020, Dec. Two electrodes in a flame (seeded to increase conductivity) with external

Date Issued Date Issued

Filing Date Filing Date

Filing Date

Filing Date Filing Date

Fig. 9. De Forest developed and pat- ented a detector formed by two or more electrodes immersed in the ion- ized gas formed by the flame of a Bun- sen burner. (U.S. Patent No. 979,275)

Patent Number Patent Number

Table 1. De Forest’s Bunsen burner oscillation-response patents referenced in the audion litigation (listed according to filing Patent Number 979,275 Feb. 2, 1905 867,876 April 4, 1906* Oct. 8, 1907 Two 867,877 June 12, 1907* Oct. 8 867,878 July 12, 1907* Oct. * These three divisional patents, while filed later, were accorded the effective date of the parent patent 979,275

Patent Number Patent Number

38 AWA Review Wenaas single electrode in the flame Fleming’s patent. Four of the six formed the second electrode. most relevant configurations from The essence of De Forest’s in- this patent are shown in Fig. 10. vention common to all his flame Like the original flame patent, this patents is captured in the last claim filing too was subsequently di- of the patent 867,877: vided—the second filing accorded “As an improvement in the art the same effective date as the origi- of receiving electromagnetic sig- nal patent application. De Forest nal waves, the method herein de- filed a two-electrode patent (No. scribed which consists in absorb- 837,901) that was mentioned in ing the energy of said waves, im- the lawsuit, but it was withdrawn pressing the resulting electrical before the trial began, and was not oscillations upon a gas main- relevant to the litigation. The three tained in a condition of molecular patents are listed in Table 2 with activity, thereby creating current the key claims that distinguished variations in the circuit of a one from another. There was one source of electromagnetic force additional two-electrode patent associated with said gas, and filed by De Forest on May 10, 1906 translating said current varia- (No. 863,071) in which an external tions into signal indications.” power source was applied to the It should be noted that certain two external electrodes in order to key claims in the De Forest flame heat the intervening gas, but this patents do not appear in the quite impractical configuration Fleming valve patent, and vice was not referenced in this lawsuit. versa. In the De Forest patents there are no claims for an incan- descent electrode from which elec- trons are emitted, or for a vacuum—both critical claims in the Fleming patent. In the Fleming patent there was no claim for “a gas maintained in a condition of molecular activity”—to the con- trary, Fleming claimed the need for “a vacuous vessel.” De Forest’s Two-Electrode Vacuum-Tube Detector Patents. Shortly after Fleming published his electric valve paper in the Pro- ceedings of the Royal Society of dated March 16, 1905, De Forest began to experiment with detectors utilizing two electrodes in a glass enclosure—a configura- tion almost identical with that of Fig. 10. De Forest’s first two-electrode Fleming, except that De Forest audion configurations were remark- claimed “a receptacle enclosing a ably similar to the Fleming valve. (U.S. sensitive gaseous conducting me- Patent No. 824,637) dium,” whereas Fleming claimed “a vacuous vessel.”18 De Forest’s first patent application was filed on January 18, 1906, well after

Volume 20, 2007 39 Marconi vs. De Forest It should be noted that in these patents De Forest departed from

ording the language of his earlier flame patents predating Fleming’s work, and instead began to make claims similar to those of Fleming regard- ing filaments and incandescence— even though incandescence was not supposedly necessary for De Forest’s detector, which he claimed required heating of the gas only to the degree necessary make it a “sensitive conducting me- dium.” In a later patent (No. 841,386), De Forest would go even further, making claims almost identical to those in Flemings patent—to include an evacuated vessel enclosing a filament (which by definition, was incandescent): “An oscillation-detector compris- ing two conductively-connected flat plates of conducting material inclosed within an evacuated ves- sel, a metallic filament sealed within said vessel and located be- tween the plates, and means for heating said filament.”19 It must have been obvious to everyone— except perhaps the oblivious patent examiner—that De Forest was appropriating piecemeal

Major DistinguishingMajor Claims Distinguishing as an Oscillation-Responsive Claims as an Oscillation-Responsive Device Device

Major Distinguishing Claims as an Oscillation-Responsive Device

Major DistinguishingMajor Claims Distinguishing as an Oscillation-Responsive Claims as an Oscillation-Responsive Device Device drawn form the lawsuit) Fleming’s valve detector patent.

De Forest’s Three-Electrode Audion Detector Patents. In the

4, 19064, Two electrodes, one heated and one excited by an external electromagnet (with- summer of 1906, De Forest began 26, 190626, Two electrodes, either both or neither of which were heated to experiment with vacuum tubes utilizing three electrodes, experi-

Date Issued Date Issued

Date Issued

Nov. 13, 190613,Nov. Two electrodes, only one of which was heated; first reference to “incandescence” Date Issued Date Issued ments which resulted in three pat- ents with filing dates between Au- gust 27, 1906 and January 19, 1907 (see Table 3). Two of the three pat- ents made claims for detection of feeble electrical oscillations, while

Filing Date Filing Date

Filing Date Filing Date Filing Date the third patent (No. 841,387) made claims for amplification of feeble electrical currents.

Patent No. Patent No. Table 2. De Forest’s two-electrode, partially-exhausted-receptacle detector patents referenced in audion litigation (listed acc to the original filing dates) Patent No. 824,637 Jan. 16, 1906 Jun. 836,070 May 19, 1906* Patent No. Patent No. 837,901 Feb. 14, 1906 Dec.

* This divisional patent was accorded the effective date of the parent patent 824,637

40 AWA Review Wenaas

The first patent listed in the table is notable because it was the first to make claims for three elec- trodes—although the control elec- dates) trode was placed outside the tube where it was not very effective. The three-electrode devices described in this patent used same two-elec- trode device of previous patents, with a third electrode formed by wrapping a wire or conductive sheet around the outside of the partially evacuated receptacle as shown in Fig. 11.20 It was also the first patent in which De Forest used the term “audion”—literally translated as “audible ion—a term which De Forest credits to his as- sistant Clifford D. Babcock.21 De Forest popularized the audion name with a two part ar- ticle in the Scientific American in late 1907 where he described sev- eral of his two- and three-electrode Major Distinguishing Claims for Three-Electrode Patents Major Distinguishing Claims for Three-Electrode Patents which were external to the tube; first use of term “audion” in a patent electrodes but no grid plate electrodes Major Distinguishing Claims for Three-Electrode Patents Major Distinguishing Claims for Three-Electrode Patents Major Distinguishing Claims for Three-Electrode Patents 1908 Claims for detection; first use of a grid structure positioned between filament and 15, 190715, Claims for both detection and selective tuning; first claim three electrodes Date Issued Date Issued Date Issued Date Issued Date Issued 1906 Jan 15, 1907 Claims for amplification of feeble signals; various configurations with two and three Filing Date Filing Date Filing Date Filing Date Filing Date Fig. 11. De Forest’s first three-elec- trode configurations consisted of a two-electrode vacuum tube with the third electrode formed by wrapping a wire or conductive foil around the ex- ternal surface of the tube. (U.S. Patent No. 841,386) Patent No. Patent No. Table 3. De Forest’s three-electrode detector patents referenced in audion litigation (listed according to the original filing Patent No. 841,387879,532 Oct. 25, Jan. 19, 1907 Feb. 18, 841,386 Aug. 27, 1906 Jan. Patent No. Patent No.

Volume 20, 2007 41 Marconi vs. De Forest detectors as , claiming they all worked on the same principle.22 In the same article he also charac- terized his early flame detectors as audions: “This early form of audion, the flame receiver, was remarkably sensitive to weak high-frequency oscillations.”23 The second patent listed in the table was the only one making claims for amplification, although none of the configurations shown in this patent including the two Fig. 13. The famous patent with shown in Fig. 12—as opposed to audion configuration containing grid the later grid configuration—made electrode between the plate and fila- a very efficient amplifier. The con- ment made claims only for detection— figurations shown in this patent not amplification, the application for were never used in practice, but the which the device became famous. claims of a filament and two addi- (U.S. Patent No. 879,532) tional electrodes connected to in- rations shown in De Forest patent put and output circuits for ampli- No. 841,387 for amplification of fication were formidable. feeble electrical currents (none of which utilized a grid) were histori- cally insignificant, while the grid configuration described in his patent for detection of electrical signals is almost universally re- garded as one of the most impor- tant inventions of all time—not because the device covered by the patent was a more efficient detec- tor (which it was)—but because it was able to amplify signals. In fact, it was only much later that De For- est and others came to understand the greater significance of the in- tervening grid as an amplifier rather than as a more sensitive de- tector. A De Forest audion with the Fig. 12. The audion electrode configu- grid configuration from this later rations shown in De Forest’s only period is shown in Fig. 14. patent for devices intended to amplify weak signals were not practical and THE CALM BEFORE THE were never used in wireless commu- nications. (U.S. Patent No. 841,387) STORM While De Forest filed his three- The third patent listed de- electrode audion detector patents scribes the three-electrode device on August 27, 1906 and January in which the third electrode—des- 19, 1907, it was more than seven ignated the grid—was placed be- years later in mid-1914 that Ameri- tween the emitting filament and can Marconi took legal action collecting plate (see Fig. 13). It is against De Forest for infringement somewhat ironic that the configu- of the Fleming patent. Why Ameri-

42 AWA Review Wenaas

While it was generally recognized as being the most sensitive detec- tor available, it was also regarded as a laboratory curiosity—unsuit- able for either commercial or gov- ernment work. According to an article by Marconi engineer Elmer E. Bucher in The Wireless Age in 1914: “While the Fleming valve is not quite so sensitive as the Audion— that is to say, will not give the same intensity of signals pro- duced by the Audion under simi- lar conditions—yet it is not so dif- ficult to adjust, and apparently Fig. 14. The grid structure of the has longer life. Consequently, audion can be seen clearly in this pho- from the standpoint of commer- tograph of a spherical audion manu- cial practice, the Fleming valve is factured circa 1914. (John Jenkins, Sparkmuseum) the more desirable of the two.” “Commercial equipments should not be encumbered with can Marconi waited seven years to ‘laboratory’ devices. The appara- take legal action, and what tus should be capable of easy han- prompted them to take legal action dling, quick adjustment and pro- when they did is an interesting tected against loss of sensitive- question that is addressed in the ness. The amplifier, in the writer’s following paragraphs. opinion, is more or less a ‘labora- The most likely reason that tory’ device of considerable scien- American Marconi did not take tific interest.”24 action for over seven years was that Marconi was not the only orga- it did not initially view the audion nization to eschew the audion. Ac- as a threat to its core business, nor cording to the Manual of Wireless did it deem the audion suitable for Telegraphy for the Use of Naval commercial work. At the time of Electricians for 1915, the crystal the invention, United Wireless was detector—not the audion—was the American Marconi’s largest com- standard for all military receivers: petitor, and while De Forest had “There is but one type of detector patented the three-electrode now in general use, viz. the Crys- audion during the period of his em- tal or rectifying detector.”25 ployment at a United Wireless sub- While the situation with the sidiary, De Forest was forced out audion itself had not changed in of this subsidiary in late 1906, and the intervening seven years since was able to take this patent with its discovery, other events oc- him. Thus, Marconi’s largest com- curred—some or all of which may petitor did not have access to the have caused Marconi to act in 1914. patent, and De Forest in his new First, the litigation against De For- enterprise, Radio Telephone Com- est may have been the natural evo- pany, did not have the economic lution of a new and aggressive wherewithal to compete directly policy of legal action against poten- with Marconi. Even more to the tial infringers instituted by Godfrey point, the audion was tempera- Isaacs when he became managing mental, fragile and expensive.

Volume 20, 2007 43 Marconi vs. De Forest director of British Marconi in 1910. ditional cost or complexity. The His first action was to sue British patent was issued on October 6, Radio Telegraph and Telephone 1914, and it was not long before Company in England for infringe- amateurs had begun to build re- ment of the now-famous 7777 ceivers with the Armstrong circuit tuner patent in 1910. The success- using the audion, a fact well known ful outcome of this litigation to American Marconi.26 Clearly, emboldened Marconi to sue Marconi would have been at a se- United Wireless in the U.S. in 1911 vere disadvantage if its commercial for the infringement of the same wireless competitors had access to patent. United Wireless had no the audion and the Armstrong cir- defense, and weakened by crimi- cuit while Marconi did not. To fore- nal prosecution of key officers, stall the use of the Armstrong re- they pleaded guilty with the result generative receiver by a competi- that the assets of United Wireless tor would have been sufficient rea- were absorbed by American son alone for Marconi to take le- Marconi in 1912, thereby making gal action against the audion—an Marconi the largest wireless com- essential part of the Armstrong cir- pany in America. cuit at that time. Litigation against the National Third, the litigation may have Electric Signal Company (NESCO) been an indirect result of Marconi began on May 3, 1912, followed absorbing United Wireless assets— rapidly by a second suit on May 23, including a number of shipboard 1912. Judgment was rendered on wireless stations. The audion March 19, 1914, in which the va- would have significantly increased lidity of Marconi patents was af- the operating range of the numer- firmed in what Marconi character- ous ship stations Marconi inher- ized as a “sweeping victory.” In ited from United Wireless, and reporting the results of this litiga- would have undoubtedly reduced tion to the shareholders at the an- the number of shore stations re- nual meeting in April 20, 1914, quired to communicate with these management made the following new stations. In support of this statement: “I may state further, in conjecture, it should be noted that behalf of the directors, that mat- shortly after taking over United ters are actively in hand to bring Wireless assets, American Marconi all infringers to terms which will sent Elmer Bucher to sea to quan- be satisfactory to the best interests tify improvement in reception ob- of the stockholders of this com- tainable with various audion cir- pany.” Since the De Forest Radio cuits. He reported in The Wireless Telephone and Telegraph Com- Age in early 1914 that he could sub- pany was the only other wireless stantially increase the effective company of any import in America range of a 2-kw transmitter at sea at the time, this announcement to 900 miles with two audions, one may have been a harbinger for the acting as a detector and the other suit against De Forest later in 1914. as an audio amplifier.27 It was Second, the litigation may have about this time that American also been prompted by Marconi began to use the audion Armstrong’s discovery of the re- in its commercial receivers—and in generative detector circuit in doing so, infringed on De Forest 1913—requiring the use of a tri- patents—as they later admitted in ode—which greatly increased the court. It may well be that Marconi sensitivity of a receiver at little ad- hoped that litigation against De

44 AWA Review Wenaas

Forest for infringement of the Marconi proposition, buttered Fleming patent would deliver the though it was by an open threat rights to the three-electrode of litigation by the Marconi Co. if audion through a negotiated settle- rejected.” ment for damages resulting from “Shortly after the above men- eight years of infringement—which tioned conference a suit was would have been substantial. brought by the Marconi Co. The For whatever reason or reasons De Forest Co. found with little dif- Marconi ultimately decided to pur- ficulty that the three-electrode sue audion infringements after al- Audion – containing filament most eight years of peace over their plate and grid was being exten- respective patent rights, the open- sively but secretly used at various ing shot occurred when Marconi important Marconi stations, the attorneys from Sheffield & Betts super-sensitive receivers there approached officials of the De For- required being simply inoperative est Radio Telephone & Telegraph without the Audion. Whereupon it Co. in the summer of 1914 in promptly brought counter-suit which—according to a later De under several of the de Forrest Forest cablegram to his factory Audion patents [841,387 and circa September 1916—Marconi 837,901]. “Caught thus with the “proposed that this company [De goods’ the Marconi attorneys Forest] give to the Marconi Co. an early in the litigation admitted unlimited license under the De validity and confessed judgment Forest Audion patents, in ex- under the ‘three-electrode Audion change for a license under the patents, but continued neverthe- Fleming Valve patent.” That less to manufacture and use Marconi proposed a cross-licens- them.” ing agreement would indicate that While not specifically men- Marconi was more interested in tioned in this memo, De Forest obtaining rights to use the audion also countersued alleging that the than they were worried about com- Fleming valve infringed other De petition from De Forest with its Forest patents including four flame audion in either commercial work detector patents (No. 979,275 and or government contracts. This is divisional patent Nos. 867,876, not too surprising since most of De 867,877, and 867,878), two two- Forest’s business was selling electrode patents (No. 824,637 and audions as part of a detector unit divisional patent No. 836,070), or receiver, while most of and one three-electrode audion Marconi’s business was leasing patent (No. 841,368). At the open- and operating commercial com- ing of the trial, Marconi confessed munication links rather than sell- judgment to certain claims of the ing tubes or equipment. other two De Forest three-elec- De Forest went on in the cable- trode audion patents (Nos. 841,387 gram to recount what happened and 879,532), and De Forest with- next: drew from issue two-electrode “Inasmuch as this proposal patent No. 837,901. The litigation smacked too much of something- was on and there would be no turn- for-nothing,—a license to use a ing back and no negotiations. device of utmost proven value in exchange for one of absolutely no THE TRIAL OPENS commercial application, the De The claims in this case brought Forest company declined the by plaintiff American Marconi

Volume 20, 2007 45 Marconi vs. De Forest against both De Forest as an indi- patents were accorded an effective vidual and the Company were rela- date predating that of the Fleming tively simple—infringement of patent. Of particular concern to Claims 2 and 37 of Fleming patent Marconi had to be claim 18 of 803,684 by the audion when used patent 867,876: “An oscillation as a detector. According to govern- responsive device comprising a ment analyst Carl A. Richmond, receptacle, two separated elec- both Marconi and De Forest trodes enclosed within said recep- quickly agreed that De Forest’s tacle, electrical means whereby potentially infringing circuit was the gaseous medium intervening represented by the one shown in between the electrodes is main- Fig. 15. According to Marconi tained in a condition of molecular memos, this circuit corresponded and ionic activity and means for to the one De Forest had been sell- impressing electrical oscillations ing in its receivers and audion con- upon said gaseous medium.” trol boxes. Counterclaims relating to later De Forest denied all allegations De Forest patents 824,637, of infringement and to resist the 836,070, and 841,386 were limited Marconi attack responded with for the most part to Marconi’s use counterclaims that were later de- of the Fleming valve in a circuit scribed by an appellate judge as with an external “B” battery be- “practically a separate action.” De tween electrodes which did not Forest alleged that Marconi in- appear in the Fleming patent. The circuit with the Fleming valve used by Marconi which De Forest al- leged infringed on its patents is shown in Fig. 16. This circuit was the one used in the Fleming valve receivers that became regular equipment in most of the impor- tant Marconi land stations and many of the trans-Atlantic steam- ers beginning in 1912.28 American Marconi was clearly Fig. 15. Marconi alleged that the De concerned about the validity of the Forest three-element grid audion in- Fleming patent because of its fringed on the Fleming valve when it used the audion as a detector in the configuration shown here. fringed on nine of its patents in- cluding four flame patents, three two-electrode audion patents, and two three-electrode audion pat- ents. Marconi confessed judgment to certain claims in the three-elec- trode patents, and they were dropped from the litigation. Of the seven remaining patents, counterclaims relating to the four Fig. 16. De Forest counter-claimed flame patents were much more that Marconi infringed on the De For- extensive than the three two-elec- est audion when it used the Fleming trode patents because the flame valve in the configuration shown here. 46 AWA Review Wenaas overly-broad claims, which in- the Fleming valve, and that the cluded not only rectification at ra- audion worked in spite of the gas dio frequencies—the real essence and not because of it. and purpose of the invention—but In order to support its conten- also rectification at low frequencies tions, Marconi staged rather elabo- which duplicated disclosures made rate demonstrations for the Court in the original Edison patent. In an involving “flame audions,” two- attempt to rectify this situation, the electrode audions, and a three- Marconi Company filed a dis- electrode audion with the exterior claimer with the patent office be- described in the pat- fore the trial began to limit the ents De Forest introduced as evi- scope of the patent to a detector of dence of infringement by Marconi. waves as follows: These demonstrations certainly “To the combination of ele- had an influence on the outcome ments set forth in Claims 1 to 6, of the trial, and are of interest to inclusive, and 10 to 15, inclusive, the wireless collector and historian respectively, of said Letters because they appear to have been Patent, except as the same are the only time and place where the used in connection with high fre- sensitivities of numerous audion quency alternating electric cur- configurations were quantified and rents or electric oscillations of the directly compared. The demon- order employed in Hertzian wave strations compared a Fleming transmission, and to the words in valve and a sensitive Wireless Spe- the specification: ‘whether of low cialty Apparatus Perikon crystal frequency or,’ at page 2, lines 32 detector under identical condi- and 33; ‘either’,’ at page 2, line 98; tions—and with the appropriate and ‘or low-frequency alternating protagonists present as observers currents of,’ at page 2, lines 98 and commentators. As far as this and 99" author is aware, the results of these The trial itself took a very long demonstrations have never been time, and there was a large volume chronicled or published—appear- of expert testimony. Evidently the ing only in the trial transcript. judge was exasperated with “the floundering of the experts, who MARCONI DEMONSTRA- became somewhat tangled up, so that it was possible by selecting TION EXPERIMENTS PRE- remarks made by the experts on SENTED TO THE COURT either side to make rather a strong In preparing for trial, Marconi showing against that side.” The engineers performed a large num- key technical issue was whether or ber of experiments, a subset of not the audion principle was fun- which was presented to the Court damentally different from that of during the month of February the Fleming valve. De Forest in- 1916, and documented in the tran- sisted that the audion required a script of the trial.29 The experi- gas which was ionized by heating, ments included a wide range of whereas the Fleming valve needed flame electrode configurations and only a vacuum gap across which materials, flame salts, vacuum electrons were projected from the tube configurations, tube elec- incandescent filament emitter to trodes and filaments, gas types, the collecting plate. Marconi in- and vacuum levels. In all cases, the sisted that the audion principle basic approach was to 1) generate was essentially the same as that of an RF signal with a buzzer circuit

Volume 20, 2007 47 Marconi vs. De Forest of the type used to adjust the crys- phone was assigned the number tal detector on commercial receiv- “1.” To determine the audibility of ers, 2) couple the signal to a sec- a stronger signal, the resistive net- ond circuit containing inductance work was adjusted until the signal and capacity which connects to the was once again just audible in the usual secondary of a radio circuit headphones, and the number on capable of tuning to the frequency the dial indicated how much stron- of the excitation, 3) connect the ger the signal was relative to one detector under test to terminals that was just audible. For example, provided for that purpose and if one had adjusted the resistive quantify the resulting rectified sig- network such that a stronger sig- nal appearing across a set of stan- nal disappeared at an audibility dard earphones paralleled by an level of 10 on the dial, then the sig- audibility meter, and 4) vary pa- nal would have been 10 times the rameters as appropriate to quan- minimum signal in the head- tify the variations in detector sen- phones discernible to the listener. sitivity. In order to relate the audibility The audibility meter such as the results to a known quantity, the one shown in Fig. 17 was relatively Wireless Specialty silicon-arsenic standard device consisting of a crystal detector identified in the variable resistive network placed in transcript as type P162B, No. 121 parallel with the headphones so was used as a means of compari- that some fraction of the audio sig- son (without battery bias, which nal flowed through the resistive according to the testimony of WSA network. The resistive network president Greenleaf Whittier was adjusted until a particular sig- Pickard, would have increased the nal was just audible, and the audi- sensitivity by a factor of two to bility level could be directly read four). The audibility for this crys- from the calibrated scale. The dial tal detector using the strongest was calibrated such that a signal buzzer signal was approximately that was just audible in the head- 1000. According to the testimony of Marconi engineer Roy Weagant, the loudest buzzer signal corre- sponded to a very large signal with a level typical of what would have been produced by a large 100 kilo- watt station at a range of about 200 miles. Test Results for Flame Audions of Parent Patent 979,275. All of the audion flame configurations cov- ered by De Forest patents were defined in the parent patent 979,275, the figures of which are reproduced in Fig. 9. Demonstra- tion tests for Figs. 1, 3, 4, 5 (but not Fig. 17. Marconi used an audibility Figs. 2 and 6) of the patent were meter similar to the one shown here performed for the Court at 57 to quantify the detection efficiency of Duane Street in New York City, the the Fleming valve, various audion con- shop of the Marconi Company.30 figurations, and a Perikon crystal de- When asked why he did not test tector used as a reference. (Wireless Specialty Apparatus Catalog, 1919) configurations of Figs. 2 and 6,

48 AWA Review Wenaas

2. Audibility could be in- creased to a maximum of about 10 with the introduction of salts to increase gas conductivity. 3. Audibility was not affected by whether or not the electrodes were heated to incandescence. 4. Audibility was only moder- ately sensitive to external voltage. 5. Maximum audibility for the reference WSA crystal detec- tor using the same excitation cir- Fig. 18. The Bunsen burner used by cuit—but with no external bat- Marconi in the demonstrations to the tery—was 1000. According to court during the litigation now resides in the Museum of Science and Indus- Picard testimony, audibility would try in Chicago, IL. (Archives Center, have been 2 to 4 times higher if a National Museum of American History, battery had been used to adjust the Behring Center, Smithsonian Institu- bias point. tion) Two major points were made to Weagant testified: “Well, I was not the Court about the results of these able to figure any way of building tests. First, the sensitivities of all the apparatus and I could not con- the flame detectors were extremely ceive of any arrangement of low. The signals from the unsalted things which had the slightest pos- flame detectors were three orders sibility of working; so I finally of magnitude less sensitive than gave it up.” The Bunsen burner the crystal detector, and they were used in these flame demonstra- barely audible—registering a “1” on tions is shown in Fig. 18, and cur- the audibility meter—in spite of the rently resides in the collection of large signal from the buzzer. The the Museum of Science and Indus- salted flames were marginally try in Chicago, IL. more sensitive by at most a factor Weagant testified that tests in- of ten, but still not sensitive volving a very large number of enough for use in a wireless sys- variations in parameters had been tem. The Marconi side emphasized made, and best of the results were the low sensitivity of the flame de- to be presented to the Court. Pa- tectors by pointing out that virtu- rameter variations included elec- ally any dissimilar conducting ma- trode materials (copper, brass, sil- terial would produce about the ver, and platinum), dimensions same results. Roy Weagant then and geometries (from a small wire demonstrated that by pinching the of No. 24 diameter to a large plate), electrodes with his fingers he was electrode locations, battery voltage able to use them as a detector of (0 to 500 volts), and salt type used RF signals which registered the to increase flame conductivity. same on the audibility on the The results of these flame meter as any of the unsalted flame audion tests can be summarized as audions! follows: Second, the sensitivity of all 1. Audibility for all flame con- flame detectors was found to be figurations without salts was ap- absolutely independent of whether proximately 1—that is to say, or not the electrodes were heated barely audible—even with the to incandescence. This point—well strongest buzzer signals. noted by Judge Mayer—was im-

Volume 20, 2007 49 Marconi vs. De Forest portant because the Fleming valve tested because Marconi’s expert required incandescence for opera- glass blowers claimed it was im- tion whereas incandescence had possible to make. no part in the operation of the Results for the four remaining “flame” audions. In stark contrast, configurations were most interest- incandescence was required for the ing. The configuration represented tube audions, suggesting the op- by Fig. 4 of the patent did not work eration of the flame audion as a under any circumstances—which detector was significantly different was not surprising because there from that of the tube audions— was no source of heat for either quite contrary to De Forest’s con- electrode or for the gas in the tube. tention. The results for the remaining three De Forest never made a claim configurations (Figs. 1, 2, and 3 of in any of his flame patents that the the patent) were virtually identi- flame audions required incandes- cal—independent of whether one cence of the electrodes, and indeed or two electrodes were heated: Weagant testified that the elec- 1. None of the three configu- trodes shown in De Forest’s flame rations worked as a detector under patents were too large to be heated any pressure conditions unless at to incandescence by the flame. least one of the two electrodes was Weagant was able to achieve in- heated to incandescence (i.e., there candescence only by using fine was no audible signal in the ear- platinum electrodes that could not phones). conduct heat away so rapidly, or in 2. None of the configurations one case a small platinum cup worked as a detector when the gas holding salts which was inserted in pressure was significantly higher the center of the flame. Despite the than that used with incandescent fact that there is no evidence what- lamps. Furthermore, the filaments soever that De Forrest ever heated quickly burned out. Tests at higher his flame electrodes to incandes- pressures were only possible with cence, nor was there any mention nitrogen gas, which precluded the of incandescence in his flame pat- filaments from burning out, but ents, by mid-1906—not long after never produced any signal detec- the Fleming patent issued on April tion. 19, 1905—De Forest began to claim 3. The three configurations that the electrodes in his earlier only worked as a detector at low experiment had been heated to in- gas pressures when one or two of candescence.31 the electrodes were heated to in- Test Results for Two-electrode candescence. Audions of Parent Patent 824,637. It is notable that the only two- Four of the six two-electrode electrode audion configuration audion configurations represented that worked was the one virtually by Figs. 1, 2, 3, 4 of the parent identical to that of the Fleming patent 824,637 (reproduced previ- valve. It is also ironic that Marconi ously as Fig. 10) were used in the did not switch the polarity of the tests performed by Marconi. The battery across the electrodes to configuration represented by Fig. emphatically make the point that 5 of the patent was not tested be- detection was due to the Edison cause Marconi engineers testified effect. For some reason, audibility that it would give the same results levels were not recorded in the as that of Fig. 4 of the patent. The sworn testimony for these four configuration of Fig. 6 was not configurations.

50 AWA Review Wenaas

It was absolutely clear to the three-electrode patent with claims Court that the physical processes as to which Marconi did not con- involved in the flame audions were fess judgment. This configuration significantly different than those of was essentially the two-electrode the two-electrode audions. First, configuration of the Fleming valve incandescence was not necessary— with a third electrode in the form or even used by De Forest—in the of a foil wrapped around the out- flame audions, where as the two- side of the tube as shown in Fig. electrode audions would not func- 19. In this patent, De Forest tion without incandescent elec- claimed that in addition to the trodes. Second, the two-electrode usual functions of a detector, this audions would not function at gas audion without a separate tuning pressures significantly above the circuit (referred to in the patent evacuated levels used for lamps and the trial as the “detector per with incandescent electrodes, se”) could be made sensitive to spe- whereas the flame audions were conceived of and tested ex- clusively at atmospheric pressures. Test Results for Exter- nal Three-Electrode Audion of Parent Patent 841,386. The final Marconi demonstration test was for the configuration of Fig. 3 of patent 841,386, the only Fig. 20. Roy Weagant developed a clever demonstration board with switches to allow a direct real-time comparison of the detection efficiency of the De Forest audion, the Flaming valve and the Perikon crystal detector for Judge Mayer in court. (Author’s collection) cific frequencies—and therefore acted as a tuning device without re- quiring the usual inductors and ca- pacitors used in conventional tun- ing circuits. Roy Weagant of American Marconi constructed the rather elaborate demonstration board shown in Fig. 20 for the Court in Fig. 19. De Forest’s three-element order to directly compare the op- audion used by Marconi in the dem- eration of this audion configura- onstration tests for the Court was es- tion with both the Fleming valve sentially a Fleming valve with a metal configuration shown in his patent foil wrapped around the exterior of the glass bulb (reproduced here from Fig. and the reference crystal detector 3 of De Forest’s patent No. 841,386). manufactured by Wireless Spe- (Archives Center, National Museum of cialty. This demonstration was dif- American History, Behring Center, ferent than the pervious demon- Smithsonian Institution) strations in that there were no pa- Volume 20, 2007 51 Marconi vs. De Forest rameter variations and no attempt to show how the two circuits worked. In- stead, this was a di- rect comparison of three designs—the audion with an exter- nal electrode, the Fleming valve, and the crystal detector. This demonstra- tion was clearly de- signed to support the contention that De Fig. 21. A schematic diagram of the Weagant dem- Forests audions in- onstration board appearing in the trial transcript ex- actly matches the demonstration board pictured here. fringed on the (Marconi v. De Forest trial transcript) Fleming valve. De Forest had made the mistake of by moving the external shield was testifying that all of his audions he able to produce small variations (flame, two-electrode, and three- in the time constants of the cir- electrode) were sensitive detectors cuit—but not sufficient for tuning and all operated on the same prin- in a receiver. ciple. Consequently, Marconi was Next, he demonstrated for the free to choose virtually any audion Court the sensitivity of the audion configuration for purposes of this in comparison to the Fleming valve comparison. The external-elec- and the crystal detector using the trode audion was particularly at- demonstration board. The sche- tractive because it had three elec- matic diagram of the test board trodes—closely related to the grid appearing in the trial transcript is audion of most concern to shown in Fig. 21, which corre- Marconi—and yet it had only two sponds exactly to the demonstra- interior electrodes, almost identi- tion board shown in Figure 20. He cal to the Fleming valve.32 This was able to change from one con- configuration also had the advan- figuration to the other by using the tage that it was actually a slightly switches on the demonstration less-sensitive detector than the board so that the Court could di- Fleming valve, and therefore rectly compare the three configu- hardly constituted an improve- rations. The key results of the dem- ment in the art. onstrations tests can be summa- Roy Weagant first testified that rized as follows: he had first tried to make the 1. The Fleming valve was al- audion detector per se operate as ways slightly more sensitive than a tuning device as described in the the three-electrode audion with claims of patent 841,386. De For- the third electrode on the outside est had claimed that the audion per of the tube se could be tuned by varying the 2. The signals received on the heating current and/or the poten- Fleming valve were “even stron- tial between the filament and plate. ger” than signals received using Weagant testified that he was to- the crystal detector tally unable to tune the audion by While the Court used the audi- any means whatsoever, and only bility meter in these demonstra-

52 AWA Review Wenaas tions, quantitative results for the audions required electrode incan- three configurations were not re- descence. The flame audions corded in the transcript of the tes- worked at atmospheric pressure, timony. but the two-electrode audions It is surprising that De Forest would work only at significantly did not try to explain or demon- reduced pressures typical of those strate how his audion could be used incandescent lamps. The ex- used for tuning, and the transcript ternal “B” battery was demon- of the trial apparently indicated strated to have only a minimal ef- that he did not even try, a fact fect on the flame and two-electrode noted later by the appellate court.33 audions (the principal effect was Judge Mayer was obviously moved an adjustment of the operating by this demonstration because in point on the characteristic curve). his findings, he stated: “The so- The two-electrode audions per- called selective per se patent No. formed no differently than the 841,386, is so utterly useless that Fleming valve, and the three—elec- it might well be declared invalid; trode audions with the external …” Obviously the appellate court grid were not as sensitive as the agreed with Judge Mayer because Fleming valve. it took the unusual step of actually invalidating this patent: “It follows THE JUDGMENT from the foregoing that we hold On September 20, 1916, Judge patent No. 841,386 void, ….”34 Mayer handed down his rulings in This and the other Marconi a well-reasoned opinion.37 His demonstrations were apparently opinion addressed the three key devastating to De Forest side. De questions before the Court: the Forest had characterized this validity of the Fleming patent, in- audion of patent 841,386 in par- fringement of the De Forest pat- ticular to be “tremendously more ents by the Fleming valve as used sensitive” than the other detec- by Marconi in a detector circuit tors—including the Fleming with a “B” battery, and infringe- valve.35 This representation was ment of the Fleming patent by the proved to be false—and in a highly De Forest three-electrode audion public forum. In fact, none of the when used as a detector. audions used in by Marconi in its Validity of the Fleming Patent. experiments and demonstrations In its countersuit, De Forest had were proved to be as sensitive as argued that the Fleming patent the Fleming valve, and none was should be invalidated given the capable of amplification. Not only state of prior art—not only prior art was the sensitivity of all these early of others but also the four De For- audions less than that the Fleming est flame patents with effective valve, but effects of the electrode dates preceding that of the Fleming incandescence, electrical field, and patent. Judge Mayer ruled against gas pressure on audion operation De Forest on its argument to in- and sensitivity was demonstrated validate the Fleming patent based to be at odds with both De Forest’s on prior art of others, stating in claims and his theory of audion part: operation.36 The operation of the “Nothing in the prior art urged flame audion had been demon- by defendants in negation of in- strated to be virtually independent vention [Fleming valve] calls for of electrode incandescence, while extended discussion. The Tesla the operation of the two-electrode Patent (No. 645,576) and the

Volume 20, 2007 53 Marconi vs. De Forest Fessenden patents (Nos. 706,742, audions began to refer to incandes- 706,743, and 706,744) were far cence. It was not until a presenta- removed from the incandescent tion to the American Institute of lamp and were commercially use- Electrical Engineers (AIEE) in less; and nothing could be learned June of 1906 (published in the Sci- for this purpose from the entific American) that be began to Valbreuze and Zehnder tubes.” mention the role of electrode in- “Rectifiers of low frequency candescence in the operation of the oscillations, such as those of earlier flame audions. His incan- Wehnelt and Cooper-Hewitt, descent electrode theories were taught nothing. These are rectifi- rebutted by the Marconi experi- ers for commercial power compa- ments which demonstrated con- nies, and it was not common clusively that the audion flame de- knowledge, as of Fleming’s date tector operation was independent that rectifiers of low-frequency of whether or not the electrodes are likewise rectifiers of radio were incandescent. high frequencies; nor is it a fact The judge saw through De that all rectifiers of low frequen- Forest’s ruse, and further recog- cies are likewise rectifiers of ra- nized that the flame detectors were dio frequencies. Further, it was totally impractical and had never not common knowledge, as of been used as a detector. He ruled Fleming’s date, that a rectifier of that nothing in De Forest’s flame radio frequencies would also act experiments constituted prior art as a detector. For instance, relative to the Fleming patent— Pickard first attributed the action and therefore held the Fleming of crystal detectors to thermo- patent to be valid. electric effect, but when Pierce Infringement of the De Forest published his investigation in Patents by Marconi. Apart from 1907, Pickard amended many of the issue of the validity of the his patent applications to conform Fleming patent, De Forest argued with Pierce’s theory of rectifica- that Marconi had infringed his tion.”38 patent by using a “B” battery in the De Forest had also argued that external circuit which appeared in the Fleming patent should also be the De Forest patents, but not in invalidated because of the disclo- the Fleming patent (refer to Figure sures of his four early flame pat- 16 for an example of the alleged ents in which he attempted to read infringing circuit used by incandescence into the flame Marconi). De Forest had argued it patent specifications, and to infer was the “B” battery that imparted much about the role of incandes- invention to the audion which cence that was not known until made it a true relay (i.e., amplifier) later. In fact, Judge Mayer pointed and not just a rectifier. out that nothing was said in any of The two problems with De the four De Forest patents about Forest’s argument were first, that incandescence, and indeed a num- the battery used by Marconi in the ber of De Forest’s flame configu- Fleming circuit was used only to rations were arranged such that adjust the operating point to maxi- electrode incandescence was not mize sensitivity and not for ampli- even possible. It was not until fication—just as De Forest had 1906—well after the Fleming used it in his two-electrode con- patent was issued—that De figurations—and second, that such Forest’s patents on vacuum tube a battery had been used for the

54 AWA Review Wenaas same purpose in other oscillation tery that made the audion act as detectors prior to the earliest of De relay rather than as a rectifier. Forest’s audion patents. In es- Judge Mayer heard conflicting sence, De Forest was trying to testimony and demonstrations claim the battery used in the from both sides, but in the end be- Fleming circuit was his invention lieved the explanation given by when clearly it was not. Judge Armstrong in the Electrical World, Mayer quickly dispatched De which the judge stated in his writ- Forest’s argument by referring to ten decision “may be regarded as the Vreeland patent No. 780,842 being read into this opinion.”39 In (with an effective date of August 1, this article Armstrong contended 1904 preceding the earliest De For- that tube operated by the attrac- est patent effective date of Novem- tion of electrons emitted from the ber 4, 1904) in which a battery was incandescent filament to both the placed across an electrolytic detec- grid and plate, and acted both as a tor for purposes of obtaining the rectifier (because of the Edison ef- most sensitive operating point. The fect) and a relay (amplifier) be- judge ruled that by using a battery cause of the grid. Armstrong in the external Fleming valve cir- plainly stated: 40 cuit, Marconi did not infringe on “It will be seen that the funda- De Forest patents. mental detecting action is that of It is true that the battery pro- a [Fleming] valve, the high-fre- vided additional energy in the quency oscillations being rectified three-electrode audion configura- between the filament and grid, tion with the grid—and therefore thereby causing a charge to accu- gave it invention—but not so in the mulate on the grid and in the grid two-electrode audion where it condenser. The charged grid then functioned only to adjust the op- exerts a relay or trigger action on erating point. De Forest had con- the wing [plate] current so that the fused the function of the second audion is at once a rectifier and battery in his two-electrode audion an amplifier.” versus his three-electrode audion, The judge correctly concluded and by claiming that all his audions that the audion worked in spite of worked on the same principle, he the intervening gas and not be- doomed his own case. cause of the gas. Further, the judge Infringement of the Fleming recognized that while the grid gave Patent by De Forest. Marconi al- the audion invention, he also rec- leged that the audion, while a pat- ognized that when used as a de- entable improvement that in- tector, the three-electrode audion creased the detection efficiency was more an extension of the when used as a detector, still in- Fleming valve than an entirely new fringed on the Fleming patent be- invention—and as a result, ruled cause it used the basic structure of that the audion when used as a the Fleming tube (incandescent detector infringed on the Fleming filament, a plate, and evacuated valve. vessel)—and the Edison effect as claimed by the Fleming patent. De COMMENTARY ON THE Forest countered by contending TRIAL AND FINDINGS that the audion worked on a totally There were three points made separate principle involving heat- in the trial that are worthy of some ing of the intervening gases in comment. The first is the equiva- combination with an external bat- lent circuit appearing in Judge

Volume 20, 2007 55 Marconi vs. De Forest Mayer’s opinion [cited by footnote the two circuits shown in Fig. 22 to 256 Federal Reporter at 953] in his opinion. Referring to the two which played prominently in his circuits, the judge quotes the decision. This equivalent circuit, Marconi argument: “These are which has baffled many historians, two well-known, standard, is explained in the first paragraph equivalent circuits. You may use below. The second is the insistence either device in either circuit.”41 Of by De Forest that all his audions course, the two circuits shown are were the same and worked by the not equivalent, and the reader can same principle. This position, only wonder what the judge was which was proved to be untrue, thinking. was by far De Forest’s worst mis- The problem is that in the origi- take, and was undoubtedly a ma- nal Marconi argument there were jor cause for the adverse outcome. actually four circuits as shown in The third was De Forest’s insis- Fig. 23, and the comparison was to tence that he was totally unaware be made between the upper and of Fleming’s vacuum tube valve lower circuits, not between the cir- during the time he claimed to have cuits from left to right. Apparently, invented his vacuum tube audions for sake of brevity, the judge omit- based solely on his Bunsen burner ted the two circuits shown at the experiments. This claim—which bottom of this figure, and in doing Judge Mayer concluded to be un- so has forever confused historians true—almost certainly raised cred- and readers of one the most impor-

Fig. 22. Judge Mayer in his written opinion used this incomplete figure to illustrate Marconi’s argument that one could use either the two-electrode Fleming valve or the three-electrode audion in either of these “two well- known, standard equivalent circuits”— which, of course, is not true—and in Fig. 23. Marconi’s argument that the doing so forever confused historians, Fleming valve and audion were leaving them to wonder what he was equivalent was based on proposition thinking. (The Federal Reporter, that either device could be used in the Vol.236, 1917, p. 953) Fleming circuit as shown in the upper and lower left, or they could be used ibility issues with De Forest. just as well in the De Forest circuit as The Equivalent Circuit Conun- shown in the upper and lower right. drum. The Marconi engineers made a clever presentation show- tant infringement decisions in ra- ing the equivalence of the Fleming dio history. It should also be noted and De Forest grid audion in a ra- that while the Marconi proposi- dio circuit. The judge was obvi- tion—one may use either device in ously impressed with this equiva- either circuit—makes sense in the lence because he excerpted por- context of all four circuits, it does tions of the argument along with not mean that the two devices were 56 AWA Review Wenaas equal or that they performed tually Fleming valves with an ex- equally well. ternal foil or wire coil placed Equivalence of all Audions. De around the glass enclosure, and as Forest testified that all his audions, Marconi demonstrated to the including the flame audions and court, were incapable of either in- the two- and three-electrode tube creased detection sensitivity (or audions, were all equivalent and tuning) as claimed by De Forest. operated on the same principle—a The three-electrode audion with principle different than the the internal grid was the only one Fleming valve. Apparently he did that was actually able to amplify so to distinguish his invention signals, and was the only one that from the Fleming valve, and to give constituted a more sensitive detec- credence to his claim that the tor than the Fleming valve by aug- audion was derived from his flame menting the received signals with detector prior to the Fleming valve energy from the external battery. rather than from the Fleming (Wisely, Marconi had confessed valve. In doing so, he hoped not judgment to this configuration, so only to escape Marconi’s claims of it was not at issue in this suit.) The infringement, but also to support three De Forest audions—the his claim that the Fleming valve flame audions, the two-electrode infringed on the audion. audions, and the three-electrode The contention that all three grid audion—all worked on differ- types of audions operated on the ent principles. same principle was both wrong As an aside, most of De Forest’s and overreaching. The flame claims and theory of audion opera- audions operated without incan- tion contained in the Scientific descent electrodes, while the two- American article in 1906 were spe- electrode audions required incan- cious. Despite the large body of descent electrodes and knowledge that had accumulated just like those used in the Fleming regarding vacuum tube operation tube—as the experimental results since his 1906 article showing Marconi presented to the court these theories to be wrong, De For- clearly showed. In fact, the only est clung to them virtually his en- two-electrode audion that actually tire life. In his autobiography pub- functioned as a detector was the lished in 1950 he writes: one that duplicated the Fleming “The audion, even in its crud- valve configuration—a point an est form [two-electrode version], appellate judge specifically noted was intrinsically very much more in a later appeal. Neither the flame than a simple rectifier of high-fre- audions nor the two-electrode quency current. The addition of audions—even with an external the plate battery made a great dif- battery—could possibly act as a ference in the intensity of the sig- relay (i.e., amplifier) as claimed by nals received, for I was employ- De Forest. The only effect of the ing the high-frequency energy, external battery in the two-elec- not to actuate a telephone dia- trode audion circuit was to adjust phragm, as Fleming had done, but the operating point—and thereby to control very much larger quan- the resulting rectification sensitiv- tities of energy from the local bat- ity—not to augment the received tery.”42 RF signal as De Forest claimed. Obviously, he still had it all The three electrode audions wrong in 1950. For some reason, with external electrodes were ac- De Forest clung to his erroneous

Volume 20, 2007 57 Marconi vs. De Forest theories about the audion all his ence may be had for a more com- life—long after the rest of the sci- plete description thereof than entific world understood. Appar- need be set forth herein.”43 ently he could not bring himself to Clearly, De Forest was aware of admit that he had been wrong, and the Fleming valve well before he at the same time he wanted to dis- applied for his two-element detec- tance his invention from the tor patent on January 18, 1906— Fleming valve. certainly earlier than December 9, De Forest Gives no Credit to 1905, the filing date of the patent Fleming: While De Forest devel- referencing the Fleming valve. oped the two-electrode audion de- How much earlier would have de- tector after Fleming published his pended on how long it took De For- paper on the Fleming valve, he tes- est to do all the work leading to the tified under oath that had no patent application: conceive of the knowledge of the Fleming valve at new antenna invention using the the time he developed the audion, valve, perform whatever experi- and instead insisted that the ments may have been necessary to audion followed directly from his support the patent application, work on the flame detector. How- prepare the drawings and docu- ever, evidence in the trial indicated ment the invention for his attor- that he was well aware of the ney, generate a draft of the patent Fleming at the time he developed application, review the patent ap- his first two-electrode audions— plication, and submit it to the and indeed they were very similar patent office. to the Fleming valve and very dif- In addition to the direct evi- ferent from the flame audions. dence that De Forest knew of the First, there was a letter to his at- Fleming valve while he was devel- torney dated December 25, 1905 in oping the two-electrode audion, which he gave an instruction to there is another compelling reason “look out for Fleming’s recent why he would have known about patent,” presumably in reference the Fleming valve—namely that in to his first two electrode patent this time frame (circa 1902), De (No. 824,637) with an application Forest methodically researched date January 16, 1906. Second, and the scientific literature looking for more damaging, was his patent No. a better approach to detectors. 823,402 with a filing date of De- From his autobiography:44 cember 9, 1905 in which he actu- “And there now I began a seri- ally used the Fleming tube in his ous systematized search through invention for minimizing static Science Abstracts, Wiedemann’s electricity disturbances in a wire- Annalen, Comptes Rendus, and less telegraph system. The follow- other journals seeking to ing sentence appears in the speci- find some hint or suggestion that fications of this patent: might possibly be a clue to the de- “The device V1, (shown in fig. velopment of a new device which 1,) connected between the antenna could be used as a detector for the at the point b and earth, is an receipt of wireless signals.” asymmetric resistance or electric For De Forest—the consum- valve which has been fully de- mate inventor and researcher—to scribed by J. A. Fleming in a pa- assert that he (and his colleagues) per published in the Proceedings overlooked Fleming’s paper en- of the Royal Society of London, titled “On the Conversion of Elec- March 16, 1905, to which refer- tric Oscillations into Continuous

58 AWA Review Wenaas

Currents by means of a Vacuum November 21, 1913, is a just com- Valve” published in the venerable ment on the cause: Proceedings of the Royal Society ‘We think that Dr. De Forest of London, March 16, 1905 is just might be more generous in his ac- not credible. knowledgment of the work of Dr. Despite overwhelming evidence J. A. Fleming. Our readers gener- to the contrary, De Forest stead- ally will probably agree that the fastly maintained during his whole audion [the three-electrode grid life that he never knew of the version], although differing Fleming valve while he was devel- widely from the Fleming valve, is oping the two-electrode audions. an offshoot of it.’” He wrote in his 1950 autobiogra- phy:45 DE FOREST’S LEGAL “During the time I was devel- TROUBLES CONTINUE oping the two-electrode I had Judge Mayer had decided never heard of the Fleming valve, wholly in favor of Marconi and en- and was therefore surprised when tirely against De Forest. Marconi I later learned that my invention 46 was enjoined from infringing on was being confused with it. The the three-electrode audions to Fleming valve was only a device which it had confessed judgment, for rectifying the alternating cur- and De Forest was enjoined from rent generated in a receiving an- infringing on the Fleming patent tenna by incoming high-fre- by using the audion as a detector. quency waves. It did not—and The judgment against De Forest could not—do anything to aug- was suspended pending appeal, ment the energy of the wireless provided he posted a $10,000 signals it received. It made there- bond. De Forest made application fore a very weak detector.” to reduce the bond to $5,000, a Why De Forest maintained this request which was granted. How- untoward position in the face of ever, on or about March 16, 1917— overwhelming evidence to the con- before the case was argued on ap- trary is not clear. There was no real peal—De Forest granted a compre- legal benefit from misrepresenting hensive license to Western Electric this claim, and the judges may have under many of his patents includ- been more disposed to the De For- ing the three-electrode audion.48 est position had he not done so. Marconi attempted to have De Both Judge Mayer and the appel- Forest cited for contempt of court, late judges clearly took umbrage contending that it was an alien- with his claim, which almost cer- ation of property that may be re- tainly contributed to the adverse quired to satisfy the judgment for decisions De Forest received in the many infringements. Judge both the circuit and appellate Mayer refused to impose sanctions courts. Consider the following but raised the bond back to paragraph which the appellate $10,000. De Forest was to provide court judge used as the closing to 47 monthly reports to the court, and his opinion: meet several other conditions, but “Among the curiosities of evi- later he defaulted on these condi- dence in this record are numerous tions and so the injunction became extracts from technical periodi- effective. cals giving the opinions of the au- Judge Mayer Rules on the thors on the subject-matter of this Audion Used as an Amplifier. In suit. One from The Electrician, of January of 1917, American

Volume 20, 2007 59 Marconi vs. De Forest Marconi started proceedings in tions with a two-electrode tube, Judge Mayer’s court to have its one of which used a coil placed decree construed to cover the outside the tube, reproduced here audion when used as an amplifier as Fig. 24. Consequently, it would and oscillator—the audion as an have been difficult for De Forest to oscillator having been designated refute Marconi’s assertion that a by De Forest as the “oscillion.”49 two-electrode tube could amplify The prevailing opinion of the court since De Forest’s own patent was that a patentee was entitled to claimed a two-electrode vacuum all the benefits of his invention, tube—indistinguishable from a whether or not known or foreseen Fleming valve—was capable of am- by him—and so if Marconi could plification. Indeed, De Forest re- demonstrate that the Fleming counted in his autobiography that valve could function as an ampli- “Roy Weagant, chief engineer of fier and oscillator, it would have the Marconi Company, easily per- provided a sound legal basis for a suaded the Judge that the Fleming ruling favorable to Marconi. valve was an amplifier.”51 Marconi was in fact able to Another argument Marconi demonstrate to the Court that the made was that an ultra-audion Fleming valve actually amplified beat receiver that De Forest was feeble signals. The method by selling at the time contained a cir- which the Fleming valve was made cuit which De Forest claimed was to amplify before Judge Mayer was simply an amplifier—but in reality referenced in the findings of the was a detector proscribed under Court of Claims as “by the ioniza- Judge Mayer’s injunction.52 De tion of the residual gas in the Forest claimed that the amplifier tube,” and “by means of a field out- in question was isolated by an iron- side the tube controlling the elec- core and therefore tron flow in a specially shaped unable to pass radio frequencies, tube.””50 In fact, the De Forest and so it could not possibly be act- patent 841,387 entitled “Device for ing as a detector. However, Amplifying Feeble Electrical Cur- Marconi countered that the cou- rents” pictured several configura- pling was essentially electrostatic (bypassing the transformer core), so that radio frequency signals were able reach the amplifier, which in fact did act as a detector. Since there were no other known devices in the De Forest radio that could act as a detector, it was pretty obvious that one of the tubes had to be performing the detection function. De Forest ultimately ad- mitted that his ultra-audion beat receiver radio fell within the scope Fig. 24. While De Forest belittled Marconi’s contention that the two-elec- of the injunction—and he also ad- trode Fleming valve could amplify, De mitted that it was an infringement Forest’s patent 841,387 made claims of the Fessenden heterodyne that a two-electrode vacuum tube such patent, as well. This violation of the as the one shown here—which was vir- injunction could not have set well tually identical to the Fleming valve— with Judge Mayer. The Judge cer- could, in fact, amplify. (U.S. Patent No. 841,387). tainly must have known of De

