Lee de Forest claimet\he got the idea for his triode h ''audion" from wntching a gas flame bum. John Ambrose Flen#ng thought that story ·~just so much hot air.
lreJess telegraphy held excit m WIng promise at the beginning of the twentieth century. Peo ~With Imagination could seethe po.. tentlal thot 'the rematkoble new technology offered forworlct1Nk:le com munfcotion. However, no one could hove predicted the impact that the soon-to-be-developed "osdllotlon volve" ond "oudlon" Wireless-telegra phy detector& wauid have on elec tronics technology.
Backpouncl. Shortly before 1900, · Guglielmo Morconl had formed his own company to develop wireless telegraphy technology. He demon strated that wireless set-ups on ships eot~ld~messageswlth nearby stations on other ships or on land. t The Marconi Company hOd also j transmitted messages across the EhQ Jish Channel. By the end of 1901. Mar coni extended the range of his equipmenttospantheAtlan1tcOcean. It was obvlgus. 1hot ~raph ~MeS with subrhatlrie cables and ihelr lnber ent flmltations woUld soon disappear, Ships· at sea would no longer be iso la1ed. No locatlon on Eorth would. be too remote to send and receive mes sages. Clear1y; the opportunity existed tor enormous tiOOnciql gain once Jelio ble equlprrient was avoltable. To that end, 1Uned electrical circuits were developed to reduce the band width of the signals produced by ,the spark transrnlf'tirs. The resonont clfcults were also used in a receiver to select one signal from omong several trans missions. Slm~deslgn principles for resonont ontennas were also being explored and applied, However, the senstflvlty and reUabltlty of the devices used to ctetectthe Wireless signals were still hln cterlng the development of commer () cial wireless-telegraph nefWorks. Inadequate Detectors. The detector most commoniV UWd at the rum of the century was theCOherer. developed by Edouord Branly In 1890. Basico~ the coherer consis1ed of a hollow non-con ducting tube .filled with metallic filings. Normally. ·the coherer acted as .an open circuit. However, when a voltage Ual6 ... Way. Thomas IWa produced by a spark dlschdrge was EdisOn cOntributed substantially to the applied, the coherer became con ~of the thermionic diode ductive and remained that way until ft:lrPugh his 'YOrk with the incandescent the filings were shaken loose by tap lamp. In 1fJ79, he had produced a ping the tube. procflcallathp with a carbon filament ·The coherer was not a very sensitiVe that operated on DC. or reliable detector and the tapping After each lamp hdd operated for a required be1ween each telegraphic WhHe. Edison noticed that a black de dot or dash meant that messages posit of carbon formed on the Inside of could not be transmitted very rapidly. the gloss envelope. The carbon Despite its limitations, the coherer was ~to be corning from the nega the best detector of wireless signals -side ot the filament, sinOe that was available at that tlme. the Side that battery ti1Ul needed ~the flew ot carbon particles. ready, he had modified the Initio! no filament cUI"fent. leCMr)Q o ciN' line or "ShadoW' of the design of the coherer to Increase its ~on the glQ&s. sensitivity and reliability as much as SineEJh carbon particles obviouSly possible. Still, thecohererhadtoo many placed across the Slit, and electrical came from N filament, Edison lNOil fundamental limitations to ever be connections were mode to the endS of dered If they might be electrfcally come o satisfactory detector in com the foil. The water drop acted as a ~; He hdped that there might mercial use, partleularly on board a short-circuit in the abSerlcEt of a ... be a "'lf0(1:>f preventing the deposits ship rolling in a stormy sea. and as an open-circuit When o signal 1I0m ~ and reducing the fight In 1902, Marconi developed and was present. outpUt Of the lamp. patented 1wo types of magneflc de Tapping was not required ~ Edison found that a current could be .tectors that overcame the need for .each telegrQPhlc dot or daSh, but the l"''leeSUted by a galvanometer con mechanical tapping. They permitted drop Of~ became electric<:JIIyde nected ~ 0 wire el$Cirode Jn.. messages to be sent at a more rapid composed and unable to -serve It's side the bulb (but not touching the rate than did the coherer. One design function dt$1' o minute or 1wo Of use filament) and the positive el.ld of the utlltzed a constantly ro1oting mOgnet, due to the minute eleCtrical currenls power S\,IPply. 