- VACUUM TUBE VALLEY.
Issue 7 published Quorrerly Spring/Summer 1997 Celebrating the History and Quality of Vacuum Tube Technology Price S8.00
The 6DJ8 and other Frame Grid Tubes
V.cuum Tu".Computen The Savage Art
al.-Tlm. Th ...... Sou.d A'tee Lansing Pov#erhouse
. Th. P.ter ...... to ry Early Sound Amplification
__ Ien.en ELECTRO-OYNAMIC SPEAKERS .�" EDI TOR S P AGE AND INDUSTRY NEW S
NewTube!i from S.-etlana VTV Issue # 7 $vedana Elecuon Devices has recently intro Table of Contents: duc",d their version of the EF8G. Thi� TUbe k:l.ture� low noise, high gain, internal shield All About the 60J8 and Other ing and gives exrTcmciy low dislOrrion in Frame Grid Tubes ...... 3 either pcmode or triode conneccion. Retail price of the EF8G is S 14 US. listening to the 60J8 .8 VIV recei,'� pre-producrion wnp les of vetlana's SV6L6GC, which is a very dose: Oscilloscopes· Part 10 S 2 copy of the original Sylvania 6L6GC/STR387. These tubes work very well in guirar :unps, 13 A/tee1500 Series both new :and vimage. 50nically, they :Ire sim ilar to the original RCA 6L6GC blackpb.lcs The Great Voice - Peter Jensen 18 and the STR387 Sylnnias.Retail price is 518 US for single tubes and 524 each in marched Build A Comparison Preamp .. 24 pairs or quads. L�rgc qU:l.milies of Ihis tube will be availabk in I.m, 1997.
Antique Sound Lob .. 26 Swriallll AQ-l 002 Svedana has �ho introductd :;I.n amazing power triodt. The new 3CXJOOAI looks like 3008 Tubes ...... 28 Computing With a lran�mit!ing tube, but is actually a low-mu audio.The OESKB is an affordabk trans audio triodt. Pbte diuiJnlion is 300 W.U1S 3 former with a 5000-ohm primary, inltndtd to Ceramic Capacitors . and plate rcsisrana is 4}O ohms. It is an 1 match tubes such as 45, 50, SV811-3, 300B exttrnalanode typeand requirc:s fOruG-air and the likt, running up 10 65 mA pblt cur Computer Video Review . .. , 32 cooling. If driven correctly, in Cl:lS$I, AB a .tnt. The similar TM3KB has a 3000-ohm pair of Iho:sc: tubes would produce more than primary, intended for a single 2A3, 6B4G or VTV at CES .34 400 wam! 1997 300B running at up to 70 mA. Bandwidth of Pre-produetion samples of the long awaiTed both types is 18 ]'h 1040 kHz, _I dB, with Tube Dumpster ..36 - 417A/5842 Svedana version of the 3000 power rriode maximum power raring of 10 waItS and with werc received by vrv for evaluation.The output apabiJides for 2-ohm through 16-ohm tube is well made, with heavy glass and inter spt'akers. nal components thal appear idential to the Fo ••e allr big po�r, the Electr;J.-P.int original item. It is clearlysuperior 10 the OTher KJ..IOKB is a J>C'rfect match for a single lcss-expcnsi\'e 300B5 on the market. [t has SV572, 211 or 845, In 10k-ohm primary and greot bass, extended high frequency response 3O-w:m r;J.ting go along with a bandwidth of and is extremely music..1.The $\'tTlana 300B 19 Hx to 24 kHz at -I dB--world-c!:lS$J>C' rfor will not be available umil lare 1997. Price has mance for such a high primary impedance. not been determined yet. For more informa ti on, contact: Svetlana Electronic Devices, And finaU)" Eleclr;J.-Prim has inrtTSnge 8200 Memorial Parkwar, J-\untsviHe, Alabama transformers for SE and push-pull applica 35802, FAX 205-880-8077, Phone 205-882- tions. The 3533-2 has a primary of 8.5 1344. KOhms and a 1:1 ratio. The primary can han dle up to 20 mA of DC plate eurrem.For Electra-Print Introduces New push-pull, the 3634-P has a 9k primary and Transformers cemer-tapped secondary, with a 17 mAp rima EJeclr;J.-I'rintAudio Co. has stan«l. manu ry currem. Contact: Electra-Prim Co., 4117 facture of thre.:: new single-en d«l. OUTpUt Rox.anneD ri"e, l2s V�as NV89108. phone (r;J.nsfonners intended for maximum-quality (702) 396-4909, ru (702) 3%-4910.
EDITORIAL STAFF Vacuum Tube Volley is published quarter Copyri�t C1997 Vocuum Tube Valley 1.W. ly for electronic enthusiasts interested in the and Big Tone IM All rights reserved. Charles Kittleson - Editor and Publisher colorful post, present and future of vacuum Unauthorized copying or duplication of this tube electronics. John Atwood - Technical Editor publication or any materials within it is strictly forbidden without permission from publisher. Annual subscription (4 issues) is $32/US Eric Barbour • StaffEditor $40/Canodo, $SO/Europe $60 Asia. Send circulation and editorial correspon Terry Buddingh - Guitar Amp Expert Cosh, US Bank check or Credit Cords dence to: Vacuum Tube Valley ore accepted as payment. Sieve Porr - Art Diredof Phone/FAX 408-733-6146 1095 E. Ouone Ave., Suite 106 Julie P. Werner ·Copy Editor Svnnyvale, CA 94086 USA Web Site: http://www.vocuumtube.tom email ·[email protected]
VACUUM TUBE VALLEY ISSUE 7 ALL ABOU T THE 6 D J 8
has to have high uansconducrance; and to use it at UHF or higher frequencies, All About the the tube strucrure has to be small. The limit given for conventional rube manu facture is usually the 6AK5 pemode. with 6DJ8 and other a rated rransconductance of 5000 to 5500 pmhos and a grid-plate capacitance of Frame Grid Tubes 0.02 pF. Ra)'theon's GAHG bettered that in 1946. reaching a Gm of 9000 p.mhos By Eric Barbaur © 1997 All Right!. Re�rved with some difficulty.
1. In tro The developmem of the frame grid took place simultaneously on both side:o; In previous issues ofVIV, wc have of the Atlanlic. in 1948-49. Tungsram introduced you to many tubes that arc (UK) engineer John A. Sar rove was the now popular for audio use, and nothing g forgotten British pioneer of this tube but. And we have told you that many of type. He developed the Tungsram UA-55 these popular lypes weren't even intended as a dual-section beam tetrode for audio for the current applications when they power applications. ironically. BUI it was first introduced. So, just to warn wefe apparently a commercial failure. It was you in advance, it's time to do that again. sensitive and was also designed to operate watt on only 90 volts Many audiophiles today swear by the efficiently, giving 1 60J8/6922 dual uiode family. They lend plate power. This was made possible by the dose spacing that frame grids allow. to think of these as the most detailed sounding tubes of all. Yet lhal whole technology and design was intended for low-noise RF appliColtions, usually in receiver front ends. Audio was hardly considered until the 19705. These types are consnucted with a complex kind of ri d which was pioneered in 1948, g known as a frame grid. Sophisticated e([uipment is required to make frame grids, whereas the convemional tube grid is juS[a spiral of wire wrapped around twO soft metal POSts. The idea of using frame-grid tubes for hi-fi would have been unthinkable in the early Fifties, but WE 416A Muliard/OM 6DJ8 evemually their technical superiority and common manufacture made them viable for home audio equipment. Special tubes were developed for UHF The credit for the frame grid usually before World War 11; they were usually 2. History goes to J.A. Monon and R.M. R)'der of made as physically small as possible to BeI[ Te lephone Labs. They introduced the reduce capacimnce and parasitic induc Unlike the conventional tube grid, the 41GA rriode at the winter 1949 meeting tance, but were nOt outstanding in frame grid is JUSt what it seems. Rather of the AIEL This remarkable (and odd transconductance. This family indudes than being wound around a swaged pair looking) metal tube was intended as a the Western Elecuic �doorknob" IUbeS of postS, very fine wire is wound onto a microwave RF amplifier for telephone such as the 316A. 380A thm 387A, 713A rectangular frame of stamped metal, often long-distance relay C
VACUUM TUBE VALLEY ISSUE 7 ALL ABOU T THE 6 D J 8
systems and specialized military equip ment. AT&T wanted to expand long-dis tance networks after the war, and microwave was seen as the only COSt effective way to do it. So the develop ment of frame-grid cubes was sparked by telephone application more than any other. RF transistors were awful unlil the THEN 1 960s, and so frame-grid tuhes like the exactmg \'enerable and popular 417A ruled the military VHF and UHF bands. Some high-perfor mance 2 meter amateur receivers and .ystem •... conveners of this era used 417A front ends.
The race for more tran5conducl'ance continued through the 19605. GE held many of the records with their ceramic planar uiodes, starting with the 7077 in 1954 (Gm-IO,OOO) and moving all the way up to the 7768 (Gm.,SO,OOO) and the 8917 (Gm.65,QOO), which has been unbeatable since the late 1960$ except by some bipolar transistors and very exotic MOSFETs.
I should also note that the cubes are mueh more linear than the MOSFETs. Tubes follow a 312 power relationship, while MOSFETs follow a square law and bipolar transistors follow an exponential rdationship. The transconductance of the Tile frame grid is the closest transistors varies considernbly with drain or collector current, as we!! as with the a1Jpl'oach to tile ideal "Phys ambient tempcnlUre; while tubes vary in icist's grid" -electrical clial' transconductance only moderately with acteJ'-j,'�tics but no physical plate current, and hardly at all with tem dimensions. It results in:' higher perature. The input capacitance of tran transconductance per milliampere sistors (espccia!!y MOSFET5) changes drnmatically with bias. In spite of 40 • tighter Gm and plate current tol· rears of aggressive R&O on semiconduc erance • Iow transit time· Iow ca tors, frame·grid tubes remain very dose pacitances • lower microphonics in RF noise performance to lhe world's rugged construction best transistors.