60 AWA Review Wenaas

Forest’s repeated violations of sartorially gotten up. Anyway, court orders associated with ear- that lordly ‘Duke of Ulcer’ soured lier trials in the same United States on him, thoroughly and through- District Court (for the Southern out—and also exhibited an inexo- District of New York) regarding rable animosity against me.” infringement of the Fessenden De Forest argued that American electrolytic detector patents— Marconi could not logically confess which included a fine for contempt judgment under two patents—as it of court. Marconi was certainly did in the first trial—and yet deny aware of De Forest’s past history, infringement of the earliest inven- and may have made the argument tions, because they all “constituted that De Forest could not be trusted a connected, logical, coherent de- to proscribe the use of the audion velopment of a single inventive as a detector, and therefore its use thought of application of a scien- as an amplifier and oscillator tific theory.” The court did not buy should be prohibited under the this argument, pointing out that De same decree as the detector. Forest started with the “heated gas The judge ruled that the audion theory” and ended with the three- came within the scope of the main electrode audion employing “the decree and its use as an amplifier commercial vacuum.” was enjoined under the Fleming According to Navy patent ex- patent as well. In retrospect, it does pert Richmond, De Forest was less seem like this ruling was punitive concerned about enforcing his own rather than justified by the evi- patents in this appeal, but rather dence, but perhaps Judge Mayer with avoiding infringement of the was predisposed to this position Fleming patent.54 De Forest’s prin- because he was fed up with De For- cipal argument was that he used an est and his history of ignoring de- external battery which gave rise to crees of the court. The judge did the “relay” action whereas Fleming however rule that the question of did not. However, the Court found the oscillion infringing the Flam- that that the two-electrode audion ing patent would have to be settled was nothing more than the by another trial rather than by in- Fleming valve with an external cir- terlocutory proceedings and affi- cuit including a battery, something davits under the main decree of the not new with De Forest. detector trial. Finally, De Forest argued that De Forest Appeals the Lower whatever the merits of the Fleming Court Ruling. De Forest’s appeal patent, the disclaimer filed by of Judge Mayer’s decision was Marconi just before the trial was heard by Circuit Court of Appeal unlawful and therefore voided the Judges Hough, Cox, and Rogers in Fleming patent. The judges ruled early 1917. De Forest recounts his that this argument was without initial encounter with Judge substance, noting that the dis- Hough as the trial began:53 claimer discarded something that “When the case finally came to was not needed, without adding court in September, 1916, Phillip anything to expand the claims. Farnsworth, now delectably in Not surprisingly, the opinion morning frock coat, striped pants, written by Judge Hough dated May and white spats, must have had 18, 1917 affirmed the lower court the same irritating effect on decision, and in doing so was grouchy old judge Charlie Hough clearly critical of De Forest. The that he had on Mayer when less opinion rendered was particularly

Volume 20, 2007 61 Marconi vs. De Forest cogent:55 ment to the government without It is not often that any case concern to patent positions. Dam- contains so much history as does ages for infringements by manu- this one. It is true that Dr. De For- facturers supplying wireless equip- est, through the whole line of the ment to the government during the counterclaim patents, sought af- war would be sorted out by the In- ter a commercially useful detec- terdepartmental Radio Board and tor, and ultimately produced one; paid by the Navy and War Depart- but it is not true that he consis- ment after the war. tently followed with the heated The war dragged on beyond the gas theory; he ended with the original trial date of February 6, three-electrode audion, employ- 1918, and so the trial did not actu- ing the commercial vacuum, and ally take place until early 1919. before he produced that success he Again at this trial, Judge Mayer’s learned of Fleming’s invention position was that a patentee was and the latter’s address before the entitled to all the benefits of his Royal Society. invention, whether or not known or foreseen by him, and so the key MARCONI SUES DE FOREST issue in the trial was whether or OVER OSCILLION IN- not the Fleming valve would actu- ally oscillate. Judge Mayer re- FRINGEMENT quired that Marconi prove in court Shortly after Judge Mayer’s de- that the Fleming valve could be cision to include the grid audion made to oscillate, and that ap- when used as an amplifier under peared to be the focus of the trial. the scope of the main decree, Roy Weagant had done his American Marconi filed suit for homework on this issue in prepa- infringement of the Fleming patent ration for the first trial. In a memo by the audion when used as an os- to General Manager N. J. Nally cillator. However, by this time, the dated October 18, 1915, Weagant Unites States had entered WWI had outlined a method by which a (April 6, 1917), at which time all vacuum valve could be made to amateur radio activities were oscillate.57 Ironically, the tech- banned and all commercial wire- nique he outline was based on two less stations were appropriated by patents by none other than De For- the government. As a result, De est—Nos. 943,969 and 979,276. In Forest responded with affidavits his memo, Weagant called atten- showing that he could not get any- tion to Fig. 5 in the first patent, re- one to act as an expert witness on produced here as Fig. 25, where his behalf because all competent the valve is used in an RC circuit expert witnesses and testing labo- to produce repetitive arc dis- ratories were engaged in work sup- charges each time the voltages porting the war effort. After con- built up to a point exceeding the siderable argument with Marconi breakdown threshold of the valve. attorney L. F. H. Betts on October Weagant did qualify this approach 24, 1917, Judge Mayer perempto- in his memo, however, as follows: rily set February 6, 1918 as the trial 56 “I might say that the action, by date. means of which oscillations are Marconi, De Forest and other expected to be produced under the wireless companies had to put instruction of this patent, is fun- aside their differences to support damentally different from that the war effort, supplying equip- which is today employed by the

62 AWA Review Wenaas

have been much more vitriolic. Relying on the court room dem- onstrations, and invoking the prin- ciple “that a patentee is entitled to all the benefits of his invention, whether or not known to or fore- seen by him,” Judge Mayer once again decided in favor of Marconi by ruling the De Forest audion op- erating as an oscillion infringed on the Fleming patent. It should be noted that the judge’s opinion was based on his belief that “the two- electrode valve possesses inher- Fig. 25. Roy Weagant of Marconi got ently the same capacity for gen- the idea for demonstrating how the erating radio waves as possessed Fleming could be made to oscillate by defendant’s three-electrode de- from Fig. 5 of De Forest patent vice”— and in retrospect this find- 943,969 (reproduced here) revealing an RC circuit with a diode in which arc ing is an error, as Weagant well discharges occurred repetitively each knew based on his memo refer- time the voltages built up to a point enced above. exceeding the breakdown threshold of the diode. (U.S. Patent No. 943,969) VACUUM TUBE PATENT successfully operated vacuum SITUATION IS RESOLVED generating valves.” In June of 1919, just before That Roy Weagant used this Judge Mayer rendered his decision technique in the trial to demon- on July 17, 1919 in the oscillion liti- strate that the Fleming valve would gation, Marconi and De Forest oscillate is supported by De reached an agreement that allowed Forest’s recollection:58 a third party—Moorhead Labora- “Roy Weagant, chief engineer tories of San Francisco—to manu- of the Marconi Company, easily facture triodes restricted to ama- persuaded the Judge that the teur and experimental use as de- Fleming valve was an amplifier. tectors and amplifiers, and made (Every radio engineer knew that Marconi the sole sales and distrib- it was not.) Then he argued the uting agent.59 Advertisements for valve was also an oscillator. (Ev- a new V.T. tube began to appear ery radio student knew that this almost immediately referencing its was untrue, except that when use as a detector and amplifier, but gassy and in series with a high the ads were silent on its use as an resistance, the ‘RC’ effect could oscillator—almost certainly pend- make of it a controllable high-fre- ing the outcome of the oscillion liti- quency interrupter.)” gation (see Fig. 26). It was about It is doubtful that De Forest six months after the July 17, 1919 ever knew that Weagant’s court- decision in favor of Marconi that room demonstration of the ads for the V.T. began to reference Fleming valve operating as an os- the oscillion, and so the agree- cillator was based on an idea from ments by all parties must have one of his own patents—otherwise been extended to cover the triode De Forest’s description of this when used as an oscillator. court room vignette surely would The ink had hardly dried on the

Volume 20, 2007 63 Marconi vs. De Forest tanglements—most notably with regard to vacuum tubes.60 Immediately after the forma- tion of RCA and merger with American Marconi, the Navy con- tacted GE and AT&T urging that RCA, GE and AT&T pool their wireless patents so that wireless hardware could be marketed with- out fear of infringement of the most important patents. Excerpts of a letter dated 3 January 1920 from the Acting Chief of the Bu- reau of Steam Engineering, Rear Adm. A. J. Hepburn, USN, to GE, and AT&T with a copy to Western Electric read as follows:61 “Referring to numerous recent conferences in connection with the radio patent situation and particularly that phase involving vacuum tubes, the bureau has con- Fig. 26. Early advertisements for the sistently held the view that all in- V.T. tube using both Marconi and De Forest patent numbers indicated that terests will be best served through Marconi and De Forest had come to some agreement between the sev- an agreement on manufacturing the eral holders of pertinent patents triode when used as a detector and whereby the market can be freely amplifier, but not as an oscillator. (The supplied with tubes, and has en- Wireless Age, July 1919, p. 35) deavored to point this out with concrete examples for practical June 1919 agreement between consideration.” Marconi and De Forest regarding “In the past the reasons for de- the audions when agreements siring some arrangement have dated November 20, 1919 were been largely because of monetary signed between the newly form consideration. Now, the situation Radio Corporation of America has become such that it is a public (RCA) and American Marconi necessity that such an arrange- merging a portion of Marconi into ment be made without further de- RCA and another portion into GE. lay, and this letter may be consid- The government had been fully ered as an appeal, for the good of behind the creation of RCA and the the public, for a remedy to the situ- acquisition of Marconi to prevent ation.” American Marconi—a company It was no accident that the Navy the Navy perceived to be a foreign approached AT&T for they had ex- entity—from becoming a wireless clusive licenses to the De Forest monopoly in the U.S. In fact, the audion patents for radio applica- Navy had ulterior motives for their tions which they had purchased action: extreme distaste for from De Forest with agreements in Marconi personnel and policies, both 1914 and 1917—not to men- and a desire to consolidate wireless tion their own patents based pri- patents under single company to marily on research during the war. avoid patent infringement en- It is interesting to note that the

64 AWA Review Wenaas

1917 agreement with De Forest were selling vacuum tubes with ads specifically included the right for such as the one for AudioTron re- AT&T to grant a license to the produced in Fig. 27, and believing Marconi Company.62 Anticipating they were infringing on its rights, these cross-licensing agreements, RCA took immediate legal action. RCA immediately instructed RCA also took out ads such as the Marconi to cancel their agreement one pictured in Fig. 28—some on with De Forest, which, because of the page opposing the AudioTron a six month notification clause, ads—claiming the AudioTron and would not become effective until the Liberty valve were not licensed July 30, 1920. The cross-licensing under the Fleming patent.63 agreement was signed on July 1, E. T. Cunningham, who had 1920, only a month before the ef- been manufacturing the fective termination date of the AudioTron triode since 1915, took Marconi agreements with De For- a novel approach worthy of men- est, and as a result General Elec- tion to fend off RCA. Cunningham tric obtained unilateral rights to arranged to have an “article” pub- manufacture the triode without the lished in the February 1920 issue need for De Forest approval, and of Pacific Radio News entitled RCA obtained exclusive rights to “The Vacuum Tube Situation.”64 In use or market the triode for com- this article, Cunningham first took mercial radio communication. issue with RCA attorney state- RCA took note that manufac- ments to the effect that Judge turers other than Moorhead— Mayer had decided that triode most notably E. T. Cunningham—

Fig. 28. RCA placed numerous ads such as this one warning potential cus- tomers that the AudioTron and Liberty Fig. 27. Ads for the AudioTron in 1920 valve were not licensed under the limited its use to amplifiers in audio Fleming patent; some of these ads circuits specifically to avoid Marconi were placed in magazines on the page patent infringement claims. (QST, opposing the AudioTron ads. (QST, March 1920, p. 79) March 1920, p. 78) Volume 20, 2007 65 Marconi vs. De Forest amplifiers and oscillators—in ad- between the Fleming valve and the dition to detectors—infringed the triode. RCA was now in something Fleming patent. Nevertheless, he of a bind because it almost surely announced that he intended to would have lost the litigation with comply with RCA’s position as it Cunningham over the AudioTron related to detectors, RF amplifiers when limited to use as an audio and oscillators—but not audio am- amplifier—and there was a good plifiers: chance they would have lost other “However, as the use of our gains for high-frequency applica- device as an amplifier in audio tions Marconi had so remarkably frequency circuits is clearly achieved in the De Forest litiga- within the specific language of the tion. On the other hand, RCA could Fleming disclaimer and therefore not drop the litigation without en- cannot infringe such patent, we couraging other tube manufactur- are selling same for that purpose ers to follow Cunningham’s lead. only, although we have a license In the end, RCA made an ac- from the DeForest Company commodation with Cunningham which permits us to sell it as a de- by signing two agreements on June tector or amplifier for amateur 15, 1920 in which GE supplied purpose only.” tubes to Cunningham for sale un- This was a very clever ploy be- der his brand name.65 cause the disclaimer filed by Many believe that Cunningham Marconi in previous litigation with got such a good deal because he Deforest was very specific in dis- must have had something on RCA claiming use of the Flaming valve and/or GE. While this may be true, in all but high frequency applica- it may also be that neither RCA nor tions. To further pique RCA, GE wanted to take a chance on Cunningham closed his article with loosing their virtual monopoly on the following taunt: vacuum tubes in risky litigation, “I have not attempted to go into and saw the accommodation with detail except in respect to an au- Cunningham as a win-win situa- dio frequency use in radio work tion for both entities. After all, un- of an improved from of the old der the Cunningham agreements, Edison device and to show that GE—who essentially owned RCA at such use is not an infringement on that time and controlled its Fleming. In the Marconi-DeForest board—would manufacture litigation, DeForest was unable to vacuum tubes for Cunningham, prove a difference in mode of op- and Cunningham would have to eration between a two-member spend its own resources to adver- and a three-member tube and tise and distribute its tubes. Also, since such a difference does exist it must be remembered that RCA’s and we believe it can be proven primary mission at that time was beyond reasonable doubt, the use erecting and operating a world of the three-member audion is not wide commercial communication an infringement of Fleming.” network—sales of radio equipment In what amounted to an open to amateurs and experimentalists letter to RCA, Cunningham in es- was only incidental, and primarily sence said that De Forest may have for the benefit of GE. been a pushover, but he was not, and he would surely upset the EPILOG applecart in any ensuing trial by This story should have ended proving in court the differences quietly with the expiration of the

66 AWA Review Wenaas

Fleming and De Forest patents in 1920, before the Board’s report the early-to-mid 1920’s, but like was released, but a small Marconi any good story, this one had an organization remained in place for ironic twist with a surprise ending the purpose of pursuing claims that played out decades later. The against the Government as well as thread of this story’s ending begins other entities. with legislation passed by the The Interdepartmental Radio United States Congress on July 1, Board issued its report dated 12 1918 which specified that the gov- February 1921, recommending a ernment would be responsible for total reimbursement of only infringements of patents by other $2,869,700 to all companies—far manufacturers who supplied short of the amount that Marconi equipment to the government, and believed they alone were due. further, that the only recourse for There were four patents in particu- manufacturers disputing the lar that Marconi took issue with in government’s findings was to be the Board’s findings: No. 763,772, through the U.S. Court of Claims. reissue No. 11,913, No. 609,154, As a result of this legislation, the and No. 803,684—the latter being Interdepartmental Radio Board the Fleming valve patent. In the was created to determine reim- Marconi case, as in all others, the bursements for infringement of Board made a detailed examina- patents in radio equipment. tion of the entire radio patent situ- According to Howeth, seven ation in the United States and to firms submitted claims totaling some extent abroad. It considered $14,860,000 after WWI for in- whether a specific invention was fringements allegedly occurring general in character and hence in- both before and during the war, volved operation of the entire sys- and 17 others requested monetary tem, or whether it could be re- awards without specifying the placed without material loss or ef- amounts.66 Marconi was one of the fect, and it reviewed prior decisions seven firms, and was seeking well of the Court of Claims to get a over five million dollars. However, sense of what constituted a cus- long before the Interdepartmental tomary royalty.67 Radio Board would release it rec- Marconi took issue with the ommendations in a report dated 12 Board findings and recommenda- February 1921, American Marconi tions, and elected to appeal the completed a merger with RCA on Board’s decision to the U.S. Court November 20, 1919. The Marconi of Claims. Marconi had already Company reserved its right to pur- filed suit against the government sue claims against the U.S. arising in the U.S. Court of Claims asking from unlicensed use of Marconi for $1M in damages for alleged in- patents by the Government prior fringements of the four patents to the merger. The proceeds from mentioned above--including the any such claims were to be turned Fleming patent--on July 29, 1915, over to RCA in return for preferred even before Judge Mayer’s in- stock equal to the cash turned over. fringement decision was made on The amount of stock was to be lim- the audion.68 This whole process ited to five million dollars or some- dragged on for a very long time what less, depending upon the dis- because the case was not heard by position of the Aldene plant. the Court of Claims until March 5, American Marconi filed a certifi- 1935, and the decision was not ren- cate of dissolution on August 2, dered until June 21, 1935.69 The

Volume 20, 2007 67 Marconi vs. De Forest litigation involved a number of the reasoning behind their deci- patent claims and issues—so many sion regarding infringement is so that the results could well be the worthy of note. The Court of subject of another article. Claims found that the government Suffice it say that things did not in using three-electrode tube did go well for Marconi in general and not infringe on the Fleming patent the Fleming patent in particular. because of “the fundamental dif- The total amount awarded to ference in the manner in which the Marconi was only $42,984 plus two-electrode tube and three-elec- interest, and the Court found that trode tube function.” Further, the the Fleming patent was not in- Court stated in its opinion that it fringed by the government. Worse was not bound by Judge Mayer’s yet, the Court found the Fleming earlier decision because that deci- Patent to be invalid on the techni- sion was based on the erroneous cality that Fleming claimed more finding—supported by testimony than he invented by not limiting from both Marconi and De For- the claim to high frequencies. est—that two-electrode and three- Marconi had been well aware that electrode tubes had the “same in- the Fleming patent overstated the herent qualities.”70 invention by not limiting it to high The Court of Claims found that frequencies, which is why Marconi three-electrode audion when used filed the disclaimer prior to the as an amplifier and oscillator first trial. While the law provides worked on different principle from for such a disclaimer to avoid in- the Fleming valve, and that it was validation of the entire patent, the sufficiently different such that it void claim must have been made escaped infringement of the through inadvertence, accident, or Fleming valve. The Court also mistake, and without any fraudu- pointed out that Fleming did not lent intention—and it must also be conceive of the idea of using the made without unreasonable ne- Fleming valve as an amplifier or glect or delay. The Court of Claims oscillator, and that one skilled in ruled that the patent was invalid the art could not understand how because Fleming’s claim for “more make it work in that manner from than he had invented” was not in- his patent. It would appear that the advertent, and the delay of over 10 Court of Claims got it right, be- years in making the disclaimer was cause, while the inventor may be unreasonable. In the publishing entitled to unforeseen applications opinion, the Court stated: of his invention, in this case the in- “We find that if ever there was ventiveness was not the Edison ef- a case of unreasonable delay in fect, but rather the application this making the disclaimer it is pre- effect to radio-frequency detec- sented in what is now before us tion—not to amplification or gen- and that the Fleming patent must erating oscillations. be held invalid for that reason.” The situation was more com- The finding that the govern- plex with regard to the infringe- ment did not infringe on the ment of the three-electrode audion Fleming patent was essentially when used as a detector. The issue redundant with the finding that it before Judge Mayer in the earlier was invalid. Perhaps the Court litigation was limited to the poten- made both findings in the event tial infringement of the audion as that one or the other finding might a simple non-regenerative detector be overturned on appeal. However, represented by the circuit of Fig.

68 AWA Review Wenaas

15—not the audion in a regenera- that De Forest simply tweaked the tive detector circuit—because De Fleming valve by adding a third Forest had not been using the de- electrode to arrive at a slightly tector in a . more efficient detector,71 the Court However, by the end of WWI, the arguably got it wrong by finding government had used the three- that the audion when used in a electrode audion extensively in its simple detector circuit did not in- regenerative receivers, and so the fringe on the Fleming valve detec- Court of Claims had to consider its tor patent. use as a regenerative detector in Marconi appealed the ruling to addition to the detector circuit the U.S. Supreme Court, and the originally considered in the earlier case was heard before the Court on litigation. April 9-12, 1943. The Supreme The Court took note of the fact Court upheld the lower court rul- that the audion when used in a ing, agreeing that the disclaimer simple detector circuit gave a re- was invalid, and also agreeing that sponse about 50% greater than a since the disclaimer was not made crystal detector—not greatly more in a timely fashion (10 years after than a Fleming valve—while the filing), the entire patent was in- audion when used in a regenera- valid.72 The Court went on to say tive detector circuit produced a sig- that it was unnecessary for it to nal 20 to 30 times stronger than a determine if the government’s use crystal detector. The Court’s deci- of radio equipment infringed on sion that the regenerative detector the Fleming patent since it was operated in a fundamentally differ- invalided by the improper dis- ent manner than the Fleming de- claimer. So, in the end, the Su- tector had great merit, and was preme Court agreed with the Court undoubtedly true. The distinguish- of Claims that the Fleming patent ing characteristic of the regenera- was invalid—but only on the tech- tive detector was its ability to am- nicality that the invalid disclaimer plify (which involved a different invalided the entire patent. Ironi- principle than that of the Fleming cally, the disclaimer filed in 1914 valve), whereas the most impor- designed to save a patent which tant characteristic of the three- claimed more than it should have, electrode audion as a non-regen- ultimately played a role in its in- erative detector was its ability to validation in 1943—long after it rectify (which relied mainly on the had expired. Edison effect as did the Fleming De Forest, when hearing of the valve, but with a relatively small Supreme Court decision, took amount of amplification). great delight in writing Fleming to However—contrary to the chide him. According to De Forest, Court of Claims findings—the Fleming responded with a letter three-electrode audion when used which “evinced disdain for what as a non-regenerative detector a mere Yankee court might think does not work on an entirely dif- of his best loved child.”73 In truth, ferent principle than the Fleming the invalidation of Fleming’s valve, and in fact, the three-elec- patent had virtually no effect what- trode audion would not be the soever, except perhaps on the great discovery it was if it were only Marconi interests who were unable a slightly more efficient detector to collect for perceived infringe- and not the amplifier that it later ments both before and during turned out to be. Since it is clear WWI. The original Fleming patent

Volume 20, 2007 69 Marconi vs. De Forest first filed in England remained in- The Federal Reporter, (West tact, and Fleming was and still is Publishing Co., St. Paul, Vol. 261, credited with the invention of the January-March, 1920) pp. 393- modern vacuum tube by the scien- 395. tific community and historians— 4. Patent No. 586,193 dated July 13, 1897 entitled “Transmitting and the vacuum tube is still re- Electrical Signals” ferred to as a valve in England. 5. Rutherford, E., “A Magnetic Fleming’s disdain for the decision Detector of Electrical Waves and of the Yankee court is understand- Some of its Applications.” able, and would likely be the same Philosophical Transactions of the as the disdain an American inven- Royal Society, 1897; ser. A, 189: pp. tor with a U.S. patent might evince 1-24. See also Murray, R.P. “The if his or her patent in a foreign 1896 Magnetic Detector of Lord country were invalidated under Ernest Rutherford.” The AWA similar circumstances—particu- Journal, April 2007, pp. 26-28, 36. 6. Marconi’s patent No. 884,988 larly on a technicality of law per- entitled “Detecting Electrical taining to that country. Oscillations” which issued on April 14, 1908; the original patent was REFERENCES filed on November 28, 1902 and 1. De Forest spelled his surname “de subsequently divided. Forest,” but this spelling leads to a 7. Patent No. 307,031 entitled number of problems when “Electrical Indicator” which issued distinguishing between the man on October 21, 1884. and his many companies—his 8. The fact that the Fleming valve was companies were invariably spelled not widely used was often cited as with a capital “D.” Many evidence that it was not a publications—including The New worthwhile device, but in fact, York Times and De Forest Marconi chose to use reserve this publications—adopted the device for its commercial business convention using a capital D for all rather than selling it outright to forms including the man himself; other entities. for example, see the following 9. Patent 716,203 entitled “Wireless catalog: DeForest Radio Telegraph” which issued on Equipment, (The De Forest Radio December 16, 1902. Telephone & Telegraph Co., 2nd 10. Frank E. Butler “Making Wireless Ed., 1922), p. 4 History With De Forest,” Radio 2. Gerald F. J. Tyne, Saga of the Broadcast, December, 1924, pp. Vacuum Tube, (Prompt 211-219. Publications, 1994). 11. , The Father of Radio 3. “Marconi Wireless Telegraph Co. – The Autobiography of Lee de of America v. De Forest Radio Forest, (Wilcox and Follett Co. Telephone & Telegraph Co. et al.,” Chicago, IL, 1950) pp. 161-162. The Federal Reporter, (West 12. Ibid. Publishing Co., St. Paul, Vol. 236, 13. Frank E. Butler “Making Wireless 1917) pp. 942-955 (the legal form History With De Forest,” Radio of this citation is: 236 Fed. 942 Broadcast, December, 1924, pp. (1917)); “Marconi Wireless 211-219. Telegraph Co. of America v. De 14. De Forest’s two defenses were 1) Forest Radio Telephone & that the invention was anticipated Telegraph Co. et al.,” The Federal by De Forest patents for his earlier Reporter, (West Publishing Co., St. “responder” (Nos. 716,000 and Paul, Vol. 243, 1917) pp. 560-567; 716,203), and/or by an earlier “Marconi Wireless Telegraph Co. of Fessenden patent for the hot wire America v. De Forest Radio barretter (No. 706,744) , and 2) Telephone & Telegraph Co. et al.,” that the invention was actually

70 AWA Review Wenaas

made by Frederic K. Vreeland, who Charles E. Apgar, “Making the had become an assistant to Records from Sayville,” The Fessenden two weeks earlier. Wireless Age, Vol. 2, No. 12, 15. “De Forrest Wireless Co. in September 1915, pp. 877-880. Contempt,” The New York Times, 27. E. E. Bucher, The Wireless Age, p. May 23, 1907. 280. 16. Susan Douglas, Inventing 28. “The Marconi Valve Receiver,” American Broadcasting 1899- Modern Electrics, February 1912, 1922, (The John Hopkins Vol. 4, No. 11, pp. 779-780. University Press, Baltimore and 29. Trial transcript, Plaintiff’s Exhibit London, 1987) p. 168. 72, “Waterman Diagram Number 17. He quickly realized that the flame 3,” pp. 849-947. flickered from the noise caused by 30. Demonstrations could not be the arc discharge, not the rf performed in the court room radiation, and so he did not pursue because there was no source of gas it at the time. for the Bunsen burner. 18. See patents 824,637 and 836,070 31. De Forest so stated in a 19. Claim 21 of U.S. patent No. presentation in June 1906 to the 841,386 American Institute of Electrical 20. In addition to the three-electrode Engineers, which was reproduced configurations, there was a two- in the Scientific American electrode configuration not shown Supplement: Lee De Forest, The here in which the plate was divided Audion—I: A New Receiver for into two sections. Wireless Telegraph,” Scientific 21. De Forest credited his assistant American Supplement, No. 1665, Clifford D. Babcock with p. 349. suggesting the name: see Lee de 32. It should be noted that in this Forest, The Father of Radio – The patent, De Forest dropped all Autobiography of Lee de Forest, p. pretenses of the previous tube 214. patents by specifically referring to 22. Lee De Forest, The Audion—I: A electrodes as filaments (which, by New Receiver for Wireless definition, were incandescent), and Telegraph,” Scientific American to the use of “an evacuated vessel.” Supplement, No. 1665, November 33. The Federal Reporter, Vol. 243, p. 30, 1907, p. 348-350; Lee De 566; Judge Hough, appellate judge, Forest, The Audion—II: A New writes in his opinion: “…defendant Receiver for Wireless Telegraph,” at trial did not do it [show how to Scientific American Supplement, make the invention work], and we No. 1666, December 7, 1907, p. think he refused to try.” 354-356. 34. Ibid. 23. Lee De Forest, Scientific American 35. Lee de Forest, The Audion—I: A Supplement, No. 1665, November New Receiver for Wireless 30, 1907, p. 349. Telegraph,” Scientific American 24. E. E. Bucher, “How to Conduct a Supplement, No. 1665, p. 349. Radio Club,” The Wireless Age, Vol. 36. The operation of the flame audion 1, No. 7, p. 525. was virtually independent of 25. S. S. Robison, U. S. Navy, Manual electrode incandescence, while the of Wireless Telegraphy (Radio) for operation of the two-electrode the use of Naval Electricians, (The audions required electrode United States Naval Institute, incandescence. The external “B” Annapolis, MD, 1915) p. 139 battery had only a minimal effect 26. An amateur station using the the flame and two-electrode Armstrong circuit was used to audions, the principal effect being record signals from the Sayville an adjustment of the operating station for the U.S. Government in point on the characteristic curve. mid-1915, which appeared in The flame audios worked at Marconi’s own publication: see atmospheric pressure, but the two-

Volume 20, 2007 71 Marconi vs. De Forest electrode audions would work only Valve as a Source of Electrical at significantly reduced pressures Oscillations,” October 18, 1915. typical of those used incandescent 58. Lee de Forest, The Father of lamps. Radio, p. 325. 37. The Federal Reporter, Vol. 236, 59. “Radio vacuum-Tube Litigation is pp. 942-955. Settled,” Radio Amateur News, 38. Ibid., p. 948. Vol. 1, No. 2, August 1919, p. 77. 39. E. H. Armstrong, “Operating 60. Eric P. Wenaas, Radiola - The Features of the Audion,” Electrical Golden Age of RCA, (Sonoran World, December 12, 1914, pp. Publishing LLC, Phoenix AZ, 1149-1152. 2007), Chapter 1. 40. Ibid., p. 1152. 61. Linwood S. Howeth, USN 41. The Federal Reporter, Vol. 236, p. (Retired), History of 953. Communications-Electronics in 42. Lee de Forest, The Father of the United States Navy, (United Radio, pp. 213-14. States, Government Printing 43. L. de Forest, U.S. Patent No. Office, Washington, 1963) p. 377. 823,402, “Static Valve for Wireless 62. Gerald F. J. Tyne, Saga of the Telegraph System, filed Dec. 9, Vacuum Tube, p. 115. 1905, issued June 12, 1906, p. 3. 63. RCA ads—often placed opposite 44. Lee de Forest, The Father of the AudioTron ads—cited two Radio, p. 105. tubes including the AudioTron and 45. Ibid., p. 213. the Liberty Valve (e.g., QST, Vol. 46. De Forest is the one who was III, No. 8, March 1920, p. 78). Tyne confused, because his two- speculates that the Liberty Valve electrode audion was, in fact, the was an AudioTron with a different Fleming valve. label: Gerald F. J. Tyne, Saga of the 47. The Federal Reporter, Vol. 243, Vacuum Tube, pp.183-184. pp. 560-567. 64. E. T. Cunningham “The Vacuum 48. Radioana, Carl A. Richmond, Tube Situation,” Pacific Radio “Memorandum for Lt. Col. News, Vol. I, No. 7, February 1920. Mauborgne.” 65. Gerald F. J. Tyne, Saga of the 49. Ibid. Vacuum Tube, p. 163. 50. Marconi Wireless Telegraph 66. Linwood S. Howeth, History of Company of America v. The United Communications-Electronics in States, No. 33642, United States the United States Navy, p 376. Court of Claims (NexisLexis, 81 Ct. 67. Ibid. Cl. 671; 1935 U.S. Ct. Cl. LEXIS 68. “Marconi Company asks Millions 189), November 4, 1935, Decided Dollars’ Damages from the U.S.,” 51. Lee de Forest, The Father of Radio, The Wireless Age, Vol. 4, No. 1, p. 325. October 1916, p. 21. 52. Radioana, Carl A. Richmond, 69. Marconi Wireless Telegraph “Memorandum for Lt. Col. Company of America v. The United Mauborgne.” States, No. 33642, United States 53. Lee de Forest, The Father of Court of Claims, November 4, 1935. Radio, p. 325. 70. De Forest had testified that his 54. Radioana, Carl A. Richmond, two- and three-electrode audions “Memorandum for Lt. Col. worked on the same principle and Mauborgne.” that both were relays that could 55. The Federal Reporter, Vol. 243, p. amplify. Marconi was happy to 566. agree with De Forest’s testimony, 56. Radioana, Letter from attorney L. although it is doubtful that neither F. H. Betts to R. A. Weagant, Judge Mayer nor either party October 24, 1917. would go so far as to state that the 57. Radioana, Roy A. Weagant, letter two- and three-electrode tubes to N. J. Nally, “Re: Patents of Dr. actually had the “same inherent Lee DeForest, Covering Vacuum qualities.” In this regard, the Court

72 AWA Review Wenaas

overstated the findings of Judge President and Chief Executive Of- Mayer. ficer. 71. De Forest started with the Fleming Dr. Wenaas retired in 2002, and valve and attempted to avoid since that time has focused on re- infringement thereof by adding a searching early wireless history third electrode. He tried a variety of electrode configurations, both and completing his collection of external and internal to the valve, receivers marketed by RCA in the before hitting on the grid 1920s, as well as American and configuration by trial an error. At French crystal sets. His extensive the time he had no idea that the collection of early Radiolas in- grid configuration would act as an cludes receivers made for RCA by amplifier, and indeed it was not for Wireless Specialty Apparatus, a number of years until he and the Marconi, Westinghouse and Gen- scientific community realized that eral Electric in the 1920’s. He re- by changing the grid bias point the cently authored a book entitled grid audion would act as a liner amplifier. Radiola, The Golden Age of RCA, 72. Marconi Wireless T. Co. of which was published in the Spring America v. U.S, U.S. Supreme of 2007 by Sonoran Publishing, Court, 320 U.S. 1 (1943). LLC. It is a comprehensive book 73. Lee de Forest, The Father of covering the early history of RCA Radio, p. 457. with color pictures of virtually ev- ery receiver sold by RCA in the This article was peer-reviewed. 1920’s. Dr. Wenaas is also a past contributor to the AWA Review, a past member of the IEEE and ABOUT THE AUTHOR American Physical Society, and a current member of the Antique Eric P. Wenaas has had a life- Wireless Association. He resides in long passion for antique radios Southern California and continues beginning with his first Radiola to enjoy collecting and displaying and crystal set given to him as a radios, and researching the early young man growing up in Chicago days of wireless. by family friends. He experi- mented with radio devices and re- paired radios and as a hobby while in high school, and went on to study electrical engi- neering at Purdue University, graduating with B.S. and M.S. de- grees in Electrical Engineering. He then went to the State University of New York at Buffalo where he earned a Ph.D. degree in Interdis- ciplinary Studies in the School of Engineering. After graduating, he moved to Southern California where he first worked at Gulf Gen- eral Atomic for several years, and then moved to Jaycor, a defense company in Southern California, where he spent most of his career, first as an engineer and later as the

Volume 20, 2007 73 Marconi vs. De Forest

74 AWA Review Babcock

AWA Review First World War Aircraft Radio

©2007 Larry Babcock ABSTRACT The centerpiece of this article is a descrip- tion of my collection of aircraft This article began radios and associated equipment. I decided to with a collection of put this material in context by including de- aircraft radio artifacts scriptions of aircraft and wireless at war drawn from the First World from publications of the time. War. This collection was exhibited at the “It’s the scout of the air. It’s faster than the Annual Convention fastest ship, with a voice that can talk to the of the AWA in 2006, gunner waiting on land or sea, half a hundred and subsequently at miles away.” the Curtiss Aircraft The man who spoke these words had just Museum in 2007. To spent six months abroad. And for the first time better place the arti- in that period his words were “uncensored.” facts in context, sto- “If you’d heard the explosions—bombs from ries were drawn from a German airship, dropping down on the heart contemporary publi- of London—you’d say with me, we need such cations. Powered scouts by the thousand. And that’s the way they flight and wireless guard Paris, day and night. The scouts are al- were invented close ways up, armed with cannon—inch-and-a-half to each other in time. Hotchkiss guns, ready to fight or to warn the It is not surprising men at the anti-aircraft guns below.” that the two tech- The speaker is one of the greatest authori- nologies would soon ties on aeronautics in the country. The air scout overlap. One diffi- of which he spoke was the wireless equipped culty at first was that airplane. He had seen it in operation. That was wireless receivers did the reason for the emphasis he laid on his not have sufficient words. One could infer that events such as he volume to overcome had witnessed had left an indelible impression aircraft noise, and al- on his mind. ternative means of “The United States should have 2,000 air- communicating from planes,” says Henry Woodhouse, governor of ground to aircraft the Aero Club of America, “and it has but needed to be devised. twenty. This is our deplorable situation in a nutshell.”1 It is said that aviation radio had its incep- tion in 1910 at an aviation meet in Belmont Park, NY. Eleven private airplanes were flying above the crowd. One spectator was Captain C.C. Culver of the U.S. Signal Corps. He said it would be a good thing if pilots could talk to the ground and to each other by radio (Roberts, 1945). Prof. Henri Lanoy of France reported that the first radiotelegraph installation in an

Volume 20, 2007 75 WWI Aircraft Radio airplane was made in 1910 on a surprising that they soon came to Maurice Farman and that mes- be used in combination. Initially, sages were successfully transmitted however, aircraft were noisy and from the Blu flying field to Trappes, wireless reception was faint, so the a distance of 10 kilometers.2 first successful aircraft wireless was “Recent military operations in the direction from the aircraft to have confirmed the early experi- receivers on the ground. ence gained at the French Army On November 2, 1912, the first maneuvers with regard to U.S. radio telegraph message was aeroplanes. In a recent lecture, flashed from an aircraft in flight Captain C. J. Burke, of the Army (see Fig. 1). That airplane was one Air Battalion, claimed that the of the Signal Corps’ entire air force aeroplane was a part of the equip- of twelve machines and was piloted ment of the modern army, as vital by Lt. Henry H. Arnold, later Com- to its efficiency as motor transport manding General of the U.S. Army and wireless telegraphy, though it Air Forces. The simple quenched- was true that its value had not been spark transmitter was operated by yet been fully appraised. Already it 2nd Lt. (later Major General) had revolutionized our methods of Follett Bradley. The ground re- studying maps; rivers, mountains, ceiver was operated by Lt. (later and forests had lost considerably in Major General) J.O. Mauborgne. importance. We have no doubt that He also designed the apparatus and Captain Burke had in mind the later became the Army’s Chief Sig- striking advantages of being able to nal Officer. The purpose of the ap- communicate from an aeroplane paratus was to enable aircraft to with land or other stations by report promptly the accuracy of means of wireless telegraphy al- artillery fire.2 though he did not give prominence “This article reports the status to that in his interesting lecture. of British radio in aircraft in 1913. That this is now possible has been In a speech, Mr. Churchill said that previously shown in this journal, several hydro-aeroplanes (sea- and was strikingly illustrated in the planes) are being delivered to the case of the Wellman flight de- British Government. They are fit- scribed here recently. ted with wireless having a range of “For an aeroplane scout, for ex- 60 miles, carry guns, and are ample, to be able to communicate speedy in flight. They can rise and with his commanding officer is, of descend in comparatively rough course, of primary importance, and waters and are as good as anything it would be an immense advantage which exists abroad.4 if he could exchange messages and “From the result of prolonged receive commands. It is a matter of exercises during the past year at importance that the military au- various naval stations with hydro- thorities should be well acquainted aeroplanes and submarines and in with attempts that have been made conjunction with the patrol flotilla to perfect wireless apparatus that we have come to the conclusion will be suitable for aircraft, thus that there should be a chain of hy- ensuring the greatest possible ben- dro-aeroplane stations at various efit being derived from the points on the British coast-line for aeroplane.”3 naval scouting purposes. These sta- Powered flight and wireless tele- tions are being rapidly established graph were invented at approxi- and a number will be completed in mately the same time, so it is not 1913.

76 AWA Review Babcock

Fig. 1. Painting by William Hardnon Foster of air pilot and observer with a spark transmitter dated 1910. This needs to be regarded as somewhat fanciful since the first recorded U.S. aircraft message from flight was in November, 1912. The badge on the pilot’s collar reads USAC.

“The problem of carrying attached for this purpose. Com- aeroplanes in ships is also receiv- pared with other navies, the Brit- ing attention and a cruiser has been ish aeroplane service has started Volume 20, 2007 77 WWI Aircraft Radio very well. Preliminary difficulties the aeroplane that the Germans have been surmounted and we brought down. When it was seen shall now be able to move steadily that there was something wrong forward in several promising direc- with the plane, the spotting correc- tions. tions still continued, and the last “There are now 40 naval message read as follows: ‘Carry on, aeroplanes with 60 pilots as com- you are hitting her every time for- pared with 5 aeroplanes and 4 ward. We have been hit. Coming trained pilots in March last year. down on water. Send a boat.’ By July 1913 when the naval ma- “A few seconds later the neuvers take place there will exist aeroplane (a land machine) 75 aircraft and 75 pilots and by the crashed into the water, throwing end of this year it is anticipated Arnold out. Cull extricated himself that there will be 100 men and 100 with great difficulty under water a aircraft and that the combined minute and a half after she struck. strength of the naval and military He had a very narrow escape in- air fleets will not be far short of 300 deed. Had these two officers not machines.”4 put the guns on to the Konigsberg In 1914, a few weeks after the in time a different ending of the beginning of World War 1, whole action might have re- Mauborgne demonstrated the first sulted.”6 two-way aviation radiotelegraph. In the early years of World War Flying the airplane was Major Gen- 1, various creative means were eral Herbert A. Dargue, then Lieu- used to communicate from the tenant, in the Signal Corps. The ground to the spotter aircraft. The purpose of the apparatus was to problem of overcoming the noise not only send messages to the of the machine and similar difficul- ground but also to receive them. ties led to the adoption of visual “The first actual tests of sea- means of transmitting directions planes to spot the fall of shots took to airmen. This was accomplished place in July, 1915, when British by white strips of cloth, 6 feet long seaplanes were used to direct the by 1 foot wide, laid on the ground guns of monitors to attack the Ger- to form letters and symbols. These man cruiser Konigsberg, which were easily visible from a height of was hidden up the Rufigi River, in 3,000 feet. East Africa.”5 Assuming that the observer had When the German ship been given the general direction Konigsberg was attacked and de- and the nature of the enemy em- stroyed by the British forces, wire- placement before rising (taking less in aeroplanes was used to di- off), he would watch for specific rect the fire of the English. Chief directions as he ascended to the Paymaster Charles Spedding of the required height in safety behind Laconia, has written as follows re- his own batteries (of artillery). The garding the manner in which the strips were then seen formed in the occupants of the British aeroplanes shape of the letter Z, which meant conducted themselves: “observe for time shrapnel,” (ex- “During the second day of ac- plain) or perhaps a P, which meant tion an incident occurred which “observe for high percussion,” or calls forth one’s admiration of the LYD, for high explosives. Later, at pluck and presence of mind of his post of observation he may Flight-Commander Cull and Wire- have noted any one of these sym- less Expert Arnold, who were in bols prefixed by an X, meaning

78 AWA Review Babcock

“change to,” or maybe two strips these numbers. Each arm is pro- in parallel, which told him, “am not vided with a flap of dark blue cloth, receiving your signals.” similarly numbered on the outside, One such system was the so that all or any of the arms can Popham signalling panel for com- be covered or exposed to view from munications between infantry and the air at will. aircraft. The poanel consists of a In Fig. 2, lower examples, the dark blue cloth to which are sewn number 158 is known to indicate strips of white cloth in the shape the message “More stretcher bear- of the letter “T.” From this “T” ers required;” the number 247 is project nine arms of white cloth. known to indicate the message These arms are numbered con- “Have gained third objective;” and secutively from 1 to 9, as shown in the number 5678 is known to in- Fig. 2, and are always known by dicate the message “Am about to

Fig. 2. Popham signalling panel for communications from infantry to aircraft. From Signal Training Manual, His Majesty’s Stationery Office, 1921.

Volume 20, 2007 79 WWI Aircraft Radio advance South.” Scouting pilots operation (see Figs. 3, 4).8 were therefore required to memo- Yet another early applica- rize a large number of message tion of wireless in the First War codes.7 was its use in observation balloons. A different approach to signal- Not until the war is over can the ing to aircraft was used on the Ger- full measure of the priceless ser- man side of the hostilities. An ap- vice rendered by wireless be made paratus consisting of a sensitive known At this time it is impossible Einthoven galvanometer with a to give more than vague hints of small electric lamp below it was the improvements which have focused through lenses on to a been made. After the war the world small mirror, which in turn re- will be given improvements which flected the light through a magni- will revolutionize wireless and fying glass. The upper part through open up possibilities not even now which the observer looked was dreamed of. constructed on the prismatic bin- When the war broke out, army ocular principle, and magnified the wireless operators were as few in effect of wireless signals on the gal- numbers as were the rest of those vanometer. The description and we so sorely needed. Luckily, how- photograph of this apparatus is at- ever, there were many who had tributed to William Dubelier, a ra- taken up the art for the pure love dio engineer from New York, who of the thing and they were quick toured the Berliner factory in to volunteer. Vienna and saw the apparatus in The first time I saw a wireless operator at work was just outside Poperinghe, about five miles behind Ypres. An observation balloon was going up and into the little car below the great gas-bag climbed a young chap carrying a headpiece and a mess of wires attached to a small box. I was told he was the wireless op- erator and I stopped to watch his work. In the field below the balloon was a motor car with a pole above it with wires extending down- wards to the ground. The balloon reached the de- sired height and then the wireless machine in the car started to snap and spark, the operator writ- ing and passing his Fig. 3. This German device permits the radio op- sheets to the officer erator in a flying machine to “see” the wireless sig- standing beside him. I nals. The engine and other noises do not interfere. The Electrical Experimenter, November 1917. learned that as the ob-

80 AWA Review Babcock

flash and clatter of the instrument kept up and at last I was impelled to ask the question: “Why not a telephone line to the ground in- stead?” With a pitying smile the expert told me that a telephone line was not anyway near as good because it would need many men to follow along with it and see that it did not catch in the trees when it would be easily broken. “Besides,” said my informant, “If the gasbag breaks loose and drifts over the German lines, the operator can keep in touch to the last minute. Don’t you see?” And I learned that this had hap- pened on one occasion and that the operator had “kept in touch.” The “last minute” was when the balloon was exploded by a shell!9

WORLD WAR ONE RADIO ARTIFACTS Since the theme of the 2006 AWA conference was military ra- dio I decided to enter my collection of World War One military aircraft radios. I was overwhelmed by the interest and I ended up not only winning the blue ribbon for mili- tary radio but also the “Best of “Show” award! Next I called the Fig. 4. Drawing of German flyer “see- Curtiss Aircraft Museum in ing” a wireless message while in flight. Hammonsport, NY and told them The Electrical Experimenter, Novem- about my collection and offered to ber 1917. bring it in to show them since they serving officer in the balloon lo- have no early aircraft radios dis- cated certain positions he gave the played there now. When they saw information to the operator who, it they asked me to plan a display in turn, transmitted it to the man of it in their museum for 2007. below. Below the balloon’s car “Airplane radio apparatus does trailed wires about forty feet long not differ fundamentally from or- and these, I was told, had some- dinary radio apparatus. It must be thing of the function of the ground especially designed, however, to wires of a stationary equipment meet a number of special condi- The balloon carried by the wind tions existing on an airplane, which and held to earth by the cable are to a great extent absent in which attached it to a powerful ground radio. motor truck drifted away a mile or “A primary consideration so towards Ypres. The ceaseless is that all apparatus shall be built Volume 20, 2007 81 WWI Aircraft Radio of carefully selected material to 10A aircraft interphone box and insure minimum weight and bulk, cables, SCR 57 Aircraft interphone as the carrying capacity and avail- and a generator to provide electri- able space of the airplane is lim- cal power to the radio system. ited and heavy auxiliary apparatus There are two DR 2 antenna reels, is not thought of kindly by the air- a Western Electric T1 microphone, plane design engineer. The design a large model of the Jenny JN 4 and construction of all parts must aircraft and much documentation also have great ruggedness in or- including five volumes of informa- der that the apparatus may oper- tion taken from early magazines, ate properly and without changing the , books, pictures and its adjustments under the con- other sources. tinual vibrations of the airplane The SCR - 65 shown in Fig. 5 due to the engine, wind, etc. For is the earliest but one of the more this reason, all adjusting handles common aircraft radios from the are usually locked in position after World War One era. It is a key op- the set has been tuned and ad- erated transmitter built by the justed. Vacuum tube sockets are Connecticut Telephone and Elec- mounted on soft sponge rubber tric Co. and modeled after the Brit- cushions to absorb the shocks, and ish Sterling set which was used al- the set box itself is frequently most exclusively in artillery work. mounted on shock absorbers. Both loop and open type an- “In addition to these mechani- tenna circuits were used on air- cal considerations of the installa- planes. The loop antenna was pri- tion, great care must be taken in marily used for direction finding. preventing all possibility of pro- The “trailing antenna” found the ducing an open spark, as the dan- widest application. It consisted of ger of fire on an airplane is ex- a 300 ft. length of bare wire at- tremely great. This will be appre- tached to a reel on one end while ciated when it is noted that the the trailing end was weighted by a “dope” or varnish painted on the small lead weight often shaped like linen or cotton fabric of the wings a fish and called by that name. The and fuselage is a highly inflam- antenna was let out of the airplane mable composition. The fire haz- through a tube called the “fairlead” ard is also increased by the likeli- mounted on the fuselage. The an- hood of gasoline vapors being tenna was reeled in before landing present in the cockpits. For these to avoid breakage.10 various reasons, all wiring on an airplane is made with high insula- tion and is securely attached along the airplane struts so that it will not shake loose. Special telegraph keys are used, having enclosed contacts, and all spark gaps are either totally enclosed or covered with wire gauze.”10 The collection includes an SCR 65 spark transmitter, an SCR 68 transmitter and receiver, a BC 11A transmitter and receiver, and an SCR 59 receiver. There is also a BC Fig. 5. The SCR 65 spark transmitter. 15A power distribution box, BC Larry Babcock collection.