'However. no current while the other employed a rapidly that flOWed through it. eould be measured when the gal moving endless belt of iron wire. lee de Forest refined the basic de vanometer was connected be1ween Both magne11c-detector designs in sign and tried numerous substances os theetectrode and negative pole of1he volved more mechanical compieldty a replacement for the water drop. He power SlJPPI¥ than was desirable and provided less never succeeded In producing a 1ruiv This unidirecti9nal current Indicated sensHMty than was needed. The Mar reliable antl-coherer detectot for his that the current ~ng particles coni Company. however. did use mag- telegraphic responder, or ..sponder" os were negatively charged. Edison as . netic detectors In Its network of he called lt. SI.lfl'IGd the negatively chdrged partl telegraph staflons for many years until a However. Reginald Fessenden de· eteswe.e carbon particles; The flaw of better detector was available. veloped a sensitive fluid-based dEJteo elections that he was actually measur Which brings us to lee de Forest. He tor in 1902. which he called his "'lqUUd Ing WOUld not be "dlscoveredH until also had dreams of establishing a com barretter." Because It utilized a fine wire 1897 by J.J. Thomson. mercially successful wireless-telegra making shallow contact with a Jl(1uld Edison soon replaced the wire cur phy company of his own. He, too, (acid) surface, the barretter was not rent"Collecting eleCtrode with a flat realized the need for a more SensitiVe suited for shipboard use. Fessenden~ metal plate located between the legs and reliable detector and spent seN patent on the barretter was also an of the filament. The plate proved to be era! years trying to deVelop a workable obstacle to Its commercial land use. an even mqre effecflve current collec telegraph receiver. or responder, Both the Marconi Company and de tor. based on what was called the "antl Forest continued the search for a .better Edison noted that the Current flowing coherer-detector" principle. detector. Their goals were the same to the metal plate varied with the tem lee de Forest had gotten his Idea for and their searches led each to virtually perature (brightness) of the fllar(lent the detector from reading a report of a Identical thermionic devices-by follow and, hence. with the appJieO OC volt laboratory experiment that had taken Ing what de Forest claimed were to1olly age. He proposed using the modified place In Germany. The device initially different approaches. The story a the iocandescent lamp as a voltage-mea- ;:: consisted of a small piece of metal foil development of the thermionic vac suring deVIce and obtained a patent ~ cemented to a glass plate. The foil was uum tube that follows clearlY •ctemon on the design In 1883. . i slit with a razor. a drop of water was strates how mast •slgnfficarlt ~ ~belngaconflrmedbe- ..... 41 liever in the use of DC voltages only, He named his detector the "oscilla Edison never connected an AC voltage tion valve." It is important to remember to the current-collecting electrode and that being a diode, Fleming's device . never placed any greater significance could only rectify oscillating currents; it on the unidirectional current flow he could never produce them. More com had observed. Hence, he never real monly today, we know his detector as ized that his device could have been the thermionic "valve" or "tube." developed into the first thermionic rec After this first wireless-detector experi tifier diode. Nonetheless, Edison's de ment the unheated current-collecting vice was later to have a significant electrode in the lamp was replaced by influence on the development of the a hollow metal cylinder surrounding, thermionic rectifier in England by John but not touching, the filament. Fleming Ambrose Fleming. immediately applied for patents on his William Preece, Chief Engineer for oscillation valve in England, the United the British Post Office, came to America States, and Germany. The artwork at the in 1884 to attend the International Elec beginning of this story is a reproduction trical Exhibition in Philadelphia. While of the drawing of the oscillation valve there, he saw some of Edison's inven and detector circuit he included in his tions on display. Before returning home, application for an American patent. Preece visited with Edison and ob The British and American patents tained several of the incandescent Fig . 2. The second style offlame detector were granted to Fleming on Sep lamps with the added metal-plate had a local battery circuit which allowed tember 21.1905 and November 7, 1905, electrode. · it to operate better. respectively. Those dates would be Preece experimented with the modi come significant in later priority dis fied lamps for a while. He was the first to putes. use the term "Edison effect' to describe current. However, at the time, he never the thermionic emission from the fila suggested this or any other practical A Flame Sparks de Forest. In Sep mentthat resulted in unidirectional cur use for the device. tember, 1900 de Forest was working in rent flow to the metal plate. Later, Fortunately for the future develop his room by gas light with his spender Preece apparently gave the lamps to ment of radio technology, Fleming was one night and noticed something very the Edison Electric-Light Company of hired in 1899 as a scientific adviser by strange. When he operated the key of London. the Marconi Wireless Telegraph Com the spark transmitter, the gas light pany. He spent the first several years flickered. He immediately called his as