And frame.grid tubes have other advan· tages over the current crop of semicon ductors: high tolernnce of £SO and EMP, grio;l' � p TM O-<:;}II'IOCIe s acln, \ole,. wider dynamic range due to high-voltage ance i s �It'm; � oy th' ure1ullJ cotUr Ol1ed diameter 01 Irid $UPPG'I operation, and a tendency towards lower roel �n Ihrse rOdS. ul,em,IJ fine l"cI w i re tll",. tronics industry has declared tubes obso lI'Iale s Ihe "Island elleet'" usually lete. See the Rohde article in the biblio entoun l e.t'(I In CCIII.. cnh onal luon witr. eQu. U y o;lOie l 'k H a VACUUM TUBE VALLEY ISSUE 7 • A LL ABOU T THE 6 D J 8 rated some special manufacturing processes derived from framc-grid meth ods. Sylvania developed a few frame-grid power tubes for TV sweep and audio applicaTions; this included the 6FG6 and GOY?, which we will cover in a future article. By rhe 19605, frame-grid tubes became common in TV sets. Single-sec tion variantS of the 60)8 such as the 6EH5, 6GH5, etc, replaced the dual-tri ode cascade as rhe front-end RF tube. Frame grid pentodes such as the 6EH7, 6EJ7 and 12GN7 were used for IF and video amplifiers. The 60)8 was introduced in 1957 by Ampcrcx, the American division of Philips. It was developed by PhiJips in Holland, under the European standard designation ECC88, and intended as a cascade amplifier for television VHF and Ampmx Tubn: Bugk Boy 6D)8, g!qb"logq 6D)8, 6922 and 7308 PQ UHF tuners, nothing more. The series string version, 70)8/PCC88 was also introduced at this time. It was apparently a higher-performance descendant of the 6BQ7 dual triode, which was also intend ed for RF cascade circuits. The 60J8 became a popular eype, especially when the engineers at Tektronix discovered it. They found that it was consistent enough for their oscilloscopes, and gave excellent pulse fidelity. So, it can he found in the sweep and vertical amplifiers ofTek's major tube scopes, Hafting in 1959. Since the 60)8 was low in cost, it also ended up in a great deal of military and com mercial radio equipmeor. h first appeared in home hi-fi equip ment as a front end for FM tuners. The classic Fisher FM-90X tuner was the firSt \0 use the 60)8 in 1957, although the tube was a devclopmenral unit at the MuUaTd Tubl'$.· CV2493, 6D)8 - !OM, ECC88 (!97M and E88CC (1960s) t"ime. Fisher called it a "Gold Cascade", and for a time it was exclusive to Fisher. Bur everyone wamed this hot tube, so it was issued a standard RETMA code: 60)8. For many YC3fS, Ampetex/Philips was the major source, with Siemens, Telefunken, GE and Sylvania weighing in with their own versions later. Premium 60J8 types appeared during 1959 co 1961. The 6922 was an industri al version, introduced in the USA by Amperex. The rare and expensive 6922- PQ version was from Amperex, with gold-plated pins and having the {WO tri odes in each envelope carefully matched. This was an early example of a "super premium", although this was done by the original manufacturer and for use in criti cal applications, not audiophile equip ment. Many i'Qs were used by Los A1amos National LaboralOry in custom built electronics. Siemens innoduced Sil'mtm TubtI: 7308, cCa, E188CC and ECC88 their own super-6DJ8 at the same ti me, VACU UM TUB E VALLEY ISSU E 7 ALL ABOU T THE 6 D J 8 in the form of the type' uCCa�. Apparently this was a contraction of tht' original European designation for the � 6922, E88CC, with the MaM indicaring a . premium version. In 1961 . Amperex introduced a special version, the 7308. h was Ihe first 10 have gu,mlntccd low microphonia, suggcsting use as the input stage in high-gain audio as wdl as RF prcamps. Nevertheless. these tubes were used pri marily in VH F receivers, TV seu, tcst equipment and nuclear instrumentation. Onc of the first audio uses for the 60)8 was the Marant'"l9 power amp of 1960. This was onc of the first � high.end � audiophile amps cveT made, long before the [eflll Uhigb_end" existed. Its use in the 9 guaranteed that the 6D)8 would have a cerrain cachet with audiophiles. 6DJ8 Companion: Simum, Sylvania.Amp�rrx and Tungsram The 9 is now a roue coll«lor'$ item, so popular in Asia that Marant"t rec�ndy r�introduccd it, with th� same circuitry 6N23Pf6922 Russian 90s. 075 .047 ·1.07 60J8 Ampercx bugl( 60s' .092 .052 -0.81 and most of the same components. 6922 RCAlSiemens1973 .075 .052 ·1.04 60)8 Am]X"re: bugle 60s' .092 .067 ·0.78 6922M PQAmpI965 .075 .052 ·1.29 60J8 HPfAmpcrex 60s' .092 .070 -0.73 In the late 19705, Audio Research 60J8 West.lSi�mens1962 .077 .065 ·1.24 60J8 HP/Amperex 60s' .092 .072 -0.66 introduced the SP-6E preamp with {Wo 60}8 GoIdcn Dngon 90s' .080 .052 ·0.82 60J8 Tclefunken 60s' .095 .047 -0.85 .095 .072 -0.73 60J8s in the phono nage. This was fol 6922 Heinz-Kauf.1 972 .080 .055 -0.95 60J8 HPfAmpere: 1974 ' lowed by the SP-B. then by the massive 60J8 AmpefCX bugle 1962 .082 .060 -0.79 cCa Siemens60s' .097 .072 -1.16 SP-IO whieh used them in all the gain 60J8 Tdcfunken60s .082 .060 -0.82 6922 Srlvani�JAN 1972 .097 .Q75 -0.88 .100 .060 -0.67 Stages. Since then, the 60)8 has pushed 6922 Siemens 60s .082 .060 ·0.84 6922 PQ Amperex 60s' Ampcrex .082 .065 -0.90 60J8 Ampcfa: 60s .100 .062 -1.20 the l2AX7 OUt of most high-�nd pr� 60J8 1977 bugle 6922 Sovlek 90s .085 .050 -0.75 60J8 Amperex 60s' .100 .065 -1.03 amps. Currently, th� 6D)8 (Of 6922--the 60J8 Mullard (Bril) 60s .085 .057 -1.30 60J8 Gr 70s .100 .Q70 -1.30 {Wo types arc basically interchangeabl�) is 60J8 Gold Aem 80s ' .085 .057 -0.97 60J8 Am crcx 60s' .102 .060 -1.04 found in reamps and oth�r equipm�m p p 60J8 West. (USA) 1963 .085 .070 -0.96 60J8 Ampcrcx bugle 60s' .102 .065 -1.02 by Audibl� illusions, Sonic Fronti�rs, 6N23P/6922 Russian 90s .090 .055 -1.10 6922 Ampere: PQ 1%2 .102 .065 -1.01 Melos, Dynaeo, Balane�d Audio, Music 6922 Ampcrex bugl� 60s' .090 .057 -0.68 60J8 HPfAmpercx 60s' .102 .065 -0.69 Reference and tOO many others to 6OJ8 RM1 Labs 90s' .090 .057 -0.80 60)8 Amperex bugle 60s .102 .070 -1.22 remember. 7308 PQ Ampcrex 60s .090 .060 -1.22 60)8 CBS/Si(m(ns1960 .105 .050 -1.23 6OJ8 SrI (USA) 60s .090 .070 -1.12 60J8 Amperex bugle 60s 105 .065 -1.12 3. Tests 7308 RCAlSicm(ns1969 .090 .070 -1.03 60J8 Amperex bugle 60s .105 .065 -0.65 Because of the ongoing popularity of 6922 Sovrek 90s .092 .045 -0.92 6DJ8 Ampmx bugle 60s' .105 .067 -0.99 the 60J8 family in high-cnd audio, we d�cided 10 gath�r som� NOS samples and subj�et them to the same distortion tesl ing that we had previously done 10 the 12AX7. After the tables, we discuss the results and some more a plications info about these popular dua rriodes. F Table 1: Distortion tC$ls al 10 yrmsOUl, 48k ohm pl�le load. I.istOO in ord�r of increasing 2nd harmonic diSlortion. '. goodused lub!:. H+ was 25oY, filamcllI6.3ydc. Type Distortion: 2nd 3rd Bias 6922 Sovr�k90s .037%.027%·1.10 CCaSiemens60s .055 .055 1.31 6922 Sovtek 90s .060 .040 ·0.81 60J8 Golden Dragon 90s' .062 .052 0.74 ECC88 Runian 70s .065 .055 -0.67 6922 RCNSiemens1965 .070 .050 ·1.11 60J8 Am perex bugle 1962 .072 .055 ·1.20 60J8 RCAlSiemells1966 .072 .055 -0.92 6922 GEfTcl�funken 70s .072 .057 -0.84 TungJram 6Dj8. MalSushira 6Dj8. Sqvuk6922. and lOOJian 6N23P VACUUM TUBE VALLEY ISSUE 7 ALL ABOU T THE 6 D J 8 7308 Raydlcon(USA) 1968 .105 .070 -1.10 60J8Tek tronixlAmp 60s' .107 .060 -1.60 6DJ8 Siemens 80s' ,107 .067 -0.97 CCa Siemens 60s" .107 .080 ·1.02 GDJ8Telcfunken 1972" .110 .057 -1.23 60J8 Ampere,.; \975 .110 .065 -0.99 6DJ8 Ampm>: bugle 1%7' .110 .067 -0.71 6922 JAN Amperex 1971' .110 .075 -0.98 60}8 GE 70s' .112 .065 -0.90 60J8 Amperex 1975 .112 .067 -0.89 60J8 Amperex bugle 60s .112 .075 -1.10 60J8 Ampere>: bugle 60s' .112 .075 -1.53 60J8 West.fSiemens 1961 .112 .085 -1.33 6922 JAN Sy!vania 1972' .112 .090 -0.58 60J8 Amperex bugle 605 .115 .072 -1.16 6D}8 MuUard/Philips .115 .072 ·1.05 60J8 Amperex bugle 60s' .115 .080 -0.68 GKN8 lEe (Japan) 70s .115 .090 _1.24 6922 JAN Ampercx 1967 .117 ,072 -O.9S 6DJ8 Amperex bugle 60s .120 .072 -0.99 Sylvania6Dj8, 6922, lWythron 7308 (AilUS madt) 60J8 Amperex bugle 60s' .120 .075 -0.73 6922 JAN Sylvania 1977' .125 .OS2 -1.01 obviously benefit from grading and selec· they arc both sold in the West as 69225. 60J8 Ampere" bugle 60s' .125 .OS7 -0.75 tion. The Sovteks were highly variable, Either one is acceptable, though the 6922 JAN Sy]vania 1970' .130 .072 -0.S2 even among the 4 samples tested. One 6N23Ps did not do as well in lisrening 60JS Ampere" bugle 60s' .140 .107 -0.73 major advanrage for ,he NOS tubes by resrs in line srages. Phono stages seem [Q 6922 JAN SylV;lnia 1973" .145 .OS7 -0.S5 Amperex (made by Dutch Philips), be less critical of sound quality than line 6922 JAN Sylvan ia 1977' .150 .100 -0.93 Telefunken and Siemens is thar they arc stages and arc better places to save 6£5S Ampoe" bugle 1967 .212 .200 -1.06 reasonably consistent. Note rhat their money, alThough noise and microphonics 6E.SS RCNMullard 70s .235 .215 -1.26 RSDs were lower than for �he Sovteks, in become more important in low-level spite of a range of many years of manu phono preamps and individual selection Table 2: Averages and RSDs (regular facture. A general consensus of users is can be necessary, even with frame-grid standard deviation) of 2nd harmonic dis- tonion for most common versio/lS. thar the Sovtek 6922 is good, but not tubes. quite as smooth as (he NOS types. Version hd RSD , The appellarion 6922/E88CC no Sovrek 6922 .06S5 .0251 4 The Sylvania JAN 69225 were nOT the longer means what it once did: a high Siemens 6922,7308,CCa .0823 .0175 7 greateSt, and my listening tests tend to quality version. The old 7308, CCa and Siemen� 60J8 (various) .0946 .01S6 5 confirm this. These are common military PQ were super-premium versions, with Telcfunken 60J8s .0957 .0140 3 surplus from rhe 1970s and 1980s. And guaranteed matching between the tWO tri Ampere" 6922 (various) .0990 .0149 6 as you've seen in Glass Audio, beware of odes and other characterisrics rested for Ampcortx 60J8 (various) .1041 .0140 29 6ES8s relabcled as 60)8 types. 6E58 is a and controlled. They could be discerned Sylvania JAN 6922 .1265 .0199 6 TemOte-Cllwff version, and will work in by the gold plating on their pins. This 6£5S RCA/Ampere" .2235 .0115 2 the socker bur will give much higher dis was unusual until recently--all kinds of Table 3: Some of the same tubes abo\'e, tortion. There is at least onc dealer rcla cheap rubes are now sold as premiums, tested as above, but with a 12k-ohm load beling 6ES85, and I have been unable [Q and have gold plating on rhe pins. In the resistor. This information is NOT intend- find out who. 50s and 60s, gold pins meant�. ed for tube selection guidance, but only Today it's JUSt a markering ploy. Given a for illustration of how distortion can vary There are a lot of Amperexes shown choice between a dirty, used Bugle Boy (or not) with ate load. Arranged by here because rhey were the biggesr sellers and a beautiful Russian with gofd pins, increasing 2n tharmonic distortion. for many years, and aTe often found in I'll clean offthe dirt, thank you. old rest equipmenT. It should be noted Type Di§tortion: 2"d 3rd Bias that not all of them had the Bugle Boy A fewnotes about the test conditions: 6922 Sovtek 90s .015 .025 -0.91 logo; it was dropped around 1970, and the load resis�nce is a relarively high 48k 6922 Sov[ck 90s .050 .060 -0.88 subsequent tubes just had the Amperex ohms, because this is seen in older 19705 6DJ8 Golden Dragon 90s' .057 .040 -0.S3 globe. preamps that t'ake 60J8s. Most high-cnd 60J8 Golden O",gon 90s' .060 .050 -0.S2 equipment since rhen has used active 6DJ8TektronixITele 60' .065 .060 -1.23 European versions were ofTen rclabeled; 6DJ8 Siemens SOs' .072 .057 -0.92 loads, and runs 6DJ8s very hard (10-12 witness the Siemens units with CBS and mAper Iriode is typical). These tubes will 60J8 Amperex bugle 60s' .082 .070 -0.60 Westinghouse marks from the 1960s. GE 60J8 RAM Labs 90s' .085 .D70 -0.80 give lower absolute distortion figures in was the only major American 6DJ8 man the actual preamps or amps, bm will tend 60J8 Ampcorex bugll967" .090 .065 -0.67 ufacrurer aside from Sylvania, and rheirs 6922 PQ Amperell 60s' .100 .062 -0.61 to wear our much fasrer. If you don'r were distincrive--inside the glass is a thin 6DJ8 GE 70s' .105 .D70 -0.83 believe me, ask any Audio Research SP- Aquadag coating. 60J8 CaS/Siemens 1960 .107 ,072 -1.19 12 owner. 60J8 Ampell'll bugle 605- .190 ,067 -0.83 [ The Sovteks arc quite different from We left out input signal levels, seen in the generic Russian 6N23Ps shown here. 4. Summary the llAX7 list in VTVI, # because cubes These cubes arc said CO be made in rhe in the 6DJ8 family seem to vary less in Even though the Russian 6922s came same plant (Reflector Saratov) bur are gain, much less than 12AX7s. Out on top in distortion, rhey would obviously different tubes, even (hough VACUUM TUBE VALLEY ISSUE 7 LISTEN ING T 0 6 D J 8 s Microphony was also not tested for in this series of tests. However, as with 12AX7s, the bias poim that gives lowest Listening Evaluations second harmonic distonion can vary gready from tube to tube; so that chang ing to an identical new tube can change of 6DJ8 Tubes the sound character. There's another argu ment in favor of mat<:hing and grading. By Charles KiHlesan © 1997 All Rights Reserved 4.0utro Over the last several months, vrv probably different and you may be per The 60J8 is rmanemly ensconced in condm;:ted what are probably the first fecdy happy with your 60)8s and the the audio worl . Although it sees little r ever published listening evaluations of the equipment that uses them. However, use: in pro-audio equipment or guitar 6DJ8 family or tubes. Using the preamp readers should be open to other opinions amps, it is dominant in high-end pre tube evaluaror built by Eric Barbour fea in order to learn more about the hobby and in some power amps. :Imps tured elsewhere in this issue, Stc:ve Parr, and potential sonic improvements to Continued manufacture in Russia and Terry Buddingh, Don Penee, Eric their syStem. The following is a summary China are assuring its availability. In an Barbour, John Atwood and J listened to of rhe listening evaluations: era when most high-gain RF tubes have what seemed like a hundred different become thoroughly obsolete by proxy. rypes and varieties of the 60)8. Amperex onc RF type has survived by becoming an 60)8 Bugle Boy (Outch -1964) � Good audio rub,. Amplifiers used included an upgraded lows and mids, but the highs were a bit Scon LK-150, an AudioNote Kit One brighter than the Siemens. Used in a lot SE-300B, a Oynaco Stereo 70 and the of rest equipment in the 19605, many Richardwn One Electron Randall Amplifier. used examples available. Speakers were KJipsch Chorus Ones and APD Buglt B &W OM [!Os with a modified Elite 60)8 Globe Logo (West German- &y6DJ8 CD player or the prororype Oynaco 19705) - Similar to early Bugle Boy. COY-Pro CO player. Program material Great bass and very good musical involve included various classical, jazl.,vocal and ment. rock COs including: �The Best of Chesky - Volume I," Chuck Locb 6922 (US-PQ 1965) - Balanced, but Simple Things� and other good record somewhat harsher than Euro 69225. �mgs. Two-dimensional and drier sounding than Bugle Boy. We noted in our listening tests that the 60J8 is nOt the most neutral sounding 7308 (US-PQ 1973) - Glorious, tube. The rypical 60)8 is lacking in bass involving and well-balanced with very and mostly tilted to the upper mids and smooth highs. We also listened to a 1968 highs. Even in this frequency range, version of the 7308 PQ which was very many examples tended ro be tWO dimen musical and live sounding. Lots of air, sional and �transisror" sounding when smooth and pleasant sounding. A great compared to a good 6SN7 or a 12BH7. sounding tube! Used with the SE 300B amplifier, 60)8$ 60J8 Bugle Boy (Richardson 1997) Bibliography were smoother sounding than when used This tube is a tested and graded with push pull ulualinear amplifiers, I. �SpcciaJ Tu bes for Broad-Band ECG/Phillips. It was cleaner sounding where they was harsher sounding. Amplifiers," Electronics, January 1952, p. than the current Chinese and Russian 203. Many modern tube audio designers types and had less distortion that typical, prefer this tube over d02.ens of others that out-of-the-box military surplus rypes. 