82 AWA Review Babcock

The trailing wire was the an- rary fading out of the signal due to tenna. The counterpoise was made detuning. With undamped wave up of the metallic parts of the air- sets with heterodyne reception it plane – the engine, stays, wires, results in a variation of the pitch etc., which were all carefully of the signal note. bonded together. The radiation The interior of the SCR 65 is characteristics of a trailing antenna shown in Fig. 6. It is a tuned spark are such that the highest gain is coil set that produced damped forward. The plane should fly to- waves. The usual airplane trailing wards the point at which commu- antenna was 150 to 250 feet long. nication is desired. The bonded wire stays of the air- The fairlead is made of a tube craft were used as a counterpoise. of insulating material such as Wave length was from 100 to 300 bakelite which is clamped to the meters. The schematic is shown on fuselage. Inside this tube is a metal Fig. 7. The set box and 8 volt stor- sleeve through which the bare an- age battery were usually mounted tenna wire passes. The antenna on a sliding wooden tray for easy makes contact with this sleeve and access. It required adjustment or a binding post attached to the replacement after every flight. The sleeve allows the radio to be con- BB-11 Liberty lead storage battery nected to it. had a discharge rate of 3 amperes The length of antenna wire used for 3 hours. The J-7 transmitting depends on the signal wave length. keys (Fig. 8) were located one on As a rough rule using the Curtiss either side of the pilot’s cockpit and JN-4 airplane, the natural wave one on the right side of the ob- length in meters is approximately server. They were all connected in equal to the length in feet of the parallel so that any one could be trailing antenna. The natural wave used to send messages. length of such a trailing antenna The antenna wire was held in an also depends on its position rela- RL-2 antenna reel with DR-2 drum tive to the counterpoise of the air- on the outside of the aircraft fuse- plane. Therefore the antenna con- lage. A WT-1 weight was fastened stants and the tuning of the air- to the lower end of the antenna plane radio set change whenever wire. A typical installation is shown the airplane alters its direction of in Fig. 9. The antenna reel is seen flight or the wind alters the posi- on the outside of the fuselage be- tion of the trailing wire. With tween the cockpits. Later installa- damped sets this results in tempo- tions placed the reel inside the fu- selage. This early aircraft radio was a transmitter only. It was much lighter than later models fitting in a small 6” x 8” x 4” box. A small battery, an- tenna wire and key made up the rest of the system. The SCR 59 Fig. 6. Interior of the SCR 65 aircraft transmitter. Larry airplane radio re- Babcock collection. ceiving set was

Volume 20, 2007 83 WWI Aircraft Radio rate. No generator elec- trical power source is required. The pilot and observer can communi- cate with each other via the interphone. A switch on the SCR 57 interphone box allows the system to be switched from intercom to transmit when a voice transmitter is connected to the interphone. Fig. 7. Schematic wiring diagram of SCR 65 spark transmitter set. Larry Babcock collection. The only rf tuning is in the antenna circuit added to the aircraft later in World and consists of a tapped coil and War One to allow the pilot and series tuning condenser.. There is observer to have two way radio no regeneration or other special communication with other loca- circuit to increase the amplifica- tions. It consisted of a 10” x 11’ x 5” tion over straight tube amplifica- box. Accessories included an SCR tion. Other than tuning the only - 57 airplane interphone, 4 V stor- other front panel controls are to age battery for filament power, two turn on the tube filaments and to 50A head telephone sets and two increase the amplification of the 328 W TRANS microphones. last stage. The communication sys- tem schematic given inside the SCR 59 cabinet is shown on Fig. 10. The parts layout is given in Fig. 11. The rf coil and tuning condenser are on the right side of the chassis. The three VT 1 tubes are on the top left with the two black inductors between and behind the tubes. The Fig. 8. J-7 transmitting key. Larry B battery compartment is below Babcock collection. the tubes. The front panel is shown The SCR 59 was a very simple on Fig. 12. three tube receiver using type VT 1 A typical 1918 aircraft radio tubes. This was the first military communication system is shown tube type. It is interesting that the on Fig. 13. It includes the radiotele- pins on these tubes had a chunk of graph transmitter and receiver in solid gold (not just plating) at the one box in the upper center of the end of each of the four pins to as- figure. In the upper right corner is sure good electrical contact was the SCR 57 interphone unit. Below made with the tube socket. When it are the two sets of leather hel- you are buying these tubes for your mets with sewn in headphones. In World War One radio, be sure the the lower center is the generator gold is still on the tube pins. I have which provided electrical power seen tubes with it removed. This for the complete system. The dis- detracts from the tube value. tribution box in the upper left cor- The B battery is contained ner contains filter capacitors and within the SCR 59 cabinet. The an inductor to filter the DC volt- small 4 V filament battery is sepa- ages to the receiver/transmitter.

84 AWA Review Babcock

Fig. 9. Typical installation for SCR 65 transmitter. But this photo shows even BC nomenclature references the more. The box in the rear on the entire communication system con- left contains ballast lamps needed sisting of several individual units. to help regulate the generator volt- The SCR nomenclature references age as it fluctuates due to changes an individual box. The SCR 68 is in wind speed, etc. The next stack the unit seen in the picture. When of boxes seem to be VT-2 tubes reference is made to the BC-11-A used in the radio transmitter. it means you are looking at the Third is a box of VT-1 tubes used transmitter/receiver unit from the in the radio receiver. The fourth BC- 11-A system. That is as close pile of boxes is probably the TB 1 as I can explain it. There are slight type tube used as a voltage regula- differences in the two units. For ex- tor in the generator unit. There are ample there are a different num- two telegraph keys in the lower ber of taps on the antenna coil center and two microphones in the switch seen in the upper left cor- lower right corner. Apparently this ner of both photographs. Both sets system had both voice and tele- have the same three tap switches graph capabilities. on the front panel. They are visible Fig. 14 is a front view of the BC- on Fig. 15 of the SCR 68 but not on 11-A radiotelephone transmitter/ Fig. 14 of the BC-11-A front panel receiver in my collection. You will because the cover is closed here. note that it is very similar to Fig. Fig. 16 shows the interior of both 15, the SCR 68. This shows the sets. There is little difference. The two methods that were used for receiver is very similar to that de- identifying radio equipment. The scribed for the SCR 59. They all use

Volume 20, 2007 85 WWI Aircraft Radio

Fig. 10. SCR 59 receiver schematic. Larry Babcock collection. three VT-1 triode tubes with induc- nections with all of the other units tors shown to the rear. The top in the system is shown on Fig. 17. right houses two VT-2 tubes for the It is the hookup drawing taken transmitter. They are mounted in- from the inside cover on the SCR side the rf coil. I have no other 68. set with this arrangement. The Two Airplane Interphone boxes lower right corner contains the are shown in Figs. 18 (SCR 57) five pole switch used to change and 19 (BC-10-A). To me they ap- between “receive” and “transmit” pear identical except for the no- The empty tube socket to the ex- menclature on the name plate. treme right center holds an LM-1 Both are made by Western Elec- ballast lamp. The lower left corner tric. Both have a large switch on holds a B battery. The schematic the top to change the communica- for the SCR 68 and the intercon-

Fig. 11. SCR 59 receiver parts layout. Fig. 12. SCR 59 receiver front panel. Larry Babcock collection. Larry Babcock collection.

86 AWA Review Babcock

Fig. 13. A typical 1918 aircraft communication system. Roberts, 1945, p. 10.

Fig. 14. Front view of the BC-11-A Fig. 16. Interior of BC-11-A and SCR transmitter/receiver. Larry Babcock 68. Larry Babcock collection. collection. tion system between intercom (re- ceive) and transmit. Fig. 20 shows the schematic of the box as well as interconnections between it and other parts of the radio system in- cluding the radiotelephone trans- mitter and receiver set, 2 pairs of headsets, two microphones and the transmit button. Fig. 21 shows the box interior layout. Wooden slots are provided to hold stacks of Fig. 15. SCR 68 aircraft receiver. Larry dry cells for the pilot and the ob- Babcock collection. server intercom systems. “The Navy realizes the necessity of organized scientific research of

Volume 20, 2007 87 WWI Aircraft Radio

Fig. 17. SCR 68 Schematic. Larry Babcock collection. Aeronautic Station, Pensacola, FL. Aircraft radio, although an accom- plished feat, has been largely a matter of guesswork unguided by scientific data. This laboratory not only tests aircraft apparatus sub- mitted by commercial manufac- turers under both laboratory and service conditions but also gathers scientific data pertaining to the particularities involved in radio signaling between isolated points above the earth’s surface. Lts. E.H. Fig. 18. Airplane interphone Type SCR Loftin and P.N.L. Bellinger are in 57. Larry Babcock collection. charge. The lab has only been in existence a few months and is not yet fully equipped but several im- portant problems have been solved. “(1) The development of a simple inter-seat telephone for pi- lot-observer conversation on air- planes under the condition of full power flight. This work was done because commercial apparatus was entirely unsatisfactory. Two types have been developed, the Fig. 19. Airplane interphone Type BC- 10-A. Larry Babcock collection. first of which requires no external battery. It is simple, rugged and is special apparatus to meet its re- suitable for use in moderate noise quirements by the establishment conditions encountered on low of a radio laboratory at the Navy powered airplanes. The second re-

88 AWA Review Babcock

Fig. 20. Airplane interphone schematic. Larry Babcock collection. quired an outside battery and may ments of various airplane noises be used under the most severe have been made. Thousands of noise conditions obtainable on tests have been made to determine present airplanes without exhaust the noise levels of motors, the effi- mufflers. The telephones which are ciencies of muffling devices, and in daily use for instruction pur- the efficiencies of radio headgear poses reduce by nearly one half the for eliminating airplane noises, the time required for qualifying stu- effect of airplane noises on the dent pilots, by permitting constant ears, the noise making qualities of coaching from the instructor and air at varying velocities, etc. are available for pilot-observer “(3) Measurements on ev- communications in two-passenger ery conceivable form of airplane airplanes. antenna to determine the advan- “(2) An exhaustive study tages of each type. has been made of the noise condi- “(4) Development of a new tions affecting radio reception on light weight radio transmitter. airplanes. A special instrument With a weight of five pounds, a sig- called the Noisemeter was devised naling range of 10 to 20 miles is with which accurate measure- easily attained. Space consider- ations are negligible. Fifty to 75 miles range is possible with a 10 pound transmitter. These weights include the antenna. The best com- mercial apparatus weighs from one to two pounds per mile of range and occupies 1-2 cubic feet per 25 miles range.”11 A novel windmill driven genera- tor for airplane wireless service developed by L.J. Lesh promises to solve some of the problems in the equipment of airplanes with power sets. The simple expedient of belt- ing the generator direct to the gasoline engine has long since been found impracticable for the wire- Fig. 21. Airplane interphone interior less set becomes inoperative when layout. Larry Babcock collection.

Volume 20, 2007 89 WWI Aircraft Radio

Fig. 22. Windmill driven generator for aircraft sets. The Wireless Age, July 1915. the motive power gives out. Engine hour and the results were very in- trouble or exhaustion of gasoline teresting.12 while the aviator is over the ocean It was learned that power devel- would present a very definite need oped increased with the speed but for a call for help just when this set reached a maximum limit at about could not be worked. Thus far stor- 60 miles an hour. Just what hap- age batteries have been found too pens then has been a question, but heavy for satisfaction to the flyers, it was definitely established that so the windmill generator should between the speeds of 55 and 65 prove extremely useful.12 miles per hour there is a compara- Taking into account that the tively small variation of power de- aviator must always maintain con- veloped. A like phenomenon was siderable forward velocity, Mr. observed in experiments with pro- Lesh turned naturally to securing pellers at high speeds. The blades power from the speed of the air- got in each others’ way and failed plane and made his first windmill to take a proper grip on the air (see experiments with an automobile. Fig. 22).12 The fan was belted to the genera- My Type GN4A radio genera- tor, a 500 cycle machine with di- tor is shown in Fig. 23. They were rect connected exciter. At a speed required for higher powered trans- of 50 miles per hour, the fan ro- mitter systems and where longer tated at 1,000 rpm. Ammeter, volt- flights were made that would have meter and tachometer were ar- required excessively bulky, heavy ranged so as to give complete data batteries. A number of different on the tests. Readings were taken types of units were made. Some at speeds from 30 to 65 miles per generated up to 325 volts, a diffi- 90 AWA Review Babcock

Fig. 23. Type GN4A radio generator/ propeller. Larry Babcock collection. cult amount to achieve with batter- ies. Mine has two voltages (five Fig. 25. Power distribution box inside wires). This is one of the rarer view. Larry Babcock collection. items of World War One aircraft A TB-1 vacuum tube (see Fig. radio equipment. I have never seen 26), was called a Kenotron and another aircraft generator avail- with a three pin base it was used able for sale except at the auction as a regulator inside the radio gen- of early radio equipment at the erator. About 4500 Kenotron Ford Museum in October of 1995. tubes were made by G.E., mostly The output of the radio genera- in 1918. It was extremely rugged to tor is connected to the receiver/ withstand the aircraft environ- transmitter unit through a power ment.13 distribution box, Fig. 24. It con- tains the filter to smooth the out- put voltage. Notice that my unit still has the Inspection tag. It is signed 2/11/19. I think this is “NOS” never used because the ter- minals were never crimped or sol- dered. The unit interior is pictured in Fig. 25. It contains a 3 ampere filter choke, two condensers the larger of which is 1 mfd., and two wire wound resistors.

Fig. 24. Airplane wireless power dis- Fig. 26. TB-1 Kenotron regulator tube. tribution box. Larry Babcock collec- Larry Babcock collection. tion.

Volume 20, 2007 91 WWI Aircraft Radio My World War One pilot or observer’s helmet for radio opera- tors is displayed in Figure 27. It is leather and has pockets by each ear where “head telephones” can be sewn in. The director of the Curtiss museum in Hammonsport, Rich- ard Lisenring, said that these hel- mets continued to be used for a time after World War One. He was not sure just when this one was made but certainly it is pre-World War Two. A photo of the Navy fly- ing uniform is shown in Fig. 28. Although my exhibit is of World War One military aircraft radio, the documentation continues on to 1920. I include, for example the first flight across the Atlantic Ocean. Did you think that it was by Lindbergh? No, it was 8 years ear- lier by the NC-4. (Navy Curtiss) It was neither non-stop nor solo but Fig. 28. Helmet with built in ear phones still very interesting. Four planes used by Navy pilots in World War One. were built for the effort. Only the a dream among aeronautical en- NC-4 completed the trip. The year thusiasts for many years, and we was 1919. of the present generation are in- “Trans-Atlantic flying has been deed fortunate in seeing this great feat of aerial engineering and ex- ploitation carried to a successful conclusion by the U.S. Naval sea- plane NC-4. Four aircraft started this trip (NC-1, 2, 3 & 4). Only the NC-4 was successful. As an ex- ample of the efficiency of this type of flyer for such form of travel over vast water distances, the NC-3 had shown that it could land on the ocean and stand considerable buf- feting about without sinking for several days. “These aircraft carried two communication systems. Radio- telephone apparatus with a range of 30 to 40 miles could operate from plane to shore of ship stations and between other planes. Radio telegraph capable of transmitting and receiving up to 150 miles was Fig. 27. World War One pilot or also carried. A new record for air- observer’s helmet. Larry Babcock col- plane radio transmission was lection. made when a message was picked

92 AWA Review Babcock up from the NC-4 at a distance of changer adjustable for two wave 400 miles at the Otter Cliffs radio lengths—425 and 600 meters. It station in Maine. consists of two generators with “Not only was the regular radio both armatures on the same communication made available at shaft—one a direct current genera- every turn, but the U.S. Govern- tor supplying excitation current, ment exercised special precautions and an alternator which develops to render certain that the aviators, the primary power used in the in case they should have to descend transmitter. The weight of the gen- to the ocean, should not lose their erator complete is 28.5 lbs.”15 lives unnecessarily. It placed war- Another transmitter is a com- ships, including cruisers and de- bined telephone and telegraph stroyers, in a complete chain from transmitter which operates on a the American coast to the Azores, small 12-volt storage battery. It is as well as from the Azores to this set which is used for telephon- Lisbon, Portugal, the official termi- ing between planes, arrangement nating point of the flight. These being made so that either the ra- vessels were placed about 75 miles dio operators or the commanding apart, and closer together where officers may telephone while the night flying occurred.”14 planes are in flight. “One of the great achievements A special feature of the tele- in the past year in radio is the ½- phone sets is the anti-noise micro- kw airplane radio transmitter used phone, which is so constructed that by the United States on their the engine noises are not heard. digiribles and airplanes during the This is accomplished by having the Great War. It was designed for the back of the microphone open. The purpose of compactness and dura- exterior sound waves strike the bility as well as efficiency back as hard as the face of the dia- “The panel of the transmitter is phragm and therefore the effect is constructed of a sheet of black neutral. The voice waves strike bakelite dilecto, reinforced by a only the face of the diaphragm, and bent wood frame extending around even though the operator can not the curved section and metal hear his own voice, the radio sets angles on the sides and bottoms. receive enough effect to modulate Upon the front of the panel is the transmitted wave.15 mounted the wave changing de- The following article gives an vice, the quenched spark gap, con- indication of the advanced type of trol switch and rheostat, also the electrical equipment used on Ger- meters. All wiring connections are man aircraft early in WW1. I have made on the back of the panel. The not heard of heated clothing, in- panel complete weighs exactly 44.5 strument lights, etc powered by lbs. (Although this article appeared propeller/generator equipment in in July, 1919, the contract date on early WW1 U. S. aircraft. a photo of the matching receiver, “Lieutenant Jean-Abel Lefranc not shown, is March 24, 1917. Ed.) states that for some time enemy “The transmitter consists of German aircraft have been occa- several parts, the most essential sionally equipped with continuous being the propeller driven genera- wave receivers of the tube type. tor and the panel, which is sup- The generator is driven either by a ported within the fuselage. The set motor or a propeller rotating at is of the 500 cycle quenched spark 4,500 rpm. It generates 270 volts, type with a specially designed wave 3 amps ac and 50 volts, 4 amps dc..

Volume 20, 2007 93 WWI Aircraft Radio The alternating current is used by interview regarding the European the oscillating circuit to generate war. Hertzian waves. The Telefunken Since the beginning of the war, “sender” consists of a rectangular Mr. Marconi has had unusual op- box containing a transformer, a portunities for observing the prac- condenser, a plate discharger and tical side of the war, having early a wavemeter. Special arrange- put his scientific knowledge at the ments permit variation of wave- service of his country. As Senator length and intensity of transmis- of the kingdom he has visited En- sion. The aerial consists of a cop- gland, Belgium, France and other per wire about 40 meters long. On countries and introduced indus- the ground this wire is rolled up on trial and shipping reforms. In his a bobbin. During flight it is sus- capacity as military officer he has pended from the aircraft. The come into close relation with the range of these sets is about 20 army and the navy and given ben- miles and their weight about 57 lbs. efit of his science and business or- The latest giant aeroplanes guide ganization knowledge to munition themselves at night by factories. radiogoniometers, as the zeppelins He has also perfected the army do.”16 and navy wireless systems, and is The generator supplies current at present working on a signal sys- for several purposes. Direct cur- tem which, it is expected, will ren- rent supplies the pilots and der far more difficult submarine observer’s electrical warming ap- warfare through the readier loca- paratus, lamps on the instrument tion and signaling of the presence boad, heating apparatus on the of such craft. The details of this he camera and machine gun. Search was unable to furnish because of light power is also provided dur- its immediate military importance. ing landings at night. All ac power Incidentally, the inventor referred is used to drive the Telefunken to the position of the United States. transmitters including the aerial “I don’t think the United States ammeter, the aerial bobbin and the should ever fear any fatal disas- antenna. trous invasion.” He said. “Her seas protect her too well. She is too GUGLIELMO MARCONI ON mighty a country in population and WAR INVENTIONS force ever to be conquered. I doubt A summary of the importance if, with reasonable precaution, of wireless in World War One was even her coasts could be injured or provided by Mr. Marconi to Wire- landed upon. The experience of less Age.17 this war has shown how easy it is On the whole there have been to protect a coast by submarine, no great war inventions that occur even when the invader is a near to me. In my own field , there has neighbor.” been some advance in practical As to the prospects of peace in wireless, by which we are now able Europe, Mr. Marconi said there are to direct the artillery fire of a ship many people who believe the war by signals from an aeroplane. “This will be over by winter. has been made possible largely “To me,” he said, “The saddest through big improvements in air- fact about this war is that so much craft.” energy has been used up which This was a statement made in might have gone to a better pur- Rome by Guglielmo Marconi in an pose.”

94 AWA Review Babcock

Marconi. The Wireless Age, October REFERENCES 1916, p. 66. 1. Anonymous. Wireless equipped . aeroplanes in warfare. The Wireless Age, March 1916, pp. 409-421. ABOUT THE AUTHOR 2. Roberts, H.W.. Aviation radio. New Larry Babcock grew up in his dad’s York, NY: William Morrow & radio sales and service business acting Company,1945. as a service tech while still in high 3. Anonymous. Wireless and aircraft. school. He served 2 ½ years in the The Marconigraph, February 1912, p. Army under Gen. George Patton in the 9. 94th Inf. Division in Europe during 4. Anonymous. Wireless in aircraft. WW II and received the Combat The Wireless World, May, 1913, p. 112. Infantry Badge and Bronze Star. 5. Woodhouse, H. How to become an Following discharge he earned a BSEE aviator. The Wireless Age, August degree from the Un. of Iowa and 1917, pp. 833-837. began a 37-year career at Bell Aircraft. 6. Anonymous. War incidents. The At Bell, Larry headed the flight test Wireless Age, December 1915, pp. program to develop radar and air-to- 150-151. air microwave relay links for the 7. The General Staff. Signal training, RASCAL missile system. This was an Part II. Appendix III. Co-operation early guided rocket powered air to between aircraft and infantry, pp. ground missile with a nuclear war 283-295. London: His Majesty’s head. Next, for 5 years he was Stationery Office, 1921. Engineering Manager for the Army’s 8. Anonymous. Seeing wireless Electromagnetic Test Facility in Ft. signals. The Electrical Experimenter, Huachuca, AZ. This was a program to November 1917, pp. 442-443. solve the interference problems of 9. A British Army Officer. How Army radio communication on the wireless is being used in the war. The battlefield. Finally, back in Buffalo, Wireless Age, March 1916, pp. 394- NY during his last 20 years he was an 396. Electromagnetic Compatibility 10. Lauer, H., and Brown, H.L.. Radio Engineer and designed military engineering principles. Chapter XII, electronic equipment to be immune to Miscellaneous applications of radio Radio Interference, Lightning, circuits, pp. 283-287. New York, NY: Nuclear radiation and TEMPEST McGraw-Hill, 1919. (security) requirements. He consulted 11. Miessner, B. Development of in these fields at Sierra research in aircraft radio in the Navy. The Buffalo, NY and many other Electrical Experimenter, November companies throughout the U. S. 1917, p. 465. Larry began collecting antique 12. Anonymous. Windmill-driven radios in about 1973. He specializes in generator for aero sets. The Wireless Federal, Wurlitzer and WW1 aircraft Age, July 1915, pp. 755-756. sets. He has written a book on the 13. Tyne, G.F.J. Saga of the vacuum history of Federal radio and another tube. Indianapolis, IN: Howard W. on Wurlitzer radio. He is on the board Sams & Co., Inc., 1977, pp. 147-148. of directors of AWA, Inc. and serves 14. Secor, H.W. Flying across the as a guide at the AWA museum in Atlantic. The Electrical Experimenter, Bloomfield, NY. He also has an exhibit July 1919, pp. 214-247. of his radios at the Amherst Museum 15. Anonymous. U. S. Navy airplane near Buffalo, NY and serves as a guide sets. Radio Amateur News, July 1919, there. pp. 20-24. 16. Anonymous. German wireless apparatus on aeroplanes. The Wireless World, July 1918, p. 210. 17. Anonymous. Artillery fire directed from aeroplanes by wireless—

Volume 20, 2007 95 WWI Aircraft Radio

96 AWA Review Teague & Knight The 1902 Wireless Connec- AWA Review tion - Santa Catalina Island to San Pedro, California ©2007 Norwood Teague and Joe A. ABSTRACT Knight

INTRODUCTION One of the com- mercial ventures in It was in the early 1970’s, while browsing early wireless was the through some early radio/electronics maga- Pacific Wireless Tele- zines in the technical history collection of the phone and Telegraph Los Angeles County Museum of Natural His- Compnay. It used a tory, that I came across an article in Radio system based on the Broadcast, from May 1924, by Robert Marriott available inventions in entitled “As It Was In The Beginning”. Marriott the literature at the described his first involvements with “wireless,” start of the 20th cen- and the beginning of the world’s first commer- tury, including a sys- cial telegraph link in 1902. This link was from tem patented by Prof. the town of Avalon on Santa Catalina Island to Dolbear emphasizing the mainland at White Point near San Pedro, ground waves. In its California. For the first time a regular wireless regular wireless con- service was available to the public. A giant step nection established forward had been made from the experimental between Avalon on laboratory to the business world. Having re- Santa Catalina Island sided not too far from the White Point station I and White Point on thus began a thirty year search for more infor- the mainland, it had a mation and details about this local piece of wire- setting that combined less history. an intervening stretch In 1892 three Banning brothers acquired the of ocean and high Santa Catalina Island from the James Lick Trust cliffs at each end. In - Judge Joseph Brent, Captain William and this situation it man- Hancock were the sons of Phineas Banning. The aged to provide a Bannings had a major interest in the Southern regular service with Pacific Railroad, which brought passengers to rudimentary equip- the docks in San Pedro Harbor. They had been ment and little under- running steam ferries across the channel on standing of wireless. It excursions, on fishing trips, and to the Tent City thus managed to offer in Avalon on Catalina during the tourist sum- the world’s first regu- mer season for a number of years. The Termi- lar, commercial wire- nal Railway (later the Salt Lake Line) was com- less telegraph link. pleted to Rattlesnake Island (later named Ter- minal Island) the same year. Here was new money, new interest, and complete control of both the island and the transportation to and from this “Paradise in the Pacific.” The town of Avalon, on Santa Catalina Is- land, is twenty-seven miles from Point Fermin and White Point, San Pedro. How long it took the local Indians to row their canoes across the Volume 20, 2007 97 The Wireless Connection Channel is not really known, but it their encampment on the Island. surely took the better part of a day. However, even a slight haze on the Sailing ships and boats were able horizon made the experiments im- to cut the time to between five and possible. They were still trying as eight hours, and the steam ships late as the summer of 1901. For and launches made it in from two three years, during the summer to four hours. Today, the power seasons of 1894-1896, the Zahn boat service makes it in from one brothers of Los Angeles operated a to two hours (see Fig. 1). regular carrier pigeon service with Attempts at “instant” communi- remarkable success, carrying mes- cation, via smoke signals, were sages to and from individuals as made by the Indians long before well as Los Angeles Times news the “white man” arrived on items. Catalina. This usually required ad- As early as 1899 an editorial in vance notice and was not visible if the Los Angeles Times of 26 May the wind was strong enough to keep 1899 states: the smoke from rising or the fog ‘Why Not a Wireless Telegraph?’ and haze obscured the view. The “The season at Catalina Island silver miners of the 1860’s occa- will soon be in full swing, and there sionally used carrier pigeons to is promise of larger crowds than send messages, but this was not a ever before gathered at that attrac- regular activity. The U.S. Army Sig- tive spot. Catalina is the only place nal Corps, as early as the summer of any importance on the Pacific of 1892, attempted to communi- Coast that cannot be communi- cate by heliographic signals during cated with in any way other than

Fig. 1, Map of Southern California showing locations of Avalon, Catalina and San Pedro. White Point is just west of Royal Palm Beach.

98 AWA Review Teague & Knight mail. For a period The Times main- owner of Cripple Creek, who, with tained a carrier-pigeon service for his wife, has taken one of the news dispatches, but that is not Hutchings cottages; S. C. now working. Since wireless tele- McCummings, a mining superin- graph has proved successful be- tendent at Cripple Creek; W. H. tween England and France, and at Hardinge, a prominent mining en- other points in Europe, the ques- gineer of Denver, with Mrs. tion arises as to whether it would Hardinge; and E. T. Sutherland, be a good idea to establish a ser- owing [owning] mines in this state vice between Catalina and the and Oregon, with his wife. The lat- mainland, as the cost cannot be ter has taken the new Frank great. Apart from the convenience Hutchings cottage for the season.” to the public, the owners of the is- As a matter of fact there were oth- land would certainly find this a ers with Cripple Creek mining in- good advertisement. There are, it terests, as well as many from the is true, a few people who would like Denver area, who visited Catalina to escape entirely from business throughout the year. This was dur- cares, and take their vacation in ing the tenure of General A. W. some place where they cannot re- Barrett as President of the Tuna ally be reached, but for everyone of Club who was a principal in the this class there are a dozen who early evolution of this story. The prefer to remain in touch with the General and Mrs. Barrett were very outside world.” active in Catalina social and fish- By 1901 people were coming ing scenes. Apparently, it was here from everywhere to the Los Ange- that we have the “hidden connec- les area. The Los Angeles Times tion” between Denver and Santa reported on 14 May 1901 that Catalina Island and the interest in 30,000 people had arrived in Los establishing the “wireless connec- Angeles during the previous week tion,” as an item in the Los Ange- by way of the railroads (Southern les Times of 12 December 1901 Pacific, Santa Fe, Terminal and the would suggest: electric lines), many of them pay- Wireless Telegraph Messages ing a visit to Catalina. An estimate Will Soon Come From Catalina To of “from 8,000 to 9,000” were Los Angeles “sheltered in the little nook of Avalon, Dec. 11.—[From The Avalon” during the weekend of 17- Los Angeles Times Resident Cor- 19 August 1901, according to the respondent.] “Within sixty days Los Angeles Times of the 18th. The messages between Avalon and Los Wilmington Transportation Com- Angeles will be transmitted by that pany estimated that over 60,000 mysterious, but now practical enjoyed the pleasures of Catalina agency, the wireless telegraph. The during the season. importance and advantages of this Among the visitors were a num- innovation cannot be over-esti- ber from people from Colorado. An mated, as the only means of regu- item of future interest to this story lar communication with the main appeared under the Santa Catalina land now are slow-plying steamers. Island head in the Los Angeles Gen. A. W. Barrett has returned Times for the 16th of July: from Denver, where he made a per- “A party of practical mining sonal investigation of the system, men came a few days ago and are and has the utmost confidence in arranging to spend some time here. this marvelous method of commu- They are Lee S. Wood, a mine nication. He has accordingly ar-

Volume 20, 2007 99 The Wireless Connection ranged with the Pacific Wireless outcome of his experiments with Telephone and Telegraph Com- telegraph and telephone commu- pany to install a plant here; the nication, was Amos Emerson Banning Company has accorded Dolbear (1837-1910) of Tufts Col- the necessary privileges, and the lege in Boston, Massachusetts. His wireless people give assurance that patent application, dated 24 March within two months they will begin 1882 and granted 5 October 1886, flashing messages across the chan- stated, in part: “...and it [my inven- nel to San Pedro and on to Los An- tion] consists in connecting the geles. transmitting instrument with a This island is ideally situated for ground the potential of which is the employment of this novel considerably above the normal, method, there being bold hills at and the receiving instrument with either end of the line, and there is a ground the potential of which is little doubt of its successful opera- considerably below the normal, the tion.” result being that an impulse from the transmitter sufficient to cause PROF DOLBEAR’S INVEN- the receiver to give intelligible sig- TION nals is transmitted through the During the 19th Century, more earth without the need of any cir- and more interest was being gen- cuit.” erated in the new field of electric- Although he was using aerials, ity by the remarkable discoveries Dolbear was concentrating on the and developments both in the earth ground alone. His system United States and in Europe. By was upside-down. There was not the latter half of the century an enough energy generated in the abundant harvest of technological aerials to produce an electromag- invention and innovation had been netic wave action. The electrostatic reaped. In the United States this field that was generated in the harvest was stimulated by the pro- aerial had only a short-range effect. fessional inventors, led by Thomas However, he did succeed in detect- Edison and Alexander Graham ing transmissions over a distance Bell, the activities of “garage me- of as much as thirteen miles. He chanics,” as well as academia. By was so close. The ingredients were 1880 the electric telegraph and the there, if he had only recognized the telephone inspired many questions potential of electro-magnetic relating to the desire to do the waves transmitted through the air. same thing without the cumber- He was very close but he did not some, expensive and trouble-filled pursue the project any further and network of wires. New develop- later discoveries by others contrib- ments in the laboratories were be- uted to the closing of the gap. ginning to suggest that it was pos- Why Dolbear did not continue sible to generate electrical waves at his work is not fully known. Being one point and detect them at an- a physicist and a professor he must other point over a distance of noth- have known of Hertz and his work. ing but air. Could there be wireless It is strange that he did not relate communication by controlled in- that work to his own wireless ex- terruption of these electrical waves periments and resulting patent. He with that controlled interruption seems to have been far more inter- detected at another point? ested in his general electrical ex- One physicist, who had been periments and the litigation with working with this question as an over tele-

100 AWA Review Teague & Knight phone priority rights. He made no saw, apparently, an opportunity to further contributions to the field of get in “on the ground floor” of a wireless communication (Fig. 2). new investment possibility. He ei- It was in 1886 that Heinrich ther searched the patent records Hertz successfully proved in labo- and discovered that Prof. Dolbear ratory experiments William Clark held a patent in the field, or he was Maxwell’s theory of electro-mag- told of the patent. At any rate, netic radiation. “In 1892 Sir Will- rights to the Dolbear patent were iam Crookes wrote of the possibil- purchased, although the patent ity of communication with Hert- had only four more years of life. A zian waves. Guglielmo Marconi, new company, the American Wire- inspired by reading Hertz’s own less Telephone and Telegraph account of these new waves, sent Company, was incorporated in his first wireless message in 1895, what was then the Arizona Terri- only nine years after Hertz’s labo- tory on 10 November 1899. This ratory discovery. But Hertz him- was the first American corporation self, strangely enough, showed no founded to exploit commercially prescience about possible uses of the potential of wireless commu- his waves; he even argued that they nication. During the next eighteen months, American Wireless Tele- phone and Telegraph began a strong promotional campaign to attract investors. The publicity caught the eye of a student at Ohio State University by the name of Robert H. Marriott who was study- ing physics and specializing in “wireless”. After interviewing he was given a job as a technical spe- cialist with American Wireless Telephone and Telegraph Co.2 There he found Harry Shoemaker, as Chief Engineer, and Greenleaf Whittier Pickard. Both men made significant contributions in the early days of wireless and have a number of patents in their names. Pickard went on and became the inventor of the crystal detector. According to Marriott, stations were built at Galilee, Briele and Fig. 2, 1886 Patent by Barnegat, New Jersey, by the for a mode of wire-less communica- American Wireless Telephone and tion using a ground-plane network. Telegraph Company. These sta-

”1 tions were built for experimental could not be of any practical use. purposes and to report the However, in the late 1890’s a America’s Cup yacht races between Dr. Gustave P. Gehring, a physician Columbia and Shamrock in the in Philadelphia, Pennsylvania, be- Fall of 1901. Marconi had reported came interested in what he had the races in 1899 and was to be in- been reading about the develop- volved again. Reports also sug- ments in wireless telegraphy. He Volume 20, 2007 101 The Wireless Connection gested that Lee de Forest would be landed on another boat. there as well. Three broadband Since Dr. Gehring’s interest in spark outfits operating within a Prof. Dolbear’s patent was prima- few miles of each other posed sig- rily financial, technical innovation nificant interference problems that was left to the company techni- were both electrical and problems cians. The franchise rights to use in promotional possibilities. Big the patent protection also included egos were involved here and no one the services of the parent was going to back off in favor of the company’s technical staff. Fran- competition. chises were sold to several new Although a number of histories companies. In turn, each company state that some New York newspa- was assigned a specific region of per ran a by-line, “Received by operation. Two of the regions were Wireless Telegraphy,” there was no bought by investors in Denver, mention in the New York Times of Colorado. The Continental Wire- any wireless communiques. Likely less Telephone and Telegraph they each drowned out the other Company had the Mountain such that no one got through at all. States, and the Pacific Wireless The Times does state in one story Telephone and Telegraph Com- that an attempt was made to com- pany had the Pacific Coast. They municate from the fleet by carrier were incorporated in South Dakota pigeon but that the poor bird be- on 26 June 1901. The Pacific Wire- came so exhausted trying to fly into less Telephone and Telegraph a head-wind that it gave up and Company was also incorporated in

Fig. 3, Copy of a 25 share stock certificate for the Pacific Wireless Telephone and Telegraph Company, incorporated in South Dakota and issued in 1902. Joe Knight.collection.

102 AWA Review Teague & Knight the Arizona Territory on 1 July absorbed it in the Fall of 1903. 1901. Both of these corporations With this transaction de Forest remained on the books until statu- acquired a large number of Shoe- tory termination, twenty and maker patents. He acquired also twenty-five years, respectively, the Dolbear patent which had ex- even though they had long since pired on October 4, 1903. Appar- ceased to exist (see Fig. 3). ently, the petition for extension Robert H. Marriott was offered had died in committee in the Sen- a job as Chief Engineer of these two ate. Nevertheless, de Forest intro- companies and reported for work duced the patent as part of the de- in Denver in the Fall of 1901 to be- fense in a patent infringement liti- gin planning the stations. The gation brought against de Forest American Wireless Telephone and Wireless Telegraph Company by Telegraph Company was supposed the Marconi Wireless Telegraph to furnish instruments as well as Company of America. The patent technical aid. According to was not considered in the court Marriott this was not forthcoming, decision. and he had to design and build the In Denver some of the investors equipment himself in Denver. Ap- wanted to stay “close to home.” parently, American Wireless Tele- They could see a future network phone and Telegraph furnished between the mining communities little or no support of any kind, and headquarters. They wanted to even though Shoemaker and establish the first link between Pickard were on the staff at the Golden and Denver, a distance of time. only about twelve miles. However, Little is really known about the there already was wire telegraph operations of the American Wire- and telephone communications less Telephone and Telegraph throughout the region. This would Company. The patents of Shoe- have been simply another case of maker during the period 1901- experimental or demonstration 1902 were assigned to Gustave P. stations. Many people from Den- Gehring and/or Marie V. Gehring ver had visited Santa Catalina Is- and/or American Wireless Tele- land so it is certainly probable that phone and Telegraph Company the investors knew some of them and/or Consolidated Wireless and had heard of the splendors of Telephone and Telegraph Com- the “Magic Isle”. These visitors pany. After the franchises had been were aware of the limited means sold and the failure to get Dolbear’s of communication that existed be- patent renewed, Gehring seems to tween the island and the mainland have lost interest. In 1902 The and the interest of the Banning Consolidated Wireless Telephone brothers in improving the service. and Telegraph Company bought They could have made this known out or absorbed American Wire- to those in the company arguing less Telephone and Telegraph the “other side” of branching out. which in turn was bought out or Besides, the terrain (ocean) was absorbed by the International flat and unobstructed, and there Wireless Telegraph Company in was high ground near the water on 1903. both sides to give elevation to the According to Lee de Forest, aerial masts. Harry Shoemaker owned Interna- The “hidden connection” with tional when the American de For- Catalina was, no doubt, the very est Wireless Telegraph Company popular president of the Tuna

Volume 20, 2007 103 The Wireless Connection Club, General A. W. Barrett. Along Marconi wireless system on the with his equally popular wife, he Pacific coast. would have been known to all who Hancock Banning said today his had spent any time on the island. company has expended upward of He was also a good friend of the $100,000 in improvements on Bannings. His visit to Denver in Catalina Island since last summer. early December of 1901 would Thirty additional rooms have been seem to support the “hidden con- added to the Metropole hotel all nection” premise. having baths attached, and the The die had been cast and water system of Avalon has been Marriott was busy getting all of the enlarged. necessary equipment designed and The Banning Company has do- manufactured through the Ameri- nated a site for the proposed new can Wireless Telephone and Tele- Marconi3 telegraph station. The graph Company. Gustaf T. Los Angeles terminus of the enter- Swenson, one of the new men prise will be one of the high places hired by the company, became in the city, probably Fort Hill. The Marriott’s assistant. By April 1902 machinery for the purpose was ex- all of the instruments had been pected here Monday. It is en route completed and were ready to ship and will reach Los Angeles in about out west to California (Fig. 4). a week. Gen. A. W. Barrett is local agent for the Marconi syndicate. MARCONIGRAMS FROM In its preparation for the sum- AVALON mer season of 1902, Santa Catalina Wireless System of Telegra- Island was becoming “up-town” phy Is to Be Introduced on This with electricity, had 30 new rooms Coast. “Within sixty days Avalon, with their own baths, a new Catalina Island, will be lighted by steamer and instant communica- electricity and will have telegraphic tion with the mainland only twenty communication with the out-side seven miles across the sea. A lot world. The last-named innovation was going on to make “getting will mark the introduction of the away from it all” have “all the com-

Fig. 4, Staff at American Wireless in Denver who built the Pacific Wireless apparatus, 1901-02. Marriott is at the left with hand in pocket. Swenson is in back middle in checkered shirt. Note the large spark coil they are standing around. From Radio Broadcast, May 1924.

104 AWA Review Teague & Knight forts of home.” establish the first stations in the The advance guard, in every North, one at Seattle; but Capt. sense of the phrase, arrived in Los Banning met Gen. New in San Angeles on April 4,1902 in the Francisco and told him of the need form of General Albert L. New, of a means of communication be- Vice-President and General Man- tween Catalina and the mainland. ager of the Pacific Wireless Tele- Men of large interest have been phone and Telegraph Company4. worried at being so far from a tele- (‘General’ New - it is not known graph wire.” where the title comes from unless By the end of the month Rob- it was from the general in “general ert H. Marriott, Chief Engineer, manager.” There is no indication and Gustaf Swenson, Chief Electri- that it was a military title.) He was cian, had arrived from Denver with described as a person whose “prin- all the equipment necessary to in- cipal fault is that he works too hard stall both the Catalina and the at whatever he attempts”. In his mainland stations. Final site selec- spare little anatomy he carries al- tions got underway immediately. most enough magnetic enthusiasm The equipment for each station to transmit messages from the consisted of a gasoline engine to mainland to Catalina.” There were power the dynamo which provided many other superlatives that had 110-volts for an of to have come directly from the 750,000 volts capacity with a General, himself, since he had been three-quarters inch spark. in Southern California for only a In the beginning a mechanical week. But that seems to have been interrupter produced 70 sparks per the nature of this dynamic pro- second. The transmission of en- moter. He came in talking not only ergy through the system was con- about the Catalina project but trolled by a regular telegraph key about many other grandiose and relay. There was also a primary schemes as well. And Mrs. New and an operating switchboard. immediately became a star in her Unlike Marconi, who was using the own right “when she caught the coherer, Marriott chose a micro- only fish taken by the party” at phonic detector. He described it as Catalina. The party consisted of having “a steel needle with a her husband, her son and General slightly rounded point held against A. W. Barrett, president of the an iron oxide surface by a fine- Tuna Club. threaded adjustment screw, fitted It was General Barrett who had with a large knurled knob. The ba- made the trip to Denver in Decem- sic structure was made from a ber of 1901 to make contact with spherometer. The most satisfac- the Pacific Wireless Telephone and tory oxidized iron was obtained by Telegraph people and arrange for burning the tin off a Prince Albert their coming to Catalina Island. He tobacco can. A small drop of oil must have been sent by the Ban- protected the area of contact.” The ning brothers, for on the arrival of receivers were “two re-wound General New in Los Angeles it was Stromberg telephone receivers, reported on April 5, 1902 by the with a common head band, to get Los Angeles Times that: the most out of both ears and to “It was through the influence of exclude outside noise.” “The circuit Captain [William] Banning that is completed by the ground wire, the company was drawn down which is connected with the oppo- here. It had been the intention to site pole on the plates and the

Volume 20, 2007 105 The Wireless Connection ocean. The water carries the return became apparent that the closer to current which completes the cir- Avalon the better! And General cuit.” New, never a man of few words, is The Avalon site that had been quoted by the Los Angeles Times offered by the Banning company on April 24, 1902: on Metropole Avenue was deter- “We had it in view to locate the mined to be too close to the large mainland station in Los Angeles,” equipment at the new electric plant said General Manager New today. across the street. There was a con- “We find that the importance of cern that the heavy machinery San Pedro as a shipping center is could cause interference through much greater than we had sup- both vibration and noise. Also, it posed and it is not unlikely that the was still perceived by those in the station will be located near San early days of wireless there needed Pedro, instead of farther inland. If to be an aerial mast of about 150 the station be established near San feet and placed on as high a prom- Pedro, then we shall within the ontory as possible. near future establish another one The site finally chosen was on at Los Angeles. the inland stage road on the hill The exceedingly delicate char- above Sugar Loaf Rock (the station acter of the instruments employed site still exists above the now fa- in wireless telegraphy makes it mous Avalon Casino). This site desireable that the transmitting provided also for un-interrupted stations be isolated from the influ- “line-of-site” possibilities to many ence of the induction of electric mainland locations as well as ships lighting and power circuits. An iso- at sea. lated location such as is needed can On May 10, 1902, General New probably be secured here better laid the cornerstone of the build- than at a point farther inland. ing in impressive ceremonies. It We expect at no distant day to was 20 by 30 feet in size (an aver- see a great many, if not all, of the age of conflicting reports) contain- vessels that come into San Pedro ing two rooms: one for the heavy equipped with wireless telegraphic machinery and the other for the apparatus. With that in view it operating equipment. The sensi- seems more practicable to have the tive instruments were mounted on station close to the seaport, so that solid concrete piers in the operat- communication between the port ing room which was packed with and vessels out at sea may be more sound-deadening compound. A readily maintained. small, four by four foot, sound- Each station will be mounted on proof closet, much like a telephone a solid concrete base, so that the booth, was provided for the receiv- instruments may be disturbed by ing operator and his equipment. At outside influences as little as pos- best it was difficult to pick up the sible. No great length of time will faint sounds in the receivers. As the be required for mounting the in- time grew closer to the “opening” struments, but considerable time day, a tent was erected along side may be consumed in establishing the station for the operator. Mes- communication. That will be sages were to be received and sent largely a matter of adjustment of up to a “reasonable” hour of the the instruments. day. I fully expect that the operators Although several sites on the at the insular and mainland sta- mainland were considered, it soon tions will be in intelligible commu-

106 AWA Review Teague & Knight nication very soon after the instru- pavilion at White’s Point for a sta- ments are set up. However, that tion. This building is now being may not prove to be the case, in enclosed and instruments, which which event we shall have to work have already arrived, will be in- on the instruments till they are in stalled next week.” perfectly harmonious adjustment. The work on the hill above If we do not succeed in getting Avalon was progressing nicely. them in that condition soon after ‘Telegraph Hill’, as it was now be- they are set up, we may establish a ing called, had become a popular third set of apparatus temporarily place for a stroll, even though it on a vessel and operate back and was at an elevation of over 300 feet forth across the channel for a while above the downtown. One could until the permanent instruments look out across Descanso Bay to are adjusted by the temporary one, the Hancock Banning residence, so that they will work in unison.” with the beautiful San Pedro chan- At the same time General New nel and the mainland in the dis- continued to talk about new sta- tance (on a clear day). tions in Los Angeles at ‘Angel’s Avalon Bay was to the right with Flight’, at Third and Olive, at Point the mountains behind. The two Loma (near San Diego), and “other poles that were to make up the 150 locations” along the West Coast. By foot mast were in place, the lower the first of June, 1902, a deed had portion being a ninety foot stick of been filed between Roman D. Oregon pine two feet in diameter Sepulveda and his wife and the at the butt end. The upper portion Pacific Wireless Telephone and to be spliced to it was the old flag Telegraph Company for a lot in pole from in front of the pavilion. Sepulveda Park at White Point in Both station poles were in place by Rancho Palos Verdes, “...and will the middle of June and the work use for the present, at least, the of making the necessary adjust-

Fig. 5, Avalon Bay, Catalina with Pacific Wireless station shown in the upper middle portion. The small building and antenna can just be seen along the stage road, above what was then Sugar Loaf Rock (later the Casino location). Courtesy of Los Angeles Country Museum of Natural History (LACMNH).

Volume 20, 2007 107 The Wireless Connection ments to get the equipment “in the entire shoreline of the Bay tune” began (Fig. 5). there were colored fires and Chi- By the middle of June two tele- nese lanterns. The wireless station graph operators had been hired. joined in on the celebration by Veil V. Stevenson, formerly with lighting a string of lights all the way the Western Union Telegraph to the top of the aerial mast. “[The Company in Los Angeles, was put lights] were over five hundred feet in charge of the Avalon station al- above the town and looked as though he fades away soon after. though they might be the portholes W. R. Carroll was put in charge of of heaven.” (Fig. 6) the White Point station. He had General New continued his been a member of the Signal Corps during the Spanish-American War and had been employed by the Marconi company in their unsuc- cessful attempt to establish a wire- less communications network in the Hawaiian Islands. By the end of the month trial messages were being exchanged between the two stations and ad- justments to the instruments had begun. Many problems still needed to be overcome. The operators, who were accustomed to listening to the clicks of a telegraph sounder and the reading of recording tapes, had to “tune in” their ears to the faint buzzes in the telephone re- ceivers. And the faintness of the buzzes was the reason for the sound-proof reception room. As a matter of fact, at times the gaso- line engine with their spark igni- Fig. 6, Finished White Point Station tion system driving the dynamos with the spliced-pole antenna. The In- had to be stopped to receive sig- dependent magazine, October, 1903. nals and then started again to talking about other stations being transmit. There was also the learn- considered. If nothing else, he was ing of the delicate process of how never at a loss to paint the big pic- to adjust properly the screw of the ture. Whether these were to some detector needle to get the best re- extent illusions of grandeur or sults. whether they were ways to pro- The Fourth of July was a spec- mote the project, the result was tacular event at Avalon. The new publicity. He was, however, con- steamer, Hermosa II, had been put stantly faced with one challenge in service and the old and leaky after another. After all, this was a Hermosa I had been retired. It was period in which all of the working decided that a proper end for the necessities were not readily avail- “old girl” would be to burn her in able. One of the local challenges Avalon Bay and what a bonfire she was how to get the messages from did make! The Hotel Metropole White Point to the closest tele- was outlined with lights and along graph office in downtown San

108 AWA Review Teague & Knight

Pedro, a distance of almost three progress or lack thereof. miles. Negotiations with the local By the end of July some suc- Postal Telegraph and Western cesses were being recorded. The Union companies for the extension first real piece of news transmit- of their lines from San Pedro was ted for public information was the needed to connect Avalon effec- result of the Jeffries-Fitzsimmons tively with the outside world. Be- fight held in San Francisco on the sides, he said he was investing night of July 25th, 1902. about $28,000 to get these two “Jeffries wins fight in eight station in operating readiness. rounds. No particulars at this One of the their most irritating hour,” was the message received at local challenges began with the the wireless station on the hill at July 7th 1902 edition of the Los 12:04 o’clock this morning. Angeles Times. The Avalon “Resi- “It was the first message trans- dent Correspondent” (J. S. mitted for the public and was ad- Mathes) of the Times wrote the dressed to “guests and residents of first of four articles that were Avalon,” and thus was set at rest “tongue-in-cheek” negative. These all doubts as to the ability of the articles were headlined as: “The company to send and receive mes- Wireless is Hoodooed”, “Wireless sages. It was passed back and forth is Still in Air”, “Wireless Worked between stations twice in order Tuesday Says New”, “Got it There that there would be no error, as by Wireless”, and “Wireless ‘Open- those who were at either end of the ing’ Remains Unopened”. “line” were amateur telegraphers It was in the last of these articles although expert electricians, and that the reporter attempted to add they wanted to be absolutely cer- insult to injury. His lead sentence tain with the first dispatch. quoted the first few words from Knowing the anxiety of many “Wireless”, Rudyard Kipling’s new- on the island to hear news from the est short story. “It’s a funny thing ringside, General New arranged to this Marconi business, isn’t it?” have a messenger in waiting in San The Los Angeles Evening Express Pedro, and the instant the word reporter was at it also. He had been was received there as to the out- trying all sorts of legal and illegal come of the fight, he was to make methods of getting the first “scoop” a Paul Revere ride on horseback to on the wireless opening. It is inter- the station at White’s Point, and esting that the Express ran no fur- from there the message would be ther articles covering the progress sent to Avalon by the wireless of the wireless after June 12, 1902, route. apparently feeling shut out of the This program was carried out to loop. the letter and at 12:04 this morn- This was a difficult personal ing, almost twelve hours earlier period too as the operator than the first boat arrived, the re- Stevenson had apparently de- sult of the conflict was known parted and the other operator, here.”5 Carroll, could be at only one place The tone of W. J. Rouse, the Los at a time. Marriott or Swenson, Angeles Herald reporter, was far neither of whom were telegraph more supportive of the efforts of operators, had to man one of the those attempting to put the wire- sites alone. It was a tense period, less connection into functioning and they did not need any of the order than that of the Times re- negative press needling their porter. That support was not lost

Volume 20, 2007 109 The Wireless Connection on General New, Marriott and the The “pony express” was just not others. It was the Herald that was the way to go. At Avalon the tele- given the privilege of sending the phone line had been completed first commercial messages ex- from the Hotel Metropole to the changed between the two stations. station on the hill. However, mes- Actually, they were not “commer- senger service was still to be pro- cial” but, rather, “courtesy” mes- vided for those not living at the sages. As reported by the Herald hotel. Young boys were hired to on August 3, 1902: run messages back and forth from ‘First Message Sent’ the town below to the station on No. 1 the hill, possibly summoned by flag San Pedro Station, Aug. 2, 1902. signals. The first wireless byline To Avalon Station: [by Wireless Telegraph] appeared Sawyer of The Herald is here. in the Los Angeles Times on 20 (signed) L.A. Herald August 1902. This “First Official” Time. 1:10 p.m. Code, Morse. message was sent by General New Words, 6. “The message was repeated sev- eral times and in less than five min- utes after the motor was stopped to allow the receiving apparatus to work the answer began to arrive. It was: “Have no message for Saw- yer,” followed by dots and dashes that could not be made to mean anything. A call was sent to repeat the last sentence, and at 1:30 p.m. the last sentence was repeated. It proved to be: “Rip and roast the Times for us.” (signed) Marriott [Insert image #7] There were at least a dozen messages sent and received during that first session. It is interesting that soon after this message was published, the Herald reporter and the Times reporter, who was also a telegrapher, were given the free- dom of the plant through the cour- Fig. 7. Finished Avalon Pacific Wire- tesy of General New. After that the less station, c. 1905. Courtesy of the tone of the Times stories im- Catalina Island Museum. proved. Much of August was spent to President Theodore Roosevelt: sending and receiving “courtesy” Avalon (Catalina) Aug. 23.— messages as Marriott and his crew “President Theodore Roosevelt, labored at the further refinement Oyster Bay, L. I.: Santa Catalina of the system. Island, heretofore isolated, sends The Company had made the greeting to her great American decision not to open until the President, on the occasion of her Western Union line was completed being put into communication from San Pedro to White Point. with the entire world through the

110 AWA Review Teague & Knight medium of the Pacific Wireless $1.00 for the first ten words and Telegraph, a purely American in- $0.07 for each additional word. vention, now in successful opera- This was as much as it cost to tele- tion.” graph New York - plus the local (signed) A. L. New Western Union charges, to and General Manager from White Point, were on top of And on the 27th the following that. The speed of the first mes- reply was received: sages seemed to be on an average Ellsworth, Me., Aug. 27.— “A. L. of about ten words per minute. It New, Avalon: Please extend my was expected, however, that as the greeting to the citizens of Catalina operators became more familiar Island and my cordial congratula- with the system, the speed could tions upon the establishment of easily reach an average of twenty telegraphic communication.” words. The cost and the speed, in (signed) Theodore Roosevelt and of themselves, were insignifi- It seems that the President or cant to the fact that something very whoever received and responded revolutionary was occurring here. to General New’s dispatch did not The reporter for the Herald sums fully comprehend what had hap- it up succinctly on August 24, pened — that the message from 1902:

Fig. 8, Completed 1902 interior of the Avalon wireless station with trans- mitter, showing the Denver spark-coil with the motorized interruptor. Santa Catalina Island had been “This was a notable day for sent over a wireless connection! Catalina Island, for, this morning The first message to be received at 8 o’clock, she was rescued from at Avalon from the mainland was the isolation of ages by commer- the announcement by a proud fa- cial communication with the entire ther in Riverside to relatives on world, when the offices of the Pa- Catalina of the arrival of a fine, ten- cific Wireless Telephone and Tele- pound boy; “Both doing fine.” (see graph [above] Sugar Loaf was Fig. 9) thrown open for the transmission The tariff between the two wire- of business. less stations was set at a rate of This morning when the island-

Volume 20, 2007 111 The Wireless Connection within two weeks when two suspects in a bur- glary, who had been under surveillance, were allowed to slip away on the early morning steamer. However, the authori- ties in San Pedro were appraised of the situa- tion by wireless, and the two were arrested on arrival. No longer Fig. 9, The first commercial wireless message, by could hoodlums prac- the Pacific Wireless T&T Co., announcing a new baby tice their trade on the boy, 23 August 1902. Courtesy of the Catalina Is- island and get away land Museum. with it on the steamer ers awoke they were greeted with - a safe practice in the past. the signs of these two great invad- Business continued to be brisk. ers, the old familiar blue and white Businessmen were keeping in of the Western Union Telegraph touch with their businesses and in- company, hung end to end with the dividuals were sending messages, black and gold of this latest won- whether they needed to or not! It der of the world, the Pacific Wire- was the thing to do. All was going less Telephone and Telegraph well with the system, experienced company. That combination of operators were at each end of the business seals scattered the last “line,” and there was little need to doubt of the successful operation make major changes. However, of the new invention, for the ap- Swenson did continue to work on proval of the Western Union, illus- improvements to the receiving ap- trated in its business cooperation paratus (Fig. 10). in a visible form, was conclusive. Marriott went to Denver in Sep- All day the little station on the tember 1902 and on to Washing- hill was a scene of special interest, ton, DC, where he was involved in and everybody wanted to send a message to the mainland. Those who got in on the first day are proudly boasting of the message they sent to “the folks.” At noon a stack of messages still awaited transmission, and these were being rapidly cast to the etheric waves, to con- nect at White’s Point with the waiting wire of the Western Union Tele- graph company.” Just how rescued from isolation Catalina Fig. 10, Interior of the San Pedro station. This is had become was proven the receiving end where the messages would be typed up. c.1903.