Fleming Enters the Picture. John developing better transmitters and· sistant to observe this phenomenon Ambrose Fleming would be knighted power-generating equipment for and together they pondered its signifi "Sir Ambrose" almost a half-century them. In 1904, Fleming was assigned cance. later by King George V for a lifetime of the task of developing an improved Their first reaction was to believe that scientific achievement. In 1884, how detector. A suitably sensitive and rug the flickering flame represented a new ever, Fleming was a consultant to the • ged detector was still sorely needed by type of detector action which might Edison Electric-Light Company of Lon the Marconi Company. overcome many of the problems that don. Fleming soon decided to try one of plagued the sponder. Fleming had noted the same the modified incandescent lamps as a The two excited experimenters darkening of incandescent lamps and detector, remembering its ability to pro watched the flame respond to the key the formation of a filament "shadow" duce unidirectional current flow from ing of the transmitter and took detailed observed by Edison. His initial experi low-frequency AC. The question in notes for several weeks. However, when ments with the newly obtained modi Fleming's mind was whether or not the the transmitter was moved to an adja fied incandescent lamps confirmed rectification effect would work at wire cent room and the door was closed, the other findings of Edison and Preece less-telegraphy frequencies. the flickering of the flame stopped. It as well. However, Fleming made an ad An induction-coil based oscillator cir then became clear that the flame was ditional and important contribution: he cuit was built to produce electromag responding to the sound waves gener explicitly reported the unidirectional netic waves. A second resonant circuit ated by the spark transmitter and not to flow of current in an 1896 scientific pa was located some distance from it and the electromagnetic waves. per. was tuned to the oscillator's frequency. Very disappointed, de Forest He also observed that qment flowed The second circuit included one of the stopped the flame experiments. He when the filament was heated by an modified incandescent lamps, to later maintained, however, that the U) AC voltage with a frequency between gether with a galvanometer to indi flickering flame started a train of () z 80 and 122 cycles per second. With an cate the hoped-for rectified current. thought in his mind that ultimately led to ~ AC filament voltage, it obviously didn't When the oscillating circuit was acti the development of the triode audion. t) · matter which side of the filament was vated, the needle on the gal By 1903, de Forest had developed a w _J connected through a galvanometer to vanometer deflected, indicating rec flame detector that worked. He inser w a: the current-collecting electrode. tification of the high-frequency wireless ted two platinum wire electrodes into :5 Fleming had thus come close to rec signals. Fleming had found, in that Oc the flame of a Bunsen burner. He next :::> Cl. ognizing the potential use of this modi tober of 1904, the wireless telegraph connected a battery and a telephone ~ fied lamp as a rectifier of alternating detector he was seeking. receiver in series with the wires. One 42 electrode was connected to an anten He later would argue that substantial na and the other to a grounded water differences existed between the two pipe. This arrangement as it was used, is devices. shown in Fig. 1. Signals from the antenna changed A Control Element is Added. Lee de the conductivity of the flame which, in Forest was convinced that his audion's turn, changed the amount of battery sensitivity as a detector could be in current flowing. That produced an au creased even more. Remembering the dible sound in the receiver's headset. improvement the separate local circuit The flame detector actually acted as a had made in his flame detector, he form of relay with the wireless signal added an additional plate external to triggering the flow of current from the the glass envelope as shown in Fig. 4. Fig. 3. Lee de Forest's diode audion · battery. Lee de Foresfs use of the word The external plate provided what de detector allowed him to eliminate the Forest called "electrostatic control" of "relay" rather than "rectifier" would be flame by using a low-vacuum glass tube. crucial in his later legal battles. · the detector. He then built a similar de With the flame detector, de Forest vice with a coil on the outside of the was able to copy transmissions from a glass which produced "electromag ship in New York harbor. Two problems netic control." were immediately obvious, however. .The next audion de Forest built had The ability of the flame to detect the two separate internal plates. The plates signals was affected by air currents. were located on opposite sides of the Secondly. the two common electrodes filament and the second plate was used for both the antenna current and used as