2. A(h'enisemen[ for Raytheon GAH6, are available. Witness the significant However, it was not up to the musicality Proceeding1l of the IRE, December 1946, number of modern tube preamps using of the tOp-rated Dutch manufactured p.20A. 60)8s. They are probably used because Amperexes and Siemens tubes. of their high rransconductance (gain), 3. Advertisement for RCA 6J4, 8223/E288CC (Euro-1970s) - This is of abiliry to use low plate voltages and read)' Proceeding1l the IRE, July 1944. the taU version of the 7308. More gain availabiliry. However, mOst modern and slightly more forward than a 7308. 4. J. L McKain and R. E. Schw;lb, "A 60J8s can be prone to microphonia and Frame-Grid Audio Pentode for Stereo are short-lived due to modern designs GeneralElectric OUtput; IRE Transactions inAudio, July that push the tube hard. The beSt sound August 1959, p. 101. ing European 6D)8s are rare and expen 60J8 (1963 US) - Fat mids and upper sive. lypical prices of European-made mids, a little light on the baSS. Not 5. Ulrich L Rohde, �Oifferences Between Amperex Bugle Boys, Amperex PQs, as balanced as a Te1efunkcn or Siemens. Tube-Based And Solid-State Based Mullard 10Ms and Siemens rypes now Receiver Systems And Their Evaluation range from $50 to 75 each and climbing! 6BQ7 (l960s US) - Not really a 6DJ8, Using CAD," QEX ARR.L but has the same pin-out. Balanced Experi.menter's Exchange, NOTE: When reading the following lis and smooth sound, but lacking any sig April 1993, p. 3. tening evaluation, remember that this is a nificant high frequency detail. subjective rest and our opinions may be Very microphonic and [Wo-dimensional different than yours. Your system is sounding. VACUUM TUBE VALLEY ISSUE 7 LISTENING T 0 G D J 8 s Vacuum Tube Valley Tube School Special Announcement The first Praetical Tube Audio Seminor was held in Son Mateo, CA on June 1. As anticipated, the event was 0 huge success. If there is regional demand for this class, we will hold it on the Eost Coast, Midwest or other areas of the country. NOTE: We hove a few class manuals I ond tube school t·shirts ovailable for $50 ppd in US and $75 ppd Europe or Asio. Send payment to VTV in Sunnyvole, CA. The Bm of Class:Ampem: 7308, Wminghouse (Simums) 6Dj8, Mullard6Dj8 and Siemam E188CC The Winners Matsushita 60J8 (West German 1964) - Highs not Arc: as For the most musical and involving 60J8 Uapanese 1970s) - Balanced, as extended other Siemens tubes. l performance, we recommend: clear sound. Very listenable with Not a ot of instrumental separation and sl d congested so din translucem highs. igh y un g. I. Amperex 7308 PQ - Very dimensional Mullard 60J8 (West German/ Westinghouse and musical with smooth, balanced highs. Brand 1961) - Sweet highs with a nice 10M - 60J8/ECC88 (1960s British) musical presence. 2. Sjemens 6PI8 (1960s West German) - A natural, musical and sweet sounding Balanced sound, smooth and detailed. tube. Highs were smooth and nOt sibi 60J8 (West German/RCA Brand This tube does almost everything right. lam. 1969) - Typical Siemens sound with rich mids and good musical separation. 3. IOM Mullard ECC88 (1960s British) E88CC (1970, British) - Another - Natural and sweet sound with good b smooth and musi l tube, but a little thin E188CCl7308 (West German 1980s?) balance. Minimal, if any sibilance. on the top. - Very detailed, but highs nOt super extended. Bass is not as deep as Amperex Tht ahof)� tuhts are hecoming rare and CV2493 (British Military 1983) - 7308 PQ. costly, so getyour liftrimesupply now.' Lively, musical and 3D sounding. A bal· anced musical presentation. Appears to Sylvania (US Manufacture) Fot those on a budget, the jAN/Philips be a good value. 60j8 and 6922 are capable of good 60J8 UAN 1966). Smooth sounding, sound at a reasonable price. To get the but slightly congested top end. Raytheon best performing and select versions of Other than that, a competent performer. 7308 (US 1968)· Smooth musical these tubes, try the newly introduced RichardsonfAPp Bugle Boy 6DI8s·. performance with balanced mids and 6922 (Ph ilipsfJAN 1986) - Better bot These tubes are burned in at full load for highs. corn end than Russian types. High 24 hours on the original Amperex burn fr equencies not as offensive. Midrange is in rack. Then, each triode is matched Russian Types acceptable. Due co price and availability, and they are selected for lowest discor Sovtek 6922 - Bright and tizzy sound this tube is a best buy. tion. They in ing, but has better bass than other are nOt made Holland like Russian types of 60J85. (Most common 7308 OAN 19 70) - Bright and sibilant the original Bugle Boys, but they are still reasonable c s and type used in modern preamps) sounding with a raw top end. available at pri e arc good performers. 6H23P/6922 - Harsh, spitting highs Tclefunken -NOTE: Richardson/APD has purchased with etched, sibilam sound. Reasonable 60J8 (West German 1964) - Smooth the rights to use Amperex and Bugle Boy low frcqucncy performance. (Note: This and realistic highs. Lots of air and trademarks on their products. tube is very often remarked as �Made in depth with excellent imaging. England" and "Mullard," so be aware! The fat envelope is a give-away that this a ECC88/6DJ8 (West German 1963) Russian tube). Great imaging with smooth, detailed highs and balanced presentation. VTV ISSUE #8 Siemens Featured articles: Tungsram CCa (West German 1962) - A tube -6B05/EL84 Tests E88CC (Hungarian 19805) - Very with great balance and imaging. -3006 listening Evoluations Three-dimensional sounding with bal· detailed, but forward sounding. More -Tube Mfg.in New Jersey anced, deep bass performance. gain than other 6DJ8s and a slight amount of upper-end sibilance. ond much more!!! VACUUM TUBE VALLEY ISSUE 7 • o S C LLOSCOPES PAR T 2 existent, phase meters were expensive, The Audio Test Bench and oscilloscopc:s were nOt calibrated. Nowadays, phase and freq uency measure by ments can either be done ofT the screen of a calibrated scope, or using frequency John Atwood counters and network analyurs. Oscilloscopes, Part 2 Figurt 4 In rhe last issue, rhe theory and choice peak-to-peak signal level. It is imporr3nt ot oscilloscopes were covered. We now to remember to convert peak-to.peak val· [Urn OUT :memion to typical audio appli ues to RMS values when comp3.ring cations of oscilloscopes. By no means all against meter measurements. possible applications can be covered here, but the basic ones will be. Once the read In the audio. RF, and AC power distri er becomes familiar with these II:'ch bution worlds, RMS voltages are nearly niques, other uses should spring 1'0 mind. universa!ly used, since the mOSt common Not included here are techniques specific waveforms used for testing are sine waves. to digital television, industrial and some In the digital logic and pulse world, though, p�k-to-peak measurements RF uses. . used, since the waveforms arc far fromare Frequency and Phalc Mea..surements being constant. Figure 4 shows the rela One of the simplest uses of an oscillo tionship between peak, peak-eo-peak, scope is to compare rhe phalic or frequen average, and RMS (Root-Mean Square) cy of tWO signals by the use of Lissajous voltages for a sine wavC'. patterns. The oscilloscope is put into the Waveform ysis �X-Y" mode (no time base is used), one AnaJ signal is fed into the X (horizontal) input, The most valuable use of the oscillo the Olhcr is fed into the Y (venical) scope is the analysis of signal waveforms. input, and the gains are adjusted so that By tepresenting an image of a voltage both signals hare the same amount of over time, many effects and subtleties can deflea:ion. If the signals are exactly iden be quickly seen. In audio testing, a repeti. tiad, and in-phase. a 450 line will be tive waveform is used to allow an easily formed. If the signals are sine waves, and viewable ima e on regular oscillos<:opes. they arc out of phase, but the same fre Where a sing! C' event is ro be analYLcd, a quency, a circle or ellipse is fo rmed. If the Storage scope, either analog or digital, can signals are nOt the same frequency, but be used, although the old anal og storage harmonically related (i.e. 2: I, 1:2. 3:4. scopes arc fi nickyand can be difficult to etc.) then .:I. more complex p.:l.trern is pro use. duced that indicates the harmonic rel.:l. Sine, Triangle Waves tionship These Lissajous patterns have been cataloged. to show these relation One of the most common rest wave ships. Figure 3 shows a practical use of fo rms for audio work is the sine wavC'. It Lissajous patterns in adjusting the 38KHz has special meaning for distonion tests oscillator of an old stereo de-multiplexer. y and spectrum analysis because it is com The 19KHz pilot tone is into the ver Figurt J fed posed of onc frequency. For visual tests tical inpllt, and the 38KHz oscillatOr out on a scope, it is not ideal, but because it put is fe d into the horiwntal input. The is so common, it helps 10 get familiar 19KHz amplifying stages are adjusted for with it. A common use for sine waves is maximum horirontal amplitude, and then to test the maximum RMS power OUtput the 38 KHz oscillatOr transformer is of an amplifier. A quick way to do this is x adjusted for a stable pattern. The phase of to increase the signal level until JUSt the patrern depends on the make of the before the output starts to �cJip�. A de-multiplexer, but figure 3 shows the 8 = sin B result for a Sherwood S3MX adapter. The upper trace in Fig. 5 (page 11) This is an example of a 2: I Lissajous pat shows the Antique Sound Labs AQ- I002 o = phase angle tern, although with some waveform dis amplifier driven somewhat into clipping. tOrtion. If the 38KHzoscillator falls OUI The power mcasuremem would be made Figure I shows how the phase differ of sync, the pattern bttomes blurred. with the signal level backed-ofT so that ence of two sine waves can be calculated. the flat spotS on the top and botlom of Amplitude Measurement Figure 2 shows rypical Lissajous patterns. the waveform JUSt disappear. The lower Lissajous patterns were very popular in While not as accurate as AC meters for trace is the distortion analyzer OUtpUt, the early days of electronics when fre sine-wave level measurements, calibrated showing the WlIvcform of the Output with quency counters were expensive or non- oscilloscopes can be used to measure the the fundamemal sine ....>av e removed. The VACUUM TUBE VALLEY ISSUE 7 • o S C LLOSCOPES PAR T 2 outpm afterthe onset and absence of rhe sine wave burst. A clean transition is desirable. This same technique is used for testing audio compressors for minimum thump or control voltage fe ed-through. Parasitic Oscillations An insidious problem in audio ampli fiers is high-frequency parasitic oscilla tion. It am be either caused by poor phase margin in feedback circuits or by high-transconductance tubes breaking meaning of this will be explained in an into oscillation by themselves. A common RF alignment techniques will not be cause of power amplifier feedback insca . , upcoming anicle on distortion :.Inalyurs. covered here (maybe in a fu ture article), bility is the practice of setting the com but to show the power of the oscillo for a quick visual check of linearity, a pensation network experimencally using a scope, twO examples of frequency sweep triangle wave is hdpful. The slope: of the resistive dummy load. Everything looks alignments will be shown. Figure 7 is the uianglc wave is perfectly straight in a good on the test bench, but when it is display using a Heathki[ IG-52 TV purdy linear system. Problems such as hooked up 10 speakers (with their various Alignment Generator on the IF strip of a amplitude clipping or cross-over distor inductances and capacitances) the amp Heathkit I>T-I tuner. The horizonl'al lion cause a deviation in this straight line oscillates, often well above human hear input to the scope is a 60Hz sine wave and are readily apparent. ing_ The subjective effect of ,his can be from the generator and the vertical input JUSt generally poor sound. Things like Square Waves is fronl an RF detector probe on the out tube currems and voltages can be erratic_ put of ,he IF strip. The 60Hz sine wave As explained in the first oscilloscope arti Tu be oscillation can happen at quite high also sweeps the frequency up and down cle: (VrV Issue 6, p. 26), squ:.trc waves are fr� uencies (1 to 100 MHz), and can rrom 10.7 MHz. Barely visible in the an excellent way of quickly checking the cause similar poor sound, as well as center of the trace is a squiggly variation radiofTV interference. Sometimes the fre quency response of a system. Figures 1 in the trace caused by � 1 0_7MHz marker and 2 in that article show the effects of amplifier will oscillate all the time, but, signal. This set-up is typical of consumer high- and low-frcGuency problems on the often, oscillation only happens during service equipment, and, while far better square wave. pan of the waveform. than simple meter-based adjustments, is uncalibrared and rather unstable. The A quick way CO find the "flal� position vertical axis is in approximately linear of tone controls in a pre-amp is to apply units, which emphasize the ripple within a