112 AWA Review Teague & Knight promoting new business, inter- There was no mention of anyone viewing the Navy Department and who was involved in any form of following up on the new patent research and development. Again, application of Gustaf Swenson’s it seems there were only the inves- for an improved receiver. He also tors who were interested exclu- visited the American Wireless sively in the revenue received in Telephone and Telegraph offices in building operating stations for oth- Philadelphia and found that they ers and not their future operating were essentially moribund. When return. he returned to Denver in Novem- All was going well and General ber the same thing had happened New was looking to new horizons. to Pacific Wireless Telephone and The wireless connection had been Telegraph. He learned that the in successful operation for four company had been sold and that, months and several thousand mes- apparently, he was no longer sages had passed across the chan- needed. He was paid $100 and nel. “We are no longer experiment- then no more.6 ing, said Gen. New. Our experi- General New also left for Den- ments are closed and we are now ver early in September. When he going to erect stations for commer- returned in the middle of Novem- cial use.” ber, F. W. Armstrong, the general As this phase ends, the death superintendent of the company, knell is sounded. The leader of the accompanied him and stated that company says it is only a business it was the intention of the company venture - not an endeavor! to move its general offices to Los Angeles. This seems to bear out the THE BUSINESS VENTURE statement Marriott had made re- At the beginning of the 1903 garding the sale of the Company. new year, a large self-winding On the same boat that took New Western Union clock was installed and Armstrong to Catalina on the at the Avalon wireless station so 15th were George S. Patton (a Di- the correct time could be an- rector of the Banning Company) nounced on the island every day at and wife, their son George S. noon. This development was fur- Patton, Jr. (the future general) and ther proof that the eireless connec- daughter Miss Patton. tion had indeed put this remote Meanwhile, the “store” had outpost twenty-seven miles in the been left, presumably, in the hands Pacific Ocean into the world scene. of Mr. Fred New (the General’s Instant information was becoming son), Gustaf Swenson, and the op- commonplace! (Fig. 11) erators. Mr. W. R. Carroll joined The initial phase had gone so the Navy in October, 1902 and C. well. The second tourist season E. Howell was hired to replace him. was expected to offer even greater And it must have been about this possibilities. The company decided time that William Sener joined the that an office, outside the Hotel company as an operator. However, Metropole, with a constant atten- with the departure of Marriott the dant, was in order. The old post stimulus for technical develop- office, next to the new one near ment in Pacific Wireless dried up. wharf gate, was remodeled for this Although Gustaf Swenson re- purpose. A room was added to the mained on board and was the station on the hill to accommodate holder of three patents, no further the operator on duty (Fig. 12). significant progress took place. General New was still at it. The

Volume 20, 2007 113 The Wireless Connection tions in Alaska as well as to connect with the Philip- pines, China, Ja- pan, and other Asiatic countries. It is also possible that a station will be erected in New York City in the course of a few months. Already the company has planned to build 153 stations throughout the in- terior of Alaska.” If only the Gen- Fig. 11, Interior of Avalon station with Western Union clock on the wall. Early 1903. Courtesy of the Catalina eral had spent Island Museum. more time being Catalina link was doing reasonably “far-sighted” with regard to the well. By the end of January 1903 technical improvement of the sys- he was expanding his horizons in tem. The only one in the company order to continue the promotion. with any ability or interest in mak- As reported by the Los Angeles ing fundamental improvements Times on January 28, 1903: now was Gustaf Swenson7. All in- “It was said at the White’s Point dications are that he was kept busy station of the Pacific Wireless Tele- working on essentially “standard” phone and Telegraph Company equipment. There was one experi- yesterday that within ten days ment of a most interesting nature. work will be commenced upon the Communication was being con- stations for a system of transoce- ducted between Catalina, San anic wireless communication be- Pedro and the Soldiers’ Home, tween some point along the north- near Sawtelle, California. What ern coast and Honolulu. The appa- these experiments were to prove is ratus is nearly ready, and will be not known, but perhaps they were shipped north in a few days. As yet, testing the effects of transmission it has not been determined over land. Or it was to test the new whether the mainland station will aerial system that was soon to be be located in San Francisco or Se- put into use. For by the end of Feb- attle, but indications are more fa- ruary the old round wire that led vorable for the former port. from the station to the top of the Gen. A. L. New says that though mast was replaced by ten thin by his system the Pacific company brass bands one-inch wide. It was has not yet transmitted messages announced that the new system for more than one hundred miles, would be able to transmit over a he believes he can go still farther, much greater distance without an and is satisfied that messages will increase in power. be sent any great distance around The Los Angeles Times re- the globe. After communication is ported the following progress on established with Honolulu, the February 25, 1903: company plans to establish sta- “The Pacific Wireless Telegraph

114 AWA Review Teague & Knight

station to San Pedro. Accord- ing to Armstrong, a lot had been leased on the bluff above Timm’s Landing. This was the first of several locations that were announced before the fi- nal spot was decided upon. In the meantime, the op- erator at Avalon had been busy with more than just the trans- mission and reception of mes- sages by way of the wireless connection. “Charles E. Howell of Avalon was married today [14th], to Miss Amy E. Fowler, also a resident of Avalon, at the home of the groom’s parents, No. 1150 Maple Avenue, Los Angeles. Mr. Howell is a tele- graph operator. The young couple have fitted up an aerie Fig. 12, Pacific Wireless new downtown Avalon office for receiving messages. home on the hill at the wire- 1903. Scientific American, 1903. less station, where they will be after next week.”8 Company is bound to be in the front. Today it opened a downtown While the young couple was be- office on the front street (Avalon) ginning to feather its nest high on immediately adjoining the post of- the hill, the Los Angeles Times was fice, which has been fitted up in making preparations to launch fine style with handsome furniture another ‘first’ in the field of wire- and two telephones, one a long- less telegraphy. On March 25th distance phone, reaching Middle 1903 the first edition of The Wire- Ranch, Empire, and Isthmus, and less, the first newspaper in the the other the local circuit and the world “publishing sure-enough wireless station on the hill. Noth- dispatches transmitted by wireless ing is being left undone for the ac- telegraph,” was issued in Avalon. commodation of the public, and to No longer was the island public show their appreciation seventeen dependent upon boat traffic and telegrams were received for trans- bits and pieces coming over the mission up to noon today.” wireless. A daily paper devoted to There was a need across the communicating the major news of channel, also, to make improve- the day was a reality. It was a four ments. The station house at White page, three column spread, 8 x 11 Point was, at best, a make-shift inches in size and contained not arrangement, having been a modi- only news from “abroad” but also fication of the old Sepulveda Pavil- local news and advertising. ion. The wires of the Western “The first budget of news for Union were also a problem, being The Wireless was filed with the frequently out of order or blown Western Union in this city [Los down. It was announced by Gen- Angeles] in the early morning. At eral Superintendent Armstrong 3:12 o’clock a.m. the whole of it had that the company would move its been received at White’s Point. At Volume 20, 2007 115 The Wireless Connection 5:30 o,clock a.m. it had been trans- ing. The first edition of 1,000 cop- mitted to Avalon. Before 10:30 ies was disposed of in less than half o’clock it was in print, and the first an hour, and a second edition was edition of the paper exhausted and gobbled up with equal haste with- a second edition underway. This out succeeding in supplying the was quick work, considering the demand. As much as $1 was of- disadvantages under which the fered for single copies (regular first edition was gotten out. As price of three cents) when the sec- soon as everything gets to working ond edition was exhausted.” (Fig. smoothly, it will be possible for The 13) Wireless to go to press before This was from the Resident breakfast. Correspondent of the Los Angeles “Surely Avalon will put on met- Times [J. S. Mathes] who was the ropolitan airs this summer with it’s little newspaper’s first editor. own morning paper. No event in Mathes was the reporter who had the history of the island has cre- given the Pacific Wireless Com- ated such a stir as the appearance pany such a bad time during the of The Wireless yesterday morn- development period before the of- ficial opening of the wireless connection. The commentary con- tinued in subsequent is- sues with the most flowery of language. All intended, obviously, to sell subscriptions, or at least as many single copies as possible. “Salute, then, to the men who have accom- plished on these west- ern shores the actual operation of a wireless telegraph system com- mercially, a system that enables the production of the news of the day across a wide channel of open sea beneath which no cable lies and above which no wires are strung. Persiflage aside, we believe this achieve- ment of The Times in presenting to the peoples of all lands the first daily journal to publish actual dis- patches sent by wireless Fig. 13, First issue of the The Wireless, March 25, 1903, 2nd Edition, after the 1st was sold out. Los telegraph to be one of Angeles Times. the notable triumphs of 116 AWA Review Teague & Knight this notable age. It marks another only a few who were directly chal- brilliantly-illuminated mile in the lenging him, such as Slaby-Arco daily march of human progress. It and Telefunken in Germany and de is but the forerunning of greater Forest and Fessenden in the things. ‘Tall oaks from little acorns United States on the East Coast. grow.’”9 At the same time the cable com- While all of this was being ac- panies were expanding their wire complished, Marconi was making systems all over the world, both on preparations to send the first news land and under the sea. The dispatches from New York to Lon- Farallon Islands off the entrance to don. On March 30,1903, two 100- San Francisco Bay, long a danger word messages were sent to the to navigation in that fog prone re- London Times, and it was stated gion, had a cable installed. A new that the era of commercial of lines had been com- had begun! Although there were pleted throughout much of Alaska five cable systems operating under and a connection to Seattle was the Atlantic Ocean, their monopoly soon to be completed. With a enabled them to fix the rates be- proven means of rapid communi- ing charged. With the advent of cation at hand, even with the ini- competition from an entirely new tial costs and the problems of means of communication, with far maintenance, investors were not less overhead, a rate war was in- so inclined to gamble on this new, evitable. The cable companies were and as yet unproven, method. The charging 25 cents a word for regu- Western Union was so sure of it- lar telegrams and 10 cents a word self that it released a statement for news dispatches sent at off that said, “The Western Union ex- business hours. It was being pre- perts have examined the claims of dicted that the rate by wireless the wireless companies, and are would be about two and a half convinced that there is nothing to cents and one cent for the same fear from any such threatened service. Obviously, a few 100-word competition, so far as overland dispatches does not make a suc- transmission is concerned.” How cessful commercial operation. The content, complacent and over-con- New York Times said, “The impor- fident one can get when they have tance lies in the promise, almost been in the “driver’s seat” for such the certainty, of greater things to a long time. come.” However, not all was compla- The promise was being realized cence with the Pacific Wireless in small bits on the “Enchanted Telegraph. Gustaf Swenson had Isle” in the Pacific off the coast of installed his improved method of Southern California. Why there receiving, which included his was no more publicity beyond the “sound intensifier”, that made it local environs is not really known. possible to do away with the It could be that the promoters were sound-proof booth. And a new somewhat provincial and did not building had been erected for the have a vision of the “big picture.” little town newspaper. (Fig. 14) Or it could have been that the East In the meantime, San Pedro still did not take the West seri- had become very impatient with ously. Or it could have been that both the Western Union and the Marconi had so publicized his ac- telephone company. Their policy complishments that all the focus of suspending operations from was on his activities. There were early evening until early morning

Volume 20, 2007 117 The Wireless Connection

Fig. 14, Pacific Wireless station in San Pedro, CA, c.1904-05. Courtesy of the Catalina Island Museum. was leaving it “entirely without his office at Port Townsend but it communication with the rest of the will be also under the control of the world.” This situation also affected secret service officers of the Trea- the wireless connection. Part of the sury Department, to further their reason that Pacific Wireless efforts in preventing smuggling wanted to move into San Pedro along the Sound, which is carried and have Western Union share on constantly. To this end, the rev- part of its building may have been enue cutter Grant, now on duty in to pressure it to remain open all or the Sound, will be equipped with a part of the night. complete apparatus for sending The site finally chosen was next and receiving messages.”10 to the water tank on the south side This involvement in the State of of Ninth Street between Palos Washington was the reason, ap- Verdes and (today adjacent parently, for the need to incorpo- to the old YMCA building). Mr. rate the new company there. The William Muller, of the Banning informal change, from the Pacific Shipyards, was involved in the se- Wireless Telephone and Telegraph lection, and Mr. Henry Weldon Company to the Pacific Wireless was given the contract to build the Telegraph Company that had station. On Sunday night, June 21, taken place in California with the 1903, the equipment for the station sale in the Fall of 1902, was not was moved from White Point, and acceptable in Washington. It is not before dawn communications had clear what all this really meant. been re-established with Avalon. The Washington corporation filed General New’s expansionism a ‘Designation of Agent’ in Califor- had finally paid off in a contract nia in 1904, and the Company in- with the government to establish corporation papers were filed in stations in the Puget Sound area. 1905. This filing marked the de- “At these stations the govern- mise, essentially, of the original ment will use the wireless, not only company, although it’s incorpora- for the purpose of issuing orders tion papers remained on the books concerning the construction work in South Dakota until 1921 and in over which Maj. Penrose, Arizona until 1926. U.S.Army, has supervision from On Catalina, The Wireless 118 AWA Review Teague & Knight newspaper continued to flourish most of the novelty had worn off. through the 1903 season, with Mr. The focus of the people of the Pa- J. B. Elliott becoming the editor. cific Wireless Telegraph Company On September 30, five months af- had shifted to their government ter its debut with but one issue contract work going on in the without its wireless budget, it took Puget Sound area. These were in its last curtain call for 1903, the fact “combination” stations. They tourists season having ended. Like were communication links for the a proud parent, the October 1st Los Customs branch of the Treasury Angeles Times summed up the Department and for the War De- success of the paper by stating: “No partment. They handled some paper ever published has gained public traffic as well. more notoriety for itself, its pub- The land based station at Port lishers and the wireless telegraph Townsend and the water based sta- in so short a time as The Wireless. tion on the revenue cutter Grant There is scarcely a spot in the civi- were working fine. Experiments by lized world where it has not pen- these stations with Tacoma and etrated.” Seattle had proved successful. A The wireless, itself, had estab- station at Seattle and another at lished its link with the newspapers Friday Harbor were well on their on the mainland. News items from way to completion with plans to Catalina were headed as Wireless expand to Fort Casey and Cape Isle Items, Wireless Channel, and Flattery. If all went well with these Wireless Sea Ripples. While the stations, the hope of reaching leaders of the company were in the Dutch Harbor could be realized. Northwest or in the East supervis- This made the possibilities of an ing the installation of new stations Alaskan network easy to reach. The or attempting to drum up new world of the West seemed to have business, the wireless connection no limits. was becoming “old hat.” As noted General New spent the first six by the Los Angeles Times on 26 months of 1904 in the East on December, 1903: business for the company and, pre- ‘Twas a Merry Christmas sumably, in the Northwest super- Avalon. Dec 25.— “Catalina has vising the installations going on had a merry Christmas. The fes- there. Without him in Southern tivities began last evening with an California there was absolutely no entertainment at the Congrega- news in any of the newspapers tional Church for the Sunday- about the wireless connection. Ap- school. There was music and reci- parently, he was the news, not the tations, an address by the pastor, wireless telegraph, or, at least, he and then the feature of the most was the news maker. Once he ar- interest to the little folks. Little rived back in town, however, “Pinkey” Moore, the first wireless things began to happen. messenger boy on earth, came The first thing he did was to stalking in arrayed in his uniform make plans to move the Avalon and delivered a message direct station from up on the hill down from Santa Claus, announcing that to the office in town. This was to he would arrive on the island at 8 eliminate one office and simplify o’clock sharp, accompanied by his the whole business operation. The wife and two assistants.” new receivers, developed by The Catalina-San Pedro link Swenson, had already made it pos- was working “to perfection” and sible to do away with the sound-

Volume 20, 2007 119 The Wireless Connection proof booth. It had become clear six-month run in 1903 lasted only that a shorter mast of about fifty one month in 1904. feet would do as well as the 150 foot At the same time progress was mast installed at the station on the being made in moving the Avalon hill. station. A wire had been strung The Wireless newspaper, which from the mast on the hill to the had not resumed publication with office downtown and successful the opening of the 1904 tourist sea- experiments in transmission had son, announced that it would de- been conducted with San Pedro. By liver its first 1904 issue on August the middle of September a short the 2nd. Why it was so late in re- mast was erected behind the office starting was not explained, but the and the move was made. return of General New may have Plans were also in the works to had something to do with it’s re- establish a new station in down- appearance. As it announced in the town Los Angeles, and by the end new season first edition: of November work was well on its “..At the opening of the present way. A very high pole, 212 feet tall, season it was the ambition of Gen. had been erected as the aerial A. L. New, manager of the Pacific mast. Wireless Telegraph Company, to “The plant is being installed make a new record in speed of down on Seventh and Alameda transmission, and he agreed to de- streets on a vacant lot between the liver 400 words of press news at Angelus Flour Mills and the Los the office of the morning Wireless Angeles Organ Company build- by 5 a.m. ings. The offices will be in the or- He was better than his contract, gan company building, to which a for at 4:05 a.m., A. F. Krenke, the heavy three-strand cable will be operator at the Avalon station, had carried from the top of the high received the first word of the bud- pole.”12 get and in twenty-five minutes had This station was put in opera- received the entire 425 words. tion sometime in January of 1905. There was not a “break” - not a This brought the wireless connec- word missed or repeated - and it tion closer to the center of things marks the Pacific Wireless System and eliminated the need to use the as absolutely perfect and estab- Western Union wires from San lishes a world’s record for wireless Pedro to Los Angeles. It did require speed in transmission.” the station in Avalon, however, to Gen. New was so well pleased move back up on the hill in order that he sent the following message in recognition of the services of those who handled the lightening: “To F. W. Lilley, San Pedro: Ac- cept congratulations on your quick service. You and Krenke have es- tablished a world’s record and have the thanks of our company. Keep up your good work.”11 The paper finally ceased publi- cation on September the 1st “for Fig. 15. Pacific Wireless Telegraph the season”, never to reappear, and Co. complimentary wireless telegraph with not one word of explanation. ‘frank’ for free messages of a personal What had been a very successful nature only, 1905. Joe Knight collec- tion. 120 AWA Review Teague & Knight to use the higher mast. How much jor stock promotion campaign was the Company planned to use the mounted by Marconi. At the same San Pedro station after the Los time the American DeForest Wire- Angeles station was put in opera- less Telegraph Company estab- tion is not known. In March 1905 lished offices in Los Angeles. No a storm snapped off the pole on the doubt there were some private, Ninth Street station, and there is experimental stations operating on no evidence that it was ever re- a limited basis, and there were at paired. least ten stations established by General New made only two the United States Navy on the Pa- appearances in print during 1905: cific Coast. But there is no evidence once in April on a visit to the Is- there were any other commercial land and again in December when stations established. he spent two weeks recuperating The last time General New is from an attack of pneumonia. mentioned in the newspapers was There was no mention of any de- in September, 1906. Then he just velopments in the Northwest or seems to fade away. He is not listed plans for any further expansion. All in the 1907 Los Angeles City Direc- seems to have become status-quo. tory as he had been since the 1903 Whether this was due to the Gen- edition. In November, 1906, Eben eral having burned out or the Com- Smith died and his son, Frank, be- pany having lost interest in the came President of Pacific Wireless endeavor is not known. It certainly Telegraph Company. However, it marks the beginning of the end for became clear, almost immediately, both. This was somewhat apparent that Frank Smith was more inter- from a Los Angeles Times news ested in other things, especially item on February 10, 1906, when real estate investments. Most of the stations that the Avalon opera- 1907 was a quiet year. Mr. F. W. tor was “visiting” with were not Lilly [Lilley] was made division identified as Company stations. superintendent in April, but in Avalon, Feb. 9.- “Operator November the first sign of real Hickman at his “wireless” aerie on competition was reported. the hill now has the opportunity of “The Queen deserves her enlarging his calling list. Since this crown, for yesterday she sent the station was erected it has been iso- first wireless message from out at lated with the exception of its com- sea ever received at Redondo from panions at San Pedro [first] and a vessel of the merchant marine. Los Angeles [later], now he can The Queen is one of the large visit with Santa Barbara, San Jose steamships of the Pacific Coast and San Francisco. From each of Steamship Company, and arrived these places he picks up messages last evening at Redondo from San daily.” Francisco. She has just been Competition had begun to equipped with wireless appara- “raise its ugly head” as early as tus.”13 March of 1904. A representative of Redondo (Beach)?? There was the Marconi Wireless Telegraph no company name mentioned and Company of America gave private no reference has been found to a demonstrations of its system to station at Redondo. However, sub- “test the waters” of the Pacific sequent events would suggest that Coast. A year later there was a brief it was the United Wireless Tele- Marconi stock promotion. Then graph Company who may have es- the following year, in 1906, a ma- tablished a temporary station

Volume 20, 2007 121 The Wireless Connection there as a publicity event. It was in other major shuffle in the local January, 1908, just two months world of Wireless. On April 6, later, that United began its inva- 1917, the U.S. Navy took over all sion of Avalon. radio station operations as a de- “Avalon, Jan. 13.—At the Hotel clared war with Germany was an- Grand View another wireless tele- nounced. In addition, all amateur graph station is being installed. radio stations were shut down. By Whether the town can support two April 17th the Avalon station was stations is a matter of much specu- back on the air under Navy person- lation.”15 nel control, handling both naval and commercial traffic. The gov- EPILOG ernment went on to acquire all the United States radio station assets Thus in 1908, United Wireless of the Marconi Company as there Company also established a wire- was great national concern about less connection between Catalina a foreign company (although En- and San Pedro with the Avalon glish) having such a widespread wireless installed in the Hotel monopolistic control over such an Grand View. Subsequently, mes- important piece of military tech- sage rates began to fall due to this nology as wireless communica- new head-to-head competition. tions. Thus the Avalon station be- This marked the eventual end of came a regular operating station the Pacific Wireless Telegraph of the U.S. Naval Communication Company as their management Service (call sign “NZL”). was now at arms length from their Sending messages by wireless less profitable day-to-day wireless telegraph had always been slow operations, and having found new and combersome. New technology financial ventures to promote. A advancement in voice communica- valiant effort was made to keep the tion would bring about another Pacific Wireless operation afloat ‘first’ for the Avalon to San Pedro for as long as possible but in 1910 wireless connection. Bell Labs had the Corporation was finally sus- developed a new wireless tele- pended, apparently for not paying phone system at the request of the their State taxes. Thus the sad de- local telephone company. On July mise of Pacific Wireless came 15, 1920, the first commercial ra- about almost without a wimper. dio-telephone call was made from No further mention in any news- Catalina to Mr. William Wrigley in papers was ever made of it. Chicago. The local telephone ra- United Wireless carried on with dio station at Avalon communi- local telegraph operations until cated to it’s mainland twin in Long they were eventually caught in Beach and from there was tied into their own legal net of stock mail the local and national phone ex- fraud and patent litigation with the changes. The local naval radio sta- Marconi Company. In the end the tion at Avalon was now hardly Marconi Telegraph Company of needed anymore so the Navy dis- America took over the entire mantled that equipment on July 1, United Wireless operations includ- 1921. ing ship and land stations in 1912. The advent of so many new Thus the Avalon station (PI) and amateur ‘hams’ after the war made Los Angeles station (PJ) were then “listening in” prime time entertain- under Marconi Company control. ment for the local radio crowd. World War 1 would cause an- This lack of privacy lead to the

122 AWA Review Teague & Knight eventual laying of a submarine telephone cable from Avalon to BIBLIOGRAPHY San Pedro in June of 1923. Addi- 1. De Forest, Lee. Father of Radio: The tionally, in the late 1920’s and Autobiography of Lee de Forest. 1930’s smaller broadcast radio sta- Chicago: Wilcox & Follett Co., tions operated from Avalon play- 1950. ing the sounds of the Big Bands 2. Fahie, J. J. A History of Wireless performing at the famous Avalon Telegraphy. New York: Dodd, Casino Ballroom. These were then Mead and Co., 1901. 3. Hager, Anna Marie. Winged Mail: connected to radio networks from Avalon to Bunker Hill. Los across the country for ‘live broad- Angeles: Dawson’s Book Shop, casts’. 1985. After World War 2 another 4. Overholt, Alma. The Catalina ‘first’ in wireless communication Story. Avalon: Catalina Island came to Catalina. In 1946 the Museum Society, Inc., 1962. world’s first commercial high fre- Ms. Overholt lived in the quency “micro-wave” radio-phone abandoned Pacific Wireless station system was installed on a tall peak up on the hill for many years. She overlooking Avalon and on a tall was the Publicist for the Santa Catalina Island Company and died building in downtown Los Ange- in 1970. les. This provided additional phone traffic capacity to the over- loaded telephone cable laid in NOTES 1923. Again, the unique topogra- 1. Morrison, Phillip and Susan, “Heinrich Hertz”, Scientific phy of these two locations, mostly American, Dec. 1957. over open water, helped provide 2. Marriott, Robert H., “Radio direct but private ‘line-of-sight’ Ancestors - An Autobiography”, A communication that was to be- condensation by , come so prevalent later in the IEEE Spectrum, June 1968. 1950’s as our radio and TV net- 3. “Marconigrams From Avalon,” Los works all switched over to cross- Angeles Express. March 20, 1902, country micro-wave transmission. p. 5. The report had one common Thus ends our story of two error of the day in calling the small towns where wireless history wireless a “Marconi system.” After all, the most publicity by far in was made on three different occa- wireless was being given to the sions, almost unknown to the rest work of Guglielmo Marconi. As a of the world, due mainly to their result, all wireless had to be the unique geographical locale and Marconi system! their human social connection. En- 4. “Snap shots,” Los Angeles Herald, tertainment, personal commu- April 12,1902, p. 6. nication and recreational necessi- 5. “First Wireless Message”, Los ties had indeed become the mother Angeles Herald, July 27, 1902. of invention - wireless that is! 6. Marriott, Robert H. “Radio Ancestors-An Autobiography,” A Condensation by Haraden Pratt. IEEE Spectrum. 5, June 1968, p. 55. 7. “Improving The Wireless”, Los Angeles Times, April 16, 1903, p. 13. This may have been the installation of improvements embodied in all three of the patents Swenson was ultimately awarded:

Volume 20, 2007 123 The Wireless Connection they were applied for in December, long and dedicated search for the facts 1902. a) Patent No. 745,463, and the story of this unique piece of “Wireless Telegraph Receiver,” wireless history. His position as a granted December 1, 1903. b) museum curator back then enabled Patent No. 751,818, “Electrically him to gain entry to places Produced Sound Intensifier,” inaccessible to most researchers granted February 9, 1904. and c) today. While his complete manuscript Patent No. 801,118, “Wireless was condensed for this publication, Telegraph Receiver,” granted the core essence of this wireless story October 3, 1905. has been left intact and Norwood is 8. “Catalina Island”, Los Angeles proud to offer this article as a Times, March 15, 1903. dedication to all those early unsung 9. “Ether Wings the Wonder”, Los wireless pioneers. Angeles Times, March 27, 1903. 10. “Wireless Going North”, Los Angeles Times, August 9,1903. Joe A. Knight was born and raised 11. “Wireless’ A Wonder”, Los Angeles in Southern California. He held his Times, August 3, 1904. first ham Novice license in 1958. Joe 12. “Wireless is Highest Yet”, Los settled down in Riverside after serving Angeles Times, November 26, overseas in the U.S. Army. He first 1904. As it turns out, Eben Smith, became interested in old radios about President of the Pacific Wireless 1977 and focused on written wireless Telegraph Company, was also history and it’s ephemera, compiling President of the Los Angeles Organ the Bibliography of Known Wireless Company. Books to 1922, commonly referred to 13. “First Wireless”, Los Angeles as “Knight’s List”. After moving to San Times, November 16, 1907. Pedro, California, where he was 14. “Avalon”, Los Angeles Times, involved with construction projects January 14, 1908. for the Port of Los Angeles, he became 15. “Avalon”, Los Angeles Times, acquainted with Norwood through February 26, 1908 the ham radio station operated at the San Pedro Maritime Museum. Over ABOUT THE AUTHORS time, he found additional materials and images that Norwood has Norwood Teague was born and incorporated into his final wireless raised in Goldsboro, North Carolina, manuscript. A couple of years ago Joe and moved to California after serving volunteered to assist Norwood in in WW II. He began a career as a finding a much needed publication professor at Cal State Los Angeles, outlet for his fine historical work. where he taught Industrial Design and When the opportunity arose, he gladly the History of Technology. Norwood agreed to edit it down to a size retired in 1977. He spent much of his appropriate for the AWA Review. He life, along with his wife Elouise, still hopes some day to find a venue amassing a complete and for the publication of the whole comprehensive collection of all the manuscript. Brownie Box Cameras. He has donated this collection to the California Museum of Photography in Riverside where you can go on line to view this fine and rare accomplishment. After retirement he became the Curator of Technical History at the Los Angeles Museum of Natural History, one of the nations largest. This is where he came across the first evidence of the Catalina Wireless Connection and so began his

124 AWA Review Jenkins

AWA Review The Early History of the Electric Loudspeaker

©2007 John D. Jenkins ABSTRACT It was a cool Saturday afternoon at San Not the result of a Francisco’s Golden Gate Park. The park was happy accident or busy as usual, perhaps even busier, for a Satur- flash of brilliance, day in December. There were the usual dog modern loudspeakers walkers and couples holding hands, and an oc- and headphones are casional clump of giggling children would pass the result of the contri- on roller skates, their laughter trailing off be- butions of literally hind them. hundreds of individu- It was only a week since San Francisco’s als, going all the way worlds fair, the Panama-Pacific International back to 1837. Much of Exposition, closed its gates for the last time. The their early develop- fair had opened in February, 1915 with the first ment can be traced to trans-continental telephone call, made by the telephone, the re- Alexander Graham Bell to the fairgrounds. An sult of a continual actual Ford assembly line was set up in the Pal- quest for louder, ace of Transportation and turned out one shiny higher quality sound black Model-T every 10 minutes for three hours reproduction. Along every afternoon. The entire area was illumi- the way, a quirky and nated by the latest developments in indirect often bizzare assort- lighting by General Electric. Thomas Edison, ment of inventions Henry Ford and other greats were seen fre- were offered up as po- quenting the grounds of the fair. On opening tential solutions. day, President Woodrow Wilson used a wire- More often than less apparatus from his office in Washington not, invention is the D.C. to start the Diesel-driven generator that result of a process, not supplied all of the direct current used in the Pal- a single event. While ace1. To the 18 million people who attended the history tends to record fair, anything seemed possible. Even on this the events, it is the day, more than a week later, excitement and process that truly tells wonder still hung in the air. the story, and it is the Perhaps that’s why no one took notice, at story that gives us in- first, when the “slender tone of a single violin”2 sight into the minds of began to weave its way through the narrow the men and women leaves of the Eucalyptus trees, sending them whose cumulative ef- dancing. It was a distant sound, yet strangely forts result in achieve- powerful, a sound that refused to yield in the ment - certainly that face of the pacific coast winds. is the case with the in- Nearly a mile away, Edwin Pridham made a vention of the loud- few final adjustments as he and Peter Jensen speaker and head- began the first major public demonstration of phone. a new kind of loudspeaker, or “loud-Talker” as it was called at the time. Loud-Talkers were well known in the telephone business, and while

Volume 20, 2007 125 Electric Loudspeaker they could sometimes be heard a address in San Diego.4 It was the few feet away, no one had imagined first time a president spoke over a one that could be heard for miles, public address system—more on at least not until this cool day in De- that later. cember. The scene is best described by THE BEGINNING the first person account of Edgar The earliest loudspeakers were Gleason, writing in San Francisco’s developed for use with telephones. Bulletin of December 11, 1915: In fact, some historians credit “Tetrazzini’s voice on a phono- Ernst Werner Siemens with the in- graph record resounding from end vention of the loudspeaker in 1877, to end of the vast park stadium! only a year after Alexander Graham The words of Attorney Thomas W. Bell invented his telephone, and Hickey reading Lincoln’s nearly thirty years before the in- Gettysburg Address reverberating vention of radio telephony. Others through the air like the roar of a credit Siemens for the loudspeaker giant! A piano solo resembling the in 1874, but crediting Siemens with chimes of Westminster Abbey, the invention isn’t accurate or fair. played by the Colossus of At the time of his first patent, all Rhodes!”3 that he really had in mind was a The demonstration was a huge way to “obtain the mechanical success. Jensen and Pridham be- movement of an electrical coil from came local celebrities as they con- electrical currents transmitted tinued to demonstrate and improve through it.” It wasn’t until after their wonderful new invention. A Bell’s 1877 telephone patent was major milestone was reached on filed that Siemens re-filed and September 19, 1919, when their claimed his invention as a tele- system was used by President Wil- phone receiver.5 son to deliver his League of Nations So, who really invented the

Fig. 1. 1907 photo showing headphone, and 1908 illustration of a “loudtalker.6.7

126 AWA Review Jenkins loudspeaker? What Page was observing was a Even among scholars there re- property later called “magneto- mains some disagreement. If you striction”, in which an iron bar look in the encyclopedia, for ex- changes its length somewhat when ample, you’ll likely find: Jensen & magnetized. Magnetostriction is Pridham, 1911, or maybe Rice & what causes transformers to hum. Kellogg, 1925. Look for the inven- Page also noticed that the pitch of tion of headphones and the most the tone depended on the size of the common answer will be Nathaniel horseshoe magnet he used. Later, Baldwin in 1910. But if the loud- this would prove to be a very im- speaker was invented in 1911, and portant discovery and a major in- the headphone in 1910, how does fluence on Alexander Graham Bell one explain the pre-1910 photos in his work on the telephone. (fig. 1) that clearly show head- phones on the desk, and an illus- THE SIEMENS PATENT tration of a loudspeaker? Was This brings us back to the Ger- there time-travel involved? Are the man inventor E.W. Siemens, who encyclopedias wrong? obtained US patent number The answer reveals a common 149,797 in 1874 for an “improved limitation with the popular histori- magneto-electric apparatus.”9 This cal record of invention: a lot of im- was the first “linear motor” – A portant details are left out. Ency- small coil is suspended between the clopedias tend to record inventions poles of a permanent magnet. as events, but more often than not, When current is passed through inventions are born of a complex the coil, the resulting magnetic field process, not a single happy accident interacts with that of the perma- or flash of brilliance. While nent magnet, and the coil is forced Baldwin certainly did invent a new to move either toward or away headphone technology in 1910, he from the magnet. But as mentioned didn’t invent the concept of a head- earlier, Siemens didn’t envision his phone, or even the first telephone invention reproducing sound, at receiver attached to a headband. least not until Bell was granted his Jensen, Pridham, Rice and Kellogg patent for the telephone. all made unique and important contributions to loudspeakers, but THE SINGING CONDENSER they didn’t invent them. Instead, In the early 1870s Varley, Pol- headphones and loudspeakers, as lard and Garnier invented the we know them today, were the re- “Singing Condenser” (fig. 2), a de- sult of the contributions of literally vice for reproducing the human hundreds of individuals, going all voice using electricity. Made from the way back to 1837. a primitive point-contact microphone, an induction coil and GALVANIC MUSIC a condenser, it roused great curi- The first record of sound repro- osity before the days of the tele- duction from electricity was re- phone because the sounds came ported by the American inventor from a flat condenser made of in 1837. In sheets of tin and paper. Augusto his paper titled, “The production of Righi, The famous Italian physicist galvanic music,”8 Page describes and future mentor to the young hearing a ringing sound in a horse- Guglielmo Marconi, explained that shoe magnet when an electromag- every charge and discharge of the net was switched on or off nearby. condenser causes a change in the

Volume 20, 2007 127 Electric Loudspeaker vice employed several steel reeds, each receiving reed (fig. 3) tuned to its counterpart on the transmit- ting end. The vibrating reed func- tioned as an “interrupter,” making and breaking a circuit, and send- ing pulses down the wire. During an experiment, Watson acciden- tally over-tightened the screws at- taching the reed, and when he plucked the reed, Bell shouted and rushed in from the other room. Instead of the expected noisy sound of the interrupter, he had heard the pure note of the plucked Fig. 2. The “Singing Condenser”10 steel reed. Instead of making and breaking the circuit, the vibrating density of the air around the con- steel created a sound-shaped elec- denser, producing sound waves. tric current in the wire which vi- Today, similar technology is used brated the reed at the opposite end, in the design of high quality elec- reproducing the original sound. trostatic loudspeakers. When Bell realized he was see-

11 ing the effects of a sound shaped BELL’S TELEPHONE current, he knew it was the key Throughout his life, Alexander piece he needed to make a tele- Graham Bell had been interested phone. He immediately dropped in the education of deaf people, any thought of improvements in and had made several attempts at the telegraph and set to work in developing mechanical models ca- earnest. pable of producing human speech. Bell’s first attempt at a tele- He was inspired by the work of phone was a simple device—essen- German physicist Hermann Von tially a vibrating reed with a tightly Helmholtz. In his thesis “On The stretched parchment diaphragm Sensations of Tone,” Helmholtz attached. It was called a “Gallows” declared that vowel sounds could telephone because of its shape (fig. be produced by a combination of 4). electrical tuning forks and resona- Vibrations in the diaphragm of tors. the transmitter were transferred to Unfortunately, Bell’s efforts to the reed, which caused corre- create electric speech were always sponding electrical vibrations to be met with the same roadblock: A way to produce a “sound-shaped” electrical wave. Eventually, Bell set the speech projects aside in order to work on improvements in the telegraph. In June of 1875, Bell and Watson were attempting to im- prove their “Multiple Telegraph System,” where multiple telegraph messages can be sent simulta- neously over a single wire. The de- Fig. 3. Bell’s vibrating reed receiver

128 AWA Review Jenkins

nology used in these receivers - a permanent magnet, thin metal dia- phragm and a small electromagnet called a voice coil - is the same as that used in radio headphones and remained virtually unchanged throughout the 1920s.

THE FIRST MOVING-COIL LOUDSPEAKER After Bell’s patent was granted, Charles Cuttris and Jerome Redding filed their patent applica-

12 tion for a new telephone utilizing Fig. 4. Bell’s “Gallows” telephone a moving coil diaphragm.14 A short sent down the wire to an identical time later Ernst Siemens applied apparatus operating as the re- for German patent No. 2355, filed ceiver. On the receiving end, the Dec. 14, 1877, for a nonmagnetic vibrating electromagnetic field parchment diaphragm as the caused the receiver’s reed to move, sound radiator of a moving-coil and these vibrations where trans- transducer.15 Unlike other tele- ferred to the attached diaphragm, phones of the time which used producing sound waves. Although magnetic diaphragms, Siemens the gallows didn’t work as a tele- used a non-magnetic diaphragm phone transmitter, Bell’s receiver did contain a diaphragm, an arma- ture, and a coil - all of the elements of a cone loudspeaker, an inven- tion that wouldn’t appear for an- other fifty years!

THE BUTTERSTAMP RE- CEIVER The next major step came in Fig. 6. Cuttris and Redding moving- 1877 when the “Butterstamp” (fig. coil diaphragm receiver - 188116 5) telephone handset was intro- duced—so called because of its re- with a small coil attached. The coil semblance to the wooden stamps was held in the field of a perma- used by nineteenth century house- nent magnet. The diaphragm could wives to produce small round pats take the form of parchment “or of butter from the larger blocks other material of trumpet form.”17 sold in bulk form. The basic tech- Because his design built on his 1874 invention and anticipated several of the key elements of a modern loudspeaker, Siemens is often credited as the inventor of the first moving-coil loudspeaker.

LOUD-SPEAKING TELE- Fig. 5. Butterstamp telephone re- PHONES ceiver13 The earliest invention that in- Volume 20, 2007 129 Electric Loudspeaker cludes all of the aspects of a mod- Graham was an early pioneer in ern loudspeaker (magnet, coil, dia- loudspeakers, demonstrating his phragm, and external horn) is de- first model in 1887 England. By scribed by J.D. Drawbaugh in his 1893 Graham was marketing loud- 1883 patent titled: “Speaking speakers for use in warehouses and Trumpet for Receiving-Tele- factories, and by 1896 for use phones.”18 See fig. 7. Drawbaugh’s aboard mercantile vessels.23 The speaking trumpet included a series company would later become of complex baffles designed to im- known for one of the premier ra- prove the fidelity of the horn. The baffles were later shown to be un- necessary provided the shape of the horn simply followed an expo- nential curve.20 The Titanic, launched in May, 1911, included a sophisticated new

dio speakers, the Amplion. With its striking “morning-glory” style, Fig. 7. Drawbaugh’s Speaking Trum- Amplion loudspeakers are a favor- pet for receiving telephones - 188319 ite among collectors. “Loud-Speaking Telephone” built AMPLIFICATION by Alfred Graham and Co. (Fig. 8)21 In 1898 Sir , Pro- fessor of Physics at University Col- lege, Liverpool, developed a tele- phone receiver that he called the “Magnifying Telephone” or “Bel- lowing Telephone.”24 As much an

Fig. 8. Alfred Graham Loud-Speaking Telephone 191122 Fig. 9. Brown microphonic relays25 130 AWA Review Jenkins

Batchelor head receiver 188329

Carey telephone support 187926

Lane telephone support 188330

Hubbard telephone support 188027

31 Chambers earplug receiver 188228 Gilliand operator set 1881 Fig. 11. Early headphone designs Volume 20, 2007 131 Electric Loudspeaker amplifier as a loudspeaker, the THE FIRST HEADPHONES Lodge device featured a sensitive As telephones became more relay made from a coil and a thin popular, practical considerations metal reed. Signals fed into the in- came into play. It was hard enough put caused the reed to vibrate, to hear the faint whispers coming which was mechanically coupled to through the line, and in a noisy a carbon microphone. The micro- room it was impossible (see figure phone then modulated voltage 10)34. But noise wasn’t the only from a large battery to produce an problem. The more sophisticated amplified signal. Multiple relays telephone networks required could be connected together in a rooms with many operators han- cascading fashion to provide addi- dling calls and the operators tional amplification. Lodge also needed their hands free to manipu- described how the final stage could late the plugs and keys used in be coupled to a “sound board” in switching. order to “magnify a speaker’s voice A host of creative solutions was by distributing it to various parts offered (fig. 11), including the dra- of a large hall.” 32 conian contraption developed by Over the following decade S.C. Carey in 1879, but they were many more designs appeared for heavy, uncomfortable and un- mechanical amplifiers, the most wieldy. popular being the Brown relay, Clearly, a smaller, lightweight developed by S.G. Brown in 1909 head receiver was needed. The lim- (figure 9.)33 iting factor seemed to be the large permanent magnet that ran the length of the receiver. In 1879, Frederick Allen Gower developed a small horseshoe magnet bent into a semi-circle, with the ends bent at right angles to the plane of the magnet. He installed the mag- net in a new telephone that looked very much like a pocket watch (fig 12.)36 Gower’s telephone could be used as a microphone or as a re- ceiver, but despite its small size it didn’t occur to anyone to attach the receiver directly to a headband. Instead the device was used with a small speaking tube. Elisha Grey holds the earliest patent (1878) for a headphone (fig. 13) which was hand-held by a handle placed under the chin.37 A Fig. 10. “It should be understood that year later Andrew Hubbard in- the telephone is difficult to hear. When vented a similar apparatus, but used in noisy places such as public with an attachment that connected areas there is a distinct advantage to the two receivers via a strap across using two telephones. One places the top of the head, as in modern them at each ear as shown, taking headphones.38 care to hold them tight against the ear in order to block ambient sounds as The first headphone to achieve much as possible”35 commercial success was designed

132 AWA Review Jenkins

Fig. 12. Gower’s telephone - 187939 in 1884 by W.L. Richards at Bell Telephone (fig 14.) Technically, the Richards “Head receiver” was a step backward from Grey and Gower, using a design similar to the Butterstamp receiver except that the long permanent magnet was located outside the case, at- tached to the headband. 40 Fig. 13. Grey’s headphone - 187842 Ten years later the headphone was improved further with the in- mitter and telegraph registers on troduction of the bipolar head- the receiving end. The first re- phone. The Bell No. 121 receiver corded use of a headphone in ra- used a folded magnet and was very dio was by Guglielmo Marconi on similar to the Gower telephone of the morning of December 12, 1901, 1879.41 It’s not clear why the Gower at Signal Hill, a rocky bluff over- design was overlooked for so many looking the town of St. John’s years. The Bell 121 was the fore- Newfoundland. runner of radio and telephone Beginning with his earliest ex- headphones that would be used periments, Marconi’s strategy was throughout the 1920s. to increase the range of his wire- less telegraph apparatus, step by SPARKS ACROSS THE step. He began the process as a SEA43,44 teenager in his father’s garden and Until the beginning of the 20th it ultimately culminated in his century, all of the efforts toward greatest achievement, bridging the improving sound reproduction us- Atlantic Ocean by wireless in 1901 ing electricity were taking place in at the age of 27. the telephone field. Radio was still Marconi set up his Signal Hill in its infancy and the devices in use receiving station with instruments were borrowed from the tele- much like those he had used in his graph—Morse keys at the trans- earlier experiments: A simple re- ceiver using a mercury oxide detec- Volume 20, 2007 133 Electric Loudspeaker

Bell #121 receiver - 1894 Richard’s head receiver - 1884 Fig. 14. Bipolar headphones45 tor, and a telegraph register that atlantic radio communication, but printed the dots and dashes of it was also the first time a head- onto a thin paper tape. phone was used to receive radio After several days of attempts to signals outside the laboratory. receive the signals being sent by his associate at Poldhu, England, NATHANIEL BALDWIN Marconi replaced the telegraph Headphones changed little in register with a telephone receiver the years following the introduc- (fig.15.) He believed that the more tion of the Bell model 121 receiver, sensitive receiver would allow him until 1910 when Stanford Univer- to better discern the faint signals sity student Nathaniel Baldwin emanating from the far side of the improved upon the old telephone ocean. receiver by using an armature that At 12:30 P.M. on December 12, attached directly to the center of a 1901, history was made when three mica diaphragm. faint clicks—Morse code for the The Baldwin headphones were letter “S” - were heard in the tele- more sensitive than other tele- phone receiver. The date went phone receivers, but were subject down in history as the first trans- to deterioration in performance

Fig 15. Collier-Matt earpiece - 190146

134 AWA Review Jenkins due to rust, cracked diaphragms, and other mechanical problems. Still, Baldwin phones were consid- ered the ultimate because of their sensitivity, and were prized by ra- dio amateurs well into the 1920s. Baldwin also designed radio speak- ers, including the deluxe Master- Baldwin Throatype Clarophone, said to be shaped like opera star ’s throat!48

Fig. 16. Baldwin Throatype Fig. 17. Peter S. Jensen Clarophone loudspeaker47 Their plans took a turn how- ever, when their curiosity was piqued by a telegraph device that recorded the incoming dots and dashes of Morse code onto a thin tape of photographic film. At the heart of the device was a very thin wire that passed between the two poles of a magnet. When an elec- tric current flowed through the wire, the resulting magnetic field would interact with that of the MAGNAVOX AND THE magnet, causing the wire to move 49 rapidly back and forth, mimicking TELEMEGAPHONE the fluctuations of the incoming In 1910, Peter L. Jensen and current. Jensen and Pridham were Edwin S. Pridham formed the impressed by the speed of the de- Commercial Wireless and Devel- vice. They wondered, “If it was at- opment Company with plans, in tached to a telephone, and a Jensen’s words: “To make re- heavier wire was used, perhaps it searches in radio, and to take out could be used as a telephone re- patents, if we invented anything.” ceiver?” They initially focused on the They quickly set to work, and Poulson Arc and planned to inves- before long had built a prototype. tigate the use of the new De Forest In order to make their invention three-element tube.