the control element. Lee de the battery current resulted in a by-pass Forest was amazed with the perfor path for the signal around the flame, mance of this audion. A patent ap resulting in decreased sensitivity. Fig. 4. The audion with external plication on this first triode audion was The first problem was easily solved by electrostatic control used a plate that was filed on October 25, 1906. Lee de Forest shielding the flame against air currents. external to the the tube. It was soon was granted the patent on January 15, The second problem was solved by followed by a device with an external coil. 1907. using one set of electrodes for the an It did not take de Forest long to real tenna and ground while a second set placed inside the envelope to produce ize that the control element should ex was used for the local battery circuit. · increased ionization. ert an even greater effect if it were That configuration is shown in Fig. 2. Lee de Forest filed for several patents located between the filament and The device's patent request wGJs on this two-electrode "oscillation-re plate. Obviously, a solid. control ele signed and witnessed on November 4, sponsive device" (as it was called in the ment would block the flow of current. A 1904 and filed on February 3, 1905. The patent applications) during the first half zigzag wire arrangement was used and 1904 signing date was twelve days be of 1906. More commonly. however, de was eventually developed into the fore Fleming's patent application was Forest referred to the device as his "au "grid" found in today's tubes. filed. Those dates would become im dion." The patents were awarded be The three element "grid audion" portant when de Forest argued that the fore the year's end. proved to be a much better detector. flame detector, and not Fleming's Initially. the two electrodes in de For The circuit in Fig. 5 is the one de Forest valve, inspired his later "audion." esrs audion were both filaments. Soon included with his January 29, 1907 grid Lee de Forest believed that the gases he eliminated the need for one battery audion patent application. The patent ionized by the flame produced the re by replacing one of the filaments with was granted on February 18, 1908. lay action. The flame, itself, was only the an unheated plate electrode. Lee de And thafs how Lee de Foresrs most mechanism for producing the gas Forest now found thatthe polarity of the famous and important audion was ionization needed. Other means of "B" battery in the circuit (shown in Fig. 3) "born." It like the earlier audions, was a heating the gases sufficiently to pro appreciably affected the operation of low-pressure device. Therefore, calling duce ionization would result in the the detector. it a "vacuum tube" is not correct. same detector action. At this point it started to become It is important to realize, however, that The best way to obtain a stable de clear that the Edison effect (thermionic the potential usefulness of the three tector, de Forest reasoned, was to emission) was involved in the operation element (triode) audion was not imme place the electrodes in a glass enve of the audion. Prior to that time, de For diately apparent. Few besides Lee de lope and heat the gases to ionization est had been convinced that the Forest initially were impressed with its with an external bunsen burner or by ionized gases in the tube were respon operation as a detector. passing an electric currentthrough car sible for its behavior. The triode audion was expensive and bon or tantalum filaments. Both heating This single-filament detector, which its filament life was relatively short. The techniques were tried. Lee de Forest de Forest also called an "audion," was bulky and expensive batteries it re l quickly realized the advantages of fila superior to anything he had used pre quired were another drawback. In con ment heating. viously. He received a patent on this trast, the comparatively inexpensive The air inside the envelope was device in the later part of 1906. The fact crystal detectors, developed at ap evacuated only to the point where the that this two-element audion was vir proximately the same time, did not re- ~ filament would not be oxidized when tually identical to Fleming's patented quire batteries. The crystal detectors <0 heated. A sodium or potassium salt was valve was of no concern to de Forest. had added advantages in that they ~
43 II
He merely used the incandescent lamp D. Edison had modified and patented in b 1883 for a new purpose-the detection of wireless signals. Edison's prior patent, I together with some very broad claims