I KHz square-wave input to the AUX the passband (not that critical), but does input (or olher non-equalized input) and not show the "skirt� selectivity outside find the positions of the tOne conrrOls the pass band. where the square wave is closest to being square. If you can't get a decem square wave, either the ci rcuit is defective, or the design is not very good. On many amps and pTe-amps with finite low-frequency response, there will usually be a slight tilt t"O a [KHz square wave, which is normal. This is caused by the slight phase shift due to a frequency roll-off below 20 Hz. The easiest way to find parasitic oscilla tion is to look at the signal on an oscillo Using a similar technique, the correct scope. Figure 6 shows a fairly bad case of response of a phono pre-amp can be oscillation on an amp driven into over checked. By using a reverse RIM net load. Us ually ,he ·fuzz� caused by oscilla work ahead of the pre-amp, the phono tion is smaller and can be quite high in equ:aliunion is cancdcd, preserving the frequency. This is one of the few cases in shape of [he square wave. analog audio work where a really high bandwidth scope helps. Tra nsient Response A more accurate sweep alignment dis RP Alignment All of the tests mentioned so far use a play is shown in Figure 8. The sweep gell continuous signal. This makes testing Although this series concentrates on erator is a Hewlett-Packard 3330B testingin the audio domain, thereare easy, but isn't typical of the real world. Frequency Synthesizer that is set up to Audio amplifiers, especially power :amps, times when one may want to align radios step the frC VACUUM TUBE VALLEY ISSUE 7 • OSCILLOSCOPES-PART 2 �he analy;;cr. For minimum distortion in Ttktronix 4658 with muhi-mfuroprioll an FM receiver, the group delay should be flat across the: IF bandpass. Group delay is the roue: of change: of phase with frequency, in othcr words, the slope of the phase response in Figure 8. For hest iineariry. the IF Strip should bl': aligned for a straight slope of the phase response: in the pass hand. Summary Only some 01 the many techniques using oscilloscopcs have been touched upon here. If you view the oscilloscope as a tool 10 allow rou "see� electronic sig nals, soon iLS usewill become s«ond nature. The power of a two-dimensional display of information is a tremendous help in understanding the-; operation or non-operation of a circuit. In rht nrXI AI/dio Tut Bmch: Dinorrion AI/a/Jurs. Bibliography 1. Rider, Uslan, Encyclopedia on Cathode Ray OsciIJ05copes and Their USC$, 2nd Ed., John F. Rider Publisher, inc., 1959. In addition to all the scope schematics, there are hundreds of draw ings of square waves with specific distor tions, many Lissajous patterns, and d�tai1c:d AF and RFalignment tech niques. 2. Tremaine, Audio Cyclopedia, 1st or 2nd Ed., Howard W Sams & Co., 1959 or 1969. A huge audio resource which includes descriptions of oscilloscopes and audio test techniques using oscilloscopes. Oriented ro\varos pro-audio. Check Out VTV On The We b! -Free Buy-Sell Tube Ads -Free Articles to Download -Cool Pictures of Tubes and Vintage Gear -Cool Links to Other Sites -Buy T- Shirts, Back Issues -Technical information and other announcements about VTV and tube-related events across the country We now get over 1000 visits to the VTV site each weekl www.vacuumtube.com VACUUM TUBE VALLEY ISSUE 7 ID A L T E C 1 5 6 8 1 5 6 9 AND 1 5 7 0 1570 CINEMA AMPLIFICATION The first of the �green wonder" ampli fiers was the 58-pound 1570A ($284), introduced in 1956. It was billed as pro Altec-Lansing ducing Uthe mOSt wattS per pound of any amplificr� available then. Using push 1 568, 1 569 and pull 811-M, it produced 165 wans RMS 1570 with less than three percent THO from By Charles Kittleson 70 IQ 20,000 cycles. Oyerall frequency response was rated at 10 to 50,000 cycles. • © 1997 The fro nt-end circuitry was Williamson All Rights Reserved type using a 12AX7 as rhe first audio and cathodyne phase inverter with a 6SN7GTB as the diff-amp driver and Trains, Pla nes and Movie Halls and IS30A amplifiers in the early 19505, phase inverter. If you eyer went to an airport, motor the later amplifiers are easier to find due Negative feedback was carried around SpOrtS race, church, movie theater, driye to higher production figures. (The all stages from a tcrtiary winding on the in-movie, or sporting eyent from the 1520A and 1530A will be covered in a outpUt transformer. The driver circuit 19505 through the 19705, you were prob future issue of VTV.) used a Peerless 17173 cenreHapped ably listening to Altec-Lansing 1500 Typically, these amps were paired up choke driven by niode-connected scries vacuum tube amplifiers. From their with Alt'CC preamps and control systems 6W6GTs. The power supply fe atured a introduction in 1956 through the last as well as the famous �Voicc of the massive Peerless 6236 nansformer and production in 1973, Altec produced more TheaterM loudspeakers. In larger theaters, four-5U4GBs that supplied 930 volts of public address and movie theater ampli A 1 or A2 systems were us ed. In small to B+ voltage. The output transformer for fiers than any other manufacturer in the mid-sized theaters, A4, ASs or A7s were the 1570A was a potted Peerless 15416 world. It is esdmated that during this used. [n sport arenas, large churches and with a Peerless 17195 power transformer. time period, more than one hundred outdoor facilities, large, suspended horn Finished in the famous Altec merallic thousand 1568-70 series amplifiers were and woofer arrays were installed to reach hammerrone green, the A featured potted produced and sold to most every country even larger crowds. Some high-profile Peerless namformers and inductors. The on {he globe. installations of these amps induded: 1570A was produced for only about two All of the 1500 series amplifiers seld by Disneyland, The Houston Astrodome, years and is somewhat scarce. Altec were conservatively designed and Cinerama Theater in Hollywood, The made for continuous, trouble-free use. Oakland Coliseum, The Seanle Circuits were simple, bur effective in pro Cineplex, Madison Square Gardens, hundreds of military facilities and most ducing {he required performance. . Although preceded by the AltCC 1520A aIrpOrts. In 1958, the 1570B ($348) was intro duced and bcrame the SEandard high power amplifier used in coliseums and drive-in movies. They were paimed the fumous Altcc Forest Green with a 10 112 inch rack panel and a full cage wiTh a high-voltage safety switch rhat cut off the B+ when the cage was removed. The rat ing was 175 wam RMS at less than 5% THD from 65 co 20,000 cycles. The circuitry and tube topology were similar to the 1570A, but the B fe atured a differ em· chassis and transformers rhat were non-potted. The rectifier tubes were changed to 5R4GYs with a P�rlcss 6410 I. power transformer and Peerless 16492 Output transformer. Another version of the 1570 was the 1570 BT ($439), innoduced in 1963, which featured a switchable meter co monjtor individual tube performance. VACUUM TUB E VALLEY ISSUE 7 A L TEe 1 5 6 8 1 5 6 9 AND 1 5 7 0 1568A 1569A Rrllr VU"W The low power version of this long lived line of vacuum tube amplifiers was the 40 w.m 1568A (5195), also intro duced in 1957. This 22-pound amplifier featured push-pull EL34/6CA7s driven by tWO 6CG7s and powered by a single 5U4GB. Frequency response was 5 to 30,000 cycles ... or - 1 dB with total har monic distortion of les!i than twO percem at Tated power fr om 40 to 20,000 cycles. The Peerless 6288 power transformer supplied 390 volts B ... and the Peerless 16432 outpUt transformer provided the smooth sound OUtput. 128A and 128B Although not in the same numbering system, the 128A ($234) amplifier was a wider band, hi fi version of the 1568A. Introduced in 1956, the 40 wart amplifi The 1570B was the most popular high fo rm!."r was a Pcerles!i 16433 and the er fe atured the same tube compliment as powered pro-audio amp ever made in the power transformer was a PCt'rless 6289. the I 568A. The 128A was intended for US. Many of (hem arc still ill use as The: OUlput impedances were: 4, 8, 16 and use in broadcast, recording and television insTalled while others have been modified 62 ohms. Another version of the 1569 studios. Frequency response was 3- with difTcrcm ci rcuitry, tubes and other was the 1569AT (5322), introduced in 30,000 cycles ... OT- 1 dB. It was finished mods for home hi fi usc. (SceTom in the: earlier Altec blue-green metallic Tutay's mod info on pages 15-17.) hamm!."rfone finish with attractive Peerless polled outpUt and power trans 1569A formers. Not very many 128As wefe sold imroduced in 1957, the 1569A ($246) and they are somewhat scarce. ....'3$ a popular mearcr amplifier used in thousands of movie houses in the Golden Billed as the "Thermoguard� amplifier, Age of Cinema of (he late 19505 through the 128B (S240), introduced in 1961, the 60s. It was also used in schools. fac used a specially designed heat shield to tories, military complexes, churches and limit the "chimney effect" when the amps were stackcd in racks. The airpons. Rated at 80 WOl([$ RMS with 2% THO from 60 to 20,000 cycles, the "Thermoguard� featured an auwmatic resetting circuit breaker located within frequency response ranged from 5 [0 the windings of the power transformer 30,000 cycles + or • 1 dB. 1963, which fe atured a switchable meter and was capable of sensing any cxces!iive to monitor individual tube performance. Power OUtput TUbeS were four Many I 569As are still in use, but several rise in temper.Hure or component fai lure. EL34/6CA7s driven by rwo GCG? dual ha,·e been modified b tube audio enthu The tube complemem was thc same as the 128A and 1 568A, except for the use triodes. The rectifier tubes were [Wo siasts with a variety 0r circuit changes SU4CBs supplying 440 voltS B+ to the and tube substitutions. of a 5AR4 slow-warmup tube rectifier. plates of the EL34s. The output trans- The power transformer also fe atured a heavy copper flux shidd to reduce the AC 128B &ar Vitw field. The amp includes a switchable meter on the faceplate to inspect the per formance of each vacuum tube by mea suring space current. Many vintage the ater amp enthusiasts claim they prefer the 128B over the 1568A, becauSl." of its bal anced sound over the music spectrum. The 128B was discontinued in 1970. Sonic Characteristics The 1500 Series amps were primarily used as PA amps, designed primarily for mid-range applications in big areas. BUI with the correct upgrades and/or modifi cations, they can sound excellent. The 1568A is very smooth and hi (I sounding due to a wideband transformer, good design and conservative rating. Its bigger brother, the 1569A, is a bit more of a VACUUM TUBE VALLEY ISSUE 7 • A L TEe 1 5 6 8 1 5 6 9 AND 1 5 7 0 Remember-SAFETI FIRST!!! The 15708 Modification 128B voltages within an Altec 1570B can Tom Tutay, located in Fort Wal ton exceed 1000 volts and a screwdriver or Beach, Florida, has completed a number finger in the plate supply can kill you. of successful and good sounding mods to Use extreme caution when working on the 1 570B over the laSt few years. Tom's this amp and always turn the power ofT 1570B mods are as fo llows: when working on it. 1. Replace the four 5R4GB rectifier Although there are several versions of tubes with one 5000 volt-500 ma fast mods to these amps, I have contacted twO full-wave diode bridge. competent amp designers for their ver sions. Bill Jones of Santa Cruz, 2. Install a slow-start time delay relay. California, has had success with the fol brute with a punchier, bigger mid-range, lowing mod: 3. Install filter choke to the negative side decent bass above 40 hz and sweet highs of rhe supply. that are a slightly rolled-off compared to 1569A Modification the big mid range. A bandwidth limiting 4. Insta[[ separate power transformer, 1. Remove the rwo 6CG7 front end device on these amps is the optional tube rectifier, choke input power supply tubes and replace with an octal Ty pe high-p:L�s input filter. for the input and drive stage tubes. 5691 first stage tube (requires chassis The 128B is probably the mos\' hi fi modification) and a type 5687 diff- amp 5. Replace VI, originally a 12AX7, with sounding of this group with the wider stage. aGE 6072A. bandwidth output uansformer and a 2. OutpUt tubes can remai n E.L34s or 5AR4 slow-warmup rectifier tube. Lastly. 6. Add a "tail" rcsiS(Qr to the diff-amp can be changed to 6L6GC or rype (6SN7) cathode. the 1570B is a brutish, voice freque ncy 59325. amp that can be modified to be an excel 7. Replace triode-connected 6W6GTs lent hi fi amp. Low fr equency power is 3. Replace all capacitors and resistOrs to with triode connected Te sla E34Ls. limited below 60 hz compared to the modern, high qualiry units. monsnous midtange response. High-fn� 8. Install B+ power supply fuses, bias quency response is very extended, well 4. Replace rectifier tubes with solid-state adjustment POtS and voltage Test point. beyond 50 KHz, with little phase shift. diodes or replace 5U4GBs with 5AR4s. 15708s can sound spectacular with clee 5AR4s will give a slow-start to the B+ 9. Upgrade bypass and coupling capaci [rastalS or other power-hungry speakers. voltage,"" draw less filament current from tors with modern, high-quality TYpes. the power transformer and B+ voltage 10. The rwo-75K V2 plate resistOrs Collectability will fisc a bit TO about 475 volts. should be matched within 1%. Many hi fi purist.� would never consid 5. Increase power supply filter capaci er buying a vintage "PAn amplifier, bur tance from 200 to 400 mfd for tighter 11. The 47K plate and cathode resisrors these Altee amps can be quite useful in bass response. on the VI splitter section should be home audio applications when restored or exactly matched. correctly modified. The rareSt types arc 6. Install separate power supplies for first rhe 128A and the 1570A with potted 12. Upgrade most resisrors to one-half and second tube (Q reduce voltage swings i'eerle$S transformers. In excellent condi from idle (Q full power operation. (No watt metal-film. The rwo wart resisrors rion a 128A can fetch $200 to $400 and schematic available for this mod) afe either metal-film or metal-oxide, an excellent 1570A can fe tch $400 to 650. On the Other hand, 1568As, 1569As and I 570Bs are relatively com 1570B mon. In excellent shape, a 1568A is Tutay Mod $125 to 200, a 1569A is $200 to $450 and a 1570B is $300 to $600 each. (1997 US collector prices.) Repairs and Modifications All of rhe amps mentioned in this arti de have big chassis, are roomy and easy to work on and best of all, very simple • and dependable. If you come into some of these amps, never plug them in until you are sure there are no shorts, missing pans, or dried out filtercapacitors. Check these items, replace and repair as needed. Then bring the amps up slowly on a variable AC transformer and check for other problems. Fix as needed. Basic rt:storation for these amps includes upgrading capacitors, resistors, tubes, selenium rectifier and general clean-up. VACUUM TUBE VALLEY ISSUE 7 • ALTEC 1568, 1569 AND 1570 " • ." , " " . ,, " .. lll� • ...'''''' ., ... I. ...