Volume 20, 2007 135 Electric Loudspeaker cause, according to Jensen, “the name sounded good to us.” They were awarded US Patents in 1914 and 1923.51,52 Their new receiver was a vast improvement over anything cur- rently available. Unfortunately, it was also vastly larger; too heavy and unwieldy for use as a practical telephone receiver. Wireless com- munication was still in its infancy Fig. 18. Sports announcer using and it would be several more years megaphone to announce game before would scores50 create demand for loudspeakers. They had a great product but no produce sound, they attached the market, and were unable to locate wire to a thin diaphragm using an a buyer. ordinary kitchen match. When it One day in early 1915, they were was attached to the telephone, visited by Jensen’s blacksmith both engineers were astounded by uncle, Ray Galbreath. They showed the strength and clarity of the him their invention and politely sound coming from their device. asked if he could think of any use Unable to accept that such an ob- for it. He thought for a moment & vious approach had not already said: “Foghorn Murphy.” “Fog- been tried, they poured through horn Murphy?” they asked. “Yes, their library searching for mention Foghorn Murphy – the fellow who of a moving string or coil tele- announces the baseball games phone, but could find neither. En- through his cardboard mega- couraged, they enthusiastically phone. Maybe if you put a big horn filed a patent application on it like Foghorn Murphy’s, you What they didn’t know was that Ernst Siemens and Oliver Lodge both held patents on moving coil telephones dating from 1877 and 1889, respectively. So, while Jensen and Pridham’s device did introduce new ideas, it wasn’t enough to convince the patent of- fice, and much to their dismay, their patent was denied. But they did know they were onto something. Neither of the previous patents had ever left the laboratory, so they improved their design further by replacing the heavy wire with a small coil, and the permanent magnet with a strong electromagnet. These changes boosted the performance of their receiver greatly. They named their new approach the Fig. 19. Jensen and Pridham with their “Electro-Dynamic Principle” be- “Telemegaphone” loudspeaker53

136 AWA Review Jenkins could use it to announce the scores the incredible power that came at the baseball game.” from the unique design of their re- They had an Edison phono- ceiver – truly an “Electro-Dy- graph with a big, 22-inch goose- namic” wonder. neck horn, and it didn’t take long for them to attach it to their tele- MAGNAVOX IS BORN54 phone receiver. They hooked up a At first they named their new microphone and powered the invention the “Telemegaphone,” receiver’s electromagnet with a which sounded better and rolled large 12 volt storage battery (fig. more easily off the tongue than 19). When the power was applied, “dynamic loud-speaking tele- there was such a loud and horrible phone.” But the name didn’t sit squeal that everyone had to run well with them – the image of a and cover their ears. Being famil- distant megaphone didn’t really iar with the phenomenon of feed- convey what they were looking for. back from their years of telephone Eventually they settled on experience, Pridham and Jensen “Magnavox,” for “great tried to increase the distance be- voice.” Perfect! tween the microphone and horn, This brings us back to that but were unable to find anywhere memorable day at Golden Gate in the house where they could be Park, December 10, 1915, when sufficiently separated. Finally, Pridham and Jensen demon- Pridham climbed up on the roof strated their new invention. Two and attached the horn to the chim- years later Jensen formed the ney, running the wires down to the Magnavox Company. microphone, which remained in On September 9, 1919, the house below. When he spoke Magnavox gained national atten- into the microphone, a voice tion by installing and operating the boomed from above with the thun- first public address system – in this der of Zeus himself. Jensen case for a speech by President Wil- couldn’t contain his excitement son at Balboa park in San Diego and began to run across the wide (fig. 20). Pridham had come to San field surrounding the house, to see Diego to supervise the installation how far the sound would carry. At and Jensen was in the glass plat- one-half mile he was amazed to form with Wilson. Rather than use hear the voice could be heard with a visible microphone, Wilson stood great clarity. Even at a mile, he was on a circle drawn near the front of able to make out the words of the stage and spoke into two large Edwin Pridham as they caromed horns suspended over his head across the countryside. that directed his voice to the mi- The idea wasn’t new, loud- crophone. The loudspeakers were speaking telephones had been hidden behind flags and bunting around in the telephone business around the platform. Although a for some time, but they were weak tube in the amplifier shorted and and couldn’t be heard without had to be replaced just before the some effort. Even in the radio field, speech, the system worked well still in its infancy, there were early during the hour-long speech. The amateur attempts to create loud- audience, who could clearly hear speakers, but these were usually Wilson’s words, applauded and simply a standard telephone re- shouted their approval. Wilson’s ceiver with a small voice could be heard one mile from horn attached. What was new was the stadium.56

Volume 20, 2007 137 Electric Loudspeaker came commonplace.

Fig. 22. Phonograph horn adapter58

HORN LOUDSPEAKERS In 1920 Westinghouse began broadcasting music from KDKA, the first commercial broadcast sta- tion. Over the next several years, Fig. 20. President Wilson at Balboa sales of radios into the home ex- Park, San Diego.55 ploded, and with them, the sales of Until Wilson’s use of this new horn loudspeakers. Most of the technology, the business for pub- loudspeakers were external horns lic address systems had been slow. with a simple telephone receiver Public speaking was considered a (by now called a headphone) hid- fine art by proud orators, certainly den in the base as a driver. The nothing to be messed with through Baldwin headphone became one of electric augmentation. Wilson’s the most popular for use as a loud- speech changed that, and the fu- speaker driver because of its higher ture of Magnavox was secure as the fidelity reproduction of sound. In use of public address systems be- addition to Magnavox, several companies including Brown, West- ern Electric, and Graham (Amplion) developed horn loud- speakers with improved perfor- mance drivers.59 At the same time there was an explosion in the market for add-on products that took advantage of the large installed base of head- phones and phonograph horns. These adapters came in a wide va- riety of creative schemes for at- taching a radio headphone to a phonograph horn in order to make Fig. 21. Clipping from March, 1925, a loudspeaker. One example is Radio Broadcasting showing Winston Churchill demonstrates broad accep- shown in figure 22. tance of public address systems.57

138 AWA Review Jenkins

CONE LOUDSPEAKERS The first coil driven direct radia- As mentioned earlier, we owe tor loudspeaker sold in the US was much to the telephone for early the Farrand Phonetron, intro- developments that led to the horn duced in 192162 (fig. 24). The loudspeaker. Similarly, many of Phonetron consisted of a moving the earliest innovations in cone coil armature attached to the cen- sound reproducers came to us by ter of a paper cone and doubled as way of the phonograph. a loudspeaker and a microphone. In 1901, John Stroh of London filed a patent for a new diaphragm invention.60 Although the dia- phragm was intended for use with a phonograph, the Stroh patent foresaw several of the key features that would later become hallmarks of the cone loudspeaker: A cone shaped diaphragm, vibrations in- troduced at the center, and a dia- phragm attached via a corrugated edge to allow for back and forth movement of the cone. In 1908, Louis Lumière developed a new phonograph diaphragm featuring Fig. 24. The first coil-driven direct ra- folded pleats radiating from the diator loudspeaker - 192164 center instead of a cone.61 The ap- proach worked sufficiently well for A number of armature style its use in a commercial radio loud- cone loudspeakers followed, in- speaker (the Sterling Primax) in- cluding the Western Electric troduced in 1924 (fig. 23). Model 540AW “loud speaking tele- phone,” the popular Pathe series, and many others from small com- panies who produced loudspeak- ers with no research—their work based on earlier patents which had by now expired.

THE ELECTRO-DYNAMIC CONE LOUDSPEAKER A huge step forward came in 1925 when Chester W. Rice and Edward W. Kellogg of G.E. pub- lished their landmark paper, ”Notes on the development of a new type of loudspeaker” in the Transactions of the AIEE.65 The impact of their work was felt far beyond the loudspeaker, directly influencing the design of the radio tubes and receivers that would fol- low. They began in the early 1920s, Fig. 23. Sterling Primax loudspeaker63 when General Electric engineer

Volume 20, 2007 139 Electric Loudspeaker 1920. A key innovation of the Rice- Kellogg design was the use of an amplifier. Radios of the early 1920s were under-powered and most loudspeakers relied on resonance (using a horn) to obtain sufficient loudness. Unfortunately, reso- nance occurs over a fairly narrow frequency range, so reliable repro- duction of sound is impossible. Rice and Kellogg realized if they could provide sufficient power to the loudspeaker, they could elimi- nate the problem of resonance and at the same time allow them to fo- Fig. 25. Rice and Kellogg dynamic cus their design efforts on reliable loudspeaker - 192566 sound reproduction.68 Their invention was introduced C.W. Rice came up with the idea to the commercial market as the of creating the best “hornless loud- RCA Radiola 104 (fig. 26).70 Unlike speaker.” He and colleague E.W. the other loudspeakers of the time, Kellogg pitched the idea to G.E. the new dynamic loudspeaker re- management, and were given liably reproduced sounds across much latitude in their investiga- the entire audio spectrum, and tion. They conducted a thorough within a few years had driven vir- scientific examination of a wide tually all other types out of the range of materials and technolo- gies for loudspeakers, testing all manner of sound reproduction from electric arcs, to a little known technique not far removed from rubbing two stones together. 67 Their research led them to con- clude that the best sound quality came from an electrostatic loud- speaker. At the time, however, an electrostatic loudspeaker was im- practical to manufacture, so they eventually settled on a more tradi- tional electromagnetic design. Their invention used a moving coil similar to that of the Lodge and Siemens approach, but attached directly to the diaphragm of a pa- per cone (fig. 25). The outer edge of the cone was loosely held to the supporting frame by thin rubber. Instead of a permanent magnet, the loudspeaker used a large elec- tromagnet, similar in design to the Magnavox horn loudspeaker of Fig. 26. RCA Radiola 104 loudspeaker 192569

140 AWA Review Jenkins market.71 It remains the principle literature, during the twenty year technology used in modern loud- period beginning in 1928.72 speakers. Clearly, loudspeakers have come a long way since cardboard THE FUTURE megaphones were used to an- Of course, loudspeaker devel- nounce the scores at baseball opment didn’t end with the Rice- games. Foghorn Murphy would be Kellogg design in 1925. Many ad- impressed! ditional improvements were to come. Fidelity over a wide fre- REFERENCES quency range was improved sig- nificantly later the same year, 1. “A Sense of Wonder : The 1915 San when Minton and Ringel proposed Francisco World’s Fair” http:// www.moah.org/exhibits/archives/ a dual loudspeaker system utiliz- 1915/ , June 20, 2005 ing a cross-over filter network. The 2. Gleason, Edgar, “Magnavox tested network divided the signals into at the park” The Bulletin, two parts, the high frequencies December 11, 1915 being directed to a “tweeter” and 3. ibid the low frequencies a “woofer,” 4. Jensen, Peter “The Great Voice,” each being designed for optimum The Havilah Press (Richardsen, fidelity within its more limited fre- Texas), 1975 quency range. 5. Schoenherr, Steve “Loudspeaker The invention of the alnico al- History”, http:// home.sandiego.edu/~ses/ loy (iron, nickel, and aluminum) 6. Ruhmer, Ernst “Wireless permanent magnet in 1930 pro- Telegraphy”, Crosby, Lockwood & duced loudspeakers with power Son (London), 1908 similar to that of the electro-dy- 7. Modern Electrics, January, 1910 namic speakers, but in a much 8. Page, C.G. “The Production of smaller package due to the use of Galvanic Music,” American a small permanent magnet in place Journal of Science, 32, 1837, p 396- of the large electromagnets used in 7. the older design. Modern rare 9. Siemens, E.W. U.S. Patent No. earth magnets have taken this even 149,747 filed January 20, 1874. 10. John Jenkins collection, further, facilitating the design of www.sparkmuseum.com very small but powerful low-volt- 11. Fagen, M.D. “A History of age loudspeakers. Engineering and Science in the Bell Finally, the electrostatic loud- System,” Bell Telephone speaker, which traces its history all Laboratories, 1975. the way back to the 1870 “singing 12. Jenkins collection condenser” (see figure 2), has been 13. ibid perfected by Janzen, KLH, Sound 14. Cuttris, Charles, and Redding, Designs, and others. Jerome, U.S. Patent No. 242,816 As the ultimate testimony to the filed November 28, 1877. 15. Siemens, E.W. German Patent No, extent to which development oc- 2355 filed December 14, 1877. curred in the years following Rice 16. Cuttris & Jerome patent 242,816 and Kellogg, the cumulative index 17. Siemens patent 2355 of the Journal of the Acoustical 18. Drawbaugh, Daniel, U.S. Patent Society of America contains more No. 272,866 filed December 5, than 1,400 entries referring to 1882. publications in the general field of 19. ibid transducers, appearing either in its 20. Webster, A.G. “Acoustical own Journal or the contemporary Impedance and the theory of horns

Volume 20, 2007 141 Electric Loudspeaker

and of the phonograph,” 49. Jensen, Great Voice Proceeding of the National 50. Radio News, February, 1924 Academy of , No. 7, July 15, 51. Pridham, E.S. and Jensen, P.L. 1919. U.S. Patent No. 1,088,283 filed 21. Hood, A.G. “The White Star Liners March 19, 1912. Olympic and Titanic,” The 52. Pridham, E.S. and Jensen, P.L., Shipbuilder, 6, 1911, p. 117. U.S. Patent No. 1,448,279 filed 22. ibid April 28, 1920. 23. “The Amplion Radio Annual, 53.Courtesy of Patricia Ann Jensen 1927”, Graham Amplion Ltd., Schindler (Daughter of Peter London Jensen) 24. Lodge, O.J., British Patent No. 54. Jensen, Great Voice 9712, filed April 27, 1898. 55. Courtesy San Diego Historical 25. Jenkins collection Society 26. Carey, S. C., U.S. Patent No. 56.“Woodrow Wilson in San Diego”, 224,073 filed December 23,1879. http://history.sandiego.edu/GEN/ 27. Hubbard, A.C., U.S. Patent No. recording/wilson.html, June 21, 230,779 filed August 3, 1880. 2005 28. Chambers, J.C., U.S. Patent No. 57. Radio Broadcast, March, 1925, 258,989 filed August 25, 1881 p.890 29. Batchelor, J.F. U.S. Patent No. 58. Radio, June, 1923 281,000 filed February 23, 1882. 59. Balbi, C.M.R. “Loud Speakers – 30. Lane, T.W., U.S. Patent No. Their construction, performance 287,301 filed June 14, 1883 and maintenance” Sir Isaac Pitman 31. Fagen, Bell System History & Sons, Ltd. 1926. 32. Lodge patent 9712 60. Stroh, John M.A., British Patent 33. Brown, S.G. British Patent No. No. 3393 filed February 16, 1901 4162 filed February 19, 1909 61. Lumiere, Louis, U.S. Patent No. 34. Niaudet, Alfred “Téléphones et 986,477 filed June 30, 1909. Phonographes” ,1884 62. Hunt, Frederick V. 35. ibid “Electroacoustics”, Harvard 36. Wormell, R. “Electricity in the University Press, 1954 Service of Man,” Cassell & Co. Ltd. 63. Jenkins collection (London), 1890. 64. Radio News, July, 1921 37. Gray, Elisha, U.S. Patent No. 65. Rice, Chester W. and Edward W. 204,027 filed January 25, 1878. Kellogg, “Notes on the 38. Hubbard, A.C., U.S. Patent No. Development of a New Type of 220,839 filed July 8, 1879. Hornless Loudspeaker,” 39. Wormell, Electricity in the Service Transactions of the American of Man Institute of Electrical Engineers, 40. Fagen, Bell System History 44, 1925, p. 461-475. 41. ibid 66. Briggs, G.A., “Loudspeakers”, 42. Gray patent 204,027 Worfdale Wireless Works Ltd, 5th 43. Baker, W.J. “A History of the ed - 1958. Photo credited to Marconi Company”, Methuen & American Institute of Electrical Co. LTD, 1970. Engineers. 44. Bussy, Gordon “Marconi’s Atlantic 67. Rice & Kellogg, Hornless Leap”, Marconi Communications Loudspeaker (London), 2000. 68. Briggs, Loudspeakers 45. Fagen, Bell System History 69. Jenkins collection 46. Courtesy Marconicalling.com 70. Briggs, Loudspeakers 47. Baldwin Throatype Clarophone 71. ibid photo courtesy of Williiam (Buck) 72. ibid Nielsen. 73. Alexander, E. “Loud speakers and 48. “Utah History to Go,” http:// how they work,” Radio News, April, historytogo.utah.gov/ 1924. hmruin.html, April 17, 2005. 74. Bain, A., U.S. Patent No. 32854

142 AWA Review Jenkins

dated July 23, 1861. hooked on antique radios and other 75. Baldwin, Nathaniel, U.S. Patent interesting objects related to the No. 957,403 filed July 1, 1909. history of electricity. During the next 76. Brown, S.G. British Patent No. 40 years, he collected thousands of 4162 filed February 19, 1909 objects, books, papers, and ephemera 77. Ganfield, Fred “What is a good that depict the scientific exploration loud speaker?” Radio News, of electricity from the 16th century August, 1928. forward into the early years of radio. 78. Hyde, Stanley “Loud Acoustical Among the most exciting components Reproduction of Wireless Signals,” of his collection are original books and Modern Electrics, June, 1912. scientific papers that chronicle crucial 79. Lodge, O.J. “Improvements in milestones in the development of Magnetic Space Telegraphy”, radio and electricity, by authors such Journal of the Institution of as Gilbert, Galileo, Benjamin Franklin, Electrical Engineers, 27, 1898, p Faraday, Volta, Hertz, Marconi, and 834 – 851. many others. 80. Siemens, E.W. British Patent No, Much of John’s collection is 4685 filed December 10, 1877. on display at the American Museum of Radio and Electricity, located in This article was peer-reviewed. Bellingham, Washington.

ABOUT THE AUTHOR

John Jenkins was born in Bellingham, Washington, and lived in the area until 1974, when he began his professional career as an engineer with Hewlett Packard. John worked for Hewlett Packard Corporation for 12 years, where he held various technical and marketing positions at the company. In 1986 he joined Microsoft Corporation and managed several organizations, most recently as General Manager of Worldwide OEM Sales and Marketing. John retired from Microsoft in February 2001 and now serves on several boards including the American Museum of Radio and Electricity, where he is board chairman. He is a member of several professional associations including the History of Science Society and the Society for the History of Technology. He is married with two daughters. John’s interest in electricity started when he was very young, but at 13 his focus switched to radio when he discovered his great-uncle’s long- abandoned ICS radio correspondence course gathering dust in the basement. About that time, John discovered a 1927 radio stored in his grandmother’s home. He got it to work, and from that moment was

Volume 20, 2007 143 Electric Loudspeaker APPENDIX

Major milestones in the development of loudspeakers and headphones.

Year Development Inventor 1837 Galvanic Music C. Grafton Page 1861 Telegraph Headphone Alexander Bain 1870’s Singing Condenser Varley, Pollard & Garnier 1874 Linear electromagnetic motor Ernst Siemens 1875 Vibrating reed receiver Alexander Graham Bell 1875 Gallows Telephone Alexander Graham Bell 1877 Butterstamp receiver Alexander Graham Bell 1877 Moving coil loud-speaking telephone Ernst Siemens, Cuttris/Redding 1878 Head receiver (headphones) 1879 Head receiver w/ headstrap Andrew Hubbard 1879 Bipolar telephone receiver Frederick Gower 1882 Loud-speaking telephone w/ Horn 1898 Bellowing Telephone Oliver Lodge 1901 Cone diaphragm John Stroh 1901 First use of a headphone in radio Guglielmo Marconi 1910 Balanced-armature headphone Nathaniel Baldwin 1912 Electrodynamic telephone receiver Jensen & Pridham 1920 Public address loudspeaker Jensen & Pridham (Magnavox) 1921 Armature Cone Loudspeaker Farrand (The “Phonetron”) 1925 Dynamic cone loudspeaker Rice & Kellogg (RCA Radiola 104)

144 AWA Review Kleinman et al.

AWA Review Emil J. Simon: A Busy Life Indeed

©2007 Russ Kleinman, Karen ABSTRACT Blisard, A.J. Link and Warren Berbit Who was “Emil J. Simon” Emil J. Simon? is not a name most Much of the an- people recognize, al- swer to this though he was a very question can be controversial and found in the de- colorful character. tailed autobiog- The story of his early raphy which is life was routinely currently under smeared across the the care of the headlines of newspa- Bancroft Rare pers, including his Manuscript Li- short stint in prison brary at the Uni- in Europe. After at- versity of Cali- tending Columbia fornia at Berke- University, Simon ley.1 The simple worked for Lee De answer is that Forest and then Emil J. Simon started his own com- was an im- pany. He success- mensely impor- Fig.1. Emil J. Simon photo- fully organized a huge tant government graph from the Inter-City Radio network of contrac- contractor for brochure, about 1920. tors to construct ra- the United States, supplying millions of dollars dio transmitters and worth of wireless equipment during and after receivers during World War I (Figs. 1 &2). It is only after read- World War I, and ing through the thousands of pages of autobio- then lost a fortune graphical material, much of it in Simon’s own trying to fight big handwriting, that his life seems more fictional business in the than real. 1920’s. Unlike other He subtitled his autobiography “Notes on a radio engineers of the Busy Life.” That really is an understatement— day, Simon does not movies are made from material much less ex- have a long list of pat- citing than Simon’s autobiography. He crossed ents to his name. paths with a great number of famous scientists, Rather, he was a politicians, and high-powered businessmen business entrepre- during his lifetime. Most of these acquaintan- neur who could get ces considered themselves either his friend or the job done. his foe, with little middle ground. For example, 27 or so lawyers represented him during the seemingly endless legal disputes in which he was entangled. His business dealings were con- sidered brilliant by some people and danger-

Volume 20, 2007 145 Emil J. Simon

Fig. 2. Advertisement in the Consolidated Radio Call Book of 1919, showing some of the products manufactured by Emil J. Simon. 146 AWA Review Kleinman et al. ously unethical by others. Some of met Milton May, who became his peers considered him a wonder- Emil’s guardian three years later. ful engineer and inventor. Others She married Milton when she was who knew him felt that he was at 18 years old, giving up her college best a second rate engineer, and at career at Barnard College. worst a crook. Jeannette and Milton May had four children of their own, one of whom BORN INTO WEALTH was named Robert L. May. Robert, Emil J. Simon’s colorful life, full Emil’s nephew, graduated Phi Beta of spectacular successes and fail- Kappa in 1926 from Dartmouth ures, began when he was born into College where he had been a class- a wealthy family in New York City. mate of Ted Geisel, also known as In his own words, “I am told that I Dr. Seuss. Robert May worked af- was born in a brownstone house at ter graduation for the advertising 12 W. 131st Street on Oct. 25, 1888 department at R.H. Macy & Com- with a silver spoon in my mouth!” pany from 1926 to 1928. He was His birth was apparently not offi- working in the advertising depart- cially recorded, as a search made ment at Montgomery Ward in 1939 for births in the Borough of Man- when he was commissioned to hattan for the calendar years 1888 write a Christmas story to be given and 1889 for the name Emil Jacob away to children during the holi- Simon did not yield any result. His days as an advertising ploy by the father, Sigmont Simon, was a self- company. He wrote the story made successful clothing salesman “Rudolph the Red-Nosed Rein- from Philadelphia. His mother was deer,” which was then published by Jennie Sampter, who was the only Montgomery Ward and later put to girl and next to youngest of seven music by Robert May’s brother-in- children. law, the famous holiday songwriter The Sampter family was from Johnny Marks.2 Europe. Jennie Sampter’s parents Emil began collecting stamps as Michael and Rachel Sampter had a child, a hobby that he continued moved from Germany to New York through his early adult years. He City in 1848. The Sampter name is used some of the profits from his from the town of Sampter, in East wartime business to buy stamps Prussia, where the family had been when he was about 30 years old. successful grain merchants and Later, when he was short of cash innkeepers. In the United States, in 1921 as his postwar business Grandfather Sampter formed failed, he sold the collection. His Sampter & Sons, makers of boy’s mother insisted that he and his sis- clothing. This business made the ter Jeannette take up musical in- family wealthy and became the struments, so he took violin lessons largest manufacturer of boy’s (Fig. 3) and she learned to play the clothing in the United States at the piano. Emil was taught how to play time. Jennie Sampter met Sigmont the violin by the noted violinist Simon eventually through one of Bernard Sinsheimer, who later per- her brothers, who had become one formed as the first violinist for the of Sigmont’s business acquaintan- Boston Symphony Orchestra. As ces. soon as he could, Emil put down Sigmont Simon and Jennie the violin for good. This was a de- Sampter had two children, Emil cision he would later regret, after and Jeannette. Jeannette was two he learned to enjoy classical and years older than Emil. In 1904 she operatic music.

Volume 20, 2007 147 Emil J. Simon household. His mother and father parted in 1894 when he was six years old. Their nasty high society separation dispute and the pro- tracted legal proceedings over the custody of their children that re- sulted from it were splashed all over the papers. Eventually the courts decided that there were no legal grounds for separation, but granted custody of the two chil- dren to the mother with permis- sion to visit the father. The sensa- tional case went all the way to the Supreme Court in New York. At least 18 New York Times articles were published about the quarrels and separation, which dealt with an argument over the firing of a maid. Fig. 3. Emil J. Simon with violin, about The maid became known as the age 8. “French maid” in the newspapers and was the governess of the chil- One memorable early experi- dren. Emil’s father felt he was ridi- ence occurred about 1897 when the culed by the maid, and he fired her young Emil Simon met President without his wife’s permission while McKinley. Emil, his sister she was away. Judge McAdam, one Jeannette, and their mother of several judges to rule on the Jennie were on a trip through the case, declared after trying in vain Adirondacks at the time. Simon to reconcile the husband and wife later wrote that he remembered that, “Wives, like facts, are stub- President McKinley pinning a born!” In spite of the separation small American flag on his shirt of his parents, Simon felt that his and telling him, “You will never go early childhood had been a happy through this world with your big one. His early school years were brown eyes shut.” The incident spent at Horace Mann School, a made a noticeable impression on private coeducational college-pre- young Emil, who was saddened paratory day school. His schooling when President McKinley was as- and almost his entire life were sassinated by a Polish anarchist by spent in New York. During his life- the name of Leon Czolgosz four time, he had 22 different resi- years later at Buffalo in 1901 while dences in New York City. Much of visiting the Pan American Exhibi- this time he lived with his mother. tion. Emil Simon also mentioned Emil J. Simon was 13 years old that he considered Theodore in 1901 when he enrolled at Roosevelt to be his hero, although Mohegan Lake Military School. they probably never met in person. His mother felt that he needed dis- Simon lived his life by Roosevelt’s cipline and respect for authority. precept, “Hew to the line and let This was the same year that the chips fly where they may.” Marconi reported sending the let- Unfortunately, much of Emil’s ter ‘s’ across the Atlantic, and two childhood was spent in a broken years later the Wright brothers

148 AWA Review Kleinman et al. would announce success with hu- ing house with supplies he ob- man flight. Both these events tained from MESCO. His career at would inspire Simon later in life. Mohegan had come to an end The automobile in 1901 was still shortly after that, when his called the ‘horseless carriage’ and grandfather’s clothing firm of M. was a curiosity which prompted Sampter & Sons became overex- boys to yell ‘get a horse!’ Jennie tended and went bankrupt after 40 Simon, Emil’s mother, had to go to years in business. With money the drugstore at the street corner tight, he could no longer afford to to make a telephone call, since tele- attend Mohegan and he finished phones were not yet commonly his pre-college education at Horace found in homes. Mann School. Mohegan Lake Military School was quite small and was located COLLEGE, JAIL, AND A near Peekskill, New York. Emil at- CAREER IN RADIO tended the school there for three Emil J. Simon entered Colum- years with 70 other cadets. He bia University in the fall of 1905 as didn’t like marching, and became a freshman in Electrical Engineer- a drummer so that he could stand ing. His college education was at the side of the field as the other partly financed by selling a five- cadets paraded carrying their passenger Winton motor car he heavy rifles. He was one of only had won in a contest. Before he had two boys at Mohegan who were of turned 18 years old, his family doc- the Jewish faith, and he felt that tor as well as the Pierce Arrow rep- there was considerable prejudice. resentative in New York taught In spite of this, he won two med- him how to drive their cars. In late als for public speaking while there. 1904, Motor magazine held a con- It is during this time that he devel- test that required its readers to oped an intense interest in motor correctly identify the names of 30 cars, in the progress of wireless te- cars illustrated in groups of 10 cars legraphy, and in the development in each of three issues. At that time of the airplane. automobiles were still a novelty, These great advances in the and there were many obscure realm of science fired Emil’s imagi- makes and models, especially ones nation and he decided on a career manufactured overseas. The first in engineering. He read Scientific prize in the contest was the $3000 American each week and anx- Winton motor car. iously followed the articles as they When all the entries were presented technical progress to the opened, 53 contestants had layman. In his room at Mohegan, guessed the identities of all of the he assembled wet batteries, mo- cars correctly. A lottery was then tors, and lamps. He reproduced held at the business office of Mo- the experiments presented in the tor magazine, and Emil was se- articles in Scientific American. He lected by preliminary lottery to be started placing orders for electri- the first to pick one of 53 envelopes cal parts from the catalog of the from a cigar box. He picked first Manhattan Electrical Supply Com- and luckily selected the winning pany of New York. In the winter envelope. The Winton Touring car of 1904 to 1905, at the age of 16, was red with four cylinders, and he he was awarded his first electrical drove it for about a month (no contract of any kind—he electri- driver’s license was required in cally wired an entire large board-

Volume 20, 2007 149 Emil J. Simon those days) before his mother de- ratory, and some was given to the cided he should sell it. It was ap- station by De Forest. The antenna parently a sign of extreme wealth consisted of two horizontal wires to drive such an automobile, and separated by spreaders and strung it was embarrassing to his family. between parts of the uncompleted By this time, much of the family University Hall. The apparatus was wealth had been lost in the failure located in the electrochemical of the Sampter clothing business. laboratory, in the basement of one The car, which he had been so ex- of the science buildings. From cited to win, was sold for $1600 to here, the students at Columbia a car dealer on Broadway in New communicated in code over the York. wireless apparatus with a station Emil’s interest in wireless teleg- operated by students at Princeton raphy was already substantial by University. the time he was a sophomore at Simon also studied with Profes- Columbia. In October of 1906, he sor Francis Crocker, the Head of attended a reading by Dr. Lee De the Department of Electrical En- Forest of a paper on the “Audion: gineering at Columbia. Simon vis- a New Detector for Wireless Teleg- ited Crocker in his office fre- raphy.” Dr. Michael Pupin, one of quently. Crocker was the co- Simon’s professors in 1908 and founder of the Crocker-Wheeler 1909 and president of the Ameri- Company which made motor gen- can Institute of Electrical Engi- erators for wireless equipment. neers (AIEE), had befriended Crocker was a friend of Thomas Simon because of the young man’s Edison. Outside of Crocker’s office, experimentation with radio. Pupin, under a large picture of Edison, a teacher of mathematical physics, was written in Edison’s own hand- and himself the owner of several writing, “All things come to him patents regarding telegraphy and who hustles while he waits.” telephony, introduced Simon to De It was as if fate had decided to Forest. De Forest told him how to take a hand in Simon’s life. Two build his own Audion, which he years after the first contest was then successfully did. When held, Motor magazine conducted Simon himself then wrote an ar- another contest nearly identical to ticle for Scientific American about the original one which Simon had the Audion, he received a note back won. Again, the prize was a $3000 from the editors stating that they automobile, and once again ten did not consider the manuscript of pictures appeared in each of three sufficient importance for publica- issues of the magazine. These were tion! Still a teenager, Emil took his pictures as before of obscure first job with De Forest during his makes and models, many of which junior year at Columbia in the were from Europe. Simon tied summer of 1908. He started at a with three other contestants for seven dollars per week salary, then the first prize this time, and it was after a two week period he received decided to give each of the winners $15 each week. a $750 car of their own choosing. Simon went on to help form the Emil Simon selected a Ford, and Radio Club of Columbia, and he he decided not to sell it, a decision and others built the first radio sta- that would have important conse- tion at Columbia University. He quences. built some of the apparatus him- He and his mother traveled to- self, some came from Pupin’s labo- gether to Europe in 1907, sailing

150 AWA Review Kleinman et al. across the Atlantic on the Graf imprisonment. After spending 3 Waldersee. The Ford had been de- months in jail in Germany (in spite livered in late May, and they left of his mother’s efforts to have him on the trip when school ended in released early), he returned to the June. They brought the car along United States on November 22, on the journey so that they could 1907, on the maiden voyage of the tour through Germany. During the Mauritania from Liverpool to New voyage, Simon spent much time York. He had missed October and with the radio operator in the ra- November of that school year, and dio room. After landing, the teen- this eventually made it impossible aged Emil and his mother began for him to graduate from college their drive through Germany. with the class of 1909. He was told However, his mother did not like he would have to attend classes for the way he had driven on the first a fifth year at Columbia in order to part of the journey, feeling that he graduate with a degree in Electri- drove “too fast and too reckless.” cal Engineering. She took the train to Dresden Emil decided not to return to rather than driving with him. On Columbia University in the fall of the road to Dresden, he struck a 1909 and he never did graduate pedestrian while driving the Ford. from college or receive a degree in He claimed that he was trying to Electrical Engineering. Instead, he maneuver around a hay wagon that went to work for Lee De Forest. He was traveling slowly in the oppo- enjoyed doing experimental work site direction on the wrong side of at De Forest’s Radio Telephone the road. The unfortunate young Company with De Forest and his woman, who had been picking ber- second wife Nora Stanton Blatch ries minutes earlier, suffered a De Forest, the granddaughter of fractured skull and died. Elizabeth Cady Stanton of the Emil did not stay at the scene, Women’s Suffrage cause. Simon but traveled on to Dresden, where worked on the further develop- he met up again with his mother ment of the three element Audion but did not tell her of the accident. and on an oscillator for radiotele- The German authorities arrested phony. It was during his employ him and he stood trial in Germany with De Forest that he first met for the hit-and-run accident. Bail George Seibt, a noted German sci- was denied as he awaited the trial entist. He also met Fritz which lasted one day. Witnesses Lowenstein, highly regarded as from the scene testified that he was one of the most superb wireless driving faster than they had ever engineers of the time. In 1910, seen a car go. Although he was Simon patented an improvement found guilty of criminal negligence to quenched spark gap design (U.S. by a five judge panel and was given patent 971,935). He discovered the a three month jail term, he was necessity for making the spaces able to avoid prison hard labor by between the quenched spark gap paying three marks a day. His cell plates airtight to avoid corroding was kept clean by other inmates the plate surfaces. He later ob- who were ordered to do the work! tained another patent, improving As had happened with the sepa- this design by making each airtight ration of his parents, at least ten spark gap individually replaceable articles appeared in the New York without destroying the seal. This Times and New York Tribune type of equipment was installed in about the accident and subsequent many De Forest wireless stations,

Volume 20, 2007 151 Emil J. Simon including on the SS Avon in 1910. Company from 1913 to 1914. He While on the SS Avon Simon re- developed ½-kW wireless trans- ceived opera sung by Caruso trans- mitting sets for the Wireless Im- mitted from the mainland to the provement Company until he left Captain’s cabin. Firth’s employ to go into business During his employ by De For- for himself in 1915. est, Simon learned much that would be useful to him when he STRIKING OUT ON HIS OWN eventually struck out on his own. In 1915, the United States Navy Seibt taught him about quenched asked for bids on a contract for a spark gap systems which Simon total of 25 wireless sets of the ½- would incorporate into his most kW type, like the ones Simon had successful designs later. A few made for the Wireless Improve- years after leaving De Forest, in ment Company a year or two ear- 1913, Simon published a paper on lier. Each bidder was required to the Seibt Direct Indicating Wave furnish a sample transmitter. The Meter with the famed Alfred Gold- sets were intended for use on sub- smith, who was a professor at Co- and had to be small lumbia University and later a re- enough to pass through standard searcher for RCA. Interestingly, submarine hatches without diffi- Simon unknowingly passed up a culty. Simon, now 27 years of age, chance for fame and fortune when convinced his mother to supply he noticed an Audion in the De $1000 seed money. He designed Forest laboratory that seemed to and built a set from parts con- be producing a high musical note structed for him by local suppliers in his headphones. He did not rec- with whom he would develop long- ognize the significance of the os- term business relationships. He cillating Audion and the credit for tested and adjusted the sample set the subsequent use of the oscillat- himself in the laboratory of the ing Audion in electrical circuits Electrical Engineering Department went to others. After helping De at Columbia. When the transmit- Forest install the powerful De For- ter was ready to submit to the est Metropolitan Tower radio sta- Navy, he brought it to the offices tion in 1909, Simon continued to of the Manhattan Electric Supply work for De Forest’s Radio Tele- Company where he showed the set phone Company until it failed in off to Louis Pacent, who was then 1910. in charge of the wireless depart- After the demise of the De For- ment for amateurs. Pacent con- est Radio Telephone Company, vinced him to invite David Sarnoff Simon worked briefly for Colonel of the Marconi Company over to John Firth at the Wireless Spe- see the set also. Both Pacent and cialty Apparatus Company. In Simon considered Sarnoff a friend. 1911, he worked for the National Simon’s design implemented sev- Electric Signaling Company for eral innovations and Sarnoff ex- $95 per week. His job with NESCO pressed admiration. Simon then involved getting NESCO wireless submitted a bid for the Navy con- sets passed by the Navy. It was in tract offering 25 of his sets for this pre-war period he was elected $1,200 apiece, a contract worth secretary of the new Institute of $30,000. The sets were to be de- Radio Engineers (IRE) in May, livered in four or five months. 1912. He worked again with Firth A total of five bids were received at the Wireless Improvement for the contract and sample sets

152 AWA Review Kleinman et al. were submitted by each bidder matter of Marconi vs. Simon, the (Wireless Specialty Apparatus court decided in favor of Simon on Company, National Electric Sig- November 18, 1915. The decision naling Company, Wireless Im- was based more on public policy provement Company, Marconi, than legal grounds and it was later and Simon.). After a few weeks reversed in the Supreme Court. time, Simon contacted George H. The final judgment for Marconi Clark, senior radio expert aide at was limited to damages vs. the the Bureau of Steam Engineering United States. However, Simon which was evaluating the bids and had managed to get a foot in the sample sets. Clark told Simon that door with the Navy, and thereaf- he was not the lowest bidder and ter he had no problem winning that although the Navy liked his set many more Navy contracts. the best, he probably would not The Marconi Company contin- win the bid since he was a young ued to try to bar Simon from sell- and untested newcomer to the ing wireless sets. A few years after field. Clark also told him that Marconi vs. Simon was decided in Marconi was the lowest bidder. the United States, the Canadian With this information, Simon re- Car & Foundry Company was sued alized that he would have to pull by the Marconi Wireless Telegraph strings to get the contract. Through Company of Canada in the Provin- his associates he contacted and ar- cial Court in Montreal for install- ranged a meeting with Franklin D. ing Simon one kilowatt transmit- Roosevelt, who was the Assistant ters on French minesweepers. The Secretary to the Navy at the time suit was based again on the famous and who would make the final de- “7777” Marconi tuned circuit cision on the contract. Simon con- patent, but since the transmitters vinced Roosevelt to award the con- were installed on the ships before tract to him. they had antennas, the court ruled It took only two weeks after the that the Simon sets could not in- Navy contract was awarded for fringe the Marconi claims! Sarnoff and the Marconi Company to have Simon served with prelimi- MAKING A FORTUNE OUT nary injunction papers. Sarnoff, OF A SHOESTRING who had seen the set at Simon’s In February, 1917, Emil J. invitation at the MESCO office a Simon received a letter from R.S. few weeks earlier, alleged that the Griffin, Engineer-in-Chief, Bureau set infringed the basic claims of the of Steam Engineering of the U. S. Marconi tuned circuit patent. Thus Navy advising him that in the “case began Simon’s long associations of a national crisis” they would with lawyers—about 27 different need hundreds of wireless sets lawyers would represent him dur- from him. The Bureau requested ing his lifetime. Roosevelt was con- that he make preliminary plans for vinced to write a letter to the court this possibility with his suppliers that the Navy would be embar- and subcontractors. When war was rassed and harmed if Marconi was subsequently declared with Ger- successful in the suit. If the court many in April of 1917, the flood- interfered with the Simon con- gates opened. Simon was already tract, Roosevelt argued, vessels of getting geared up for production war badly needed and nearing when the war broke out. He re- completion would be jeopardized ceived a telephone call from Com- and the Navy embarrassed. In the

Volume 20, 2007 153 Emil J. Simon mander Hooper of the Navy tell- interested in the invention of the ing him to proceed urgently with airplane from an early age, and he the manufacture of 100 ½-kW sets had experimented with wireless at $1,700 each. Simon was the first from airplanes in San Diego, Cali- contractor to deliver on a war con- fornia. In 1916, he had traveled tract for radio sets, finishing the there to test a spark transmitter first of these sets in June of 1917. installed in a Curtiss JN4 Jenny, Simon’s organization at this time an airplane that became known as was very small.3 All financing, one of the leading trainers in the manufacturing, assembling and world. In February, 1917, again in testing was done locally. San Diego he sent a message from With more Navy contracts ex- an airplane 112 miles back to a re- pected on the heels of the first, ceiving station. The Navy contract Simon needed to move to a larger was for sets that were similar to the headquarters and he hired more ones he had tested in San Diego staff. Alfred Seelig joined the orga- within the past year. The Wireless nization part time. He formerly Improvement Company, directed managed the German-owned At- by Simon’s friend Roy Thompson, lantic Communication Company manufactured these sets following which operated the Sayville, New designs and drawings supplied by York, transatlantic station. Seelig Simon. These aircraft sets were was well-respected for his transla- used in Europe during the war. tion of Dr. ’s By 1918, there were at least book “Wireless Telegraphy” from twenty subcontractors manufac- German into English. Alfred turing items for the United States Grebe, who formed his own com- Navy under Emil J. Simon draw- pany to manufacturer broadcast ings.4 In Simon’s own words, receivers after the war, also joined “Never before had an organization Simon’s organization. Grebe had without a factory of its own pro- experience installing and testing duced so efficiently and so well.” Telefunken wireless sets in the U.S. The major production items of the During the war, Grebe installed organization were the ½ and one- radios in French anti-submarine kW marine sets. The team in boats and built receivers for charge of engineering for these sets Simon. At the end of 1917, the in- included Frank Hinners, Benjamin ventor and engineer Benjamin Leibowitz, Larry Tesh, Benjamin Meissner joined the Simon engi- Meissner, and Cyril Reinhard. All neering staff. The Navy had re- of these men were tried-and-true cently fired Meissner because of friends of Simon. His mother was questions about his loyalty, and in charge of the production and Simon had to again ask for favors purchasing records. Since he had from Commander Hooper and As- no factory of his own, he used sub- sistant Secretary Roosevelt before contractors such as Kilbourne & he was finally allowed to hire Clark, the Pearce Manufacturing Meissner. In 1918, Joseph Freed Company, and others to manufac- joined Simon’s engineering staff. ture equipment according to his Freed later became the successful own design. head of the Freed-Eisemann Radio During the war, Emil J. Simon Corporation. was known for quantity, not qual- Later in 1917, Simon landed a ity. He personally made a fortune Navy contract for 100 aircraft on war orders, furnishing radio transmitting sets. Simon had been equipment worth about $2 million

154 AWA Review Kleinman et al. to the United States and France. bers by the U.S. Navy’s Bureau of Interestingly, Simon himself was Steam Engineering, as were his of military draft age during World sets that differed only in the power War I, but he noted that he was supply requirement or in the type exempted from serving military of generator present. The type duty by the Navy because of the numbers for the equipment that “important nature of my work.” was designed by Simon for the During the year 1918, the Emil J. Bureau of Steam Engineering (U.S. Simon organization turned out Navy) begin with “CE”, the desig- about 1000 transmitters of ½ and nator assigned to the Emil J. one-kilowatt for submarines, Simon organization. subchasers, and airplanes. After The Simon ½-kW quenched the war, people who had been ship- spark gap panel sets designed dur- board radio operators during the ing World War I for the U.S. Navy conflict wrote to Simon telling him were the CE 6065 (also known as that they had been able to commu- the SE 606), CE 606A (Fig. 4), CE nicate back to New York from 606B, and the CE 859, CE 859A Buenos Aires, Argentina, and Rio and CE 859B6. The CE 606 type de Janeiro, Brazil, using these one- transmitters (an “A” or “B” suffix kW Simon quenched gap spark indicates a minor modification of transmitting sets. This was quite the standard design) included a an achievement at the time! In wave changer capable of a range January, 1918, a patent (number from 300-600 meters, while the 1,253,103) was issued to Simon CE 859 type transmitters had wave pertaining to the unit construction changers capable of a wider range of airtight quenched spark gaps. of transmitting wavelengths. Both This patent was infringed by nearly the CE 606 and CE 859 type trans- all the manufacturers of Navy mitters were manufactured under transmitters during the war. After contracts with the Navy in 1917 the war, Simon was awarded and 1918. They both used the same $30,000 by a Board established to assign compensation to companies for unauthorized government use of patents. Much of Simon’s success can be attributed to the fact that many of his transmitters conformed to standard designs which varied little between models. They used many of the same parts and some- times differed only in substituting components rated at higher volt- age when necessary for higher power output. This allowed Simon to produce the transmitters very quickly. His competitors made improvements and changes to their designs often and as a result they had to waste time gearing up for the new model lines. Simon’s slightly different transmitting sets Fig. 4. Emil J. Simon CE 606A ½ kilo- were assigned different type num- watt quenched spark gap transmitter.

Volume 20, 2007 155 Emil J. Simon CE 607 Simon-designed quenched ½-kW Simon transmitting sets spark gap, Dubilier transmitting were later converted to the Model condenser and protective device, TM 100 watt alternating current and Crocker-Wheeler motor gen- vacuum tube transmitters for use erator. The front panel of these in submarines when these tubes transmitters had a connecting strip became more widely available at the bottom center for the attach- around 1922.8 The submarines had ment of the CE 613 send-receive loop antennas, and the TM was switch (Fig. 5) which also had the developed to deliver as much transmitting key mounted on the power to the antenna as possible same base. While the transmitting at relatively low voltage. The insu- key for these sets is shown as the lation on the cable of the loop an- CE 611 in the Navy Type Numbers tenna made such a limited voltage Book for 1924, photographs of necessary. The peak voltages when these sets show that the Wireless using continuous wave are consid- Specialty Apparatus Company’s SE erably less than those of a spark 653 spark key was commonly set.9 mounted on the CE 613 send-re- Before the start of World War ceive switch instead. I, the U.S. Navy purchased and tested aircraft transmitting and receiving sets from De Forest, American Marconi, Sperry Gyro- scope Company, and Emil J. Simon.10 The Navy purchased 15 of the Simon ½-kW CE 615 airplane quenched gap spark transmitting sets. Although the Navy purchased a similar number of sets from the three other companies at this time as well, the Simon set (according to Howeth) was the only one that worked well and weighed about 100 pounds.11 The CE 615 had a Fig. 5. Emil J. Simon CE-613 ½ kW brake (CE 697) for the wing- Antenna Switch and Board, with Wire- mounted generator so that it could less Specialty Apparatus Company be stopped when not needed. The SE-653 Spark Key. tuning of the antenna circuit was The SE 653 was an enclosed key accomplished by varying the intended for use in submarines, as length of the insulated trailing an- were many of the transmitting sets tenna wire that was fed out from that Simon produced. The CE an antenna reel (CE 702) mounted 606A was made by the Kilbourne on the airplane. The receiver used & Clark Manufacturing Company7 a triode and a regenerative circuit. and varied from the CE 606 only Early tests of the CE 615 showed in the Kilbourne & Clark motor that it could transmit over a dis- generator with which it was sup- tance of 150 miles, a useful range plied. The CE 606B was made for for battlefield spotting and similar 120 VDC drive, and was made by missions requiring communica- the Wireless Improvement Com- tion back to the command post. pany. The modifications made to The CE 615 was modified into the the CE 859 were similarly minor. CE 700 by extending the wave- Some of the CE 859 and CE 859A length range.

156 AWA Review Kleinman et al.

Another 100 of the Simon air- manual were supplied with the plane sets were purchased in late transmitters. A post-war one-kW 1917. The transmitting key sup- set was manufactured for the Navy plied for the CE 615 and CE 700 by Simon as the CE 1221, the main airplane sets was the CE 693 (Fig. difference between this set and the 6), an unusual key design. The CE 861 apparently being the spark contacts on the CE 693 are gap units and spark gap assembly. mounted inside an enclosure, the There are relatively few ref- top portion of which is a flexible erences to U.S. Army Signal Corps diaphragm (made of kangaroo contracts awarded to Emil J. skin) carrying the upper contact Simon. Simon did, however, make that can be pushed down onto the radios for the Signal Corps. One of lower contact fixed to the base of these was the SCR112, a quenched- the key. The remarkable knob for gap transmitter and one-tube re- the key is a concave disc over two ceiver probably made just after the inches in diameter, presumably so war.13 that it would be easier to hold onto in a moving airplane.

Fig. 6. Emil J. Simon CE-693 Dia- phragm Type Spark Key. Part of the CE-615 ½ kW Airplane Transmitter. The one-kW panel radio set made for the U.S. Navy during World War I was designated as the CE 861 (Fig. 7). It is essentially the same basic set as the ½-kW CE 606, with components rated for higher power. Only about 100 of the one-kW panel sets were or- Fig. 7. Emil J. Simon CE 861 one kilo- dered from Simon by the U.S. watt quenched spark gap transmitter. Navy during the war, compared to Notice the similarity to the CE 606A about 300 of the ½-kW sets and ½ kW transmitter. This type of stan- more than 100 of the ½-kW air- dardization allowed the Emil J. Simon plane sets. Another 38 of the one- organization to mass produce trans- kW panel sets were furnished to mitters. the French government at a cost of POST WAR PROBLEMS $3075 per set. As with the other There is no doubt that the most panel sets, the front panel was productive time of Emil J. Simon’s made of 2 sections of bakelite, and life was during the war. After the the otherwise front-heavy appara- war, he struggled to compete with tus was balanced by the motor gen- bigger, more influential companies erator mounted on the rear bottom with more political clout and finan- of the iron frame.12 As usual, a box cial backing. His organization of spare parts and an instruction dwindled as orders for his trans-

Volume 20, 2007 157 Emil J. Simon mitting sets fell off and he searched place Western Union or the Postal for other sources of revenue. Many Telegraph Company landline ser- of the same people whom he had vice with a wireless service. The considered friends before the on- corporation would initially target set of the war had become com- New York, Chicago, Detroit, Cleve- petitors jealous of his success and land, Pittsburgh and Philadelphia. they tried to block Simon’s post- As time went on, plans would in- war ventures. Much later in life, clude Duluth, Buffalo and other he was nostalgic for the high life cities closer to the Midwest. and times he had experienced with The venture was officially his wartime organization. The dis- known as the Inter-City Radio banding of his wartime organiza- Telegraph Company after incorpo- tion was heartbreaking. rating in Delaware in December, In 1919, Simon made plans to 1920. The umbrella corporation create a ship-to-shore and Inter- consolidated Inter-City Radio City radio service as soon as gov- companies in Ohio, Michigan, Illi- ernment war restrictions were nois, New York, and Delaware. lifted. Simon hoped to develop a The service was officially opened national radio network in the in a formal ceremony featuring United States by the installation Mayor Hylan of the City of New and maintenance of 10-kW radio York (Fig. 8). The story was car- transmitters in as many major cit- ried in major newspapers includ- ies as possible. The idea was to re- ing the New York Times along with

Fig. 8. The Inter-City Telegraph service was officially opened in a formal cer- emony featuring Mayor Hylan of the City of New York in 1920. The Mayor is center left reading from a script, and Emil J. Simon is the short man in the bowtie with a mustache further left. Dr. Lee De Forest is right center in the light-colored suit.

158 AWA Review Kleinman et al. a picture of the Mayor and the from the Post Office to build three other dignitaries present, Emil J. five-kW wireless transmitters for Simon himself and Dr. Lee De For- the newly created airmail service. est. A few years later Herbert This was a pioneer service and Hoover, who was the Secretary of helped inaugurate the first airmail Commerce at the time, refused to service between New York and reissue Inter-City’s wireless license Chicago. The transmitters, which under the radio law of 1912. Simon became known as “Airways” trans- sued him and won. Hoover alleged mitters (Fig. 9), were to be located that he could find no vacant fre- in Bellefonte, Pennsylvania, Cleve- quency to assign to Inter-City, but land, Ohio, and Chicago, Illinois. the court ruled that it was the re- The transmitters were to be five- sponsibility of the Secretary to is- kW, 500-cycle quenched-gap sue licenses and frequencies and spark rigs. Each station had a that the Secretary had no discre- tower 200 feet high and a range of tion for whom to issue a license. about 400 miles to airplanes and The Radio Communication 700 miles between stations. These Company was another Simon busi- stations were mainly intended to ness venture. He traveled to Ger- help pilots navigate and avoid in- many to meet with Telefunken of- clement weather. Simon’s prelimi- ficials and negotiate a New York to nary specifications for these radios Berlin wireless circuit mainly in- showed one complete transmitter tended for the transmission of and receiving unit (Army Type B, news. RCA was also in competition Fig. 10) offered for a price of for the contract and tried to use $17,650. political influence and favors to lock Simon out. It was dirty busi- ness, but Simon eventually got the Telefunken contract. RCA only succeeded in winning the contract for communications involving the German government. As Simon noted, there was “nefarious plot- ting by big interests behind the scenes.” In particular, David Sarnoff of RCA tried to influence banks to refuse loans to Simon for both Inter-City Radio and the ven- ture with Telefunken. When asked his opinion by the banks, he gave Simon a poor recommendation and suggested that banks pursue an “unbiased” opinion from Dr. Alfred Goldsmith. Goldsmith at the time was on the RCA payroll as a consulting engineer! Simon had in the meantime met with the influential William Randolph Hearst and obtained his backing for the Telefunken radio contract. In the early summer of 1919, Fig. 9. Simon Airways 5 kW transmit- ter built for the new Postal Airmail Simon was awarded a contract Service.