:vI in Fleming's patent application con I cerning the rectifying capabilities of M the oscillation valve, made Fleming's 2 legal position questionable. As the markets for both Fleming's 4 valve and de Forest's audion were ini c tially relatively small, no resolution of the - ,,,, + legal questions was pursued. The Lee de Forest Radio Telephone Company A sold wireless equipment that incorpo rated grid-audion detectors to the U.S. Navy. It also sold the spherically shaped grid audions for commercial use. The Fig. 5. Lee de Forest's grid audion was the first grid-containing low-vacuum tube. Marconi Company used the Fleming Multiple-electrode tubes were soon to follow. valve, modified to include a grid similar to that in de Forest's triode audion, in its were more reliable. simpler to operate. considered to be a relay), Fleming's de own network of wireless stations. and virtually indestructible. vice could do nothing to increase the In his 1906 patent application, de effective energy of the signals. Amplifiers and Oscillators. In 1911. Forest called the original triode audion The explanation de Forest provided de Forest was hired by the Federal Tele a "device for amplifying feeble electric concerning how his detector func graph Company. The company's goal currents." The techniques necessary for tioned differently· from Fleming's was of recording high-speed telegraph sig using it as a true amplifier, however. wrong. Both Fleming's and de Forest's nals for later decoding created the would not be developed for almost six devices were rectifiers, virtually identi need to increase the energy of the re years. cal in appearance and operation. The ceived signals. Lee de Forest now battery de Forest used in the plate cir worked to get the triode audion to am Patent Problems. Other problems ex cuit of his audion put a positive bias on plify at audio frequencies. isted for de Forest. too. The Marconi the plate. but was not necessary for its Lee de Forest and his co-workers Company of America owned the pat operation as a rectifier. Lee de Forest's achieved amplification in 1912 by using ent on Fleming's valve and maintained device was not, in any sense of the audio-frequency transformers to cou that de Forest's two-element (diode) word, a relay. ple the signal to and from the audion. audion infringed on that patent. Fur While the legal position of the de For They also found it extremely beneficial ther. Lee de Forest's triode audion con est diode and triode audions was not to reduce even further the already low cept was based on his diode audion. very clear, that of the Fleming valve also pressure inside the audion ·envelope. In Consequently, de·Foresfs right to man was in question: Strictly speaking, Flem time. the audion's spherical shape ufacture and sell the triode audions ing did not "invent'' the oscillation valve. would become tubular and the gener was not without dispute. ic name for this and similar devices. Lee de Forest maintained that the very appropriately, would become the idea for his diode audion came from his Books and Articles. "vacuum tube." earlier flame audion, not from Fleming's In the course of developing the am valve. The records show. however. that "Lee de Forest and the Triode Detector;" By plifier, de Forest found that his circuit de Forest knew about the Fleming valve Robert A Chipman, Scientific Americon. oscillated, a phenomenon commonly and had some copies of it made for his March 1965, pp. 92- 100. encountered by amplifier builders own experiments late in 1905. In fact, Saga of the. vacuum Tube; By Gerald F.J. even today. The discovery that the tri one of de Forest's patent applications Tyne, Indianapolis. Howard W. Sams & Co.. ode also could be used as an amplifier (for what he called a "static valve"), 1987. or as an oscillator established it as a filed on December 9. 1905, refers to 76 ~rs ofRadio Tubes dnd Valves; By John truly important electronic device. Fleming's valve by name. That patent W. Stokes, Vestal. N.Y., The Vestal Press ltd., application was filed several months 1982. Lawsuits and Stalemate. The Mar before he applied for patents on his The Continuous Wave: Technology and coni Wireless Telegraph Company of (f) own diode audion. American Radio; By Hugh G.J. Aitken, America sued Lee de Forest and the S:2 A further claim by de Forest was that Princeton, N.J.. Princeton University Press, z 1985. Lee de Forest Radiotelephone and 0a: the use of a battery in the plate circuit Telegraph Company in 1914 for in 1- "The First Electron Tube"; By George Shlers. () of his diode audion distinguished it from fringement of its Fleming-valve patent. UJ Scientific American. March 1969, pp. _J Fleming's valve. He also argued that Lee de Forest a.nd his company filed a UJ 104-112. Fleming's valve was only a device for countersuit for infringement of their tri a: "The life and Work of lee de Forest'; On 14 :5 rectifying high-frequency wireless sig ode-audion patents by the Marconi ::::> Parts), Edited by W.B. Arvin. Radio News, Oc2 a.. nals. Lee de Forest maintained that. un tober 1924-November 1925. Company. lr like his own audion (which he still (Continued on page 85)