- �'.'�-' ..- ...... ' .... - ,. ..•. . . - .. -. �,--- --_... ALTEC LANS!NG 15708 AMPLIFIER .,._. VACUUM TUBE VALLEY ISSUE 7 at A L T E C 1 5 6 8 1 5 6 9 AND 1 5 7 0 �" c T ------��-•• � ... vi , ....111.;. ' ", ".- .�" • •• ;.������;;�����=t�0� ...... >(,...... , <.1' s� .....-s:TI' ... -r oc·:T1I.L Tilot" �iU..-." �I:,-A� C�lJ"T-syt'�C O,",Uu·"oA �"ft:; ArR"X. I���: : ,,,. Po'-'''''' \':;-...a;e.·oJt>S l� . � oK --.t.$ I Altec 1570 Mod. Copyright 1997 Dnignm Skmhbook Tom Tutoy, ft. Wolton 8eoch, Florida the positive grid bias on the SllAs, uses his 1 S70Bs full-range illla his home adjustable with the 1 K bias pot. brew speakers with 12 inch Amperex woofers, SEAS mid-ranges and Vifa silk 2. The bias pot also adjusts the idle cur domed twee ters in a three and one-half rent through the EL34s and U. Tom cubic foot box tuned to 29 cycles. He suggestS 4.1 volts DC max for the higher claims that the bass is right and d«:p. of the two. However, the amps perform excellently with smooth four ohm loads such as 3. TP4 (cathode of cathodyne splitter) Magnepans. They also work quire wdl should range from 85·100 V@arVI. with power hungry doccrosrats like 4. The circuit as shown gives approxi. Acousrats and Sound Lab Pristines as wdl matdy 29d8 of closed.loop gain. as others. 5. The three output transformer sec ondary windings cannot be connected 10 Tom Tutay, Transition Audio Design, PO achieve less than an 8 ohm output Box 553, Fort Walron Beach, Florida impedance. Four ohm sp�kers can be 32549 (904) 244-3041 used with good resuits, however. Accordin! to Tom, the moded 1570Bs arc:xt e reme y stable and have excdlent A ip uial tha/lks to Tarn Turay, Port walron square waves with minimal tilt at low fre &ach, Fwrida. Bill jon s. Santa Cruz, quencies. Using Svedana 811As, they ( although some are marked CC(carbon California and jim Long, Srn;( Up church comp) for power supply bleeders as they typically clip (very gently) at about 180 and Fra nk Wa lk" of £V! Audio, can tolerate higher voltage. waru into a 5.7 ohm load @ 1000 cycles. It is wdl known (hat power in the deep BI/chal/an, Michigan Measurcmenu after mods: bass range is limited by the output trans (o wnm of Alttc Laming). I. All test points measured chassis to former and most audiophiles use 1570Bs ground. TPI measures total idle current with $ubwoofers using a first order hi· of bo(h 81 IAs. TP2 and TP3 measure pass @70 to 90 cycles. Tom, however, VACUUM TUB E VALLEY ISSUE 7 .. THE G REAT v 0 ICE crs to success, Jensen and Pridham knew of no onc else anempting [0 do what they had done. The Great Vo ice The Early Years In Denmark Peter Jensen waS horn on the island of Tribute to Peter Jensen, Founder Falsrcr, Denmark on May 16, 1886. He A was born (0 a family of seafarers and was expected to follow tradition. His family of Magnavox and Jensen was quite poor and depended upon their Based on Jensen's Memoirs Through small farm to supplement his fa ther's 1930 income from working as a pilor. It was By Paul Bourbin expected that Peter would go to sea at 5. fourteen and then be on his own. Thus, © 1997 All Rights Reserved Peter Jensen .spent the early years of his life, fishing, farming and helping his for sportS and philosophy. He had five father with his pilot boat. lntroduction children and was married [Wice. Towards He loved (Q read and spent much of the end of his life, he spent much time in what linle free time he had reading nov This book is quite different than other Denmark, was knighted by the Danish els; many of them were aboU( the books I have reviewed in that it was never Ki ng and was offered the position of cul American West. He did so well at school offered to the public. The Great Voice is tural attache for the United States in that his teacher recommended {hat he the memoirs of Peter Jensen, co-founder Denmark (a position he declined). continue his education and enter a college of Magnavox, published by his daughter, preparatory rather than go 10 sea. After Pauicia Ann Jensen Schlindler. The print much family discussion, his father per run was limited to 500 copies and were mitted him to continue his schooling. distributed mostly to family, fr iends, and a While he was at boarding school, his few libraries and universities. This autobi father died, causing much grief and hard ography was also published in Denmark ship to Peter's family. Fortunately he fin under the tide of: JeDsen. ished four years study in two and a half, so he graduated. How I came upon a copy is an interest ing story. At the first California Historical He tOok a job in a saw mill to suP ,n Radio Society regional meet in San himself. His boss at the mill realized that Francisco, a fellow came up to me and Peter's education highly overqualified him asked if I know of anyone who might have for the menial work he was doing and a Magnavox speaker for sale. He had that he should go to the ciry to seck his asked quite a few people, none of whom fo rtune. Peter's mother finally relented had much interest in helping him. and off Peter went to Copenhagen. Someone referred him to me because I was the President of the Sociery at that An acquaintance of the family, Lemvig time. He told me that he knew nothing Fog, a prominent man in Copenhagen, about vintage electronic C VACUUM TUBE VALLEY ISSUE 7 • Radio takes another step for ward with these wonderful new Magnavox devices THE new Magnavox models here shown extend and supplement the already famous Magnavox line of Reproducers and Power Amplifiers. There is a Magnavox for every receiving set. Mognoyox Reproducers A2-M same as AZ·R but with the new Magnavox Re· R2 with lS·inch cutvex horn producer Ml . $85.00 in new acoustiC finish $60.00 R3 with If-inch curvex horn Mognoyox Power Amplifiers in new acoustic finish 35.00 At meets the demand for a M I new Magnm'Ox Reproduc� I·stage Power Am lifier. requires no battery for r Special finish meta case. 27.50 the field, thus meeting re quirements of dry battery receiving Sets. With 14- AC·2·C Magnavox 2·stage inch curvex horn in new Power Amplifier with acoustic finish . 35.00 Bakelite panel in highly finished hardwood case . 55.00 Magna'Vox Combination Sets AC·)·C Magnavox )·stage New tlfagnayox AI_R consisting of Magnavox Power Amplifier 75.00 Combination Set RaReproducer With 14_in. horn and l..stage Magna_ Ask ,our dealH for The firstpublic addrrn amplifier symm uSfd Angeles and radio telephony was possible by4 US 1919. at night. When rhe Federal Telegraph PrrsitUm (Wilson) Company purchased The Poulsen Wireless Telegraph and Telephone Company, Coburn asked Jensen and Pridham to quil Federal and join him in a venture to get the world-wide rights to the Poulsen Arc and start a world-wide radio system. When they gOt to Denmark, they found that Poulsen had already sold the rights to an Englishman. They returned to America and Jenscn realized that he was a real American and this would be his home forever. Jensen and Pridham returned to San Francisco, where Coburn introduced {hem to Richard O'Connor, a prominent San Frnnciscan. He became their finan cial backer. Jensen, Pridham, Coburn was also the only one who knew Morsc broadcastS because radio was monopo (temporarily) and O'Connor formed the Code. This earned him 11 promotion and lized by the Danish government and Peter Commercial Wi reless and Development he began to fllre much bener. had helped him to build an illegal set. Company to support, atent and exploit He sent a post card to Peter thanking him the tadio inventions 0r Jensen and The Poulscn Arc permitted much for the broadcasts--possibly the first fan Pridham. They purchased a house and sharper tuning than spark Iransmitlecs. mail. Peter then joined the Navy as a lot in Napa, (north of San Francisco) Poulscn invented a system with 11 trans wireless operator to complete his required where chey could work undisturbed. mitter and receiver tuned to twenty-eight military service. While in the service, They built a Poulsen Arc radio station frequencies corresponding to the twenty Poulsen contacted him and asked him if and proceeded to experiment with it and eight teners in the Danish alphabet. he would be interested in going tcfthe de Forest's vacuum tubes. They may have Typing on a specific key on the tt:l.nsmit United States and help with the selting discovered ,he oscillating and amplifying 11 characteristics of the tubes, but, innead, leTS typewriter would Oper:HC corre up of tWO radio stations. sponding key on the receiver's rypewriu�r. took another tack. A radio controlled boat was also invented. JelUen Comes to Cal.ifornia One of the devices that they brought These were probably the first experilllents Coburn and Elwdl had purchased the of this eype because the Poulsen arc was over from Denmark was a radio telegraph American rights to the Poulsen Arc sys recording device that photographically the only system that could mnc d,al tem and needed someone to supervise the sharply at that lime. recorded the dots and dashes by pho� construction of the stations. Peter was tographing the movement of a gal very glad to have the opportuniry to visit Range was, at fim, limited by the vanometer needle connected to a moving the United States. So in December 1909 touchy coherer, but when Iloulsen invent coil. It occurred to Jensen and Pridham he left for America. The first use of the ed rhe ticker, rhe range increased consid that this rapidly moving needle could, erably. Am:mpts were made to attach a transmitters was that of stock promotion. perhaps, fo llow the vibrntions of sound microphone to the various circuits to per At thc SaCr.lmento station Jensen met and cherefore, be made to work as a form mit the transmission of speech. For a Edwin Pridham and they became faSt of telephone receiver. By attaching che while, all attCmpts failed. One day, Peter friends. Pridham, an American from the vibrating wire 10 a diaphragm and hook gOl the idea to hook up the microphone Chicago area and Stanford graduate, ing up th e device to a telephone circuit, to the ground circuit and use the ticker in helped the immigrant learn the ways and they were able co reproduce sound quite series with a crystal detector. The crystal the language of America. Again, ama well. They were astonished that no onc detector had been previously rejeCted teurs enjoyed their test broadcasts. When else had invented such a simple device. because it was not as sensitive as the tick they built the station in San Francisco, With mueh enthusiasm and high hopes, er, but Peter thought it might help with many more pcople heard them, yet it did they named their device the "dectro voice transmissions. The circuit did not not occur co people associated with the dynamic principle" and applied for a work, bm, when he removed the ticker, ic Poulsen Wireless Telegraph and pacent. They were quite disappointed co did! Peter managed to transmit voice, 'leJephone Company that broadcasts of find out that their device had been well song and music fo r the first time with the music, entertainment and news would be covered by previous patents and that their Poulsen arc. Other than McCarry's voice signifiClnt. Most of the radio communi application had been rejected. The devices produced by others existed only transmissions using modulated spark in cation at that time was ship-co-ship and San Francisco and Fcssenden's alternator ship-to-shore. They were interested in in their laboratories and no commercial CJ[periments, this was probably the first promoting the system as a replacement version was ever produced. voice nansmission by radio. for the telephone. For two years, they worked to improve While operating the arc transminer, Eventually, in 19\0, continuous radio their invention and reduce itS size. Along they would play records and the few peo telegraphic communication was estab the way, they invented the output u:ms ple with radios setS would be astounded lished between the transmitters in former and the �Buclcing coil. They to hear music. Einer Diseau was the only Northern California and the one in Los could only envision the device as a tele ci vilian in Denmark to listen to these phone receiver, and could make it as VACUUM TUB E VALLEY ISSUE 7 • THE G REAT v 0 C E small and light as the ones in current use. connection of the e1ectro-dynamic tele to their successful completion; this inven Although their method was already phone to the horn. The next day, tion seemed to come OUt of the blue. patented, they were finally able to obtain Albertus made the fitting, attached six patents on their particular device. In German microphones to a single mouth Magnavox or Tele megaphone? 