Volume 20, 2007 159 Emil J. Simon Pensacola Navy Yard. In a bro- chure printed by Simon to pro- mote the Meissner Airfone, two of the military endorsements for the product came from a 2nd Lieuten- ant by the name of Herbert Sampter.16 Sampter was most likely a relative of Simon although the general public would probably not have known this. Many of the Meissner Airfones were sold to pi- lot training schools. The partner- ship with Meissner, named the Fig. 10. Simon Army Type B receiver, Meissner Inventions Corporation, as supplied with the 5 kW Airways also developed a noise canceling transmitter. “open back” microphone but even- tually had to concede priority to Simon’s business ventures also Magnavox. Magnavox in return included the Directive Radio Com- agreed to pay them a portion of pany, which he started as a part- royalties collected on the patent. nership with Fred Kolster. Kolster Meissner also invented the elec- was a well-known inventor who tronic organ and guitar. owned the rights to several useful patents and the company was formed to promote the inventions based on them. Most of these de- vices were used in the testing of radio transmitters. While Simon was traveling in Europe with Kolster in 1919, he met Cyril Elwell, who had been instrumen- tal in importing arc transmitter technology to the United States. Simon drew up preliminary speci- fications for the manufacture of Fig. 11. The Meissner Airfone, devel- Elwell Arc transmitters of different oped by Benjamin Meissner in 1917 power outputs.14 At least one five- so that occupants in a noisy airplane kW Elwell Arc set was manufac- could speak with each other more eas- tured and shipped by Simon in ily. The device was marketed by In- ter-City Telegraph Co. Photo from In- 1920. ter-City brochure about 1920. Simon started another business with Ben Meissner, who was a Simon was the ultimate entre- close friend and had been an engi- preneur who felt that his success neer in Simon’s wartime organiza- could be attributed to opportuni- tion. Meissner had invented a ties that the new wireless technol- speaking tube device that operated ogy had provided. He testified in acoustically and allowed the pilot December, 1918, before the House to speak to his co-pilot or a pupil Committee which was considering easily in a noisy airplane. The de- placing wireless development and vice, which was called the Meissner circuits under government control Airfone15 (Fig. 11), was developed and ownership in the United in 1917 while Meissner was at the States. He testified that competi-

160 AWA Review Kleinman et al. tion was healthy and desirable. to gov’t restrictions on wireless) Simon felt strongly that future gen- RCA, David Sarnoff: erations of scientists and engineers “Personally I do not think that should have the opportunity by he is a good businessman...” which he himself had prospered. “attained his position more In 1919, Simon placed an adver- through good luck than good tisement in the New York Times, management” stating that he had $100,000 to “Marconi is now suing him for invest in a worthy business ven- infringement…and if these suits ture. One of those who replied was are successful he will be a poor Samuel Goldwin, who wanted him man.” to invest in one of his motion pic- Alfred Goldsmith: tures. Goldwin turned down “my opinion regarding him is Simon’s offer eventually because not very high” Simon wanted to invest in the en- “I do not think that he is can- tire motion picture venture and not did/frank man or trustworthy” just one film. “he would use every loophole” Simon asked banks for loans for “his reputation is generally his businesses on several occa- rather shady” sions, and the banks requested in- “I do not know of anyone in the formation from many knowledge- radio field who would act as his able radio engineers and business sponsor” associates. These recommenda- “very nearly Bolshevistic” tions provide an interesting insight “the wireless co would be one into Simon’s personality and how more link in the German espio- he was perceived by his contempo- nage system” raries. Comments from some of “I do not have confidence in Mr. Simon’s peers regarding his abil- 17 Simon and I doubt his ability in ity and character include: the wireless field” Kilbourne & Clark Mfg Co, Mr. Wireless Improvement Co., Roy Cooper VP: Thompson: “very fine and upright gentle- “exceptionally bright & ener- man” getic man” “undoubtedly possesses great “honest and reliable” skill as a radio engineer” “he can drive a very sharp bar- “as regards his technical ability gain” he is a second rate man” F.G Pearce & Co, Walter Pearce: Cutting & Washington: Bowden “He is all work and very much Washington: of a hustler.” “I do not have such a good opin- “I regard Simon as being almost ion of him as an engineer and I on the same level with such think his ability in a technical men as Marconi…” line is really secondary” RCA, George H. Clark: “Technically I think he is a sec- Eventually even former employees ond rate man, but he certainly turned against him: letter fromE. can make money...” Deighan (Radio Sales & Service “would really warn you against Co) to EJS, Sept 21, 1922: investing any money” “I take this means of informing was told that “Simon was as you that any plans you may near a Bolshevist as anyone” make to do business in Cleve- (referring to Simon’s opposition land will be promptly

Volume 20, 2007 161 Emil J. Simon squelched. The sale of any stock allo- in this state by any organization cated 76 short wave channels for you are associated with will be point-to-point communications. sufficient cause to ask the Dis- Inter-City applied for some of the trict Attorney at Columbus for channels, as did RCA, Universal an injunction restraining you Radio, and Mackay.18 Universal from operating in this state. was granted 40 of the channels. Your protested checks and pro- RCA was granted 20 of them, and paganda furnished by local bro- Mackay received allocation of 10 of kers will materially assist in the channels. None of the new keeping you out of Cleveland as channels were allocated to Inter- we have no room for you here. City. Any man who does business at The planned Inter-City expan- the expense of his employees sion was to be funded by Warner doesn’t deserve to be in busi- Bros. The failure of Inter-City to ness. Your action in holding up successfully win further channels the sale of Inter-city assets is for expansion westward was com- the signal for every employee to pounded by the effects of the stock hurt you as much as they for- market collapse in 1929, which merly helped you. You can made prospects for additional fi- count on me to hurt you to the nancing grim. Simon held discus- tune of $2100 with interest. sions with shipping magnate R. Yours truly, E.I. Deighan Stanley Dollar regarding the acqui- PS: We are all set to trim your sition of Inter-City. Dollar initially pants off in the appeal” agreed, then backed out when Sarnoff threatened a lawsuit on 25 Most of Simon’s ventures in the patents if Dollar took over Inter- postwar period failed or were sold City! Simon was not to be denied as his luck failed. He sold his hold- easily and brought suit against the ings in the Radio Communication Federal Radio Commission and Company in 1921 to the Commer- tied up all 76 wavelengths. It later cial Cable Company for $100,000. was proven that Universal had suc- Inter-City Radio Company oper- cessfully bribed a federal official ated from about 1920 to 1930 for the wavelength allocations, but when it failed. For most of its ex- Universal went bankrupt just a istence, it was embroiled in law- couple of years after Inter-City suits involving patent rights and failed anyway. Although his law- competitors. The company was yers felt that Simon would even- probably not above bending the tually win his lawsuit, Simon was rules. For example, in a 1922 bal- unable to bear the expense of the ance sheet, Inter-City Radio listed costly litigation. The teletype, a its total assets at $1,621,429.45. Of promising new technology devel- this total, $1,600,000.00 is de- oped by AT&T, also contributed to scribed as goodwill and patent Inter-City’s failure. With Inter-City rights! Walter T. Spalding, Presi- deep in debt, the Federal Radio dent of Inter-City Radio in 1922, Commission refused to renew the resigned his position in April of licenses of the five Inter-City sta- that year over arguments about tions, and shareholder stock be- business practices related to un- came worthless when assets were sound financing and cash advances sold to satisfy creditors. The Fed- to stock promoters. The company’s eral Trade Commission eventually fate was sealed in 1927 when the brought action against RCA,

162 AWA Review Kleinman et al.

AT&T, General Electric, and only once, but the love was not re- Westinghouse for conspiracy to ciprocated. He met Louise violate antitrust laws. All this hap- Dittemore in 1919 while on a trip pened too late to help Simon’s to meet Fred Kolster in Paris as companies. part of a U.S. delegation to a radio In 1932, shortly after the fail- conference. Simon’s itinerary for ure of Inter-City Radio, Simon the trip berthed him on the S.S. conceived and developed a direc- Adriatic from New York to tion finder he called the Simon Southampton, England. Louise Radioguide, a device used in air- Dittemore was a very beautiful 18- planes. He incorporated the Radio year-old blonde from Boston. The Navigational Instrument Com- young Louise was in love with a pany to manufacture and market Spaniard with a rich family inher- the Simon Radioguide. This com- itance. Simon (who was 31 years pany failed in 1948, and after that old at the time) proposed to Louise Simon became an independent twice, once while on the ship and consulting engineer specializing in once afterwards near London. She radio patent suits. turned him down both times. She married the Spanish nobleman PERSONAL MATTERS and they gambled away his inher- Much of Emil J. Simon’s busi- itance at Monte Carlo. She even- ness and public life oscillated be- tually settled down with a second tween rich and poor, crook versus husband in Providence, Rhode Is- hustling radio inventor. He made land. This European trip was oth- a fortune during World War I, and erwise noteworthy for Simon, be- then ran short of cash trying to bail cause he met General Ferrie and out Inter-City Radio in the 1920’s. Lucien Levy while in Paris. General His personal life was similarly Ferrie was the head of the French vexed. He never married, and Military Radio establishment and never had any children. In 1915, he was as responsible for the advance- traveled to San Diego with a beau- ment of the new technology in tiful girlfriend, Queenie McDonald. France as anyone. Levy was a Queenie, whose stage name was noted French radio engineer and Queenie Stair, was an actress from inventor. Simon also visited Ber- Melbourne, Australia, who played lin on this trip—his first trip to Ber- small singing parts on stage. Her lin since he was in jail in 1907. mother was arriving in San Fran- Although Simon incorporated cisco, and Simon agreed to take many businesses after World War Queenie along with him to Califor- I, none of them achieved the suc- nia if she would spend the nights cess of his wartime organization. with him. This proved a problem, By the time of his death, he had since neither one of them wanted managed to outlive many of his the arrangement to be made detractors. He was not a top-notch known. On the way from San engineer according to many of his Francisco to San Diego Simon and peers, and this reputation is prob- Queenie encountered Roy ably responsible for the little rec- Weagant and his wife. Weagant ognition he is awarded by histori- was an excellent wireless engineer ans these days. During the last who was well-known for his work years of his life, Emil J. Simon re- with Marconi. Simon had difficulty sided at the Shelton Towers Hotel explaining Queenie’s presence. in New York, now known as the Simon apparently fell in love Marriott East Side Hotel. He died

Volume 20, 2007 163 Emil J. Simon of a heart attack in New York at and Office of Naval History, Midtown Hospital on September Washington, 1963, p. 268. 14, 1963.19 His sister Jeannette 11 Early aircraft design and power died on May 13, 1969. limitations required that the payload be as light as possible. 12 Emil J. Simon. Transmitter NOTES Manual, U.S. Navy 1 kW Panel Set 1 Most of the non-technical (undated.) autobiographical information in the 13 Douglas, Alan, Radio manuscript is derived from the Emil Manufacturers of the 1920’s, Vol. 2, J. Simon autobiographical material at Chandler, AZ: Sonoran Publishing, the Bancroft Library at the University 1995, pg. 258. of California, Berkeley.. This material 14 Preliminary Specifications for is therefore written from Emil J. Elwell Arc Sets, Emil J. Simon Simon’s point of view and no attempt (original typewritten documents in is made to present the views of others the Russ Kleinman and Karen Blisard except as noted within the text. Collection.) 2 Information on Robert L. May is 15 The Meissner Airfone was not the from the 1940 Alumni Directory at same as a home radio receiver that Dartmouth College. Simon designed and marketed in 1919 3 Simon’s organization simply did as the “Simon Airphone.” business under the business name 16 Product brochure for the Meissner “Emil J. Simon”. Airfone by Inter-City Telegraph Co., 4 Including Frederick Pierce Co., from the William J. Hammer where Frank Angelo D’Andrea (later Collection, National Air & Space the proprietor of FADA radio) helped Museum, Smithsonian Institution. design the Simon CE 957 receiver. 17 All quoted comments quite Joseph Freed, founder of Freed- amazingly as in original documents Eisemann in the 1920’s, is another and letters within the Emil J. Simon famous receiver producer who biographical material at the Bancroft worked for Emil J. Simon during the Museum, University of California, war years. Berkeley. 5 Simon CE 606 was developed 18 Mackay Radio and Telegraph was initially as the Simon design Type 500 formed when Clarence Mackay D.1. before receiving the designation acquired Federal Radio and Telegraph from the Bureau of Steam Company in 1927. Engineering. 19 From a New York Times obituary 6 From the U.S. Navy Type Numbers of the next day. Book for 1924. The CE 859B was made by the International Radio This article was peer-reviewed. Telegraph Company. 7 Transmitter Instruction Manual Photo Credits: for Kilbourne & Clark ½ kW Figs. 1, 3, 8 are from the Emil J. Simon Transmitting Sets (undated.) Autobiographical Collection, Bancroft 8 Information about the TM sets from Rare Manuscript Library, UC the George H. Clark Radioana Berkeley. Collection, American Museum of Fig. 2 is from the Jim and Felicia Natural History, Behring Center, Kreuzer Collection. Smithsonian Institution. Figs. 4, 5, 6, 7, 9, 10 are from the Russ 9 Installation and Operating Manual Kleinman & Karen Blisard Collection. Model “TM” Vacuum Tube Fig. 11 is from the William J. Hammer Transmitter 100 Watt output, U.S. Collection, National Air & Space Navy Yard, undated. Museum, Smithsonian Institution. 10 Howeth, Captain L.S. USN (Retired), History of Communications-Electronics in the United States Navy, Bureau of Ships

164 AWA Review Kleinman et al.

APPENDIX -World War One Simon Contracts

Date Contract # Description Contract Price April 10, 1916 Navy 26,106 15 1/2-kW Spark type radio transmitting $17,115.00 Sets @ $1141 each June 15, 1916 Navy 26,589 Parts contracts transformers@ $96.50 Nov. 4, 1916 Signal Corps 3 1/2 –kW radio sets, aeroplane $4,400.00 1090 Jan. 13, 1917 Navy 718- 1 ½-kW Panel Radio set $1,000.00 1917 spare parts for above $99.00 Feb. 10, 1917 Navy 28,896 30 ½-kW quenched gap, 500 cycle radio $48,300.00 Transmitting sets @ $1610 each Mar. 24, 1917 Signal Corps 20 ½-kW aero-radio transmitting sets 1229 with generating aerial etc. @ $1000 each $20,000.00 10 extra parts @ $150 per set $1,500.00 April 4, 1917 Navy 670 100 sets ½-kW 500 cycle panel type $170,000.00 @ $1,700 each May 14, 1917 Navy 30022 50 ½-kW 500 cycle panel type radio $85,000.00 Sets @ $1700 each Aug. 1, 1917 N.Y 524 50 1/2-kW 500 cycle panel type radio $82,250.00 Sets at $1645 each 10 sets same as foregoing, @1700 each $17,000.00 Sept. 29, 1917 Navy 26,106 50 ½-kW 500 cycle panel type radio sets @ $1665 each, 120 volts $83,250.00 10 ½-kW 500 cycle panel type radio sets, 80 volts, @ $1700 each $17000.00 Nov. 7, 1917 Navy 32,760 100 1-kW 500 cycle panel type radio sets, @ $2250 each $225,000.00 Nov. 26, 1917 Navy 33,259 100 Aeroplane sets @ $950 each $95.000.00 100 sets spare parts for foregoing @ $641.71 $64,171.00 Nov. 30, 1917 Navy 33,294 50 ½-kW 500 cycle panel type radio sets @ $1660 each $83,000.00 Dec. 31, 1917 Navy 3077-1918 30 Ammeters @ $52.25 each $1,567.50 54 Transformers @$80.00 each $4,320.00 60 Spark gaps @$160.00 each $9,600.00 100 Gaskets @$1.95 each $195.00 —————- $15,682.50 Feb. 21, 1918 Navy 4200-1918 2 Ammeters @$130.00 each $260.00 2 sets of brushes @$1.00 each $2.00 2 Transformers @$80.00 each $160.00 2 Transmitting Condensers @$35.00 each $70.00 2 Sets bearings @$9.70 each $19.40 2 Spark gaps (16 units in frame) @$250.00 each $500.00 2 Sending keys @$13.50 each $27.00 2 Telephone cords @$1.25 each $2.50 20 key contacts @$1.00 each $20.00 4 Protective devices @$12.50 each $50.00 ————- $1,110.90 Sept. 11, 1918 Signal Corps 83 3 1-kW 500 cycle quenched spark Transmitting radio sets @ $3,315 each $9,945.00 3 spare part sets @ $293.20 each $879.80

Volume 20, 2007 165 Emil J. Simon

ABOUT THE AUTHORS Russ Kleinman (WA5Y) became Karen Blisard (N5IMW) was born interested in the history of wireless and raised in Lubbock, TX, where she after becoming a ham radio operator attended Texas Tech University as an in 1971. He has specialized in spark undergraduate. She received her PhD era keys, transmitters, books and in Pharmacology and her MD from related material for several years. He Case Western Reserve University in is the author of the Spark Key Project, Cleveland. She did her Pathology a cooperative effort among collectors residency and Neuropathology to document spark keys and those fellowship at the University of New who made them. The Spark Key Mexico in Albuquerque. She has been Project is part of a larger website in private pathology practice in Silver about telegraph keys that can be seen City, NM, and is now semi-retired. at http://www.zianet.com/sparks/ Karen, N5IMW, was first index.html. licensed in 1984, and earned her Extra Russ and his wife Karen have now Class license in 1986. She enjoys CW written 5 manuscripts for the AWA and DX (closing in on that 5 band Review as well as articles for the Old DXCC), and has enjoyed operating as Timer’s Bulletin and the Vail a DX station. Her other hobbies Correspondent. Russ continues to include gem cutting, bird watching, enjoy the opportunity to give and most recently, dog training. She seminars on the history of wireless to and her sheltie Brandon compete at various groups in New Mexico as well the excellent level in agility and rally as for the AWA at its annual meetings obedience. Robbie the puppy is in in Rochester, New York. training. When he is not occupied by the Russ and Karen have been , Russ also enjoys married for 27 happy years and live hiking the Gila Wilderness, botany, in beautiful southwestern New and SCUBA diving. He feels that his Mexico. Their home at 6700 feet at occupation (General Surgery) the edge of the Gila National Forest is interferes excessively with his shared with four shelties, a border hobbies! collie, elk, deer, javelina, numerous species of wild birds, and the occasional bobcat, mountain lion, and bear.

166 AWA Review Kleinman et al.

August Link (AJ) was born in Tartu, Estonia, in 1940. After the Russians invaded that country and killed his Warren E. Berbit, pictured here father, he and his mother fled to with his younger son who had just Germany, and then Los Angeles, completed the NYC Marathon in where AJ attended high school and October, 2006, is an avid collector of graduated from UCLA with a B.S.E. keys and telegraph instruments. He AJ worked as a systems engineer for is licensed as a radio amateur Sprague Electric and North American (K2UVV) since 1958, possesses an Rockwell before starting his own Extra Class License, and prefers company, Surcom Associates, in 1970. operating Morse Code. He is a former Surcom Associates is a major supplier electrical engineer, with a BS(EE) of high power RF capacitors. from the City College School of From an early age AJ was Engineering, and an MS(EE) from interested in historical items, and New York University. As an engineer eventually that interest, plus years of Mr. Berbit worked for Sperry and then selling military surplus equipment Bendix on the first airborne digital while attending high school and computer controlled NAV system for college, led to collecting military the military. He flew extensively with radios. This endeavor began with a the system from 1965 to 1968. WWI Signal Corps SCR-68 aircraft Mr. Berbit, having graduated transceiver. His dedicated military with a Juris Doctorate from Fordham equipment collection began with the Law School in 1973, is the principal of addition of a 1917 Navy CF-99 tube- a law firm at which he specializes in type motorboat set manufactured by education and municipal law. He also the de Forest Telephone and helped found, and is Village Attorney Telegraph Company. of the Village of Montebello, is AJ has been collecting for General Counsel to the Clarkstown more than thirty years. Initially he Central School District, and Vice concentrated on American World War President of the Suffern Free Library I vintage equipment and earlier Association. military electronic memorabilia. Mr. Berbit, who has Later, he expanded the scope of the published poetry and articles on collection to include the period antenna design, at one time wrote a between wars, and eventually to monthly column for the NJDXA World War II equipment. His newsletter entitled White Noise. collection contains pieces dating from 1907 to 1945 and includes some representative foreign radios.

Volume 20, 2007 167 Emil J. Simon

168 AWA Review O’Neal A Portal into Radio’s Past: AWA Review Francis A. Hart and his Radio Log ©2007 James E. O’Neal ABSTRACT Locked away in a sturdy steel cabinet in a While trying to sub- storage room at the Smithsonian Institution’s stantiate Reginald A. Museum of American History in Washington, Fessenden’s claim to D.C. is a small notebook. It’s rather drab—noth- having made the ing out of the ordinary. It’s the sort of thing that world’s first radio most people would pass over completely with- broadcast in December, out ever looking inside if they saw it on a yard 1906, the author dis- sale table. The colors are dull tan and red, with covered a radio log a slight amount of gold trim—stationers didn’t from that period. The sell brightly colored notebooks 100 years ago. little-known document, It measures some 6 1/4-inches by 8 1/4-inches, referred to as the “Hart and is approximately 1 1/8-inches thick. The log,” was placed in the covers are scuffed, the edges rounded, and the Smithsonian binding is showing through the endpaper. All Institution’s custody in all, though, it is in remarkable shape for an nearly 40 years ago. object that has passed the century mark and The log is unique in that during its early life was in daily use. (Fig. 1) it was kept for a period The book is in good company. It shares space of more than three with such treasures as portions of Edwin years and offers a great Armstrong’s breadboard FM modulator and amount of insight into prototype superhet, Hazeltine’s first communications dur- neutrodyne receiver, and some of the earliest ing the first decade of Jenkins apparatus. This notebook has radio’s widespread been here since the tenth of February, 1969. It commercial use. A sur- was checked in then and assigned an identifi- prising amount of in- cation number of 329734. There is no name or formation about the log other descriptor on the cover. It’s known unof- and its author, Francis ficially at the Smithsonian as the “Hart Log.” A. Hart, has now been Apparently, few people know of its existence. located. Hart was one I discovered it last year, while seeking out of the first generation information that of young people to em- would substantiate the brace radio, first as a claim that Reginald hobby and later as a Fessenden transmit- profession. Not only ted speech and music does Hart’s log provide on Christmas Eve, a unique window into 1906. Susan Douglas the radio world of 100 briefly mentions the years ago; it also sheds Hart log in her work, light on the youth of Inventing American Fig. 1. The radio log kept Hart’s generation. Broadcasting: 1899- by Francis Hart, now in 1 the Smithsonian collec- 1922, and Elliott tion

Volume 20, 2007 169 Francis Hart Radio Log Sivowitch cites it in his examina- ceremony, as it was sent by U.S. tion of broadcasting’s ‘pre-history,’ mail. “A Technological Survey of On first examination of the Broadcasting’s Pre-History.”2 book’s contents, I assumed (incor- Aside from citations from Douglas, rectly) that the log’s keeper was a Sivowitch and myself, it appears to loner, possibly even a shut-in or have been largely overlooked by invalid who whiled away his hours researchers. by eavesdropping on radio commu- As 2006 marked the one hun- nications between commercial ship dredth anniversary of the date and shore radio stations. Hart used given for the Fessenden transmis- the notebook to log the QSOs or sions, I was rather surprised when radio conversations received at his Smithsonian curator Hal Wallace home radio receiving station. told me that I had been the only one (Technically, QSO is incorrect here, requesting access to the notebook as the “Q” series of abbreviations during that year. I have viewed this did not come into being until 1909. little book on several occasions However, the term is well under- now, each time, signing my name stood and will be used throughout on a “checkout” list, even though this paper in the sake of brevity.) the book does not leave the stor- There are thousands of entries, age room without a carekeeper in covering a period of slightly more tow. It can only be handled with than three years. While most of special gloves and is carefully these loggings are merely an abbre- guarded. What it so special about viated notation indicating that a this document? particular station was in contact with another. Hart was quick to WHO WAS FRANCIS HART? annotate the more interesting It once belonged to Francis QSOs and to add his comments. He Hart, a radio enthusiast from 100 frequently commented about radio years ago who spent a large portion propagation, meteorological condi- of his life in Sayville, N. Y., a com- tions, annoyances and anything munity east of New York City and else that captured his attention. located about halfway toward the The book also provides a descrip- eastern end of Long Island and tion of Hart’s receiving apparatus along its southern coast. Hart’s and his upgrades to the station as name is not one ordinarily associ- the three years passed. ated with the history of radio. The The first 16 pages are devoted to Smithsonian staff has no informa- the identifiers or call signs used by tion about him. Sivowitch, now re- the land and ship stations of the tired, was a museum specialist day. In 1906, there was no FCC, or when the log book first entered the even a Federal Radio Commission. museum 38 years ago, but could A station assigned its own two-let- provide but little additional infor- ter identifier, which may or may mation about its origin. He did re- not have had some significance. As call that the book was donated by an example, Fessenden’s Brant someone in Hart’s family and that Rock station used the identifier the donation was prompted by a “BO,” which may have stood for fairly prominent early radio figure, “Brant on the Ocean” or just “Brant . Sivowitch re- Ocean.” Without these laboriously members that the logbook came entered identifiers to decode Hart’s from Sayville and arrived without notations, the book would be of any sort of fanfare or presentation very limited interest. (Curiously, in

170 AWA Review O’Neal carefully examining the document, more about Hart had hit a dead there is an apparent breakdown in end. the uniqueness of the call system. “RA” is identified early on as a U.S. A DISCOVERY AND A REV- Navy Station at Jupiter Inlet, Fla. ELATION A few pages later it reappears, this However, on re-examining the time in connection with a Marconi log, I discovered something that I’d shipboard station. Still later it sur- not noticed before, a small oval of faces again, now associated with a Kraft paper inserted within blank de Forest land station in Port leaves near the end of the journal. Arthur, Texas.) On first appearance, it appeared to On the surface, the logbook pro- just a scrap torn from a grocery bag vides an intriguing amount of in- that Hart or someone else had used formation about radio and the as a bookmark, but on turning it people who operated radio equip- over, this scrap revealed the name ment during the first decade of its “Burton Hart” and gave a Sayville commercial utility. It also provides address. an unequaled portal for glimpsing After a century, finding some- 100 years back into time. one with firsthand knowledge of These were probably not Francis Hart would be far too much Francis Hart’s intentions as some- to hope for, and I reasoned that this time in the late summer of 1906 person had possibly been a grand- when he walked into a neighbor- son, perhaps a great grandson, hood stationary store and left a few maybe a grand nephew of Hart. cents in exchange for the notebook. Still, there was a chance that some- The one he selected has stiff cov- one in the family, even two or three ers and sewn-in pages. It might generations removed, might be have been intended for use as an able to provide some information, address book, ledger, or diary. The so I located Burton Hart in the tele- sewn-in pages would have made it phone directory and placed a call. valuable for someone keeping I was very much surprised, al- records of experiments for use in a most at a loss for words, when one patent application, as once the in- Burton Hart answered and identi- formation was entered, there fied himself as the son of Francis would be no way to alter it without A. Hart. Burton, or Burt, as he is such changes being obvious. known to his friends, had just cel- Hart’s purchase of the logbook ebrated his eighty-seventh birth- did not mark the beginning of his day, was in very good health and interest in radio. He had just turned was only too glad to provide me 15 in August, and it appears that he with information about his late fa- merely wanted to properly docu- ther! ment his “dxing” activities, some- Burt set the record straight thing that had started sometime about his father’s birthplace and prior to September of 1906. As upbringing. Francis or “Frank” mentioned, the assumption had Hart had been born in Brooklyn, been made that Hart was a lifelong N.Y. and did his growing up there. resident of Sayville. However, ge- The move to Sayville came later. nealogical research turned up no Burton Hart provided me with con- trace of Hart or his family there siderable information about his fa- during the first decade of the twen- ther and the Hart family and then tieth century. At this point, it steered me to another Long Island seemed as if efforts to find out resident who was now in posses-

Volume 20, 2007 171 Francis Hart Radio Log sion of several Hart family photos, precedented amount of invention scrapbooks, newspaper clippings and discovery. There was wireless, and other memorabilia. This was of course, but along with it were George Flanagan, who was to pro- the automobile, controlled and vide much additional information powered flight, motion pictures, x- in my quest to learn more about rays, radium, and the beginning of Hart and his radio log. the nuclear age and modern phys- ics. “SMART BOYS” It’s difficult to know with cer- Francis Arthur Hart had been tainty what ignited this spark in born on August 12, 1891, the same Hart, but yellowed newspaper clip- year that Tesla began his research pings indicate that he and his in the field of radio, and some four friends were all greatly involved in years before Marconi transmitted activities at the new Brooklyn Sci- his first signals at the family’s es- ence Museum (an organization tate in Italy. Hart was part of the that exists to this day).3 The head first generation of young people to of the science department there, grow up with the new science of Mary Day Lee, taught courses on radio and was one of those capti- science and radio, including the vated by it at an early age. copying of code, to all comers. By The newspaper people of his 1906, Francis Hart could copy with time called him and his contempo- the best of them (Fig. 2). raries “smart boys.” Today, we Sometime earlier, Hart had de- would refer to these young people voted one side of his bedroom to as “science geeks” or “techno- his radio and telegraphy hobby, nerds.” Be it 100 years ago, or now, equipping it with apparatus for these are the sorts of individuals both landline telegraphic ex- who, due to their curiosity and all- changes and wireless (Fig. 3). absorbing interests, sometimes al- Doubtless, he crossed the Brook- low the human race to make quan- lyn Bridge into Manhattan and vis- tum leaps in quality of life, com- ited Hugo Gernsback’s Electro fort and health. Importing Company on multiple Hart and several other lads in occasions. Outside the family his Brooklyn neighborhood were home on St. Mark’s Avenue, he these sorts of young people who strung an antenna as high as pos- were discontented with the ordi- sible and installed a proper ground nary. They were the type that system for his equipment. would rather be buried in books, His slightly older friend, Lloyd tinker with homemade machines Espenschied, another “smart boy”, and contraptions, or stay up all had similarly equipped a portion of night listening for spark transmis- his grandfather’s attic for radio sions from ships far out in the At- operations. Espenschied was the lantic. They would rather do these envy of the neighborhood, as he activities than play stickball, go to possessed a small lathe and other neighborhood dances, attend band machine tools that could be used concerts, or spoon with girl friends to transform scraps of metal and in parlors or on unchaperoned bits of wire into radio components. park benches. Rounding out the trio of wired Indeed, there was much to at- and wireless friends was Austin tract a curious young man’s atten- Curtis, who was described by tion. As the new century rolled Espenschied as, “…something of a round, it brought with it an un- naturalist and was quite a genius

172 AWA Review O’Neal in making his own apparatus in his too happy to help her set it up and mother’s kitchen. He was in fact make it operational. Hart volun- quite original and analytical and teered to design and construct the discovered …the zincite crystal that necessary antenna. This took the later formed part of the famous form of a tall mast about 85 feet Perikon detector.”4 above the ground and supporting The boys talked, ate, slept, and some 1,500 feet of wire. Hart, in dreamed radio and electricity, large part fabricated and erected it passing information about these (Figs. 4. 5).5 subjects among themselves at ev- Radio activities weren’t all of a ery opportunity. The central point group nature, however. Hart con- of their existence was the Museum, tinued to hone his code copying and Miss Lee was their mentor. ability and sometime in the late She realized that these youths were summer of 1906, decided to keep obsessed and driven by the com- score. This was his purpose in ob- mon interests and did everything taining the notebook. Beginning on she could to see that their consid- September, Hart made it an almost erable energies were channeled in daily habit to don headphones and the right direction. “listen in” to see what he could Even though space was at a pre- hear. mium in the former private resi- His logbook begins with a de- dence that had been gifted to the scription of station equipment. The borough of Brooklyn for the mu- entry is done in red ink: “Call Let- seum, Lee managed to convince ters and Log Book of Francis her superior to set aside a room for Arthur Hart, using the Fessenden a wireless station. Curtis, system of tuning and Electrolytic Espenschied and Hart were only detector” (Note: for clarity, all log-

Fig. 2. Brooklyn’s “Smart Boys”—Lloyd Espenschied, Austin Curtis and Francis Hart—with their mentor, Mary Day Lee, head of the science department at the Brooklyn Children’s Museum, 1907

Volume 20, 2007 173 Francis Hart Radio Log at Cape Cod, Mass. This is a distance of about 220 statute miles. Before Hart had finished for the day, he recorded a total of 10 con- tacts. For slightly more than three years, this notebook was used on almost a daily basis to record the results of Hart’s experimentation with radio and his moni- toring of the commercial and military radio stations existing then. As he started a new day’s logging, Hart would use a rubber stamp to establish that date within the log’s pages. Hart frequently made notations about radio propagation conditions, the weather and anything he deemed related to his hobby of copying transmissions or Fig. 3. Francis Hart’s radio station, circa 1907. “reading the mail” as some An antenna transformer is seen mounted to the called the sort of legal wall at the left. Ball spark gaps on the trans- eavesdropping he was do- former are connected to an induction coil on the ing. As such, the book that table below. Headphones and both electrolytic and crystal detector stands are seen to the right he created provides a first- of the coil. The calendar indicates that the date hand look at radio in its in- was May 15, 1907. A pocket watch hanging near fancy that apparently does the center fuse panel indicates that the photo not exist anywhere else. (It was made at 2:40 (probably a.m., as there is no must be kept in mind that light entering through the window) when Hart began his en- book entries reproduced here are tries, less than five years had in italics.) passed from the time that Marconi received the fabled three dots in HART’S FIRST ENTRY Newfoundland.) After spending some amount of Hart’s 1906 receiving station time on the initial pages of call sign was constructed around a then data, Hart got down to the business state-of-the-art Fessenden electro- at hand and started to use the book lytic detector. This device con- as he intended. The first entry is sisted of a miniscule platinum wire dated September 1, 1906. The first barely making contact with a small radio transmission he heard is re- receptacle of acid (usually nitric). corded as follows: A polarizing voltage was applied across the detector and high im- “R.B. by P.H. loud” pedance headphones, forming tiny bubbles in the acid around the con- R.B decodes as North Truro tact. In the presence of radio fre- (Mass.). P.H. is a U.S. Navy station quency energy, resistance between 174 AWA Review O’Neal

Fig. 4. The Brooklyn Children’s Museum radio station circa 1907. The station was set up by Frank Hart, Lloyd Espenschied and Austin Curtis.

Fig. 5. The Brooklyn Children’s Museum building. Hart was instrumental in designing the antenna system and installing the 60-foot high antenna mast on the building’s topmost point. The mast supports six 250-foot long conductors.

Volume 20, 2007 175 Francis Hart Radio Log the acid and contact varied, pro- prises had only one option—an in- ducing a signal in the headphones. crease in power. Some stations As resistance varied with RF inten- during this period, especially those sity, the device could demodulate of the navy, did resort to gargan- amplitude modulated signals. It tuan transformers and spark dis- would not be until the following charges. However, many (particu- year that the more sensitive larly shipboard installations) were carborundum and crystal detec- more modestly powered. Hart tors would be placed in general use, seems to have had little trouble thus obsoleting the electrolytic de- reeling in the majority of transmit- tector. ters in all power ranges. There were no electronic ampli- Indeed, during the first month fying devices available when Hart of his logging activity alone, Hart began his logging of signals. The logged some 80 QSOs, hearing sta- invention of the de Forest audion tions as far away as Cape Hatteras, was several months into the future N.C., a distance of some 400 air- and samples would not be available line miles. An undated newspaper until well into 1907. Prevailing an- clipping assumed to have been tenna theory in 1906 was to get as printed around 1907 and pasted in much wire up as high as possible. a family scrapbook says, “There is In essence then, Hart’s receiving not a message from any station station consisted of nothing more within a radius of 1,400 miles that than an antenna and ground sys- he does not get well.”6 tem, an unamplified detector, a tuning system and a pair of high NOT JUST A LISTENER impedance headphones. In light of On first reading and rather cu- the radio developments during the riously, the log seems to indicate past century, we might be tempted that Frank Hart had contented to look upon it as something of himself with only receiving radio schoolboy’s toy. However, in the signals. The equipment station in- autumn of 1906, no one on this ventory preface to the log men- planet really had anything better. tions no transmitting apparatus. In our age of extremely low- Nowhere in the log is there a di- noise receiver front ends, exotic rect mention or a transmitter, nor antennas, noise blankers, com- of Hart’s “working” any of the sta- puter algorithms for pulling infini- tions that he hears. However, tesimally small bits of intelligence newspaper clippings and photo- from overwhelming amounts of graphs have survived that indicate noise, and similar tools for receiv- he was indeed “on the air” as an ing and decoding radio signals, amateur operator.7 It is possible Hart’s log offers wonderful testi- that Hart kept a separate record of mony to what can be done with his QSOs, but no direct evidence nothing more than a simple of this has been found. unamplified detector. On many There is other evidence that occasions, he was able to copy Hart had an operational spark transmissions from well over transmitter. Espenschied men- 1,000 miles away. At the time the tions it, and the surviving photo- log begins, there were really no graph of young Hart’s “shack” improvements to be made to the shows an induction coil connected receiving equipment. If there were to a ball spark gap across the pri- difficulties in attaining reliable re- mary of an antenna transformer ception, commercial radio enter- (Fig. 3). Also, a brief biography of

176 AWA Review O’Neal

Hart published in a company log, there is ample evidence that newsletter states that “In 1905, he Hart constructed and experi- was one of the first three to have a mented with a small arc transmit- transmitting and receiving ama- ter at his Brooklyn “shack”. “Frank teur station in Brooklyn, N.Y”.8 has within the last two months In light of this discovery, the log constructed an electric arc and was re-examined to see if there connected it with his wireless in- might be any clues to transmitting struments, so that he may con- activities. One of these—and very verse eight blocks away with a obscure—surfaces in a December friend. It is an odd sight to see this 27, 1907 entry: boy stand in his bedroom, one wall of which is covered with tele- “I am working to-night graphic and electrical apparatus, about as good as any station on and by playing the blaze from a the coast, and a hundred times Bensen (sic) burner on an arc light better than any private station that he has rigged up, talk with a or boat in and around New friend a quarter of a mile away York City & Brooklyn.” without the use of wires.”9 The ref- erence to the Bunsen burner by the The reference is quite subtle reporter is interesting in that he or and could equally well refer to a she is probably describing Hart’s receiving station. use of a hydrocarbon (illuminating gas) to enhance ionization within RADIOTELEPHONY the arc chamber. This was done in In 1906, brute force spark was much larger arc transmitters, with state-of-the-art transmitter tech- the selection of hydrocarbons nology, but there were alternatives. eventually becoming something of When he began the log, Francis an art. Hart’s knowledge of radio trans- missions consisted of the violent HOW DID IT GET TO THE crashes of sparks, or occasionally, SMITHSONIAN? sparks modulated by a rotary gap Even though this is intended to and possessing slightly musical be a chronicle of Francis Hart and tones. his radio log, it wouldn’t exist had Some experimenters, especially it not been for the intervention of Reginald Fessenden, were trying to the previously mentioned Lloyd convince the radio world to break Espenschied. A word or two about away from the concept of violently him is certainly in order. discharging electricity into an an- Espenchied was the oldest of sev- tenna and to move on to sinusoi- eral young wireless enthusiasts liv- dal carrier wave technology. ing in Brooklyn at the turn of the Fessenden pressured General new century. He and Hart knew Electric for several years to pro- one another, possibly having met duce an AC generator (alternator) in school or through the Children’s that could operate at frequencies Museum. The two remained upwards of 100 KHz. Valdemar friends throughout their lives. Poulsen showed that he could gen- Espenschied was the first of the erate a close approximation to a group of “smart boys” to take a full sine wave by using a high current, time job in radio, becoming a ship’s low voltage arc operating into a wireless officer. (Due to poor aca- series tuned circuit. demic performance and an episode Though not mentioned in his involving the principal of

Volume 20, 2007 177 Francis Hart Radio Log Brooklyn’s Boy’s High School, of Hart’s station, what else can be Espenschied was discharged from gleaned from the log? Static is an that institution. He then enrolled enemy of radio reception and al- in the Brooklyn Manual Training ways has been. It frequently High School, but later dropped out plagued Hart. After making nu- to seek employment in 1907 at the merous entries in the log about age of 18.)10 static conditions, he decided that The Brooklyn “smart boys” he would “automate” this notation knew each other so well that they and had a rubber stamp made up could readily recognize each in late March or early April of 1907 other’s style of sending code, or with the word “STATIC” in capital “fist.” This is captured in the May letters. From then on, he used the 16, 1907 entry in the log. stamp to designate bad radio days Espenschied has taken a job as a and nights. shipboard wireless operator and A number of arcane radio terms Hart has picked him out of the gen- surface in reading the logged eral radio traffic that day. QSOs. One of these is “magnetic storms.” We know this today as “NY by JH - Espenschied intense solar activity. Another went out on her for his trip an term that Hart used quite fre- officer. Was somewhat ner- quently in describing radio receiv- vous, making a few mistakes.” ing conditions was “induction.” (NY is de Forest’s New York “Inductance all a.m.” – June land station; JH is the identifier for 21, 1907 the Maine Steamship Company’s “Horatio Hall.) “Induction and static In 1943, Espenschied, then troublesome all evening.” – working at Bell Labs, had a work- September 12, 1907 place meeting with Austin Curtis and decided to set down some of “Induction so bad I couldn’t his early memories—Hart is men- read Q.I.” – November 26, 1907 tioned several times. Espenschied drafted another edition of his boy- “Induction really terrific to- hood recollections in 1963, lament- night so fierce I couldn’t use ing that of the group of boys from long wave tune at all.” – Janu- Brooklyn, Curtis was now de- ary 3, 1908 ceased and that the other two (he and Hart) were on “borrowed “Induction something ter- time.”11 Espenschied was aware of rific. I had to take off my the existence of the log book kept phones and once it actually by his friend, and after Hart’s pass- made me jump.” – March 9, ing in 1967, pleaded with the fam- 1908 ily to donate it to the Smithsonian. Without this intervention, the The meaning of “induction” as document may have well found its used here only came clear after dis- way to a junk dealer or landfill. cussing these entries with radio historian Alan Douglas. What Hart WHAT’S IN THE LOG? calls “induction” is a reference to Aside from the information on man-made interference—noise in- operational radio stations in the duced into his receiving antenna 1906 time frame and a description from power lines or streetcar

178 AWA Review O’Neal cables. After the arrival of the “Q” operators of the day. signals in 1909, this sort of receiv- ing problem was referred to as “The different kids around QRM or QR Mary. This was “man- here raise an awful noise—all made interference (as opposed to try to talk at once—call when QR Nancy, or naturally occurring anybody is in and never use interference). The man-made ele- any sense; half can’t read 4 ment is driven home with his en- words a minute & sit calling try from March 23, 1908. everybody within 20 miles & can’t hear 800 ft. from another “Could hear him without station that distance. Among any receiver or antenna (prob- the troublesome ones are RT, ably) by induction from mains, LM, GW and FH. Although FH the receiver not being con- is a very good reader he tries nected with the mains at all. He to say too much at one time. uses an electrolytic interrupter Then (?) the poor reader makes and the Edison circuit by which him repeat it and they keep the kick back probably travels that blooming buzz up for back to my house and I heard hours.” – November 23, 1907. the make & break – only the dashes go off in a high buzzing Hart is not referring to himself sound (but?) weak. Dots good. here, despite the identifier “FH.” Very queer business, eh!” (Espenschied’s writings list “call signs” for himself and the rest of It wasn’t all static, magnetic his radio pals. Hart signed “HA;” storms and induction though. The Espenschied used “XY;” and Aus- evening of October 1, 1907 was par- tin Curtis went by “OS.”) ticularly rewarding for Hart—up until a point. He has been chasing “Amateurs are becoming DX all evening and says: worse and worse each day. Ab- solutely impossible to work “One of the best wireless other stations through them nights in a long while; every- except ((long waves))”. - June thing came in v. strong com- 18, 1908 pared to other nights.” “A crazy kid by name of However, this is to change a bit Stokes at Ansonia Hotel, N.Y. later, as an operator at a certain breaks up everybody with 5 Marconi station (on the vessel Kw set. He can neither read “Caruana”) gets out of line: (n)or send; It ought to be stopped.” - November 1, 1908 “A.A. l(ou)d. 9:27 pm—he broke things up by sending ‘v.’ (At the time, the Ansonia was A plague on the Marconi sys- probably New York’s most luxuri- tem. The Marconi operators ous residence hotel. It was home have a (fabulous?) habit of to its owner, W.E.D. Stokes Sr. and jamming other stations when his son, W.E.D. Stokes Jr.—some- business is on the rush list.” times known as Weddy. In 1907, the two Stokes helped to found the Rather hypocritically, the log Junior Wireless Club Limited. In reflects that one of Hart’s greatest 1911 the name of this organization pet peeves is the amateur radio was changed to the Radio Club of

Volume 20, 2007 179 Francis Hart Radio Log America.) curiously, Hart makes no mention Hart was an experimenter, fre- of experimentation with the de quently adding more and better Forest audion. However, equipment to his station from time Espenschied recalled that Hart to time. As previously mentioned, beat everyone in the Brooklyn when he began his log, Fessenden’s group in obtaining a sample of the electrolytic detector technology new device. was “cutting edge.” However, this “Frank Hart was the first to get was to change with the invention one of the bulbs. Through ‘the un- of the crystal and other detection derground’ he learned where they methods. By mid June of 1907, were made, went there and pre- Hart was experimenting with a vailed upon the proprietor to sell carborundum detector. him one. There was some fixing to do in the way of batteries and cir- “NY by JH came in loudest cuits, but Frank got the thing on carborundum in ’Commer- working all right.”12 cial service’ I ever heard…” - In addition to recording techno- June 11, 1907 logical changes, Hart’s log also re- flects changes in wireless practices. “CC v.v. good on short This first occurs in a QSO logged aerial; came in fine on Silicon on September 8, 1907: crystal also” - August 17, 1907 “PT by WR in Continental “New detector ‘Perikon’ code. Don’t know who he is?”

ZNO + CuFeS2 works better than ‘Electrolytic’ for majority A similar entry is made on Sep- of stations; not so good on ‘CC’ tember 18: though. Another new one also,

Hessite = Ag2Te which works “HA by PT in Continental to advantage. Carborundum is code.”

nowhere these days.” - June 18, 1908 Hart’s entries made at the end of 1907 provide a bit more infor- “Using new detector which mation about this shift from Morse seems more sensitive than to Continental code. After the last Perikion. Sulphate of Copper contact of the evening (and the and Zincite. Also Chalcocite year), Hart makes a prominent log and Zincite, the Sulphate “Note” with the following informa- works to better advantage that tion: (than?) either Chalcocite or Copper Pyrites.” – October 30, “The Continental code is to 1908 be used throughout in the navy beginning Jan. 1 & according Not long after the addition of to the International Confer- the carborundum to his shack, ence of wireless Telegraphy Hart first mentions the use of “jig- beginning with the first of ger” tuning at his station. “Jigger” January, 1908.” was Marconi terminology for an antenna transformer. Hart also Hart’s entry for November 22, mentions the use of a Fessenden 1907 sheds light on the times in “Interference Preventer,” or which he lived. wavetrap, from time to time. Very

180 AWA Review O’Neal

“WN doesn’t, as a usual Hart also chronicled news thing, get his juice before 7 pm, events of the time. One of these as the Electric company don’t was the public slaying of New York turn on power until around architect Stanford White by the then; but this was (a) very crazed millionaire Harry Thaw on cloudy & dark morning, so I June 25, 1906. guess this is reason.” “The 2nd trial of Harry K. (In the early part of the century, Thaw is now on for the mur- commercial electric power was not der of Stanford White. This is always delivered on a 24/7 basis. & has been the longest trial in A photo of Espenschied’s 1907 many years.” – January 14, Brooklyn radio station and accom- 1908. panying text appearing in the May 1990 issue of the AWA’s Old RADIOTELEPHONY AND A Timer’s Bulletin illustrates this. A MEETING WITH DE FOREST wall-mounted lighting fixture seen During Hart’s teen years, the in the photo is equipped for both world of radio began to undergo a gas and electricity. This dual func- great deal of change. Inventions in tionality was necessary, as the ‘cen- the art came with increasing rapid- tral (power) station’ typically op- ity. Among these were the first at- erated only during the evening tempts to transmit speech and hours, closing down around 10 music. History shows that p.m. It was noted that the Brook- Reginald Fessenden was probably lyn generating plant also closed the first to experiment with the use down during electrical storms.) of an alternator in generating a si- nusoidal or continuous carrier Every so often, Hart marvels at wave upon which to modulate the a particularly good bit of DX. spoken word. By December 21, 1906, he was able to give a public “PT worked SL with his 2 ½ demonstration of this technology. kw a distance of 2281 miles Unfortunately, Hart’s log offers over land and water—the Gulf no support for the widely made of Mexico; couldn’t tell if I hear claims that Fessenden made a gen- SL or not, heavy snow storm eral broadcast of speech and mu- including static & induction; sic on both Christmas Eve and it’s fierce. December 3, 1907 New Year’s Eve that same year. In fact, Hart does not record his first In light of Fessenden’s Brant reception of speech and music un- Rock work with high frequency al- til March 20, 1907. The entry is of ternators as transmitters operating interest for two reasons. It is the below 100 KHz, Hart’s January 11, first time that Hart has had heard 1908 commentary is particularly anything except code and static, interesting: and it also ties in with a meeting reported by Espenschied that he, “BO loud 11:20 am. I had to Hart, and Curtis had with de For- put in inductance amounting to est and his assistant, John V. L. fully 2000 meters before could Hogan. read him to his full strength. The prelude to the event oc- Could get him much better on curred a few days before—March Fessenden interference 14, 1907. Espenschied had learned preventer than on direct tune.”

Volume 20, 2007 181 Francis Hart Radio Log that de Forest would be speaking ies” in this city Thursday Evening that evening at the Brooklyn Insti- 14th inst and after same walked tute of Arts and Sciences. He quite with you to the Elevated station — naturally decided to attend and in- You requested that I should inform vited his pals Curtis and Hart. As you if at anytime I should hear part of the presentation, de Forest your music and I received it quite passed around one of his new clearly to-day at 5:27 P.M. on my audions. As Espenschied recalled own private wireless system. Hop- the event, “What does stick in my ing to meet you again, I am, memory, in light of subsequent events, was a strange little tube that de Forest displayed, without Very truly yours, operating it, his latest form of wire- less detector, the Audion. He then Francis A. Hart passed it around the audience. It didn’t amount to much, more of a Hart’s log reflected the event: curiosity than otherwise. Little did we realize that it potentially was “Music at 5:27 from De the revolutionary electron ampli- Forest’s; good 3rd time.” – fier, nor did the speaker.”13 March 20, 1907 Espenschied further relates that he and his pals stayed after the In the weeks and months that lecture, wanting to make the ac- followed, the novelty of hearing quaintance of de Forest and wireless telephony began to wear Hogan. The three “smart boys” off and became something of a walked with the inventor of the audion and his assistant to the Fulton Street “elevated.” Along the way de Forest undoubtedly spoke of his latest interest—wireless te- lephony. He had constructed an arc transmitter a few months be- fore and was experimenting with it. Espenschied relates that he, Hart and Austin began to listen for the station, each sending de For- est a reception report when they logged it. These reports were saved in a scrapbook by Hogan, who, many years later showed them to Espenschied. Hart’s was hand written and dated March 20, 1907. The text is reproduced here (see also Fig. 6):

Dr. Lee De Forest, Fig. 6. Letter dated March 20, 1907 Dear Doctor from Francis Hart (age 15) to Dr. Lee de Forest acknowledging de Forest’s I had the pleasure of listening radiotelephone transmissions of speech and music that day. The event to your lecture at the “Art Galler- is recorded in Hart’s radio log.

182 AWA Review O’Neal nuisance factor in Hart’s world of conditions, current events, tech- wireless. nology or bad operators, occasion- ally editorializes in his log. This “De Forest’s blooming tele- excerpt from January 3, 1909 is phony buzz raised the deuce one of his longer commentaries. with the L.W. & everything else.” – May 9, 1907 “SL v. l(ou)d, been breaking everybody for last three nights, By 1908 radiotelephony had while lying at dock—sending ‘v’ moved forward a great deal. On and general nonsense, which is February 11 of that year, Hart cop- not generally done by the ies a very long speech QSO involv- Marconi people in the harbor; ing Fessenden’s Brant Rock Sta- the Gov’t should put a stop to it tion. under two arguments that (1) It keeps others from doing paid “1:16 pm wireless phone at business, (2) The English law Jamaica (N.Y.) & other must be prevents any other company at Brant Rock, Mass.; phone (Marconi Co. excepted) from very clear except for a rasping operating within (3) English noise that mingles with the miles of an English coast, and voice and is hard to B(?). I under those conditions why managed to get the following should we allow them to break and probably could have ob- us up and refuse to do any con- tained more except for ‘9’ and versing with out Gov’t ships, etc.: ‘How’s that now?’ ‘You etc. when we are treated in that came in louder than that yes- absolutely pig headed English terday’ Could hear music as way over there.” plain as voice from weaker sta- tion but couldn’t make out (SL is the St. Louis, a ship words from other station al- equipped with a Marconi wireless though they came in fair.” operation.) This is the last entry in the log Hart copies enough of this ex- for that month. Logging does not change between Fessenden work- resume until February 14. This gap ers to fill a page and a half of his is explained by a newspaper clip- log. ping preserved in one of Hart’s scrapbooks and dated January 14, During the opening months of 1909. It describes the sailing of the 1908, wireless telephony transmis- S.S. Prince George, a Bermuda-At- sions become more and more lantic Co. steamship. This was the prevalent, but the technology is not maiden voyage of that company’s quite ready to replace the tele- regular service between several phone just yet. U.S. ports and Bermuda. Hart an- notated the clipping, “Made first “Wireless fone fair all day. trip on this boat.” A gurgling sound produced by Hart had signed on as a seago- the fone sounds like water run- ing wireless operator, and in doing ning at times.” – March 2, 1908 so embarked on a professional ca- reer in radio. (His resume cites EDITORIALS employment with the United Wire- Hart, when not commenting less Telegraph Co, as “operator on on radio receiving and propagation various vessels” in that time frame.