44 SLOT MACHINE VACUUM TUBES (Continued from page 39) (Continued from page 44) diameter of the fiber-optic cable for a The Marconi Company admitted its snug fit, and the other holes should fit guilt and was prohibited from further the hardware you're using. infringement. On September 20, 1916, a Attach a U-shaped mounting brack U.S. District Court ruled that de Forest et to the coin slot as shown in Fig . 8. Fit and his company similarly had violated the emitter assemblies on one side of the Fleming patent. In short. neither the EARN YOUR the coin slot and the detectors on the Lee de Forest Company nor the Mar other. Now simply twist and solder them coni Company could manufacture tri B.S.E.E. to the leads coming from the circuit ode audions without infringing on the board. Follow the polarity indications other's patent. DEGREE supplied with the exact parts you use in This legal impasse was not resolved THROUGH HOME STUDY your circuit. until after World War I. During the war, Our New and Highly Effective Advanced-Place Make sure that the circuit is counting however, de Forest was granted legal ment Program for experienced Electronic Tech nicians grants credit for previous Schooling and change properly before installing it in immunity to manufacture grid audions Professional Experience, and can greatly re side a cabinet. Since this article is really for the United States Government. duce the time required to complete Program and reach graduation. No residence schooling re about the electronics, we will not go quired for qualified Electronic Technicians. into the details on how the cabinet was Legacy of the Audion. Once the tri Through this Special Program you can pull all of the loose ends of your electronics background assembled, but the basic measure ode audion's capabilities for amplifying together and earn your B.S. E. E. Degree. Up ments are shown in Fig. 9. The elec and oscillating were discovered, the grade your status and pay to the Engineering Level. Advance Rapidly! Many lin ish in 1 2 tronic circuitry and the slot assembly development of regenerative and su months or less. Students and graduates in all 50 mounts inside the top of the unit. The pefheterodyne receivers quickly fol States and throughout the World. Established Over 40 Years! Write for free Descriptive Lit top and bottom half of the bank are lowed. Commercial radio broadcast erature. attached with hinges to allow empty ing began shortly thereafter. In just a ing of the change compartment and few years, the triode vacuum tube was access to the reset button. The top half joined by its more sophisticated tetrode of the bank has a fairly large cutout in its and pentode relatives to make possi base to allow .the coins fall through into ble even more sophisticated electronic the bottom compartment. • circuits. •
This publication is LASER COMPUTER (Continued from page 66) available in microform from University lot more things than I need or want. I Microfilms make use of the wordprocessor be cause of the convenience in writing International. when and where I feel in the mood, as opposed to being captive at a desktop location. The documentation is clear, the programs are largely intuitive, and file transfer is fast and easy. A privately produced quarterly news letter devoted to the PC4 is available from Michael Posner, LUGNuts (Laser THE MONEY MAKING OPPORTUNITY User Group Nuts), 4200 Community OF THE 1990'S IF you are able to work with common small hand Drive #1514, West Palm Beach, FL tools, and are familiar with basic electronics (i.e. able 33409, for $10 a year. You can request a 0 Please send information about these titles: to use voltmeter, understand DC electronics) ... . free sample copy. IF you possess average mechanical ability, and have a VCR on which to practice and learn ....then we can The Laser PC4 is available directly teach YOU VCR maintenance and repair! from Laser Computer, Inc. (800 North FACT: up to 90% of ALL VCR malfunctions are due to Name ______simple MECHANICAL or ELECTRO-MECHANICAL Church St., Lake Zurich, IL 60047-1596), breakdowns! for $249.95 plus shipping. The order Company/lnstitution ------FACT: over 77 million VCRs in use today nationwide! Average VCR needs service or repai r every t2 to 18 phone number is 800-LASER43, or you months! can call Laser at 708-540-8911 for your Address ------Viejo 's 400 PAGE TRAINING MANUAL (over 500 pho tos and illustrations) and AWARD-WINNING VIDEO local dealer. If you get a PC4 and need City ------TRAINING TAPE reveals the SECRETS of VCR mainte technical support. call Laser at Sta te ______ZiP·------nance and repair-"real world" information that is NOT available elsewhere! 800-551-5742 or 708-540-5022. For Phone ,_(---'------ Also includes all the info you'll need regarding the more information contact the man Call toll-free 800-521-3044. In Michigan, BUSINESS-SIDE of running a successful service op- Alaska and Hawaii call collect 313-761-4700. Or ufacturer or circle No. 119 on the Free mail inquiry to: University Microfilms International, eration! FREE INFORMATION 300 North Zeeb Road, Ann Arbor, MI 48106. CALL TOLL-FREE 1-800-537-0589 Information Card. • Or write to : Viejo Publications Inc. 5329 Fountain Ave . Lo s Angeles , CA 90029 Dept. PE 85 CIRC LE 20 ON FREE INFORMATION CARD II