1914, they took their invention to the piece, attached a transformer and twelve After hearing the device, the sharehold Western Electric Lahorarorics in N(.'W volt battery to the circuit. They hoped ers told Jensen and Pridham that they York in an ancmpr to sell the patent that they could hear it across the room. would not have to worry about financial rights to them. At the same time, When the battery was connected, they backing. This was important because We stern Electric was examining de heard a loud crack followed by terrible they were in a serious financial condition Forest's triode vacuum tube. Thus, they howling. The first sounds were that of at the time. The next problem was to had two of the seminal inventions of the feedback! They had exp erienced feed give the device a name. "Dynamic loud electroflies industry in their lab al the back before, in telephone work, and knew speaking telephone� was not practical. . same time. that the cure was to separate the trans 'Ielemegaphone was considered, but mitter (microphone) from the receiver rejected. They tried Latin instead of They did not realize that before them (loud speaker). Greek and thus was born Magnavox, were two devices, if put together, would meaning great voice. During 1915, they have given them control of mOSt of the They decided to try again, howl or no continued to perfect their speaker. electronics industry. Western Electric howl, and the voice.: came thundering out bought de Forest's patent for a fraction of in spite of the howl. Excited by the pos They redesigned their dynamic tele what i{ turned OUt to be worth and sibilities of their invention, they discon phone to give the device a better appear rejecu:d Jensen and i'ridham's offer to sell nected the SYStem and put the horn on ance. They improved their microphones, them their patent. Jensen felt that he got top of the chimney with Peter's brother which were the limiting fa ctor at that the better deal because he finally got what Karl, holding on to the horn. Pridham time, by using four elements in a hand his invention was worth. Jensen and started talking into the microphones. held unir. A phonograph �pick-up� was I'ridham returned 10 Napa wondering This time his voice was quite clear, but invented. It consisted of an artachment what they were going 10 do. extraordinarily loud. Albertus and Jensen that conneCted to the phonograph, con ran across the fields to find OUt how far taining a diaphragm and a microphone. The First Electrical Loudspeaker the sound would carry. It was quite loud Jensen and I'ridham were at the end of a quarter mile away, and finally died Out Ir"5hould be nored that in 1916, Jensen their rope. They were still thinking along a mile away with a slighr crosswind! and Pridham patemed an electric phono rhe lines of wired or radio communica There mUSt have been less background graph. They often gave concerts to the tion. They had JUSt about run Out of noise then. residems of Napa by attaching their ideas. Early in 1915, Ray Galbreath speaker to the chimney, poiming it (uncle of Jcnsen's wife Vivian), a black Jensen later figured that their output toward town, and playing records. smith visited their laboralOry and saw the was about ten to twenty-five watts. They Similar events occurred when early com electro-dynamic telephone receiver. After repeated the experiment then called their mercial Magnavoxs were available in the the demonstration, Ray said, "If you can't backer, Dick O'Connor, with the exciting Sunset District of San Francisco. A radio sell it to the telephone company, what news. He could not believe what he was enthusiast would point his speaker out chance have you? But now, if you can IOld; finally he agreed to bring some of his window and make early radio broad make it talk a little louder and put a horn the shareholders to Napa the nextday. casts available to residents within several on it, like Foghorn Murphy's, and if you Everyone thought they were crazy. Most put enough of them around a ball park, OIher inventions had been alluded to prior maybe we can understand what is 1921 ukmq;aphonrAd heing said a little better, and maybe they don't nced Foghorn Murphy any more. That'.� the only chance J ean see for you." Foghorn Murphy was a San A�@HAIW@X Franciscan who rode around the area on horseback with a megaphone announcing thc time and place of the RADIOWht., u.O They starred working on the loud speaking telephone that night. They had, in their laboratory, a goose 270J-Z765 EAST 14'"" STJlttT necked Edison phonograph horn. OAKLAsn, CALlFOJlNJA They designed a fitting to allow the VACUUM TUB E VALLEY ISSUE 7 • THE G REAT V 0 ICE blocks. [One wonders what would hap bur there was much imerest in the ampli· casting became fairl),well established and pen if this were done tOday!] Jcnsen doe fied phonograph. In 1917, the there was a huge demand for speakers. umentN events which showed the fideli Commercial \Vireless and Development ry and strength of the sound emanating Company merged with the Sonora Since things were going well, Jensen from their speaker. Apparently it could Phonograph Company of the Pacific decided to go to Europe to visit his be heard as far as seven miles! Coast to fo rm the Magnavox Company. mother, see European speaker devclop This allowed th� to expand without mC':nt, and market Magnavox etjuipment. He was approached by Guy Burner of They were concerned about exhibiting having 10 depend on capital infusions their apparatus at the 1915 World's Fai r from shareholders. the Sterling Te lephone Company who in San rr.mcisco because they were not wanted exclusi\'e rights 10 the Magnavox surI': of the strength of their patents (yet Jensen's Contributions in WWI in England. Burney was granted the they were nOl bothered for five years). Soon the United States entered into rights. A problem existed in markeling However, they did play a few phono World War One and Magnavox went radio in England. graph concerts from thl': Tower of Jewels into wa They tried to make a r work. first The wealthy thought the phonograph that were even heard by ships on the Bay. system that would allow people in air and radio were for ,he lower classes and Finally they were: pressed to demonstrate planes to talk to people on the ground. the system to the press. This occurred on refused to purchase them. However, This was unsuccessful because the micro radio's popularity throughout the world December 10, 1915 from SHawbcrry phone picked up all of the loud noise Hill, in Golden Gate Park, San Francisco. finally convinced the wealthy English to from the airplane. They also wanted to acquire sets. Jensen demonstrated his Witnesses were impressed by the nCt that make an s m for airplanes. intercom yste equipment as he had dont in America. the sound did nOt appear co be that loud, Again they had the same problem. They but carried ex(raordinarily wdl. People He met Marconi and talked with him at tried many ways to block the stray Itngth about the early days of radio. could stand ten feet away from the speak sounds from entering the microphone, cr and be comfonablc, yet hear the sound without success. Jensen went to Denmark and demon· over great distances. Strated his apparatus. He visited his Pridham came up with the idea of leav mOther and enjoyed seeing his old home [t was customary in San Francisco to ing the structure of the m cro o as i ph ne and haunts. Ht realized that he should have a large civic celebration on open as possible. Since the offending keep his pride in his Danish heritage. Christmas Eve. Jensen and Pridham wert' sounds would hit both sides of the asked to provide their equipment for this diaphragm at almost the same time, the Jensen returned to California in t'Vem planned 10 occur at Civic Center sounds would cancel out and allow the September 1922 to see the great radio Plaza, a larger venue than in previous words spoken into the microphone to be rage at itS fullheighr. He saw the speak years. It so successful h y was that t e were heard. This was quite sucCC':ssful and was er used to make talking pictures in 1927 invited 10 provide sound reinforcemen! adopted by the Navy and others to be and become universally used in phono for the opening of the Civic Auditorium used throughout the War. After the war, graphs in 1930. Tht U$ts farexceeded on Dccember 30, 1915. This event was they continued to produce microphones, what he had expccttd when he first heard the first time amplifiedspeech used was including a WOIterproof type, for use on The Great Vo ice. He wondered, �How in a building and the first time that it ships. The loudspeaker business was not did we get along without it?" was used for a speech from a remote loca at all developed, but they continued to tion. The organizers wanted a speech work on it. Pa ul). Bourbin is a San Francisco, made at the Governor's house 1'0 be heard California bmtd rodio and phonograph col at the even! and Jensen and Pridham had The Presidential PA I«tor and historian. He is Past President the only way possible 10 accomplish this. They developed the vacuum tube of the California Historical Radio Society The use of telephone lilies not possi was amplifier to use with their equipment. and hat wriuCI'I articlts for uveral related ble because of the weak signal Strength pub/icatiom including Anti ue Radio This made it much more useful. AI last q (remember, this had to be done without their break came. An ail ing President ClAssified, RAdio Age and orhm. the aid of vacuum tube amplification). Wood row Wilson was slated to speak in They fo und OUt that there was a very San Diego about the League of Nations. Peter Jensen sold hi s portion of Magnovox heavy cable usC':d for StrC':et lighting run He neede d to be protectC':d from the stock to 0 group of investors in 1925. In ning down Van Ness Ave. Since both weather, so the Magnavox was an ideal 1927, he founded Jensen Radio Manufac· locations were close to Van Ness AvC':., solution. Pridham set up the equipment tur;ng Company. By 1929, the company this made the transmission possible. and il worked perfectly. Suddenly, a few They hooked their microphone and seconds before the speech was due to hod grown so lorge that he moved the recC':ivC':r to this cable and useda water begin, the sysrem went dead. Smoke rose company ond key employees to Chicago. pipe fo r a ground. Except for a slight from the amplifier. He pulled out a tube He remained President of Jensen Radio hum, the hook-up worked perfectly and and rhe smoking stopped. He inserted a Manufocturing Compony until 1940 when provided plenty of sound. Even though new tube and all worked well--just in resigned. Many improvements to lood· this event was a success, Jensen and time. This installation in a stadium por � speaker design and performance were Pridham knew that they would need vac tended things to come. Even though this uum tube amplification to make their and other events occurred successfully, ocomplished byJ ensen this pe riod ond ore device practical. they worktd in relative isolation. They reRected in the 1950s advertisement on the had no competition until 1920. At thal next page. VTV Editor Late in 1916, they moved their labora time, AT&T tntered the field. Still, tory to San Francisco. They found that business was good and much of it came there was a limited market for their from radio amateurs who bought speak equipment for the amplification of voice, ers for their stations. By 1922, broad- VACUUM TUB E VALLEY ISSUE 7 .: Precedents are a habit with Je nsen Since 1927, world's fo remost manufacturer of fine loudspeakers The ]ensen organization came into being a quarter-century ago, st based on a new concept in loudspeakers. Since then ]ensen has made history by consistent contributions to better sound reproduction through improved speaker performance. This record of achievements means that the ]ensen speaker you buy today is a product of disting�ished experience ...your assurance that Jensen speaker performance is always way ahead. lensen Manufacturing Company-6601 South Avenue, Chicago 38, Illinois Diyisi�n of I�e Miller Company-In Canada: Cappnl"aml, Wire Products, Ltd., Ucensee • 1953 JmsmAd B U L 0 A COMPAR SON PREAMP or input jacks and one pair of output jacks. We DO have a mute switch, which disconnecu the OUtpUt fr om the OUtput Build A Comparison jacks. The resiscor R3 provides a "blecd� for any DC levels that appear on the cou pling capacitor, so when the tube is Preamp plugged in (muted) and allowed to warm up, by the time you hit the mute switch, no DC means no thump to the power By Eric Barbour C>1997 All Rights Reserved amp. So long as you let the tube S[abilizc, ------�------ithere is no da nger of damaging your amp. If you get some kind of idea that liuen types, and 6BQ716DJ8 types. That covers Coupling capacitors have an effect on ing to differen r tubes for vrv is a load the vast majority of medium-mu duo-tei sound; dielectric, foil, winding method ofmonkeys, think again. We sit there for odes used in audio to dare. all have their sonic artifacts. Since the hours, doing the same thi ngs over and sound of the tube is what wc are listening The cathode bias resistOrs are fixed 1.5k over...... plug in tuhe, wait for warmup, fo r, wc chose a neutral-sounding 0.22 uF hit mute switch, hit pause on CD player, ohms with Black Gate bypass capacitors. Mylar capacitor for this preamp. It's an listen and take notes, hit pause, hit back Previous experiments had involved an Illinois Capacitor unit wc: use here at space, hit mute, remove tube, repeat. adjustable grid-bias supply and grounded VTV for restorations and general work. cathode, to allow nulling of second har Charlie bought a large batch of them at We do this bCCluse nobody has ever monic distortion. That's the best way 10 . OEM prices. They are excellent fo r...all conducted such tesrs in a systematic fash test a tuhe on the bench to develop dis kinds of tube amps and preamps, anCi ion. And they certainly haven't published tortion charts. But in listening, I found they are very affordable. unlike the exoti the results! Anyone who previously did that fixed bias was adequate. Because ca. Depending on your taS[e and budget, this promptly started a manufacturing small tubes like this vary all over the you may want to try some of the audio company. That's where the gorgeous place in bias paim, you'd have 10 adjust phile grade capacitors in this circuit. mesa-buck stereo preamps come from. the bias with a distortion analyzcr. And usually, these guys don't even invite Lis tening to them, however, showed that The power supply is in a separate plas the opinions of others, and may use parts this was illU necessary. tic box, with wires and banana jacks con of mediocre quality. You don't need to necring to the preamp. The 150v plate spend over $5000 to get a good sounding Individual tubes of the same brand and supply and filament supply are each preamplifier. type tended to have the SAME sound heavily filtered, so the preamp is very quality, regardless of individual bias varia quiet. No regulation is used, and none is So, in Ihe interests of freedom of choice tions. So we chose 1.5k ohms, a common needed. J tried all kinds of tuhe and and of honesty, we don't mind at all shar value. Believe me, hundreds of hours of solid-state regulation, and was unable [0 ing the resulrs of our listening tests. Nor tests have verified that fixed bias does nOt verify that these kinds of small-signal do we mind sharing me preamp circuit affect the sound markedly over self-bias. I tubes really benefited from regulation at we used for these same testS. Using a suspect rhat cathode materials have FAR line levels, perhaps 1-2 volts p--p OUtput. commercial modern design is nOt work more effcct on the sound character than If you are really anal, a low-dropout regu able--ncarly of them use 6DJ8/6922 all biasing arrangemems. Every manufacturer laror for the 6.3 volt filament supply ty es. If you want to listen to, say, 6C4s, p had (or has) their own recipe for cathode would insure accuracy. what then? The preamp shown here is oxides for each tube family, and that is one that I constructed specifically to lis The filament circuit shown varies about the quality that you hear most of all. ten to the differences in the sounds of 0.3 vain depending on the load; a 6CG7 Except for real junk tubes, each brand medium-mu miniamre triodes. It's simple draws the voltage down a bit more than a tends to sound consistent with itself. (although nOt recommended as a con 5670, for o:ample. It hardly maners, struction project fo r beginners; space is The input is JUSt a lOOk stereo volume since the sound quality changes only tight in this box) and it is repeat'able and control. No input switching, JUSt one pair slightly with such filament variations, and reasonably easy to use. This is the minimum scheme for a line stage. All the sockets are wired in parallel, so as to dispen se with complex switching. I did that fo r the preamp in GLASS AUDIO 3/97, and it worked but was (ahem) less than optimaL Since we are doing medium-mu triodes only, the only plate load is 47k ohms, shared among all the sockers. Only ONE ty peof tube can be plugged in at once, and both sides of the dual uiodes are used for the stereo channels. The exceptions are a pair of7-pin sock ets to accept 6C4s or 9OO2s, and a pair of 9-pin sockets to accept triode-connccted EF86s. The only other sockers are wired CompariS()1J Prtamplifitr and PoWtrSupply fo r 12AU7 types, 5687 types, 2C51 VACUUM TUBE VALLEY ISSUE 7 B U I L D A COMPARISON PREAMP PI�c<> dl,,,,,,1y .1 grid pin. One +l�OV may tt.ne_d fat NCI! luN, JUSt show jt to your (a vQrice audio ••p&eieJly high rt�n.COfIdU<:IQnCl! " guru, and watch him flap his arms hyster onN like the 6DJ8. 