Volume 20, 2007 183 Francis Hart Radio Log Hart’s log entries again become ciple itself. In this case, Hart’s work sporadic with the coming of the really amounts to using tuned cir- 1909 summer, stopping com- cuitry to reduce or eliminate radio pletely with the July 11 entry and interference. The patent was ap- not resuming until September 18. plied for on May 6, 1910 (when It is to be presumed that he was Hart was 18) and was granted on serving as a ship wireless operator February 21, 1911. There is no in- during this period. dication that he made any money from this invention. END OF THE LOG Hart tested for his commercial By the autumn of 1909, and radio operator’s license on June three years of entries, Hart’s inter- 20, 1911. The test was adminis- est in keeping his log was waning. tered at the Brooklyn Navy Yard. There are fairly complete entries The document issued shows that for September of that year; how- Hart was tested on a “combina- ever there is nothing for October 1 tion” of apparatus and his knowl- (Friday). The October 2 entry is edge of systems and regulations indicative of Hart’s usual level of was found to be “excellent” (Fig. 7). interest (including mention of a Between 1910 and 1911 he magnetic storm). On October 3, he worked for Lee de Forest, as op- logs even more traffic. But after erator in charge of the De Forest these entries, there are no more. Radio Telegraph and Telephone The log ends rather tersely with Company’s Metropolitan Tower these words: station. Hart’s resume also men- tions working with F. A. Kolster in “No more records kept.” testing operations while at de For- est. By now, Hart was18 years old, Between 1911 and 1917, Hart had completed school, and had was employed by the American entered the workplace. In light of Marconi Company, serving as as- this, it is understandable that he sistant to the chief engineer, chief chose this time to end his logging inspector and assistant superin- activities. According to Hart’s re- tendent of construction (Fig. 8). sume, he worked until 1910 as an During this period, he found time operator for the United Wireless to co-author a book, List of Radio Stations of the World, with H.M. Telegraph Company, at both the 14 company’s 42 Broadway location Short in 1915. Short is identified in Manhattan and also on sailing in an advertisement for the book vessels. as a resident inspector for the (U.S.A.) Marconi International Marine Communication Com- AFTER THE LOG pany, Ltd. What became of Frank Hart af- Hart became a charter member ter ended his log and became a of the Institute of Radio Engineers fulltime ship’s radio officer? in 1912. His membership number Records show that he was not idle. was 31. (Hart’s membership status A patent search disclosed Hart’s was that of “associate member,” as invention of an “Apparatus for Re- he had no college degree.) The ceiving Electromagnetic Waves” June 1914 issue of Wireless Age (U.S. patent 984,762). As with a lot featured a story on radiotelephony of patents, the name is not really testing being carried out by Hart— descriptive of the patented prin- “New York to Philadelphia by

184 AWA Review O’Neal

Wireless.”15 One of Hart’s associ- ates during his years with Marconi was none other than David Sarnoff. George Flanagan says that Hart initially handled traffic for Marconi at the company’s first New York area station at Sagaponack on Long Island. He later moved to the Sea Gate station at Coney Island. At the time, Sarnoff was managing the Marconi station at Wanamaker’s Department Store in Manhattan. Flanagan says that while he has no hard evidence, he suspects that Hart and Sarnoff must have worked together some- where during this period, as Hart gave Sarnoff’s name as a reference on an employment application in 1919.16 Hart’s son, Burt, acknowledges that the two men were fairly well acquainted, but said that his father had no use for Sarnoff, due to the Fig. 7. Francis Hart’s commercial ra- latter’s inflated claims about work- dio license issued in 1911 by the De- ing distress traffic at the time of the partment of Labor and Commerce. Titanic’s sinking. The younger Hart Hart’s level of proficiency was rated said that his dad claimed to have as “excellent.”

Fig. 8. Wireless message copied by Francis Hart during his employment by American Marconi. The date is August 4, 1914 and the message reads “En- gland has declared war to Germany.” Volume 20, 2007 185 Francis Hart Radio Log been equally or more involved in Sayville radio station, but in the message handling in connection capacity of a civil engineer. He re- with the doomed ship and its af- vealed that operations there began termath.17 to be curtailed there after WWII, During the war years (1917- with his father eventually being 1919) Hart worked as a radio in- reassigned to New York City. spector and expert radio aide for Rather than relocate or face a long the U.S. Navy’s Bureau of Engi- daily commute, Frank Hart elected neering in Washington, D.C. It was to leave the company. here that he met his bride to be, He found employment for a pe- Mary Gorman. The two were mar- riod at nearby Islip Transformer (a ried in Washington, D.C. in 1918, subsidiary of Islip Radio Manufac- with James Edward Parker, an- turing Corporation). However this other of the Brooklyn “smart boys” business went bankrupt around serving as best man.18 1948 and Hart found himself out Following WWI, Hart served as of work. It was at this point that superintendent of construction Hart severed ties with radio com- and operating manager for the In- pletely and never looked back. dependent Wireless Telegraph There is no indication that he ever Company, building stations in New pursued radio again, even as a York and Texas. He left Indepen- hobby. After the transformer dent Wireless in 1925 to join Com- company’s failure, Hart found em- mercial Wireless, Inc., which had ployment as an auditor with the leased from the U.S. Navy a radio Sayville Federal Savings and Loan station originally built by Association (later Eastern Federal Telefunken in 1911-12. (The station Savings and Loan Association) in was seized by the U.S. government Sayville. He remained there for the with the beginning of WWI hostili- next 17 years. ties.) The station was located near Sayville, N.Y. and this career move EPILOGUE brought Hart to that city. He was Francis Arthur Hart died of a to remain in Sayville for the re- heart attack on October 12, 1967 mainder of his career and life. at the age of 76. Hart’s son noted To keep the record complete, that both his father and grandfa- Hart’s employer at this point, ther died at the same age. Burt Commercial Wireless, was ac- Hart remembers his father as an quired by Mackay Radio and Tele- active individual and a person who graph, which later became IT&T enjoyed playing tennis, but also as World Communications. Hart was someone possessing a “short fuse.” manager of operations and con- Also, according to Burt, Frank Hart struction (officer-in-charge) at the didn’t have much faith in doctors Sayville station for a long period. and rarely, if ever, saw one. “My He and his family occupied a home dad believed that there was noth- on the station grounds from about ing that couldn’t be cured by qui- 1928 to 1936, later relocating to a nine,” the younger Hart observed.19 private home in Sayville. Lloyd Espenschied, the oldest and last surviving member of the END OF RADIO CAREER Brooklyn “smart boys” outlived his Hart’s son, Burt, was not par- friend Frank by almost 10 years, ticularly interested in wireless and dying in June of 1986 at the age of did not follow in his father’s foot- 97. steps. In time, he did work at the

186 AWA Review O’Neal

REFERENCES 1. Susan Douglas “Inventing American Broadcasting 1899-1922,” Baltimore, MD: The Johns Hopkins University Press, 1987, pp. 208-209. 2. Elliot Sivowitch “A Technological Survey of Broadcasting’s ‘Pre- History,’ 1876-1920; Journal of Broadcasting, Vol. XV, No.1 (Winter 1970-1971, pg. 1-20). 3. “Tap Wireless Messages – Brooklyn Boys Have Unique Apparatus of Their Own” – undated newspaper clipping from Hart family scrapbook. The newspaper is not identified, but is likely the Brooklyn Daily Eagle. Hart is described as being 16 years old, so the story would have been published in 1907 or 1908. George Flanagan, who is now in possession of the Hart scrapbooks notes that the first clipping kept was dated September 27, 1907 with a rubber Fig. 9. Francis Hart and his wife Mary. stamp (probably the one Hart used This photo is undated, but was prob- for his radio log). Flanagan also ably taken in the early 1950s. notes that some of the scrapbook clippings had been removed by tearing or cutting them out, AUTHOR’S NOTE: destroying other information in the process. As Francis Hart’s log was an in- 4. Lloyd Espenschied – unpublished formal document—much like a di- manuscript dated July 9, 1943. ary or classroom notes—and not Hart family copy, now in George meant for anyone but himself, his Flanagan collection. sentence construction was not al- 5. Newspaper clipping from the April ways perfect. Hart frequently ab- 19, 1908 Brooklyn Sunday Eagle. breviated words and omitted punc- F. Hart annotated the clipping with tuation. Some words are illegible. the following information: “Picture In transcribing excerpts of the log of new antenna on museum for this paper, I have tried to be as building. Composed of 6 wires 1 1/ 2 –ft. apart and 250 ft. long, abt. faithful to the original document as 85 ft. from ground; erected by FAH possible, but have taken the liberty & others during 1907-08. The from time to time of adding small antenna was designed by FAH and amounts of punctuation to make erected under his direction.” these excerpts more readable and 6. “Boy Holds Key to Wireless” – understandable. In some cases ab- undated newspaper clipping from breviations have been spelled out Hart family scrapbook. It is and in others, extraneous words undated, but is probably from early and phrases were deleted for com- 1907, as it mentions Lee de Forest’s pactness. In every case, I have tried experiments in wireless telephony which began in the first months of to be as true as possible to the idea that year. The article also mentions or ideas that Hart attempted to set the date December 19. As Hart’s down. Volume 20, 2007 187 Francis Hart Radio Log birthday was August 12, this 3. Elliot Sivowitch, retired further strengthens the case for Smithsonian Electricity Collections publication in early 1907. The curator - for sharing his knowledge newspaper is not identified, but is of early 20th century wireless probably the Brooklyn Eagle. technology and for reviewing and 7. Ibid. critiquing this manuscript 8. Hart scrapbook magazine clipping; 4. George Flanagan – for providing identified as from the IT&T News invaluable background – June, 1930 information on Francis Hart and 9. “Boy Holds Key to Wireless” – for the use of photographs and undated newspaper clipping cited other material reproduced here above. from his extensive collection. 10. Lloyd Espenschied – manuscript 5. Alan Douglas, radio historian – for dated July 9, 1943. Espenschied supplying me with answers to my has titled this “Radio Amateur Days many questions concerning in Brooklyn.” Hart family copy, wireless technology and now in George Flanagan collection. terminology 11. Lloyd Espenschied – manuscript 6. Jane Johnson, reference librarian dated March 3, 1963; an edited Charlotte, N.C. Public Library – for version of this was published in assistance in tracing the Hart May 1990 The Old Timer’s Bulletin; family vol. 31, no. 1, p. 12 – “An Early 7. Jim Haynes, retired educator and Chapter In Radio-Electronics.” electrical engineer – for reviewing 12. Ibid. this manuscript 13 . Ibid. 14. Frank A. Hart and H.M. Short, “List of Radio Stations of the ABOUT THE AUTHOR World,” New York, NY: Marconi Publishing Corporation, 1915. 15. Information provided by G. James E. O’Neal, AG4DH, is Flanagan in e-mail correspondence technology editor at Television from March 18, 2007. Technology. He’s held this position 16. George Flanagan – e-mail to since late August of 2005, three weeks author dated March 20, 2007. after retiring from the Voice of 17. Telephone conversation with America’s television service. Mr. Burton Hart, March 12, 2007. O’Neal had been with VOA Television 18 . Undated newspaper clipping from and its precursors, WORLDNET Hart scrap book. George Flanagan Television and United States provided the information about Information Agency Television and Parker in e-mail correspondence Film Service for more than 30 years. with the author dated March 18, He is a graduate of the University of 2007. Arkansas at Fayetteville. 19. Burton Hart, telephone Mr. O’Neal’s career in conversation cited above. broadcasting goes back to the 1960s and his teenage years, when he was first employed as a weekend operator This article was peer-reviewed. at the radio station in his hometown of Hope, Arkansas. Mr. O’Neal was ACKNOWLEDGEMENTS: bitten by the radio bug early in life and 1. Electricity Collections – had previously constructed and Smithsonian Institute National operated a pirate radio station. The Museum of American History, for manager of the local station, KXAR, access to the Hart Log visited his facility and within a week 2. Hal Wallace, Smithsonian issued an invitation to come to work Electricity Collections curator - for for the “real” station. It was sometime his patience and special assistance before the reason for his hire surfaced. with this project The manager was later quoted as saying, “I was afraid that if I didn’t hire

188 AWA Review O’Neal him, he would go into competition with me and start selling spots.” Mr. O’Neal’s part-time career in broadcasting continued throughout his high school years, and on into his college years. He transitioned to television during his senior year at college, being an early hire at the city’s first television station, which went on the air in early 1969. He spent five years in New Haven, Conn., working at WNHC-TV/ WTNH-TV. During his government service he traveled extensively. One of his trips to Moscow resulted in the discovery of a 1930’s RCA television receiver labeled in Cyrillic. His account of this find and the story behind it was published in Antique Radio Classified. He has also published articles in the IEEE Broadcasting Technology Society Newsletter, Satellite World, Radio World and other periodicals. His writings frequently appear in Television Technology. He is a self- described inveterate collector of broadcasting equipment and related memorabilia. He has been a member of the Antique Wireless Association since 1989 and is also a member of the IEEE, the Society of Motion Picture and Television Engineers and is a life member of the Society of Broadcast Engineers. He has presented papers on early radio and television at SMPTE and National Association of Broadcasters’ conferences. He holds both an FCC commercial operator’s license and extra class radio amateur license.

Volume 20, 2007 189 Francis Hart Radio Log

190 AWA Review Bart & Bart

AWA Review Exploring the Origins of the Loud Speaker

©2007 David and Julia Bart

ABSTRACT INTRODUCTION The invention and The first speakers designed specifically for subsequent mass pro- radio resulted from nearly 75 years of research duction of the loud into the reproduction and projection of sound. speaker after World Successful radio applications stemmed from War One made radio earlier work in the telegraph, telephone, pho- commercially viable for nograph, public address, motion picture and the general public. The radio industries, representing the culmination loud speaker offered a of highly varied research and inventing efforts dramatic improvement over many years. The earliest research into elec- over existing head tronic reproduction of sound dates from 1837. phone technologies by Commercial development of practical electro- providing a sound re- magnetic reproduction of voice and other production device ca- sounds began in 1876 within the telephone and pable of projecting the phonograph industries. Ideas for a device spe- radio broadcast to an cifically capable of ‘loud speaker’ type projec- audience. Conceptual tion can be traced to Siemens and Lodge as early origins for electronic as 1874. However, their concepts lacked any sound reproduction method for producing the necessary electrical trace their roots to early amplification. Only after Pridham and Jensen telegraph applications. combined newly available sound amplification As new methods for technology with their improved designs for sound reproduction speaker drivers did the loud speaker emerge as emerged, entrepre- a viable device after 1915. Both Western Elec- neurs and manufactur- tric and Magnavox pursued large scale commer- ers incorporated more cial production following World War One. The sophisticated ap- subsequent radio boom of the 1920s witnessed proaches; first develop- an explosion in the sale of speakers to consum- ing telephone receivers, ers as many companies raced to grab a share of then horn reproducers the newly created market. The public would see and finally direct radia- loudspeakers quickly convert from horn pro- tor cone speakers. This jection to direct-radiator cone technology in less article reviews the pri- than 10 years. However, continued use of horns mary inventions and in public address applications provided the im- patents underlying the petus for further development of horn projec- growth of loud speaker tion technology well into the 1950s. This ar- technology. The discus- ticle explores the early history of speakers to sion encompasses the clarify the primary historical developments that years 1837 to approxi- made the technology of sound reproduction mately 1960. possible. EARLY TELEGRAPH AND TELEPHONE RESEARCH As early as 1837, Dr. C.G. Page in Salem,

Volume 20, 2007 191 Loud Speaker Massachusetts recognized that ting needle” mounted on a double sounds, called “Galvanic Music”, resonating wooden box with a would emanate from an electro- wooden lid. His transmitter incor- magnet when the magnetic fields porated a round membrane made were electrically changed. (Page, of sausage skin as a diaphragm 1837 a,b,c; Prescott, 1879; Thomp- which was attached to a platinum son, 1883; Munro, 1891; Hunt, strip resembling the “hammer” of 1954; Fagen, 1975; Post, 1976; Coe, the human inner ear. The hammer 1995; Catania, 2004). In 1854, transferred the sound vibrations by of France con- serving as an interrupter for the ceived the basic method for trans- electric circuit. Unfortunately, his forming acoustic vibrations into non-metallic diaphragms, or mem- electrical signals and then back branes, failed under varying tem- again using a moveable diaphragm perature and humidity conditions, laid across an electro-magnet; ef- and his loosely mounted electro- fectively creating the first ‘trans- magnet receiver gave uneven re- ducer’ to convert acoustical/electri- sults. The device ultimately failed cal energy forms. Bourseul even in several courtroom proceedings. defined his concept as “Electrical Reis died in 1874, never receiving Telephony” and prophesied that acknowledgement for his discover- human speech would one day be ies. (Reis, 1862; Prescott, 1879; carried by electricity. Unfortu- Thompson, 1883; Munro, 1891; nately, Bourseul’s research focused Hunt, 1954; Coe, 1995; Baker, on making and breaking the elec- 2000; Catania, 2004 and 2006). trical pulse with sound vibrations. Cromwell Varley and Dr. Ernst Further, Bourseul never attempted W. Siemens also made contribu- practical application of his ideas. tions that remained undeveloped. (Bourseul, 1854, 1857; Van der In 1870, Varley discovered that Weyde, 1869; Prescott, 1879; sound could be emitted by a con- Munro, 1891; Hunt, 1954; Coe, denser when its charge varied. 1995; Catania, 2004). (Prescott, 1879; Munro, 1891; In 1861, Fagen, 1975). Four years later, in applied a continuous electrical cur- 1874, Siemens filed his U. S. Patent rent within the circuit. Reis, from No. 149,797 for a “magneto-electric the Garnier’s Institute in apparatus” on Jan. 20, 1874 which Friedrichsdorf, Germany, con- was granted on April 14, 1874. He ducted his initial research from was the first to describe the “dy- 1860-1861 on the electrical trans- namic” or moving-coil transducer, mission of speech, and published using a supported circular coil of his first paper in 1861 entitled “On wire in a magnetic field capable of Telephony By The Galvanic Cur- moving axially to convert electrical rent”. Reis demonstrated up to energy into acoustic energy. twelve versions of his new device Though Siemens’ patent states that from 1861-1867, and a number of his invention could be used for Reis telephones were produced and “moving visible or audible signals”, shipped worldwide. Reiss is also he never developed it for commer- credited with first using the word cial application. (Hunt, 1954; ‘telephone’ to mean a device ca- Fagen, 1975; Schoenherr, 2001; pable of transmitting human Eargle and Gander, 2004). speech. Reis created his electro- Two years later, on February 14, magnetic receiver by winding a coil 1876, Alexander Graham Bell ap- of wire around an iron rod or “knit- plied for a U.S. Patent on his “Im-

192 AWA Review Bart & Bart provement In Telegraphy”, known Gray’s electro-magnetic receiver today as Bell’s first telephone formed part of his U.S. Patent Nos. patent. Bell arrived at the Patent 166,095 and 166,096, both issued Office only a few hours before on July 27, 1875, and U.S. Patent Elisha Gray who came to file a ca- No. 175,971 issued on April 11, 1876 veat, or a preliminary notice of in- (see Fig. 2). vention, under the accepted notifi- cation practices of the period. The caveat stated that Gray intended to file his complete patent application at a later date. Ironically, Gray had already made and exhibited four different working telephone re- ceiver models by 1876 stemming from his experiments over the pre- ceding two years with the musical telegraph. Gray utilized electro- magnetic coils and various dia- phragms together with tin can and wooden sound boxes as well as wash basin reflectors to concen- trate the tones (see Fig. 1). Gray effectively developed the first tone Fig. 2. Figure From 1875 Gray U.S. oscillator and combined it with a Patent No. 166,096. (Source: U.S. parabolic acoustic sound radiator. Patent Office) (Baker, 2000; Catania, 2006). Though the Bell and Gray tele- phone devices of February 1876 demonstrated further evolution and were remarkably similar in concept, decades of subsequent liti- gation held that by law Bell right- fully owned the telephone patent, allegedly the most valuable patent in history, which was granted as U.S. Patent No. 174,465 on March 7, 1876. The famous transmission “Watson, Come Here, I Want You” was later uttered in demonstration tests on March 10, 1876. (Hunt, 1954; Coe, 1995; Baker, 2000). Bell’s U.S. Patent No. 174,465 described “employment of a [con- tinuous] vibratory or undulatory current of electricity in contradis- tinction to a merely intermittent or pulsatory current”. He specified use of a “cone” and referred to the words “oscillation”, “vibration” and “undulation” synonymously. He Fig. 1. Gray’s 1874 Wash Basin included his method for “gradually Receiver.(Source: Prescott, 1879) increasing or diminishing the resis-

Volume 20, 2007 193 Loud Speaker tance of the circuit”; carefully claiming the method for “transmitting vocal or other sounds.” Bell also named brass, copper and other metals as “bodies capable of in- ductive action”. In these early years, the transmis- sion and reception of sound were made on the same device. Bell’s cones Fig. 3. Figure From 1876 Bell U.S. Patent No. collected sound (wide 174,465. (Source: U.S. Patent Office) end spoken into), and also concen- trated the reproduction of sound from the diaphragm (using the same device with the narrow end placed against the ear; see Fig. 3). Bell followed his 1876 patent with another application filed on January 15, 1877, Bell’s second telephone patent, which was awarded on January 30, 1877 as U.S. Patent No. 186,787 for an “Im- provement In Electric Telegra- phy”. The 1877 patent focused on improvements to the transmitter/ receiver. This patent clearly de- scribes use of a “permanent mag- net with a plate of iron or steel or other material capable of inductive action, with coils upon the end or ends of said magnet nearest the plate”. It also describes use of a “sounding box” and “the employ- ment of a speaking or hearing tube, for conveying sounds to or from the telephone” (Fig. 4). After Bell’s patents were granted, Dr. Ernst W. Siemens Fig. 4. Figures From 1877 Bell U.S. filed German Patent No. 2355 on Patent No. 186,767. (Source: U.S. December 14, 1877. The German Patent Office) Patent specified a nonmagnetic necessary to increase the signal parchment diaphragm using a bal- enough for practical use. Here, the anced armature moving-coil as the diaphragm took the form of a cone, transducer. The electro-magnetic with an exponentially flaring coil connected with a radiating sur- “morning glory” trumpet form. face described as the frustum of a This is the first patent for a loud- cone. Effectively, Siemens had in- speaker type horn reproducer that vented the initial concept for the would later be used on most pho- direct-radiator cone speaker; years nographs and cylinder players in ahead of any amplification device the acoustic recording era. The

194 AWA Review Bart & Bart

German Patent was granted July Company in 1869. Western Union 30, 1878, and a second British later acquired the company and it Patent No. 4685 was granted Feb- was subsequently sold to American ruary 1, 1878. (Hunt, 1954; Fagen, Bell Telephone. In 1899, AT&T 1975; Schoenherr, 2001; Eargle took over both American Bell and and Gander, 2004). Ironically, Western Electric. (Adams and But- Charles Cuttriss and Jerome ler, 1999). Under AT&T, Western Redding of Boston independently Electric became the primary sup- invented a similar device that same plier for the Bell telephone com- year and filed their own U.S. patent panies and provided research and application in November 28, 1877. experimentation in both telephone Their device also employed a mov- and the new radio technologies. ing-coil type transducer for use in a telephone circuit. (Hunt, 1954; Beranek, 1954). By 1881, the telephone receiver had essentially acquired its defini- tive form and little change would Fig. 5. Telephone Receiver Designs occur for the next 50 years (Hunt, Circa 1895. (Source: Fagen, 1975) 1954; Fagen, 1975). In keeping By 1907, new management re- with its function, designs empha- structured and merged most of sized the sensitivity of the mag- AT&T’s remaining engineering netic driver over any acoustical staff into Western Electric. (Dou- gain that might be obtained from glas, 1991; Adams and Butler, use of a horn. (Fig. 5).Broader con- 1999). Western Electric’s re- cepts for other possible telephone vamped research projects included applications remained a novelty. developing sound projection sys- In April 1909, Modern Electrics tems for use as loud speaking de- reported demonstrations in Berlin vices in place of telephone receiv- where a telephone receiver was ers. In 1907, Western Electric pro- placed beneath a horn to make a duced its first horn speaker, a loud-talking telephone. The Model 1C “Howler”. The telephone “Lautsprecher” reportedly industry widely purchased the achieved some success in Germany Howler as the first commercially in that year. (Paul, 1989b). But it produced electrically driven horn would take new applications to projection sound device. (Western stimulate a search for ways to gen- Electric, 1908; see Fig. 6). erate enough sound to fill a room The Howler, only 6 1/4” x 6” x or even project outdoors. 4”, hung suspended above a tele- phone with the horn projecting WESTERN ELECTRIC’S horizontally. The Howler simply HORN SPEAKER CONTRI- used the horn to amplify the in- BUTIONS coming telephone signals. Al- By the dawn of the 20th century, though successfully employed in Western Electric Company oper- the telephone industry, it never ated under the control of Ameri- saw radio related applications. In- can Telephone and Telegraph stead, Howlers hung in noisy loca- (AT&T), the parent company of tions instead of a bell or ringer to American Bell Telephone. West- signal the incoming call. The horns ern Electric traced its origins to operated on either A.C. or D.C. cir- Elisha Gray and Enos Barton who cuits and used an electro-magneti- founded the Gray and Barton cally controlled vibrator to strike Volume 20, 2007 195 Loud Speaker a thin metal diaphragm at the base Navy. The CW929 employed both of the horn. Eventually, some a detachable horn and coupled re- Howlers simply used small electric ceiver mountings. The telephone motors to rotate a notched wheel receiver itself resided in a water against a hard metal cam on the tight cast iron box. The CW929 diaphragm causing it to “howl” utilized a tube driven audio fre- loudly. (Coyne Reference Guide, quency amplifier to increase the 1943). original signal strength of the tele-

Fig. 6. Advertisement For Western Electric Howler. (Source: Western Electric, 1908) Western Electric later applied phone receiver for loud speaking knowledge gained from developing sound projection. (Robinson, 1919; the Howler in its new “Loud Paul, 1981). Thus, Western Elec- Speaker”, first shown at the Bos- tric and Bell Laboratories became ton Electric Show in September one of the first to successfully com- and October 1912. “Combining the bine electronic amplification with articulating qualities of a telephone acoustical sound projection tech- receiver and the sound intensify- nologies (Fig. 9). ing qualities of a megaphone,” the Western Loud Speaker served as an “an- Electric did nouncer” for paging visitors, not enter the broadcasting a baseball game and radio market finding the parents of a lost child. until the early (Telephone Engineer, 1912). The 1920s. Strict new Loud Speaker operated as the interpretation first commercially successful of three-way sound projection system, but did cross-licens- not serve radio applications (Figs. ing agree- 7,8). ments be- By World War One, Western tween Gen- Electric produced its first balanced eral Electric armature horn speakers for US Company, Navy submarine chasers. (Paul, Radio Corpo- 1981). Bell Laboratories collabo- ration of rated with Western Electric to fur- America and ther develop the horn speaker con- American cept, focusing primarily on public Telephone address type systems. Finally, in and Tele- 1919, Bell Laboratories produced graph (includ- their own public address horn ing Western speaker, Model CW929. (Fagen, 1975). The CW929 formed part of Fig. 7. Western Electric Loud Speaker at the Boston Electric Show in Sep- the CW936 submarine chaser ra- tember and October 1912. (Source: dio set first developed for the U.S. Telephone Engineer, 1912)

196 AWA Review Bart & Bart

commercial success selling loudspeakers for radio appli- cations came in June 1922 with the Western Electric Loud Speaking Telephone Outfit Model 10-A. (Douglas, 1991). The complete outfit in- cluded a 30” tall Model 518- W balanced armature loud- speaker incorporating a coil and magnet assembly driven by a separate Model 7-A am- plifier. (Western Electric, 1923). The company also produced a Model 10-D Fig. 8. Western Electric Loud Speaker Loudspeaker in 1922 which again at the Boston Electric Show. was identical to the 518-W but in- Electric) dating from 1920 prohib- cluded an internally mounted ited Western Electric from selling transformer to allow direct con- loudspeakers outside of the tele- nection to the radio’s output tubes. phone industry. (Archer, 1939; (Paul, 1981 and 1989). In 1923, Hunt, 1954; Douglas, 1991). Nev- Western Electric brought out a ertheless, Western Electric’s first new Model 521-W Loud Speaking Telephone Receiver featuring a smaller bi-polar headphone driver assembly and goose-neck shaped horn. (Paul, 1981 and 1989; Figs. 10-12). The great popularity of the Western Electric’s loudspeakers emerged as one of many contro- versies which became subject first to attempted mediation and then to arbitration between RCA, AT&T (Western Electric) and General

Fig. 9. Bell Laboratories CW929 Speaker Circa 1919.(Source: Fig. 10.Western Electric 10-A Speaker Schoenherr, 2001). Outfit with Model 518W Speaker, 1923

Volume 20, 2007 197 Loud Speaker (Archer, 1939; Douglas, 1991). Meanwhile, by late 1924, Western Electric brought out its new “horn- less” direct-radiator double-cone speaker, Model 540-AW, as dis- cussed below. (Radio Doings, 1924; Fagen, 1975). This new Western Electric cone speaker quickly became the dominant speaker technology for radio. Yet, the home entertainment portion of Western Electric’s busi- ness remained small. (Paul, 1981). Following a 1922 decision by Ed- ward Craft, Western Electric’s re- search administrator, Western Electric focused their research and Fig. 11.Western Electric 518-W design efforts on the motion pic- Speaker Driver (Source: Western ture industry, and most of West- Electric, 1923) ern Electric’s future speaker devel- Electric under the 1920 cross-li- opment centered on public address censing agreements. Under David systems for auditoriums, broad- Sarnoff’s leadership, RCA engaged cast studios and early sound mov- in litigation to block production of ies. (Adams and Butler, 1999). wireless components made by the other companies, including loud- PRIDHAM, JENSEN AND speakers. Legal maneuverings MAGNAVOX over the next three years eventu- The Magnavox story began in ally left Western Electric with the 1910 when Peter Jensen, Edwin authority to manufacture not only Pridham and Richard O’Connor, a loudspeakers but also broadcast financier, established a laboratory transmitters and audio equipment. for wireless experimentation near

Fig. 12. Receiver Coil and Magnet System with 518-W Speaker Circa 1923. (Source: Western Electric, 1923)

198 AWA Review Bart & Bart

Napa, California. Together, they work of Siemens and others. formed the Commercial Wireless (Lewis, 1980; Sanders, 1988; Development Company which for- Futtrup, 2004). mally organized on March 6, 1911. Undeterred, Pridham and Jensen, formerly Valdemar Jensen continued to experiment Poulsen’s assistant, and Pridham, with sound reproduction. Popular an electrical engineer, had resigned myth holds that during a telephone from Federal Telegraph Company, conversation in May 1915, a 22” the American offshoot of Poulsen Edison Amberola cylinder phono- Wireless and Telegraph Company, graph horn was placed near a tele- to pursue their own inventions, phone and the sound was pro- including their attempts to perfect jected with such volume that a more sensitive telephone re- Pridham and Jensen realized the ceiver. (White, 1963; Lewis, 1980; full benefit of using a horn with Sanders, 1988; Douglas, 1989). At their receiver. (Lewis, 1980). that time, crystal detectors pro- Pridham took the horn and vided the primary means for radio mounted it directly over their reception together with telephone sound reproducer in place of a receivers for signal reproduction. stethoscope, achieving sounds that Placement of the telephone re- could be heard two blocks away. ceiver into a tea cup provided some (Sanders, 1988). They quickly de- improvement by acoustically veloped a new electro-dynamic transforming the incoming sounds moving coil circuit, combining it using the tea cup as a miniature with a horn to create a single sound horn to reflect and project the projection speaker. (Lewis, 1980; sound. (Hilliard, 1976). Hilliard, 1976; see Figs 13, 14). In the summer of 1910, Pridham and Jensen publicly dem- Pridham and Jensen recognized onstrated the new device for the that existing receiver technology first time on December 10, 1915. only provided 10% efficiency for reproducing the incoming signal, and that 90% of the signal trans- mitted through telephone receiv- ers was lost in the receiver itself. (Sanders, 1988). They achieved initial success in their search for a more sensitive and efficient sound reproducer by connecting the re- ceiver driver’s diaphragm with a toothpick to a wire thread placed between the poles of a magnet (probably an Einthoven string gal- vanometer) and used a stetho- scope and ear tubes to channel the sound. (Beranek, 1954). They called this concept electro-dy- namic. Unfortunately, they could not sell their device to the tele- phone companies due to its bulky size and weight. The U.S. Patent Office subsequently refused their Fig. 13. Pridham (Left) And Jensen (Right) In The Napa, California Labo- application, recognizing the prior ratory. (Source: Shepherd, 1986)

Volume 20, 2007 199 Loud Speaker (Lewis, 1980). This success was transmitter for marine telephones, followed by a Christmas Eve dem- leading to a marine telephone sys- onstration before 50,000 specta- tem which incorporated their dy- tors in front of San Francisco’s City namic moving coil loudspeaker. Hall. Unfortunately, their at- The Navy and Merchant Marine tempts to sell the invention to the extensively used this system, Columbia and Victor phonograph firmly launching Magnavox into companies also failed, but they the public address business. continued to make other high pro- file demonstrations. (Beranek, 1954; Hilliard, 1976; Lewis, 1980). A chance encounter and dem- onstration with U.S. Navy Lieuten- ant Commander George C. Sweet led to introductions with Frank M. Steers of the Sonora Phonograph Company of California. Sweet had been actively following the various inventions and developments in wireless radio for some years and knew many of the top company officials on the west coast. Discus- sions culminated on August 3, 1917, when Commercial Wireless merged with the Sonora Phono- graph Company forming a new entity with Steers as President. The newly organized Magnavox Company, named after the Latin words Magna Vox for “great voice”, would further develop and market the loudspeaker for commercial use. (White, 1963; Lewis, 1980; Sanders, 1988; Douglas, 1989). Fig. 14. Figures From 1920 Pridham During World War I, work on and Jensen U.S. Patent No. the loudspeaker halted as 1,448,279. (Source: U.S. Patent Of- Magnavox attempted to solve air- fice.) plane communication problems (White, 1963; Lewis, 1980; Sand- for the Navy, working with the U.S. ers, 1988; Douglas, 1989). Navy’s nearby Mare Island Ship- Magnavox first introduced its yard. The Navy first thought the horn speakers to the public in 1918 loudspeaker could overcome the using the word Telemegafone to background noise of engines and describe the moving coil driver and propellers. Instead, Pridham and horn assembly. The Telemegafone Jensen came up with the operated primarily as a public ad- Magnavox Anti-Noise Telephone dress loud speaker until January by designing a unique transmitter 1920 when Magnavox modified the that mechanically filtered out the design and advertised it as a radio extraneous noise. (Pridham and speaker in the Electrical Experi- Jensen, 1920; White, 1963). In the menter. Magnavox also intro- fall of 1919, the Navy requested duced the first generation of power Magnavox to adapt the anti-noise amplifiers to the public market at

200 AWA Review Bart & Bart this time. The new Magnavox graph markets. (Lewis, 1980). “electro-dynamic” receiver and Ironically, Pridham and Jensen “Telemegaphone” redefined the did not file for U.S. Patent No. public’s concept of a “loud- 1,448,279 until April 28, 1920, and speaker” as popular magazine ar- it was not granted until March 13, ticles explained the novel design 1923, even though it still re- and highlighted the many success- sembled concepts first pro- ful demonstrations before large pounded by Oliver Lodge (see be- audiences. (QST, May 1920; Fig. low; see Fig. 14). A second patent 15). Magnavox captured the filed May 10, 1922, issued January nation’s attention the preceding 13, 1925 as U.S. Patent No. year by providing loudspeakers for 1,523,349, encompassed further President Woodrow Wilson’s ad- simplifications and improvements dress before 50,000 spectators at to the construction, operation and San Diego Stadium. Both the Re- assembly of their original moving publican and Democratic national coil design. The original patent conventions in 1920 used the relates to the Telemegafone. It Magnavox system, and President describes the electro-dynamic Warren G. Harding’s March 1921 speaker, magnetize-able casing, inauguration in Washington, D.C. magnetic core, diaphragm, coil, air also employed Magnavox loud- spacing gaps, and circuit. Al- speakers. (Lewis, 1980; Shepherd, though the driver assembly fits a 1986; Futtrup, 2004). Fearing the horn speaker, the patent does not inevitable competition and litiga- encompass the horn. Magnavox tion from AT&T and General Elec- achieved additional improvements tric, Magnavox yielded most of the in sound quality and volume by public address business to AT&T modifying its horn designs and by after 1922 and focused almost using crystalline finishes to reduce solely on the radio and phono- the tinny sound. (Sanders, 1988; Paul, 1989a). By 1926, Magnavox had pro- duced over 400,000 speakers. Magnavox ultimately became the largest horn speaker producer of the 1920s making more designs and models than any other manu- facturer. (Paul and Sanders, 1980; Paul, 1989a). Ironically, after all the progress made in electro-dynamic horn speakers by 1924, Magnavox intro- duced its own magneto-driven per- manent magnet horn speaker that year, the M-1, to satisfy public de- mand for a horn speaker that used less power and which could elimi- nate the need for external batter- ies. Other models soon followed as Magnavox further reduced the cost and size of their speakers to meet changing consumer demand. Fig. 15. Magnavox Telemegaphone. (Source: QST, May 1920) (Sanders, 1988).

Volume 20, 2007 201 Loud Speaker

ADVANCES POINT THE WAY jection occurred on April 27, 1898 TOWARD DIRECT-RADIA- when Oliver Lodge filed for Brit- ish Patent No. 9712 for a loud TOR CONE SPEAKERS speaker based on Siemens’ origi- After 1923, most industry re- nal concepts. Lodge, however, ap- search focused on direct-radiator plied nonmagnetic spacers to cone speakers rather than horn maintain the air gap between the type loud speakers due to the im- inner and outer poles of the mov- proved impedance and greater ef- ing coil. (See Photo in Wireless ficiency derived from them under World, 1927; Hunt, 1954). John the amplification circuits now Stroh provided the next significant available. However, wide scale advance when he obtained British production and commercial accep- Patent No. 3393 on December 14, tance of cone speakers did not oc- 1901 for a conical paper diaphragm cur until after 1927. (Wente and which terminated at the rim of the Thuras, 1934; Hunt, 1954; Kellogg, speaker. Anton Pollak dramati- 1955; Paul, 1983 and 1989b). In cally improved the moving coil de- the intervening years, horn speak- sign with his voice coil centering ers dominated the markets (Fig. spider that stabilized the linear 16.) movement of the voice coil assem- Thomas Watson of the Bell bly. Although described as an “ap- companies first explored the initial paratus for strengthening weak concepts for direct-radiator speak- electric currents”, Pollack recog- ers and obtained U.S. Patent No. nized that “the apparatus can be 266,567 on October 24, 1882. He advantageously used…as a sound strengthener…[for]…cable, telegraph, wireless te- legraphy, & etc.” He re- ceived U.S. Patent No. 939,625 on November 9, 1909 for his inven- Fig. 16. Three Moving Armature De- tion. (See U.S. Patent signs: (a) Monopolar, (b) Bi-polar, (c) Balanced. (Source: Hilliard, 1977) No. 939,625; Hunt, 1954; Schoenherr, 2001), Fig. 17. developed a telephone receiver in- Western Electric’s Henry corporating a balanced armature Egerton eventually used similar with a U-shaped magnet and two concepts and those first explored bi-polar pieces plus two voice coils by Watson and Siemens to obtain connected in series. Watson’s re- U.S. Patent No. 1,365,898 filed ceiver wound the driving coil di- January 8, 1918 and issued Janu- rectly onto the armature. Later ary 18, 1921. Egerton described a innovations would leave the arma- new balanced armature with ad- ture moving freely in a fixed coil. justable air-gaps between the poles Watson’s patent recognized that of the magnet. This design permit- increased “amplification or vibra- ted greater control and allowed the tion of the diaphragm may readily sensitivity of diaphragm move- be done” to allow its use as a re- ments to be adjusted; thereby re- ceiving instrument. (See U.S. ducing distortion by improving the Patent No. 266,567; Fagen, 1975; alignment of the magnetic fields Schoenherr, 2001). and the armature. (Beranek, 1954; Further progress in sound pro- Hunt, 1954; Fagen, 1975). 202 AWA Review Bart & Bart

and stiff while the outer zones were left more flexible and more damped. Ricker and Wegel’s efforts cul- minated in Western Electric’s suc- cessful 1924 Model 540-AW Loud Speaker, among the first of the new magnetic cone speakers sold to the public. The 540-AW utilized a double-cone with the back cone frustum attached to the metal frame and with the front cone at- tached to the back cone at the outer edges. The apex of the front cone directly attached to the large mag- netic armature, rather than to the small diaphragm used in earlier horn speaker and earphone driver designs. (Beranek, 1954; Hunt, 1954; Fagen, 1975; Paul, 1989a; Schoenherr, 1999b and 2001; Figs. Fig. 17. 1874 Siemens Type Horn 18, 19). Speaker As Later Used in . (Source: Schoenherr, 2001) Egerton’s improvements achieved gains in receiver sensitivity and were used in some of Western Electric’s earliest horn speakers such as the Model 10-D and Model 518-W. Egerton’s technique achieved real success when N.H. Ricker ap- plied it to a direct-radiator paper Fig. 18. Western Electric Model 540- cone, receiving U.S Patent No. AW Driver. (Source: Western Electric 1,859,892 filed October 6, 1922 Bulletin, 1924) and issued May 24, 1932. Ricker created a hollow air chamber for sound reproduction by designing a double-cone “comprised of vibra- tory members joined at their pe- ripheries.” R.L. Wegel received U.S Patent No. 1,926,888 filed February 5, 1924 and issued September 12, 1933 for his adaptations of Ricker’s concept. Wegel’s patent described a double-cone diaphragm but went on to include a “light rigid non- Fig. 19 Western Electric 540-AW Di- resonant protective [wicker] cas- rect-Radiator Cone Speaker with Bal- anced Armature Driver. (Source: ing” to protect the speaker compo- Western Electric Bulletin, 1924) nents. Wegel further described the central portion of the cone as light

Volume 20, 2007 203 Loud Speaker Earlier in 1921, C.L. Farrand speaker a viable replacement for introduced the first electro-dy- existing horn speaker technolo- namic cone speaker ever sold in gies. The Phonetron only survived the U.S., known as the Phonetron. a few months. (Hunt, 1954; (Kellogg, 1955; Sanders, 1978; Kellogg, 1955; Paul, 1989b; Fig. Schoenherr, 2001; Multiple, 20). 2007). Farrand filed U.S. Patent The cone speaker received its No. 1,847,935 on April 23, 1921. By real impetus in 1925. That year, the time the patent was issued on two independent discoveries rede- March 1, 1932, twelve of the origi- fined their capabilities. In April, nal fourteen claims had been dis- 1925, Chester Rice and Edward allowed by the patent examiner, Kellogg at General Electric pub- leaving only two principal claims lished their seminal paper combin- remaining. (Official Gazette, 1931). ing new amplifier designs with Farrand’s Phonetron successfully their “hornless loudspeaker”. Rice combined a single cone shaped and Kellogg successfully reduced diaphragm, voice-coil driver and a the distortion caused by horn reso- field coil. Farrand’s speakers re- nance and diaphragm resonance. ceived wide acclaim, but they did They designed a coil driven dia- not employ a uniquely designed phragm with a baffle board that amplifier or baffle. These proved improved the frequency ranges of essential to achieving the kind of the sound being produced while dramatic improvements in volume simultaneously supplying more and quality of sound reproduction signal strength through a newly necessary to make the cone designed amplifier. (Rice and Kellogg, 1925; Hunt, 1954; Kellogg, 1955). Unfortunately, their use of a coil driven diaphragm resembled Siemens’ early ideas, leading them to subsequently proclaim “The an- cients have stolen our inventions”. (Beranek, 1954; Eargle and Gan- der, 2004). However, though many elements of their design such as the coil drive, cone dia- phragm, and the baffle had been proposed individually by other in- ventors, none had combined all these aspects with an amplifier to create a complete working system. (Kellogg, 1955). Oddly, General Electric allowed Rice and Kellogg to publish their research without patent protection. (Hilliard, 1977). That same year, in 1925, Ed- ward Wente at Bell Laboratories filed U.S Patent No. 1,812,389 on April 1 which was issued June 30, 1931 for his own independent dis- covery of similar principles. Wente Fig. 20. Figure From 1921 Farrand incorporated a dome shaped cen- U.S. Patent No. 1,847,935. (Source: ter portion of the rigid circular dia- U.S. Patent Office) 204 AWA Review Bart & Bart phragm with a w-shaped rim buff- including not just music but hu- ered by paper rings. Wente previ- man speech, effectively ending the ously contributed his designs for a silent movie era. (Archer, 1939; smaller dynamic moving-coil loud- Thrasher, 1946; Kellogg, 1955; speaker capable of driving a the- Hilliard, 1976 and 1977; ater horn; using an air tight sound Schoenherr, 1999; Figs. 21-23). chamber located underneath the The radio listening public began diaphragm. These earlier innova- to experience the full benefit of tions were recognized in U.S. these advances through RCA’s Patent No. 1,707,545 filed August highly successful Radiola Model 4, 1926 and issued April 2, 1929. 104 electro-dynamic cone speaker. (Hunt, 1954; Fagen, 1975; The Model 104 speaker, based Schoenherr, 1999b). largely on the research of Rice and Bell Laboratories and Western Kellogg, became available to the Electric research culminated in the public in September, 1925. This 1926 release of Don Juan by date is generally considered the Warner Brothers’ Vitaphone Cor- birth of the practical dynamic cone poration. Warner Brothers ac- speaker and marks the end of the quired the Vitagraph film company horn speaker era. (Beranek, 1954; in 1925 and formed Vitaphone to Hilliard, 1977; Villchur, 2000; develop a sound motion picture Futtrup, 2004). company that could exploit the Magnavox issued its own mag- Western Electric and Bell Labora- netic cone speakers soon thereaf- tories innovations. Don Juan was ter in 1926, and in 1927 Magnavox the first successful commercial produced its first electro-dynamic sound motion picture to offer fully cone speakers; keeping pace with synchronized projected sound. industry movement toward A.C. followed in 1927, power supply and other speaker design changes. On May 15, 1929, Pridham even applied for a patent on the hum bucking coil to de- crease speaker background hum. (Sanders, 1988). Jensen followed with his U.S. Patent No. 2,014,621 issued on September 17, 1935 for Fig. 21. 1877 Si- Fig. 22. 1925 diaphragm modifications, im- emens, Cuttris Rice and Kellogg provements to the baffle, and and Redding Di- Design.(Source: rect Radiator Beranek, 1954) structural changes to the mount- Cone Speaker. ing rings; all designed to allow the (Source: Beranek, diaphragm in electro-dynamic 1954) cone speakers to have greater am- plification. Thus, as the roaring 1920s came to a close, it seemed that the once dominant horn speaker technology had been com- pletely eclipsed by the electro-dy- namic cone speaker.

OTHER VARIATIONS IN Fig. 23. Rice and Kellog’s 1925 Free- Edged Coil Driven Conical Diaphragm EARLY SPEAKER DESIGNS Cone Speaker (Source: Rice and The first few years of the 1920s Kellogg, 1924) witnessed an explosion of new en-

Volume 20, 2007 205 Loud Speaker trants into the horn speaker mar- Eventually many types of ma- kets, and new manufacturers were terial were used in horn construc- eager to capitalize on the growing tion including brass, tin, alumi- radio craze. Although most early num, wood, pressed wood, paper manufacturers did not understand mache, shells, various early plas- the physics underlying horn tics, cement and other materials. speaker acoustics, all knew that the Variations in the thickness and horn concentrated and directed mass of the horn coupled with the the sound up the air-column to the density of materials used in con- bell. (Eargle and Gander, 2004). struction were found to dramati- Early on, Thomas Edison con- cally affect the resonant character- ducted his own research on the istics, and were combined with the amplification properties of horns use of damping materials and lami- and “megaphones” in connection nates to improve the overall acous- with telephone and phonograph tical properties. (Hilliard, 1976; sound projection during the 1870s. Sanders, 1978). Magnavox found (Prescott, 1879). By the 1920s, the best sound reproduction came popular press articles for amateurs from horns constructed of heavy, offered directions for making soft, non-resonant material such simple home-made horns that as brass which were given a spe- could operate from telephone cial rough acoustical enamel finish headset receivers. (Wireless Age, employing a heavy crystalline de- 1924). Early Western Electric re- posit on the inner surface. (Sand- search finally proved that an expo- ers, 1978). Other companies tried nential flaring of the horn’s column different methods and radio re- from the base to the bell optimized ceiver horns eventually came in a results. (Flager, 1975; Hilliard, wide variety of shapes including 1976; Paul, 1989a). But, the most round, trapezoidal, oval, curving, significant research into the acous- square and other designs. Folded tics of horn shape and design origi- and morning glory exponential nated in the phonograph industry. horns were also used in public ad- (Hunt, 1954; Wilson, 1974). Here, dress systems. (Beranek, 1954; A.G. Webster’s 1919 “Acoustical Hilliard, 1976; Paul, 1989a and Impedance and the Theory of 1989b; Fig. 24). Horns and The Phonograph” is Early products such as the King credited as the first published pa- Amplitone and the Rhamstine per to address acoustical horn theory based on research con- ducted from 1914-1917. (Hunt, 1954; Beranek, 1954; Hilliard, 1976 and 1977). Magnavox also invested heavily in research about horn de- sign and composition. (Sanders, 1988). C.R. Hanna and J. Slepian of Westinghouse Electric and Manufacturing Company pub- lished the seminal paper on this subject in 1924 addressing throat areas, chamber volumes, horn openings and damping factors. (Beranek, 1954; Hanna and Fig. 24. Comparison of Horn Speaker Slepian, 1924). Designs. (Source: Beranek, 1954)

206 AWA Review Bart & Bart

Adapt-O-Phone simply allowed nologies. By September 1925, so any headset with two earphone many speakers were available that drivers to be clamped onto the base the Electrical Record published an of the horn. (Radio News, Febru- encyclopedia with 129 illustrations ary and July 1922; Paul 1989b). plus numerous advertisements of Other manufacturers including speaker products. After the rapid American Hard Rubber Company, acceptance of magnetic and then Kent and Murdock used only one electro-dynamic cone speakers earphone driver. (Radio News, during 1926, production of horn July 1922; Paul, 1989b). Over a speakers for the radio public dozen manufacturers incorporated abruptly ended. By 1927, over 330 earphone drivers made by manufacturers had produced over Nathanial Baldwin, Inc. At least a 5 million horn speakers. (Paul, dozen more offered the Baldwin 1989a and 1989b). But, customer drivers as optional equipment. preferences changed quickly. The Baldwin’s success resulted from a October 1927 Radio Broadcast unique design dating as early as preview of “1928 Loud Speakers” 1910 that incorporated driver dia- showed 22 new speaker designs phragms made of mica, aluminum, without a single horn speaker. All fiber and other magnetic materi- the major manufacturers including als together with an unusual design Magnavox now featured cone utilizing component springs that speakers. Thus, within one year, proved capable of obtaining twice cone speaker sales had fully the sensitivity of competing prod- eclipsed the older horn speakers. ucts. (Paul, 1989a, 1989b, 1983 and Soon, cone speaker designs were available in dozens of styles, sizes and shapes from over 200 manufacturers. Cone speakers came with single or double cones, free or fixed edges, rolled cones and limitless cabinet and frame variations plus differences in cone materials such as cardboard, parchment or linen. (Paul and Sanders, 1983). Edwin Pridham even filed U.S. Patent No. 1,848,433 on September 9, 1929 which was issued on March 8, 1932 for new cone designs; enabling the cone to “vibrate freely” and offer- Fig. 25. Figures From Baldwin 1915 U.S. Patent No. 1,153,593. (Source: ing better diaphragm and coil cen- U.S. Patent Office) tering for “light, strong and exceed- 1995). Baldwin received U.S. ingly durable” cones that were in- Patent Nos. 957,403 and 1,153,593 expensive and practical to manu- on May 10, 1910 and September 14, facture. Many other patents fol- 1915, respectively, for these inno- lowed from companies and inven- vations, see Fig. 25. tors intent upon improving the But it was Western Electric and available sound quality and con- Magnavox that had paved the way struction of loudspeakers. Some for other manufacturers to copy or companies, such as Sturrup- modify the public address and then Plimpton-Newell Co. of Massachu- radio sound reproduction tech- setts, brought out giant 44 inch tall

Volume 20, 2007 207 Loud Speaker floor standing cone speakers as the phragm trace their origins to race evolved to capture profits and Page’s 1837 observations and market share. (Popular Radio, Bourseul’s 1854 ideas about sound 1928). Future technical research reproduction. Western Electric into cone speaker designs would ultimately based its first horn take on increasing complexity and speakers during the World War I sophistication, but still addressed era on over 30 years of preceding basic problems. For example, telephone industry research and Harry F. Olson applied for U.S. applications which stemmed from Patent No. 2,814,353 on February Bell’s first experiments. Although 26, 1953 which was issued on No- more advanced concepts of direct vember 26, 1957 for his “fluid filled sound reproduction originate with diaphragm suspension” mecha- Siemens and Lodge as early as nism designed to provide uniform 1874, it was Magnavox’s Pridham and smooth frequency responses. and Jensen who first successfully Yet, research into horn speak- applied these concepts to amplify ers for public address systems also radio signals in 1915 with their dy- continued for another thirty years. namic horn speaker. Others soon Folded and flared horns continued followed, including Western Elec- to be used for movie houses. After tric, as the electric horn speaker 1933, multi-cell horns by RCA and achieved success in both radio and Western Electric achieved even public address applications. Yet, greater acoustical performance by even after the horn speaker was adding fins within the bell assem- eclipsed by the electro-dynamic bly based on the earlier research direct-radiator cone speaker fol- of Hanna and Slepian. (Hanna and lowing Rice and Kellogg’s 1925 Slepian, 1924; Beranek, 1954; publications and mass production Hilliard, 1976). Bell Laboratories beginning in 1927, it continued to continued its work with horn evolve, not reaching its modern speakers well into the 1950s as the form until nearly a century after theories about sound reproduction the idea of electrical sound repro- and transmission finally caught up duction was first conceived. Many with manufacturing techniques. other advances would push the (Hilliard, 1978). Other advances cone speaker forward; propelling came in the form of acoustical its use into the 21st century of digi- labyrinth designs to channel sound tal technology, even while the horn from the driver, and directional speaker continues to evolve well lenses were applied (Beranek, over a century after its original 1954). Thereafter, the basic design conception. of high-frequency horn speakers remained relatively unchanged REFERENCES (Hilliard, 1978). Ultimately, many Adams, S.B. and Butler, O.R. (1999). of the initial design concepts first Manufacturing The Future: A tried in the 1920s would continue History Of Western Electric. to survive in public address sys- , England: Cambridge tems and other cone speaker ap- University Press. plications well into the modern Archer, G.L. (1939). Big Business And digital era of sound reproduction. Radio. New York: American Historical Company. Baker, B.H. (2000). The Gray Matter: SO, WHO WAS FIRST? The Forgotten Story Of The Credit for the concepts under- Telephone. St. Joseph, Missouri: lying an electo-magnetic dia- Telepress.