0.22 �F ically. It doesn't look impressive, but it \ 630V Output Mute shows that you actually COMPARE rube Input " tK. 114W brands on your own, and are SERIOUS "'.m. Stopper r about finding the heST tube. No guru can (!;&---- Resistor M, o----y deal with thar. Amazing that such a tiny 270K (:",.-= ' ."'" 114W lOOK. test box can have such power! It may even 8udio result in a aradigm explosion. Don't he tape, \r-, 330l'F surprised i ryou dump your fr ame-grid " . f- '" triodcs and start listening to 12BH7s. 1.SK Rub),<:on ,W 6lac� Gat .. Fig. I On, char/miof audio circuit (othu Table 2: J=- char/rid if idmtica/) Types Usable in Each Socket Note: these types vary in gain, so levels must be reset when changing to a type with different gain. Va lidity of listening tests can be assured by keeping the same listening level throughouL Some types -flr--<,�"",">,---4�[ir" ",-_"'_' ._ 'J' -_-__"�;'__�: _--.__ ,- __� ;-_._'-_�___· _" ,-F '"." �lll �,'�;� listed are remote-cutoff or otherwise nOt i7 - _ intended for audio. This list may nOt be exhaustive. -� ,;,� 12AU7: 12AE7, 12A'I7. 12AU7, 12AU7A, 12AV7, 12AZ7. l1AZ7A, 12BH7, 12BH7A. 12FV7, 5S14. 5S14A, 5963, 5965, 5965A, 61S9, 6201, Fig. 2 POWIr Supply 6211, 6211A, 6414, 6679. 66S0, 6S29, 6913, 6955, 7062. 731S. 772S. 7730. CV455. CV491, CV2007. CV20 11, CV2016, CV4003. CV4016, we are comparing similar [Ube types any for the 2C51 and 60)8 sockets to pre CV5212, CVS I54, CVSI55, CVS22 1. ESICC, way. vent possible RF oscillation. 12AU7 types ES2CC, EISOCC, E2157, E2163. ECCSI, don't seem to need them, nor do 6C4s or ECCS2. ECC1S6. ECCSOI, ECCS02. QA 2406. The preamp is in a plain aluminum EF86s. The safesT way is to put stOppers MS136. MS162. QA2407, QB309, QS 2406. chassis box with a thin aluminum bot on the grid pins of ALL the sockets_ tom. The 9-pin sockets are Sverlana Some gurus claim that stoppers are (llAX7, AY? high-mu IyPCS can be l�md by SK9As, which are very tight and can be extremely critical to the sound; don't making lh� plate resiSlOrs RI.! 50k ohms.) difficult to insert some [Ubes into; but believe them! Remember, we are compar 2e51: they grip the pins VERY tightly and give ing tuhes to each other, not changing excellent contacT. The plate and cathode stopper resistors. Keep the other compo 2C51. 396A, 5670. 63S5, 6386, 6854, 7861, resistors are good old Allen-Bradler 5% nents the same and concentrate on the 6CC42, CV2575, CV5S94, CVS247. carbon comps, which I matched channel sound of the tube. to-channel. A-B stopped making them 5687: recently, so get yours quick. For VHF Once your prealllp is built, you can sit tubes, grid-stopper resistors are necessary. down and uy NOS and current tubes 56S7, 6900. 7044, 7119, 7370. 7S92. Try I K, 1/4W resistors directly attached against each other. This is also a great CV257S, EIS2CC. . . to the socket pins. This is recommended conVerSatlOn pIece. 6Dj8, Table 1: Pin Wiring To Sockets 6AQ8, 6BCS, 6BK7INB, 6BQ7IA, 6B5S, 6BX8, GBl7. GBlS, GCC43, 6CG7, 60J8, 12AU7 2C51 56S7 60J8 6C4 EF86 60TS, 6ESS, 6FW8, GJKS, 6KN8. 7057, 7308. 7803, 8223, 8431, CD.. CV2492. CV5231. Triode, left side, grid 2 3" 2" 2" 6" 9 CV53SS. CV5331, CV5365, CVS 065, ECC85, Tr iode, left side, plate I 4 I I I 1,6 ECC8S, ECCISO, ECCI89. E88CC, El 88CC, Triode, left side, cathode 3 2 3 3 7 3,2 E28SC, X155. Tr iode, right side, grid 7 r r r 6" 9 6C4, Triode, right side, plate 6 6 9 6 I 1,6 6AB4. 6C4, GtOO, G135. 6664. 9002. CV133, Tr iode, right side, cathode 8 8 6 8 7 3,2 CVS52, CV2842, EC90, ECn, QL77. Filament, +6.3v side 9 I 4 4 3 4 Filam�nt, ground side 4,5 5,9 5 5,9 4 5 EF86 in Triode: '" stopper resistor recommended 6267, EF86, CV2901, CVS068, Z729. VACUUM TUB E VALLEY ISSUE 7 • ANT QUE SOU N 0 A 1 002 Q Antique Sound Lab AQ· 1 002 Kit One Amplifier A review by John Atwood © 1997 A new wave of low-cost tube: hi fi regulation of the driver B+ voltage and JUSt 50 watlS RMS (one channel operat equipment is reaching America. Pans negative supply to the first differential ing, 120V line). Biasing the tubes to a costs are reduced by being fabricated in (long-ta iled) stage. Individual emitter fol more normal 65 mA (30 wattltuhc dissi places like Taiwan and China. Onc might lowers buffer the bias voltage to each Out pation) raised the maximum power level expect that these cheap units would be put tube - hdping bias stability. The to 55 waIlS, and reduced the distortion lacking in sonic quality or reliability, but only real complaim I would have about somewhat, at the COSt of shorter tube life in general, this is not the casc. By being a the basic tOpology is that both channels (especially for the Chinese KT88s). factory direct dealer, Antique Sound USA share the same power supplie5, allowing has some of the lowcst COSt tube cquip slight interaction between channels. The frequency response is good: -3db menr available. Their power amps arc down at 5 w.:ms from 6Hz to 67KHz. Chinese, their preamps arc U.S. made and There are some deficiencies in the The overall squa�-wave response is quite power ratings on some of the resisrors ood, es ecially for an amplifier wilh such moSt of their products arc: .available in kit g p fo rm. We r�icw here the modd used in the AQ- I002 amplifier. R6 and low fe edback. AQI002K 50 watt/channel push-pull R30, 24K resistors, were measured to be Since the driver stages are direct-cou amplifier kit. dissipating a solid 1.4 wans, despite being what looked like 1/2 watt meral film resis pled, exact balance is critical to achieving The AQ- I002 initially appears to be yet rors. Their bodies arc getting discolored. good performance. The manual's instruc another push-pull 6550/KT88 stereo The 10 ohm cathode resistors and 150 tions for adjusting the driver balance amplifier in the tradition started by the ohm screen resistors on the 6550 output including balancing the plate voltages of Scan LK-150 and Harmon-Kardon tubes arc JUSt 1f4 watt metal film resis the first stllge. When this was done, the Citation [I. It, in fact, bean a visual tOrs. Under normal operation, little power amp had fairly high 2nd harmonic distor resemblance to the Citation [[ with its is dissipated in these resistOrs, but ifa tion in both channels: about 4% at 5 10% clean layout and thr� potted tranSform rubc got gassy or shorted, these resistors Wa[[5, rising slowly to at 50 wattS. Adjusting the driver balance for lov.·esl ers. The external physical appearance is would definitdy go up in flames. One or distortion brought the distortion figures quire nice. A heavygold-co lored plate is two wan flame-proof resistors arc: the F.m ened to the from of a chrome-plated norm in these locations. down to .15% at 5 wattS and 4.7% at 50 St�[ chassis. Good-quality gold-plated watts. Without a di stortion analyzer, The AQ-I 002 we built was supplied adjusting for DC balance on the second RCA phono jacks and speaker binding pons arc used. From a circuit poim of with the optional Svetlana 6550C output driver stage brought the performance view, it i.� similar to the Acrosound UL-2 tubes, although Chinese KT88s are also much closer to optimum than balancing or Stereo 120, with direct-coupled differ available. The driver tubes were Philips the first stage. The driver circuit is con cmial phase splitter and driver amplifier ECG 5814As, clearly from the recent US ceptually good, but in this amplifier con using 12AU7s. The Output tubes are run military tube disposal. These are ruggedi tributes morc: distortion than it really in the ulna-linear mode with about 465 zed 12AU7s, and have reasonable sound, should; thus, itS adjustment is critical. volts on the plates. Enhancements over although when vintage Te lefunkens were While the exterior finish and basic cir the classic designs are discrete transistor substituted. the sound was dearer. cuit topology are good, the internal Unusual for an mechanical design, especially regarding amplifier of this the PC board, is deficient. With the type is the rela exception of a pre-assembled bias board, tively low amoum all the circuitry is mounted on a single of fe edback - sided fiberglass PC board, which is in about 3.5 db. turn mounted on a steel sub-chassis, This gives the apparently to eliminate screw heads on amplifier an the lOp of the chassis. There arc 12 stand appealing dynam offs between the PC board and the sub ic sound, but the chassis, but the flimsiness of both lets the distortion is fairly board flex excessively when tubes are high. At the rec inserted or removed. The bias test points ommended out are kind of a joke: power plugs (like on pUl tube bias cur an AC adapter) that are supposed to be rent, 50 ma per silicone glued into snar-in LED retaining tube, the amp rings. Since these don't even fit together, starred to clip at and if glued, looked likely to come apart, VACUUM TUBE VALLEY ISSUE 7 • ANT QUE SOU N 0 A 1 002 Q these weren't installed, and biasing W3S done with the: bottom covered removed. NfU!, updated A-102 Amp The PC board had some: problems from both a construction and a reliability poim of view. It is a single-sided board with all components except the tube sockets and bias trimmers on the bottom. The board is arranged wilh the copper fo il on [he opposite side of the components, which is conventional, but in this case leads to twO problems. The rrimpou and tube sockc[S arc very hard to solder, because their bod ies block access [0 the foil. It is also impossible to change components with OIl[ removing the board - a difficult pro cedure. a In order to allow wires to he soldered to the main PC board without removal, small copper eyelets were installed in all holes whert': wires connect, as wdl as around �ch pin of the octal tube sockets. However, this is no kit for b�inners! pand that now makes bias adjustment This is a nice concept . but on power-up, There arc no diagrams identilying pans, much easier. A new circuit board without it was found that sevcral of the eyelets no diagrams showing mounting details, the problematic eyelets is being used. were not making electrical contact with no diagrams showing wiring, no diagrams These chOlnges are a good sign that the the traces under them, especially on the showing adjustment locations, in short. manufacturer is responsive, and is inter filament connections to the tubes. After no diagrams! When the instructions say ested in purring out a quality product. lots of fussing and resoldering, separate �!nStall eight (8) test point insulalOrs (sic) wires running from pin ro pin of the retaining rings," �Install 09, Zener diode The Antique Sound Lab AQ-I002K kit tubes solved me problem. However, after lOOV observe polarity: or "Mount the as built and rested was marred by insuffi a few weeks of use, another filament con assembled main PC board- (how????), you cient assembly instructions, a marginal nection opened up, requiring more better know your components and be quality PC board, and critical adjust bypass wires. It seems stable now, but resourceful in identifying which hardware ments. Most of these problems have is needed to mount already been fixed in the A-1 02. The fin ished unit is quite a decent performer, AQ-f002 Chassis, things. This kit is recom mended for �a reasonably with good, powerful sound, and for the facep/ate and'rtlmfonnm experienced person," but money, is an excellent buy. our review unir was built AI02 50W/channel Stereo amplifier. by a hi ghly experienced }Gt price: US $899 Assembled: $1195 tube amp constructor who had numcrous con Antique Sound USA fu sions and complaints 6717 NE 181st St., Seattle, WA 98155 due to the terse insrruc USA tions. Extra explanations TeI: 206-481-8866, FAX: 206-485-3836 and descriptions can only help. And, please, add some diagrams! email: [email protected] website: I We auditioned the http://www.amemura.com/sound/ _� amplifier with a wide variery of music through such poor electricJI integrity is unaccept both large, efficient speakers and smaller, Tube Amplifiers Lab able. Antique Sound USA is now having less efficient speakers. The AQ- !002 was bener boards built, but the beSt solution, dynamic and powerful sounding. It per Tubes orB baclc! learn how electron although a bit more expensive, would be formed especially wdl with jazz, rock and tubes work and some basic things 10 use a double-sided board with plated orchesual music. The overall presenta about audia electronics_ We will build through holes. tion of the AQ-I 002 was balanccd, with o superb stereo amplifier from scrokh strong bass, smooth mids and extended The kit's assembly manual was obvious highs. It works well with both high and using a step by step approach. This is ly wrinen and produced in America, low efficiency speakers, which is a plus. o unique opportunity to build your which eliminated the odd English present own state-of-the-art audio equipment. in the small �Instruction Manual�. In Antique Sound has made improve John Atwood - Instructor addition to assembly instructions, the ments 10 the design of the AQ- l 002 to assembly manual gives resisror color OIddress some of the problems in the earli Contact The Rondall Museum, codes, initial testing, technical spt.'CS, a er units. The updated amp. model A-I 02, parts liS(, and the schematic diagram. has a switchable bias meter on the front 5an Francisco (4J5) 554-9600 VACUUM TUB E VALLEY ISSUE 7 • COMPUT N G WIT H TUB E S PAR T I COMPUTING WITH TUBES-THE SAVAGE ART By Eric Barbaur © 1997 1. ABe, THE FATHER OF EVERYTHING on a small budget in a small lab at Iowa �computing� had previously been done State College (lSC) between 1939 and with mechanical devices, and the limits of 1942, made the first electronic digital �digital� electronics were: described by the computing machine:. And I mean the: bistable multivibrator, or �flip-flop: Entry FIRST one. Nor only that, they created invcnted by Eccles and Jordan in 1919 many of the thi ngs dm logic designers (and used \'ery little until the late 19405). Every age has tried (0 rewrite history to take for granted today, from the NOT The things wc now call "gates" were sui! iC5 leading citizens. So it is wi th [he gate to the dynamic memory. unkno , and even though Claude computer industry, today and all through wn Shannon published a paper in 1938 on its brief existence. People with power and Concept ways of doing Boolcan algebra wi th money manage to uncreate the paSt, even relays, Aranasoff was not aware of while they feed upon its fou ndations. In 1935, Atanasoff was an associate Shannon's work. Very, very few �innova(ors� manage: to professor of physics at lSe. He wanted to find an automated way of solving simul creatc an cmirc: new universe with their When he finally had a starting point. taneous linear equations, of the type first major invernion. Indeed, �innovatorM in Septembcr 1939 he hired a physi often found in areas such as ibration or cs is a word which is tossed around lightly v grad student named Clifford Berry as an eiasric Structures. He had the idea of today, and affixed to all kinds of person assistant, and within fo ur months they looking imo the possibility of doing il alities. Vinually none of them can claim had constructed a small protOtype of the wi th the new thing called radio circuitry. 