208 AWA Review Bart & Bart

Beranek, L.L. (1954, September). 1977, Vol. 25 No. 9. Loudspeakers and microphones. Hilliard, J.K. (1976, January). Journal Of The Acoustical Society Historical review of horns used for Of America. Vol. 26 No. 5. audience type sound reproduction. Bourseul, C. (1854, August 18). Letter Journal Of The Acoustical Society to the French journal Of America. Vol. 59 No. 1. L’Illustration. Hilliard, J.K. (1977, February). Bourseul, C. (1857). Expose des Electroacoustics to 1940. Journal applications de l’ectricite. Paris. Par Of The Acoustical Society Of Le Cte. Th. Du Moncel. America. Vol. 61 No. 2. Catania, B. (2004, October). The Hilliard, J.K. (1978, July/August). telephon of Philipp Reis. Antenna: Guest editor’s note to a reprint of Newsletter Of The Mercurians. “Auditory perspective- Vol. 17 No. 1. loudspeakers and microphones” by Catania, B. (2006, April/June). Early E.C. Wente and L. Thuras first electromagnetic telephone published in the Transactions Of receivers. Antenna: Newsletter Of The American Institute Of The Mercurians. Vol. 18 No. 2. Electrical Engineers, Vol. 53 No. 1, Coe, L. (1995). The Telephone And Its January, 1934. Reprinted in Several Inventors: A History. Journal Of The Audio Engineering Jefferson, North Carolina: Society, Vol. 26 No. 7/8. McFarland and Company. Hunt, F.V. (1954). Electroacoustics. Douglas, A. (1988/1989/1991). Radio Harvard Monographs In Applied Manufacturers Of The 1920s. 3 Science No.5. Cambridge, volumes. New York: Vestal Press. Massachusetts: Harvard University Eargle, J. and Gander, M. (2004, Press. April). Historical perspectives and Kellogg, E.W. (1955, June). History of technology overview of sound motion pictures-first loudspeakers for sound installment. Journal Of The reinforcement. Journal Of The Society Of Motion Picture And Audio Engineering Society. Vol. 52 Television Engineers, Vol. 64. No. 4. Reprinted by the Audio Electrical Record. (1925, September). Engineering Society, December Loud speaker encyclopedia. 2003. Fagen, M.D. (1975). A History Of Lewis, W.D. (1980). Peter jensen and Engineering And Science In The the amplification of sound. Bell System: The Early Years 1875- Technology In America: A History 1925. New York: Bell Telephone Of Individuals And Ideas. Laboratories. Vol. 1. Cambridge, Massachusetts: MIT Futtrup, C. (2004, October 25). Who Press. invented the electrodynamic Modern Electrics. (1909, April). speaker? Internet site. See http:/ References on Lautsprecher and /home1.stofanet.dk/cfuttrup/dpc/ illustration in Paul, 1989b. inventor.htm. Multiple active coil speaker patent A Guide To Simplified Practical invention. (2007, February 1) Electricity For Home Study And Internet Site. See http:// Reference. (1943). Coyne www.freshpatents.com. Reference Set. Chicago: Coyne Munro, J. (1891). Heroes Of The Electrical School. Telegraph. London: William Hanna, C.R. and Slepian, J. (1924). Clowes and Sons, Ltd. The function and design of horns Official Gazette Of The U.S. Patent for loudspeakers. First published in Office. (1931, December 15). Transactions Of The American Published decision dated May 27, Institute Of Electrical Engineers, 1931 on Farrand U.S. Patent No. February, 1924, Vol. 43. Reprinted 1,847,935 issued March 1, 1932. in Journal Of The Audio Vol. 413 Pages 559-561. Engineering Society, September Page, C.G. (1837a, April). Benjamin

Volume 20, 2007 209 Loud Speaker Silliman’s Journal. Vol. XXXII. airplane telephone. Telephone Page 396. Engineer, December, 1920. Page, C.G. (1837b, October). QST. (1920, May). The Magnavox Benjamin Silliman’s Journal. Vol. radio telemegaphone. XXXIII. Page 118. Radio News (1922, February). Page, C.G. (1837c). American Journal Advertisements for the Rhamstine of Science. Vol. XXXII Page 369. Adapt-O-Phone, the King Paul, F.A. (1981, June). Western Amplitone, and Reproducer Horn. electric horn speakers of the 1920s. Radio Doings (1924, October 4). A Old Timer’s Bulletin. Antique Western Electric achievement. Wireless Association. Vol. 22 No. 1. Advertisement by Braun Paul, F.A. (1989a). Radio Horn Corporation. See Douglas, 1991. Speaker Encyclopedia. Glendale, Reis, J.P. (1862). On telephony by California: All In One Publishing. means of the galvanic current. Paul, F.A. (1989b). A decade of Jahresbericht des Physikalischen electroacoustic reproduction Vereins zu Frankfurt-am-Main, (1920-1930). AWA Review. für das Rechnungs Jahr 1860- Antique Wireless Association. Vol. 1861. Pages 57-64. 4. Rice, C.W. and Kellogg, E.W. (1925). Paul, F.A. (1995, November). The first Notes on the development of a new dozen radio speaker type of hornless loudspeaker. First manufacturers. Old Timer’s published in the Transactions Of Bulletin. Antique Wireless The American Institute Of Association, Vol. 36 No. 4. Electrical Engineers, September, Paul, F.A. and Rogers, B. (1983, 1925, Vol. 44. Reprinted in Journal December). Nathanial Baldwin Of The Audio Engineering Society, earphones. Old Timer’s Bulletin. Vol. 30 No. 7/8 July/August, 1982. Antique Wireless Association. Vol. Robinson, Capt. S.S. (1919). 24 No. 3. Robinson’s Manual Of Radio Paul, F.A. and Sanders, W. (1980, Telegraphy And Telephony For December). Collectors guide to The Use Of Naval Electricians. magnavox horn speakers. Old Annapolis, Maryland: United Timer’s Bulletin. Antique Wireless States Naval Institute. Association. Vol. 21 No. 3. Sanders, W. (1978, January and Paul, F.A. and Sanders, W. (1983, February). Loud talkers. A two part October). Magnetic cone speakers. article published in Radio Age. Old Timer’s Bulletin. Antique Sanders, W. (1988, April to 1990, Wireless Association. Vol. 24 No. 2. March). Magnavox. A 13 part series Popular Radio And Television. (1928, of articles published in Radio Age. April). Running the musical scale Schoenherr, S.E. (1999a). Motion with the new SPN speaker. See also picture sound 1910-1929. Internet the advertisement: The SPN Site. See http:// speaker. history.sandiego.edu/gen/ Post, R.C. (1976). Physics, Patents recording/motionpicture1.html. And Politics: A Biography Of Schoenherr, S.E. (1999b). Sound Charles Grafton Page. New York: recording research at bell labs. History Publications. Internet Site. See http:// Prescott, G.B. (1879). The Speaking history.sandiego.edu/gen/ Telephone, Electric Light And recording/bell-labs.html. Other Recent Electrical Schoenherr, S.E. (2001). Inventions. New York: D. Appleton Loudspeaker history. Internet Site. and Company. See http://history.acusd.edu/gen/ Pridham, E.S. (1947, September 11). recording/loudspeaker.html. The Story Of Magnavox. Shepherd , G.A. (1986, Spring). When Magnavox Company Publication. the president spoke at balboa Pridham, E.S. and Jensen, P.L. stadium. The Journal Of San Diego (1920). The Magnavox anti-noise History, Vol. 32, No. 2.

210 AWA Review Bart & Bart

Telephone Engineer. (1912, Convention Of The Audio November). A novelty at the Engineering Society. Preprint No. Boston electric show. See Douglas, 1002b. 1991. Wireless Age. (1924, March). Make Thompson, S.P. (1883). Philipp Reis your own loudspeaker. Inventor Of The Telephone. Wireless World. (1927, December 21). London: E.&F.N. Spon. Photo of Oliver Lodge’s Thrasher, F. (1946). Okay For Sound: loudspeaker drive mechanism. Vol. How The Screen Found Its Voice. 21 Page 807. New York: Duell, Sloan and Pearce. ABOUT THE AUTHORS U.S. Patent Office. Patent image David and Julia Bart are both searches in 2006-2007. See from the Chicago area where they www.uspto.gov/index.html. continue to reside with their two sons, Van der Weyde, P.H. (1869, May). The John and Michael. They first met at telephone [Reis]. The the University of Chicago where they Manufacturer And Builder. Vol. 1 took a year long course in the history No. 5. of science and natural philosophy. Villchur, E. (2000, July). A short David received both his history of the dynamic Bachelor of Arts Degree in loudspeaker. Voice Coil. Anthropology and Statistics (1985) Webster, A.G. (1919). Acoustical and his Masters Degree in Business impedance and the theory of horns Administration with Concentrations and the phonograph. Proceedings in Finance and Accounting (1993) at Of The National Academy Of the University of Chicago. He is a Sciences. financial consultant and expert Wente, E.C. and Thuras, A.L. (1934). witness in corporate bankruptcy and Auditory perspective-loudspeakers commercial litigation. David has and microphones. First published maintained a strong interest in the in Transactions of the American history of science and early Institute of Electrical Engineers, communications since childhood, and January 1934, Vol. 53 No. 1. has collected radio, telephone, Reprinted in Journal of the Audio phonograph and telegraph devices for Engineering Society, July/August, over 20 years. David’s interests now 1978, Vol. 26 No. 7/8. focus on the history of telegraphy and Western Electric Company the application of this technology. His Catalogue For Telephone current research on the history of Apparatus And Supplies. (1908). teaching the Morse code in America In 2 parts. Chicago: Western is being compiled into the first Electric Company. comprehensive book on the subject. Western Electric Company Bulletin David holds a Ham Radio License and T-680. (1923, January). is a long time member of the Antique Instructions for operating Western Wireless Association. He is the Electric no. 10-A loud speaking President of the Antique Radio Club telephone outfit. Chicago: Western of Illinois and is an active member of Electric Company. the Michigan Antique Radio Club and Western Electric Company Bulletin the Indiana Historic Radio Society. T-750. (1924). Illustrations He is the Chairman of the Museum reprinted in Fagen, 1975. Advisory Council for the Museum of White, L.C. (1963). Pioneer and Broadcast Communications in Patriot: George Cook Sweet, Chicago and is Curator of its radio and Commander, U.S.N. 1877-1953. television collections. David is Del Ray Beach, Florida: The currently consulting with the Southern Publishing Company. Newberry Library of Chicago on Wilson, P. and Wilson, G.L. (1974, future exhibits exploring the impact September 9). Horn theory and the th of science and technology on society phonograph. Presented at the 49 and the humanities.

Volume 20, 2007 211 Loud Speaker Julia received her Bachelor of Arts Degree in Behavioral Sciences (1987) from the University of Chicago. She is an elementary school teacher with a Certification in Early Childhood Education and is working toward her Masters Degree as a Reading Education Specialist at Concordia University. Julia’s interests focus on literature and the arts, the sciences and the process of language acquisition. She is currently involved in designing procedures for the assessment and identification of special needs children to facilitate the delivery of early intervention educational services. For the past five years, Julia has served as a judge for the Illinois Junior Science Academy regional and state competitions. Julia is a long time member and past Treasurer of the Antique Radio Club of Illinois where she continues to play an active role as a volunteer. Julia is also an active member of the Michigan Antique Radio Club and the Indiana Historic Radio Society. Together, David and Julia with their two sons John and Michael have enjoyed building their collection Julia and David Bart of communication devices and provided numerous demonstrations and programs for the Boy Scouts of America, school groups and local historical societies.

212 AWA Review Kirsten

AWA Review The Supreme Model 45 Tube Tester and the 1933 Tube Pin Standard ©2007 Charles C. Kirsten ABSTRACT A brand new mystery package. I must con- fess that when I bought the Supreme Model 45 This is the story tube tester (Fig. 1), I was attracted by Supreme of an ingenuous vintage Instrument’s proud claims in their advertise- Model 45 tube tester de- ment of it’s switching versatility, and the fact sign by Supreme Instru- that “all socket holes are numbered in accor- ments, and how it was dance with the RMA standards”.1 The instru- rendered obsolete ment came in what appeared to be almost new within a year of the time condition, accompanied by all of the original it was marketed, when paper goods: An instruction sheet, set-up tables, the vacuum tube pin packing list, but no schematic. A “Model 45 numbering standard of Accessories Order Form” teased: “Drawing, 1933, the tester’s design Blueprint, Circuit: Stock No.717-C.. $0.40”. Un- basis, was replaced by a fortunately, Supreme Instruments was gone, so new numbering stan- while trying to locate a circuit diagram else- dard. The new code was where, I spent some time examining the tester, part of RCA’s 1934 tube and its operating instructions. chart including base Individual tube pin switches. The front drawings of the new pin panel of the Supreme model 45 tube tester has numbering standard seven rotary switches. The one marked “3-4”, published in a two page below the meter, selects the filament voltage to spread in both the Sept, connect to all pin “4”s, (while all pin “3”s are 1934 issue of Radio, and filament transformer returns}. This switch se- in the Oct 27, 1934 issue lects the proper filament or heater voltage from of Radio World . the following voltage choices: Other instru- ments (radio analyzers by Supreme, Weston, and Hickok), were also disabled by this change in radio national stan- dards, and were equally innovative products. The filament/heater pin assignments were hard wired into the Supreme Model 45 tester such that they were not readily reassigned when the new code was intro- duced.

Fig. 1. Supreme Model 45 tube tester using the 1933 RMA pin code numbers. Volume 20, 2007 213 Supreme Model 45 1.5, 2.0, 2.5, 3.3, 5.0, 6.3, 7.5, 12, sockets, showing that the tube 14, 25, 30. tester has never been used since it The control knob marked “TC” was built. switches any top cap of the tube Filaments and heaters only on just like any of the other pin con- pins 3 & 4. I finally traced the wir- nections to the tube. The other five ing in the tube tester itself (Fig. 3), are numbered for pins “1, 2, 5, 6”, producing the Model 45 schematic and “7” which are the correspond- shown in Fig. 4. When I reviewed ing socket pins which the knobs in- the circuit, I admired the simplic- dividually control. Each numbered ity of the heater/filament wiring, control, (and pin) may be indepen- but I wondered how Supreme dently switched to any of the fol- could “get away with” no provisions lowing internal voltages, or circuit for switching the heater or filament options: connections to other pin locations. 0V 67½V 250V It fascinated me that all of the pins 10V 90V -C (Tube Grid) numbered “3” were connected to- 22½V 135V -B (Meter Return) gether and wired permanently to 45V 180V the filament transformer circuit, as Since each numbered knob has were all of the pins numbered “4”. a pointer showing the value of volt- For such a filament arrangement to age or connection selected for that work, there must to be no tubes pin, it is easy to see what voltage is with heater or filament connec- being applied to each pin number tions other than the “3” and “4” of the tube during test “set-up”. I pins as identified in the Model 45 thought this whole arrangement tube tester. was really “cool”, but unfortu- I reviewed my tube manuals. In nately, the Supreme 45 did not use the manuals after 1935, the pins the approved RMA pin numbers. were numbered in the familiar oc- When I was 14, and first started to tal tube numbering system, and the repair radios in 1936, I had memo- octal tubes had filaments on pins 2 rized the tube pin numbers from and 7, though even that varied. The my RCA tube manual, and they only tube pin markings I found in were not the numbers on this 1933 tube manuals before 1933, were P, tube tester. G, F, K, or H for plate, grid, fila- Custom molded tube sockets ment, cathode, or heater respec- with pin numbers. While photo- tively. However, excepting tubes graphing the tester, my digital camera clearly showed pin numbers alongside each pin hole (Fig. 2). Since Supreme had spent money on pro- duction tooling by having special sockets made with the pin numbers molded in, they must have planned on using this RMA numbering system for a long time. Not sur- prisingly, there is no evi- dence that tubes have Fig. 2. Supreme Model 45 sockets with molded- ever been inserted in the in RMA tube pin numbers.

214 AWA Review Kirsten had to handle these heavy currents was the filament switch which se- lected the proper filament voltage transformer tap. I liked this ap- proach, especially after seeing the heater current being routed through several successive rotary wafer switches, in other testers, which were not rated for these high currents, and even in Hickok’s 539- C tester parts list, were rated at only 0.65 Amperes. Alternating current emission. Fig. 3. Supreme Model 45 inside parts Fig. 5 is a simplified circuit diagram and wiring view. of a typical test set-up on the Su- with heater connections at the top, preme Model 45 tube tester, for a no tubes of the pre-octal era had type 6C6 pentode with a 6-pin base. filament, or heater connections on The model 45 uses no D.C. poten- other than the pins which were tials, but operates solely on alter- numbered 3 and 4 in the 45 tube nating voltages applied to all tube tester. elements, including the plate and What a neat way to avoid grids. Electron flow inside the tube switching heavy heater currents from the cathode to the screen grid, required by low voltage AC tubes and to the plate give a combined such as the types ‘24, ‘27, ‘35, ‘46, electron flow through the “250” and ‘2A5 (1.75 Amp); 2A3 (2.5 and “90” taps of the anode trans- Amp), the rectifier types ‘13, and former, which is returned through ‘80 (2 Amp), ‘83 (3 Amp), and 2V3 the 1,000 Ohm resister, from “-“ to (5 Amp). The only switch which “+” through the meter, and back to

Fig. 4. Supreme Model 45 tube tester schematic diagram.

Volume 20, 2007 215 Supreme Model 45 screen voltages for test conditions. When appropriate plate and grid voltages are applied, the meter de- flection is a valid measure of the tube emission characteristics, and is compared to the tabulated range of standard or acceptable values. Grid bias and grid current (gas) tests. The model 45 has no sensi- tivity control to “calibrate” the meter deflection range for emis- sion levels from different types of tubes, but the 1,000 Ohm cathode load resistor serves to limit current through the tube as well as biasing the control grid of the tube. To change the grid bias on the tube, Fig. 5. Simplified circuit diagram of a the “Bias” push button switches typical test set-up on the Supreme the control grid return from the ‘-‘ Model 45 tube tester. side of the 1,000 Ohm self bias re- sistor, to the ‘+’ side, giving a “-B” and through switch 5 to the change of grid bias. This gives a cathode. Note that the meter “+” proportional increase in plate cur- is always connected to the filament rent, which shows on the meter as which returns the filament emis- a measure of transconductance. sion current back through the This is compared to the expected meter in series with the 1,000 Ohm value to judge the tube’s load resistor, (which also serves as transconductance. a “self bias resistor”). Cathode type A test for gassy tubes is also tubes such as the 6C6 are tested by provided; the “Gas” test button switching the cathode pin to “-B” shorts out the 100,000 Ohm grid which connects to the meter “+”, leak resistor which normally re- so the meter will then measure the turns from grid to -B. This causes cathode emission, instead of fila- a change in plate current propor- ment emission. tional to any positive grid voltage The D’Arsonval type meter of due to grid collected ion gas cur- the Supreme 45 tester reads the rent passing through the high re- average of the rectified, half-wave sistance grid leak. pulses through the tube, so it de- flects to only one-half of a full- THE SEARCH FOR THE wave current, and since the pulses are sine waves, the moving coil MODEL 45’S RMA NUMBERS meter averages these to 0.9 of the Sampling local sources. Having root mean square (RMS) alternat- exhausted my own tube manuals, ing current value. Thus, the meter and still not finding any evidence will read only 0.45 of the RMS cur- of an RMA tube pin number code rent applied to the tube plate. which matched the Supreme Since the peak applied plate tester, I came to believe it might voltage, and peak current will be be found in a 1933 tube manual, or 1.4 times the selected RMS values, in an RMA, or Institute of Radio care is taken to keep these peak Engineers (IRE) document. I values below the tube ratings when didn’t have a 1933 tube manual, so selecting the applied plate and I started searching through maga- 216 AWA Review Kirsten zines, looking for any information other sources of this 1933 pin code and now that I was looking for appear to be facsimiles of the RCA/ those little numbers next to the Cunningham tube charts. tube pins, they started showing up, Copies of the Charts of the but strangely they were few and far RCA/Cunningham tube pin and between. socket drawings for the 1933 RMA Typical examples were found in pin standard were found only in a few new tube announcements,2 the publications listed below: in a few instrument construction Feb, 1933, RCA Radiotron- articles3 and if you magnified Cunningham Radio Tube Manual, them, in advertisements for radio Technical Series No. RC-11, Pg 150 set analyzers4 , but only in some Mar, 1933, RCA Complete Rid- 1933 and 1934 magazines. ers Manual, (Vols I-III) pg X. Others gave a helping hand. As Apr 1,1933, Radio World, “New I have often done, I appealed to Tube Chart”, pages 11,13 Charles Brett in Colorado Springs May, 1933, Riders Radio Ser- for help. Gracious and helpful as vice Manual, Vol III, RCA 3-78,79. ever, Charles searched out and Aug, 1933, Radio Retailing, provided a copy of the first pub- “New Characteristics Chart”, pg 31. lished tube pin numbering stan- Sep 9,1933, Radio World, dard of 19335, from Page 150 of the “Charts of Characteristics”, pg 8,9. RCA Radiotron-Cunningham Ra- 1933-34, Official Radio Service dio Tube Manual, no. RC-11. This Manual, Gernsback, Vol. IV, pg confirmed Supreme’s statement 392 that the model 45 pin numbers Apr 20, 1934, RCA Victor Ser- conformed to the RMA tube pin vice Notes 1931-32, unnumbered. code numbers that were valid in 1933 pin code numbering rules. 1933 and early 1934. Thanks are also due to Gary Mayfield of Marquette, KS, who offered to loan his RC-11 for a cover scan, which I eagerly accepted (See Fig. 6). The RCA Radiotron- Cunningham Radio Tube Manual, no. RC-11. After the 1931 merger of the C.E. Cunningham Tube Co. with the Radio Corporation of America in 1933, these companies merged their separate R-10 and C- 10 tube manuals of 1932, to pro- duce their first combined tube manual5, The RCA Radiotron- Cunningham Radio Tube Manual, no. RC-11, copyright 1933. Subse- quent searches for any other for- mal or official announcement of the 1933 RMA pin number code left the impression that the RC-11 Tube Manual was the first and only publication to present the 1933 RMA tube pin numbering code as Fig. 6. RCA Radiotron-Cunningham an official public disclosure. All Radio Tube Manual, no. RC-11.

Volume 20, 2007 217 Supreme Model 45 Fig. 7 is an enlargement of the top counter clock-wise direction. four socket connection diagrams Counting backwards (clockwise) from page 150 of the RC-11 tube from 3, makes the plate pin num- manual. The diagrams show vari- ber 2, and the grid pin number 1. ous tube internal element connec- If a top cap is present, it is given tions for typical 4, 5, 6, and 7 pin the number 5. A similar method tubes. Note that the pin number- gives the numbers seen on the six ing shown will be the same for any and seven pin sockets. The same tube, or socket with the same num- technique is used for the five pin ber of pins. The numbering of all socket, starting from the middle tubes and sockets, when viewed two (heater) pins. from the pin end of the tube, or the Origins of the 1933 tube pin bottom of the socket as shown on numbering code. In searching for the chart follows these rules: any early clues to RCA tube manu- - Starting with “1” the pin num- facture, I remembered that bers increase in a counter clock- Ludwell Sibley had taken orders wise direction around the socket. for copies of the RCA Radiotron - Number “1” is always selected vacuum tube manufacturing such that numbers “3” and “4” are “Standardizing Notices”6 which the two filament, or heater pins of were part of the Dowd/RCA Ar- the tube. chives belonging to the Antique - Filaments or heaters of 4, 6, Wireless Association Museum, and and 7 pin tubes are the two large I had ordered a set of the standards pins. for the types 01-A, 10, 45, 50, 2A3 - Filaments or heaters of 5 pin and the generic “Complete Tube tubes are the middle two pins of Assembly”. According to the infor- the four pin cluster opposite to the mation included with the “Stan- single pin. dardizing Notices”, General Elec- - The tube cap number is one tric light-bulb factories had used more than the highest pin number. “Base Threading Standards” to To illustrate, consider the four- guide the proper insertion of leads pin socket shown in Fig. 6. The two from a glass bulb into the attached large heater pins are oriented to base. When RCA took over a ma- the bottom of the figure, and are jor General Electric factory in 1930 numbered 3 and 4 counting in a to manufacture radio tubes, they adopted the use of “Base Thread- ing Standards” as assembly guides for their vacuum tubes, and named them “Standardizing Notices”. The dates on these early RCA docu- ments came years earlier than the publication of the 1933 RCA Radiotron-Cunningham Radio Tube Manual, which includes a full description of the approved 1933 RMA tube pin connection and numbering standard. As a typical late example, the assembly standard7 for the type UX 250 (Fig. 8) was originally ap- Fig. 7. 1933 RMA tube pin and socket proved on October 10, 1929, and numbering code example. though the tube leads were identi-

218 AWA Review Kirsten

Fig. 8. RCA Radiotron Tube Standardizing Notice for Type 50. fied as “P”, “G”, “+F” and “-F”, the “3”, and “4”. At that time, I com- tube base pins were numbered on pletely missed the significance of exactly the same plan as subse- this highly useful conformity. quently shown in the RC-11 tube Benefits of the 1933 tube pin manual of 1933. In rereading the number code. Features that made information from Ludwell, I was the 1933 tube pin number code of reminded that he had pointed out value to the Supreme model 45 that all filaments were numbered tube tester, and to any instruments Volume 20, 2007 219 Supreme Model 45 that dealt with detailed tube con- to look page-by-page through ev- nections were:- ery one of my radio magazines that o The filament/heater pins covered 1932 through 1935. were always adjacent, minimizing Don’t ask, don’t tell. Although filament magnetic and electro- use, or even mention of the pin static fields. numbers was totally absent from o The filament/heater numbers most magazines, it was gratifying are always the same (#3 & #4). to find the occasional evidence that o The pin numbers and base some people had found valuable connections are standardized, and uses for the 1933 tube pin code. may always be found in the refer- Only two magazines, Radio-Craft ence manuals. and Radio World, reported on and By making the circuitry of tube used the pin numbers in a normal connection simple and lucid to the way. In all other periodicals, it took repairman, and allowing instru- a meticulous search covering many ment settings to be based on the monthly issues (up to a year) to tube parameters given in the tube find the infrequent instances of manuals, there was no need for the use, or reference to the tube pin instrument manufacturer to pro- (and socket) numbers, and I be- vide new charts for every new tube came really puzzled by the general test. pattern of non-reporting. Where Publication coverage of the editors should have been promot- 1933 code use. Having the 1933 ing use of the 1933 standard, most date for the release of the RMA were silent about the pin code. I Code, I looked for it’s appearance tried to write up these contradic- in some of my hard-back books. I tory findings from 9 different ra- found no mention of the tube pin dio magazines, but finally, I gave numbers in any of the 1932 to 1936 in to the idea that “One picture is copyrighted books, with one ex- worth a thousand words”. ception: James A. Moyer’s Radio Chronological overview of Receiving and Television Tubes,8 magazine coverage. Fig. 9 gives a which laudably showed the sche- chronological picture of the atten- matics of three different tube tion given to the 1933 pin number testers with the 1933 code pin code, which was published and numbers on every socket. The valid for 17 months, from Feb. schematics of two other testers are 1933 until June 1934, after which shown with the new number code there was a new, RMA approved and the octal metal tube socket number code. numbers. Ten months later still, in April I had discovered the source and 1935, RCA/GE introduced their confirmed the existence and pub- new metal octal tubes, and the cor- lication of the approved 1933 tube responding pin code. These key pin numbering system in the RCA dates are marked on the second Radiotron-Cunningham Radio line of the chart by “labeled darts” Tube Manual, no. RC-11, but I was below the “Year” column heads. still curious as to why it was so low- It is well to remember that this key. How wide-spread was its use? chart covers over 36 months of Who had published, or champi- time, and eight monthly and one oned the system? Who besides weekly publications, a total of 444 Supreme, Weston, and Hickok had opportunities to show, or mention used or supported the 1933 tube the pin numbers. Instead, month pin code? To find out, I resolved after month, scores of magazines

220 AWA Review Kirsten

Fig. 9. Pattern of radio magazine publications of the 1933 RMA vacuum tube pin numbering standard. were published with new tube an- code was in force for a year and a nouncements, instrument circuits, half, why did these magazines dur- and radio diagrams, without a ing all this time, follow a strange single pin number in sight! course of not including the 1933 Since only the dark dots on the pin code, but generally avoiding chart represent articles which give the subject. the code numbers for the socket or tube connections, either the pin code number code was poorly cov- RADIO MAGAZINE SUP- ered, or completelty ignored by PORT OF THE 1933 RMA PIN five out of nine magazines, during CODE the 17 months that the 1933 code The following are a couple of was in force. examples of the way radio maga- The chart of Fig. 9 is a cryptic zines dealt with the 1933 and sub- summary, so it doesn’t reveal the sequent pin codes: mostly through strange lack of editorial and reader silence. comment shown by the various QST. In the March, 1933 issue, magazines. This silence was most George Grammer, Assistant Tech- notable during the time the indus- nical Editor of QST authored try abruptly changed the estab- “Straightening Out The Socket lished and useful 1933 tube pin Connections”,9 an article showing code with a new tube pin number a chart of the base connections of code. Since the magazines pro- all receiving tubes, “with the pins vided no explanation for the new numbered according to the system tube pin numbers, this left many adopted by the manufacturers”. questions unanswered. For ex- Since the article was offered as ample, since the 1933 pin number an authority on the new tube pin

Volume 20, 2007 221 Supreme Model 45 code numbers, it was unfortunate pin tube bases with the 1933 tube to find both 4 Pin socket drawings pin code12 (Fig. 10). with blatant pin number errors; The next issue of September, e.g., showing “1,2,4,3”,instead of 1934, carried a two page center- “1,2,3,4”. fold, of the complete RCA- In June, 1933, QST again of- Cunningham Radiotron radio tube fered an article with the RMA tube chart of all non-octal tubes with code numbers; “The Dial Coded the new pin numbering system. Universal Tube Checker and Cir- Radio continued to feature the old cuit Analyzer”,10 by C.B. DeSoto, 1933 tube pin code on their front Assistant to ARRL Secretary. This cover for 17 months; from Aug, “Dial Coded” tube tester is de- 1934 through Dec, 1935. (See Fig. signed to the same precepts as the 10) This silent protest of the Supreme model 45 tube tester, but change in number codes was con- ‘free’ switches only pins “1”, “5”, tinued for almost a year and a half “6”, and “7”, as the article alleges after the 1934 RMA pin code had that the other pins are committed. been announced by RCA- A riveting detail of this article is the Radiotron tube charts; and right uniquely mis-numbered 4 pin up to the time that Radio was socket which doesn’t even match merged with R/9 13 in Jan, 1936. the 4 pin mis-numbering in QST’s Radio World. This radio indus- March, 1933 article. This 4 pin try weekly was the only periodical socket counts “1,3,4,2” instead of which consistently published all “1,2,3,4”. information on the 1933 RMA tube I found no errata or correction pin number code from before its in subsequent issues of QST. In- first appearance in the RCA stead, the magazine then elimi- Radiotron-Cunningham radio nated all use of the RMA pin num- ber code until April, 1935 when the octal metal tubes, and the octal numbering system were an- nounced. Radio. Began as “Pacific Radio News” before WW I. Renamed af- ter WW I, this Western periodical from San Francisco maintained an absolute silence about pin code numbers, not even using them in new tube type announcements, or technical articles (where they would be expected) from Jan, 1933 to March, 1934. In April, 1934, a full page advertisement by Hickok Electrical Instrument Co. featured the S.G. 4800 Hickok Analyzer us- ing the 1933 pin code.11 In August, 1934, Radio broke with their former avoidance of the 1933 tube pin code, and featured a full height 1½ inch column on the left side of their front cover which Fig. 10. Sample covers of Radio showed drawings of 4,5,6, and 7 magazine showing 1933 numbered tube socket drawings. 222 AWA Review Kirsten tube manual in Feb, 1933, through to the code’s de- mise, in October, 1934. Af- ter waxing enthusiastic over the 1933 RMA code in the June 17, 1933 article, “The Tally Method”14 by Herman Bernard, Radio World followed up in Au- gust 5, 1933 with, ”A NEW ACCESSOR Using the Tally Method of Significant Switching”,15 also by Herman Bernard. The September 9, 1933 issue showed a front page, plus a two page centerfold of the RCA tube charts16 with base diagrams giving the 1933 RMA pin number- ing code. Why was Radio World the only magazine to add an introductory article when they published the new pre-octal 1934 RMA tube pin numbering code17 in Oct 27, 1934.? The foot- Fig. 11. Sample socket numbering from Harry note on the RCA tube Murman’s article on the new 1934 RCA pin charts states: code. “Tube symbols and bottom ing system design. Ironically he views of socket connections have suggests that it is important that pin numbers according to the new the new system not change the system recently standardized by tube pin numbering. the Radio Manufacturers Associa- A summary of the tube pin tion, Inc. Courtesy of RCA numbering codes. Fig. 12 gives ex- Radiotron Co. Inc., Copyright amples of the three codes used for 1934”. vacuum tube pin identification, on An internal RCA document the early 4,5,6, and 7 pin receiv- guides tube assembly workers in ing tubes, and the 1935 RMA octal transitioning from the 1933 tube pin code for the new RCA octal pin standard to the new “RCA” metal tubes: tube pin numbers.6 No explana- A. Element name initials : Used tion of the tube charts was given from 1906 to the last use of by RCA, but in the Jan, 12,1935 is- vacuum tubes. sue of Radio World an article “New B. The 1933 pin number Tube Data”18 by Harry Murman code.(GE Historic), Jan 1933 to was featured, (Fig. 11). No creden- June, 1934 tials were given for Mr. Murman, C. The 1934 number code. (RCA and it is not known if he was spon- Code) June, 1934 onward. sored by RCA, but he seemed to D. The 1935 octal metal tube pin speak as though he was involved code. Jan 1935 onward. in considerations of the number- E. An example of what the oc-

Volume 20, 2007 223 Supreme Model 45

Fig. 12. American vacuum tube pin numbering codes. tal metal tube code could have pin “2” is filament/heater, pin “3” been. Nov, 2006. is plate, pin “4” is screen, pin “5” is Rules of the 1933 RMA tube pin control grid, pin “6” is supressor numbering code are given above. grid, pin “7” is the filament/heater Rules of the 1934 RMA tube pin return, and pin “8” is cathode. All numbering code: When viewing pins may have different, or “no” the bottom of the socket, with the connections. filament pins nearest the viewer, A mystery with the evidence in (See Fig. 10-C), counting starts plain sight. I was not only sur- clock-wise with “1” on the left fila- prised, but chagrined, to find that ment pin, “2” on the plate, and during my sixty years of radio re- ends with the right filament pin as pair and communications engi- the last numbered pin. This makes neering, I had never noticed the the right (return) filament num- scattered remains of the 1933 tube bers different for every socket with pin number code. From my con- a different number of pins. tacts with other collectors of vin- Rules of the 1935 RMA octal tage radios and test equipment, it metal tube pin numbering code: was clear that I was not the only When viewing the bottom of the one who had never known about socket, with the key of the center the 1933 code, or the way in which post positioned at the bottom (See it had been replaced by an inferior Fig. 10-D), counting starts clock- code. From my own reaction it wise with “1” on the left bottom seemed likely that other radio col- pin, and proceeds around the pins lectors might share my interest to end on “8” on the right lower and curiosity about this once well pin. Pin “1” is the shell (ground), regarded and useful pin number-

224 AWA Review Kirsten ing standard with such an obscure tions of innovators like Edwin and brief life. Armstrong, Charles Leutz, Philo Who did it? I am of the opinion Farnsworth, Charles Jenkins, that RCA Radiotron was the source John Rider, and David Grimes as of the tube pin number system, evidence of RCA’s “grind them into since the same system was used by the ground” approach to any form vacuum tube assemblers in the of competition, or even free trade. Radiotron factory. RCA I could admire RCA’s technical ac- Radiotron/Cunningham pub- complishments, but I found their lished the 1933 RMA pin number business practices deplorable. code in the their RC-11 tube What could RCA’s motive have manual, and also advertised free been? As I looked for articles on tube charts which included the the use of the 1933 RMA tube pin 1934 tube pin code. code in my radio magazines, I was Skepticism has been shown attracted by all of the business and when I expressed my opinion that legal actions being reported, which RCA was the logical leader in or- often impinged directly on the ganizing a conspiracy of silence technical items I was pursuing. around the 1933 RMA pin number For example, how did RCA dare to code. When I expressed my belief make any questionable decisions that RCA had openly destroyed the when they had just come out of a 1933 RMA tube pin numbering suit by the U.S. Department of Jus- code, putting in place a new code tice which had charged them with of their own devising, without co- unfair and monopolistic business ordination, review, or popular ap- practices? That suit had been com- proval, the usual reaction is disbe- pleted in November 21, 1932,19 but lief, and sometimes a mounting surely the U.S. Government main- indignation! The entire concept of tained surveillance. slipping an un-needed, and tech- The fact is, that RCA and nically inferior version of the RMA Radiotron came out of the consent national tube pin numbering stan- decree by the Federal Courts19 far dard by the entire radio industry, stronger and less vulnerable than where, by past tradition, the indus- when they were indicted. The Fed- try leaders would approve any such eral Judge, while separating GE, move, defies common sense, but I Westinghouse and RCA back into believe that’s just what RCA three Corporations, and awarding Radiotron did. However, I think Majestic Radio $7 million in dam- they had almost complete coopera- ages, also gave RCA licensing con- tion of all major members of the trol over not only the RCA patents, radio industry. but also the “Pool” of all radio re- There has been a suggestion by lated patents including those from a thoughtful and respected mem- GE and Westinghouse. The con- ber of the AWA, that RCA was sent decree further stated that forced to accept the code change RCA had to pursue the licensing of by a fractious RMA. The RCA all other companies under these documents shown in support of patents, for which each licensee this opinion are suggestive of such would have to pay RCA a 5% fee. a view, but certainly are not con- Impact on the Supreme Model clusive. Consequently, I take my 45 tube tester. The changing of the opinion from the overall evidence. tube pin numbers caused immedi- For me, this includes RCA’s re- ate confusion to users of 1933 pin peated prosecutions and persecu- number based equipment like the

Volume 20, 2007 225 Supreme Model 45 Supreme Model 45 tube tester. months after the change to the Since the new 1934 pin code did 1934 tube pin code, came the re- not always provide the same two quirement to accommodate the pin numbers for filament conn- new octal tubes and the octal pin ections as was true of the 1933 pin number code, compounding the code, it was necessary to provide disruption to existing tube equip- new instructions, and conversion ment designs, and expanding it to charts to relate the new and old all tube testing equipment. codes for all instruments based on The octal tube effects on the ra- the old 1933 pin number code. It dio manufacturers. The introduc- is probable that Supreme contin- tion of metal tubes to radio manu- ued to supply set-up data for the facturers was all that RCA had model 45 tube tester as new tube hoped it would be. Radio receiver types were announced, but the ad- sales escalated, and gave the radio vent of octal tubes, and changes industry the boost it needed. Not in filament pin connections quickly all manufacturers found this a sealed the fate of this “orphan of blessing from RCA. Arcturus, for change”. one, chose to make their own tubes Impact on the analyzer users. with octal bases. An alternative It is doubtful that most servicemen was to ignore the octal metal tubes, who used set analyzers found the and to continue with their own new 1934 pin numbers much more brand or purchase other types of than another pain-in-the-neck in- receiving tubes. Philco chose this convenience. Manufacturers were course at the beginning, but after notorious for not considering the five years or so, capitulated. service convenience of their radios All of the other radio manufac- or instruments, so new numbers turers who had been buying their were probably just another learn- tubes on the open market, now had ing hurdle to surmount. The cost to buy their tubes from RCA, or of buying new equipment or up- someone licensed by RCA to make grading existing testers was a more the metal octal tubes, or buy “MG” significant problem to small busi- or “G” tubes. These tubes could nessmen in a depression economy. not be bought at previous cut- Impact on radio receiver throat prices. Other large tube manufacturers. The 1933 tube pin manufacturers, like Ken-Rad, and numbering code seemed to have National Union immediately been largely ignored by radio re- bought the machinery and licenses ceiver manufacturers. Due to the to make the octal metal tubes, and long established factory practice of competed with RCA as they paid using color coded wires and visual them patent royalties for the hardware examples to guide as- privelege of doing so. semblers in the factory, almost The new octal metal tubes were none of the radio receiver schemat- not markedly better in perfor- ics showed pin numbers on the mance than the glass types, prima- tube socket drawings, so there was rily because they were almost ex- little reason to put the numbers on act electrical copies of the former their hardware. Therefore, the glass types. They were, however, change to the 1935 tube pin num- more precisely made, which made ber code had little effect on the ra- them less microphonic, quieter, dio manufacturers. and less susceptible to physical Effects of the octal base and damage. In these respects they the 1935 octal pin code. Six were better. Since every octal tube

226 AWA Review Kirsten used the same type of socket, parts tem which was compatible with procurement and spares were sim- the existing 1933 pin numbering plified. The down side of this was standard. Instead, the tube manu- that the tube positions and types facturers produced a poorly engi- had to be identified, so the tubes neered base connection arrange- would not be replaced in the wrong ment which broke with the long sockets. This was not a great prob- honored and rigorously investi- lem, for the self shielding tubes gated convention of keeping the removed all need for the former heater leads adjacent to each other. clutter of removable metal tube Without apparent reason and in shields, giving a clear open chas- violation of the widely known test sis. results on filament connections, The octal tube impact on tube the heater leads were separated by tester designs. The advent of the two other pins, increasing electro- octal tube was a curse and a bless- magnetic and electrostatic hum ing to the tube tester companies. fields caused by the heater or fila- The curse was the fact that all ex- ment power.22 isting tester designs were not ca- The New Zealand author John pable of handling the additional Stokes, in his book, “70 Years of eighth pin, were therefore obso- Radio Tubes and Valves”, ad- lete, and must be redesigned. The dresses the reason why RCA bright side of the event was that Radiotron-Cunningham, in 1938, since every tube tester was obso- released single ended versions of lete, every service-man must buy the original 6F5, and 6Q7 metal a new octal tube capable tube tubes.23 On page 98 Stokes says: tester. “The initial release of four types, The decision by RCA to sepa- 6SJ7, 6SK7, 6SF5, 6SQ7, was an- rate the filament connections on nounced by RCA late in 1938. It is the octal tubes onto pins #2 and interesting to note that of the four, #7 was probably the most unfor- two were AF voltage amplifiers, tunate in the otherwise illustrious and this may be explained by the innovation of metal tubes by GE/ fact that comparable double-ended RCA. It must be believed that the types had met with some criticism octal filament decision was made in the matter of excessive hum lev- by executives, more interested in els when used in high gain circuits. manufacturing ease and not by This hum was mainly attributable engineers concerned with tube to the use of widely spaced lead- performance. Weston and Su- out wires and base pins for the preme seemed to accept the inevi- heater connections, resulting in table changes, and took full page stray magnetic and capacitive cou- advertisements assuring their cus- plings. In the case of the 6SF5 and tomers that there would be no loss 6SQ7 this was avoided by keeping of test capability from these new the heater leads and base pins ad- developments.20,21 jacent even though this had the un- What seems most puzzling is fortunate result of breaking the that in 1935, when the metal octal (newly) established standard for tubes were announced, none of the the position of the heater pins. It companies who were developing was probably the same hum prob- the octal tubes expressed concern, lem and its solution that allowed or made an effort to assure that the British Mazda to profit from the new octal tubes were designed and American experience and keep the introduced with a numbering sys- heater connections of all their own

Volume 20, 2007 227 Supreme Model 45 octal based valves strictly adjacent *2. Complete Official Data on Four and always on the same two pins New Tubes. Base connections and where, we can say with hindsight, characteristics given for tube types; the American connections should 25Z5, 5Z3, 2A3, and 2A5. Radio have been in the first place.” World. February 4, 1933, pages 3-7. *3. Grunwold, W.J., The “Calvert After 70 years, why should we Converter”, A short wave converter to care? As always, it was a real plea- work with any broadcast receiver for sure to discover a vintage device foreign broadcast reception. Short such as the Supreme Model 45 Wave Radio May, 1934, pages 34-36. tube tester, which showed an el- *4. Weston Electrical Instrument egant simplicity in design. How- Corp., Advertisement. Radio ever, it was disturbing to find that Retailing August, 1933, page 34,35. once again, RCA Radiotron *5. RCA Radiotron - Cunningham seemed to be involved in an exer- Radio Tube Manual, Technical Series cise of their monopoly powers to No. RC-11. Harrison, NJ: RCA, January, 1933; page 150. take advantage of real or potential *6. RCA Radiotron. Radiotron competitors with no concern for Standardizing Notice, Complete the resulting long term damage Tube Assembly. Drawing No. 92S- In this instance, RCA Radiotron 36181. Harrison, NJ: R.C.A. Radiotron had created and published a really Company Inc., Circa June/July, 1934 useful 1933 USA radio receiving *7. RCA Radiotron. Radiotron tube pin number standard, only to Standardizing Notice, replace it two years later with an Complete Tube Assembly, Tube Type octal tube standard so technically 50. Drawing No. 92S-31079. flawed that later, it had to be cor- Harrison, NJ: R.C.A. Radiotron Company Inc., October 10, 1929, rected for several tube types to en- * 8. Moyer, James A. Radio Receiving able acceptable hum level opera- and Television Tubes, 3rd Edition, tion. It should be noted that the New York, NY: Mcgraw-Hill Book RMA major officers and directors Company, 1936. came from industry, and many of *9. Grammer, George. Straightening them were RCA, Radiotron and Out the Socket Connections. QST Cunningham top executives, so March, 1933, page 30. there was probably little objection *10. DeSoto, C.B. The Dial Coded to the actions taken by RCA. Universal Tube Checker and Circuit My investigations of the inspi- Analyzer. QST June, 1933, pages 21- 23. ration for, and the vicissitudes en- * 11. Hickok Electrical Instrument Co. countered by the Supreme model Advertisement. Radio April, 1934, 45 tube tester proved to be far page 34. wider ranging and educational *12. Front covers of 17 issues than I had expected. I hope that illustrated six base connection my next project will show as many diagrams with the 1933 pin number interesting facets as were revealed code. Radio August, 1934 to January, by the innocuous looking Supreme 1936. 45! *13. R/9 and Radio merged. Radio January, 1936. *14. Bernard, Herman. The Tally REFERENCES Method of Significant Switching in an NOTE: References preceded by an Analyzer. Radio World June 17, 1933, asterisk contain drawings or examples pages 12-14. of the use of the initial 1933 RMA tube * 15. A New Accessor using the Tally pin code numbers. Method of Significant Switching. *1. Supreme Instruments Corp., Radio World August 5, 1933, pages 8- Advertisement. Radio-Craft, Sept., 10. 1933, page 172. 228 AWA Review Kirsten

* 16. RCA-Cunningham. Charts of days of these transitions: vacuum Characteristics of All Tubes. Radio tube to solid state, analog to digital, World September 9, 1933, Front & L/C tuning to phase-locked loop sig- pages 8,9. nal acquisition, carrier signal to 17. RCA-Cunningham. Charts of pseudo-noise coded Characteristics of All Tubes. Radio reception, main-frame to personal World October 27, 1934, pages 14, 15. computer, battery power to solar cells, 18. Murman, Harry. New Tube Data. and domestically, from radio to color Radio World Jan. 12,1935, pages 12, television and talking cars. Charles 13. found his job as an electronic engineer 19. RCA Suit Settlement. Radio designing telemetry and command Retailing December, 1932, systems for unmanned interplanetary pages 26, 27. spacecraft demanding, but still found * 20. Weston Electrical Instrument time to earn an MSEE at UCLA. Engi- Corp. Advertisement. Radio Retailing neers ran fast just to stand still. June, 1935, page 38. Charles’ interest in collecting old 21. Supreme Instruments Corp., radios began in 1976 while he was as- Advertisement Radio Retailing May, signed to NASA Headquarters in 1935, page 32. Washington, D. C. After each of sev- 22. McNally, J.O. Analysis and eral trips to Geneva, Switzerland as a Reduction of Output Disturbances member of the Committee Consultatif Resulting from the Alternating- International Radiotelephonique, Current Operation of the Heaters of Charles took time to stop in Paris and Indirectly Heated Cathode Triodes. peruse the stalls at the Marche aux Proceedings of the Institute of Radio Puces and Les Halles Antique Center, Engineers, August, 1932, Pages 1263- looking for antique Recepteurs 1283. Telephonie Sans Fils. 23. Stokes, John Whitley. 70 Years of Recently, Charles has been spend- Radio Tubes and Valves, 2nd Edition, ing his time analyzing vintage test Chandler, AZ: Sonoran Publishers, equipment and writing articles on the 1997, Pages 98, 153. unusual and innovative approaches to be found there. He has written a few This article was peer-reviewed. short articles for ARC, and the AWA Journal. This is his third article to ABOUT THE AUTHOR appear in the AWA Review. Life Memberships: Antique Wire- Charles C. Kirsten is a retiree from less Association, Tau Beta Pi (Engi- over 50 years of electronic and com- neering Honorary, UCLA), Sigma Xi, munications engineering. His interest (Science Honorary, Caltech). Past in radio began at age nine, when he Memberships: Institute of Radio En- was given a crystal set to occupy him gineers, Inst. of Electrical and Elec- during an episode of childhood illness. tronic Engineers, Consultative Com- This led to a love of music, home ra- mittee on International Radio. dio construction, and later radio re- pair as an avocation with a toolmaker’s apprenticeship at Pratt & Whitney Aircraft. Enlistment in the US Navy during WWII brought train- ing in RADAR, SONAR, and commu- nications, and service on Guam as an ETM/2C. Post war studies at UCLA while working at Beckman Instru- ments Research resulted in a BSEE, and a career shift to the Telecommu- nications Division of Caltech’s Jet Propulsion Laboratories (JPL). The transfer of JPL from the Army to NASA came during the exciting Volume 20, 2007 229 Supreme Model 45

230 AWA Review