10 have made: a rgjI;:hangc: to our civi basic arithmetic-memory device. In a lization, as they merely offer some small By combining electronics with a mechan improvemenr to an existing body of ically-scannt-d memory knowledge. using charged capacicors, Aranasoff hoped to solve [ am going to inrrooucc: you to tWO up to 29 equations of 29 men who really DID change the: world. unknowns, us ing the ,_ _-:::; ;;::::-__--, And yet, standard Gaussian elimi they aTC nation method. Since this vinually required multiplication, a fo rgotten difficultthing to do elec today. tronically at the time, he Don't lis opted 10 subtract one row tcn to of coefficients from their those who counrerparrs in the other claim that equation, and then shift the basic the first coefficients. units of When the coefficient was tOday's removed, a change of sign computers would occur and thc were machine would StOp. The invemed operator would watch for by the Cljffimllkrry this and read off the CoUr[(SY of ioW:l St::l!( Univ(rsity ENlAC results. This rcduced team, or by most multiplications and Babbage, or by C1aude Shannon. If you divisions 10 add-subtract feel like confronting somebody about shift operations. keeping this, grab a few computer-industry pro the machine relatively fe ssionals, and ask: who were John simple. Vincent AtanasoCf and C1ifford Berry? If the ignorance machine is working at its Aranasoffread up on usual level, those �professionals� will be the previous work by pio unable to ize those [Wo names. neers such as Babbage, And they :��id know the names, as they Merrificld, Jevons, reirce, owe their careers to Aranasoff and Berry. Marquand, Hollerith and Bush. Unfortunately, For these tWO obscure men, working nearly all of what we call VAC UUM TUB E VALLEY ISSUE 7 .. ------ COMPUTING w T H TUB E S PAR T 1 major and convulsive step, AtanasolT decided {hat base-two arithmetic was ,." simple, efficient and suitable for his com - . puter. The ENlAC, which came along in Olll.,.n •• 1946, still dun g to decimal one-of-ten arithmetic. An d the I BM 603 electronic multiplier, which was also announced in 1946, used BeD (Binary Coded Decimal) rather than straight binary. ea,ry In .-I��LT-+------' --r:»--L---L�"'" Basic Blo(:ks ty capacitors for the carry memory func tion, used during addition/subtraction. Because Atanasoffand Berry were And a drum with only a set of simple working in rdative isolation, they _120V contacts, lined up with the studs on the devised their own jargon and notation for memory drums, provided the system the logical systems in the ABC. The sim dock signal. All this was driven from an plest one was what Atanasoffcalled, in ordinary AC motor. AI; a rl':sult,ABC Fig. :1 _ Th� his own logic shorthand, a (1,1) circuit in A8C NOR tale managed to do all the logic with only Figure 1. This means a single-input cir 255 rubl':s. �'" cuit having a high input to lower the OUt ' ' put. Today, we call this an inverter. Or, a 0 > NOT gate. The concept was not original with ABC, but this was the first time it '" was done with electronics and not relays or mechanical devices. 300K _ In Figure 2 is what Atanasoff called a �V·_.O_·�'_r-�6�:G (2,1) circuit. Wc call it a NOR gate. As -!lOY you can see, the resistor inputs limited the fan-in of these circuits. They were +120V optimi"led for one, two or three inputs by Fig. 3 _ The adjusting the grid resisrors. And in Figure ABC NAND g"u 3 we have Aranasoff's (2,2) circuit, better known as a NAND gate. lr is identical to 360 K the NOR, except for the grid resistors. '" '"' All of the logic in ABC was implemented 360K 1= HO wilh 6CBG dual triodes, which are akin " ' to 65N7s except for a higher mu and a �6C8G grid cap for onc triode. S70K � Memory in ABC was not done with Eccles-Jordan multivibrators. Atanasoff -!20V had the bright idea of using a mechani cal-electrical memory scheme instead, to minimize the number of tubes. 50 he had 6C8G Berry make large phenolic drums, cov ered with brass studs. Each stud had a The .0015 uF paper capacitor attached to it, add-subnact circuit was made of with the other ends of all the capacitors logical primitives likl': the above, plus a connected to a common slip-ring for "threshold" circuit, which gave a low OUt pickup by a brush. A row of smaller put if at least twO inputs were low. brushes contacted the studs, allowing a Figure 4 shows the ABC add-subuactl':r, regeneration circuit to read the charge on drawn with modern Boolean gatl':s where each capacitor, then rewrite the charge. possible. ABC did this in a bit-serial This is basically the same concept used in manner for each binary number, with today's solid-state DRAMS. ABC used carry/borrow remembered in its separate two large drums, holding thirty numbers capacitor drum. The addend was on the of 50 bits each. A smaller drum held thir- "counter" drum, rhe augend or subtra- Bonkof adJ-iubrract circuib in th�ABC compu{fT VACUUM TUB E VALLEY ISSUE 7 .. COMPUT N G w T H TUB E S PAR T 1 machine with very. very liHle previous an 10 build upon. And like many {{ue pio neers before and since, they have been rendered non-existent by ignornnce of the sort that financial empires and lnrerncu arc built upon. So where is the giant stone monument to the REAL innova tors? Final notcs: For those: who want (0 learn more about the ABC and the 1964 lawsuit that established it (not EN lAC) as the first digital computer, the Burks book is highly recommended. It is jammed with basic information, including enough to allow a smart engineer 10 reproduce ABC exactly. And, starting in 1994, staff at the Ames Labor.ltory at Iowa State University decided 10 conStruct a fully operational replica of the ABC. Their machine was finished in 1996 and displayed at the Supercomputing '96 show in PittSburgh in November 1996. You can examine it in detail at the Ames Lab web site, www.scl.ameslab.govIABCI. hend on the �keyboard� drum. Results was done in a brute fo rce way--holes were Many thanks to tf" staff o/ The went onto the counter drum. When a BURNED imo the card by a row of Computtr Museum for their invaluable variable in the linear C VACUUM TUBE VALLEY ISSUE 7 • CATHODE B A S C ERAM C CAP A C TORS ...... High K G: ramic Capaciton - __ Cathode Bias by 1 John Atwood Ceramic Ca pacitors Engineers are familiar with the pecu liarities and non-linearities of iron-cored inductors and transformers. The advan tages of fe rromagnetic materials such as iron, nickel, and various alloys are so important that ways are fo und to work around their non-ideal properties. But did you know thaT there is an decrrosratic equivalent to fe rromagnetism that is the basis of many ceramic capacitors? Ai; with ferromagnetic materials, fe rroelectric materials allow much smaller deviccs, but they also share other less desirable proper Ties such as nonlinearities, temperature sensitivity, microphonics, and aging polycrystaline form, could be used to the ceramic is heated ahove the Curie effects. make ceramic capacitors with much high temperature, the capacitance resets to its er values than the traditional NPO types. original value. Ceramic capacitors originally were Their compact construction (either tubu made from ceramic or porcelain, the same lar or disk) made them very useful for RF With this aging phenomenon in as used for insulation, for low-value RF applications, and they could be made mind, 1 decided to check a few old capac uses. It was fo und that different mixtures very cheaply. They were also much more itors. I noticed That rhe de-emphasis in a would give different controlled tempera reliable than the mica or paper capacitors Fisher 100B FM tuner (from about 1962) ture coefficients, examples being NPO they replaced. By the early 1950s, so was off. I carefully clipped twO capacitors (zero change with temperature) and called "High-K» ceramic capacitors were out of the circuit, each marked: "RMC JF N750 (negative 750 pans per million per ubiquitous in electronic equipment. .0022 10% Z5F,» Using a General Radio QC). These capacitors find use in RF res 1650A impedance bridge, the fo llowing onawrs that must stay tuned with tem These High-K ceramic capacitors had capacitances and dissipation factOrs (DF) perature variations. Still made today, a host of electrical problems, though. were measured: these capacitors range from I pF TO over They have relative high dissipation factor (1 to 2%, much worse than mica, poly Capacjtot A Upa cit o[ B 1000pF, and have very good characteris tics: low loss, low drift, low dielectric styrene, or NPO types), had very signifi Cap. OF Cap. OF cant temperature drift, change with age, absorption. As clipped from circuit: and can be microphonic. 1820pF 0.6% J645pF 0.55% In the 1 920s, crystals were discovered High-K ceramic capacitors come in that could hold an electrostatic charge. In After leads heated by soldering iron: several grades, including X7R, Z5U, and addition, they were piezoelectric (could I 880p!' 0.8% 1810pF 0.75% Translate mechanial energy to electrical Y5Y, with Z5U being the mOSt common. energy and vice-versa). The most common From 25cC to 50ce, Z5U capacitors After capacitor heated by match flame: . decrease in value by about 15%, while 21l0pF 1.0% 2040pF 1.0% type, Rochelle salt (NaKC4 H406 4 H 0) was soon being used in the first Y5V decrease by about 50%! These types 2 Re-heating the capacitors raised the "crystat» microphones and phono pick also change their capacitance with applied capaciTance fwm 16% to 24%, but nearly ups. Rochelle salt is water-soluble and has DC voltage, with up to a 25% decrease as doubled the dissipation factOr. The use of a Curie point of only 23QC. lbe Curie the maximum voltage is reached. One of a 10% pan in something critical to the point is the same as in ferromagnetic the mOSt interesting effects is aging. The sound of the tuner is questionable, but material: the temperature above which the Z5U and Y5V types lose 5% oftheir with the aging effects of ceramic capaci material's special properties are lost. Thus value for every decade increase in time. For example, at 1 hour, it is 5% less, at tors, the impact on frequency response is Rochelle salt �ould only be used in benign consumer envIronments. 10 hours it is 10% less, at 100 hours it is especially bad. Needless to say, I replaced 15% less, etc. Assuming that the manu these capacitors with mica capacitors. In 1943, a new ferroelectric material was facturer accounts for the first lOO hours Another interesting characteristic of discovered that was very stable, could be or so, a tolerance of 10% can only be high-K ceramic capacitors is the piezo fo rmed as a ceramic, and had a relatively held for abotH 2 years. With much vin electric effect. Normally, if a capacitor is high Curie temperature of 120QC: Barium tage equipment using ceramic capacitors cooled-down below its Curie temperature Titanate (BaTi0 3)' Barium ti tanate has a being 20 to 40 years old, capacitance very high dielectrlc constant, and in its shifts of up to 20% can be expected. If with no DC volrage applied, 110 piew VACUUM TUBE VALLEY ISSUE 7 • COMPUTER v o E 0 R E V E W dectrit;: activity occurs. However, if the capacitor has high-vohages applied at high temperatures, which can sometimes COMPUTER PIONEERS occur during device burn-in and testing, a permanent charge can be trapped in the dielectric and the: capacitor becomes AND PIONEER microphonic. This is more than just an academic problem; I have encountered a significantly microphonic disc ceramic COMPUTERS capacitOr in the signal path of;1n Ampex 351 tape recorder decuonia. A Video Review By Eric Barbour Ur:1mic cpa(:itors arc used in the sig © 1997 nal path of a lot of old audio cquipmcm. Aside from rhe: problem of high [ass, which usually translates to a de graded COMPUTER PIONEERS AND sound, their use in frequency-critical PIONEER COMPUTERS applications, such as mnc controls, filters, Package containing tW{) VHS tapes: and de-emphasis circuits should be avoid ed due to the problems of capacitance Tape J, "Dawn OJ £kctronic drift. Good replacements would be poly Comp uting. 1935-194r, 53 minum; sryrene, polypropylene, or mica capaci tors. Ceramic capaciroI'S used in RF or IF Tape 2, 7/u First Computers, 1946- applications should probably not be J950�. 54 minutn. replaced though, since, in most c:asc:s, (hey an: �ingused in bypass applia Produudfor The Comput" MWn4m tions, whcre the euCI value is not critical. and the Association For Comp uting Also, the parasitic inductance of the Machintry by Larry Mondi Productions. apacitor and its leads is often accounted Priu: $50. 00. Availabit from: Tht fo r in the design of the RF circuit, so VS Comp uur Mustum. 300 Congrtss arbitrarily changing a ceramic capacitor Strut, Boston MA 02210 (617) 426-2800, \'0 anmher type in these circuits can lead [email protected], www.tcm.org. to unexpected problems. In the case of N750 or similar ceramic c.tpaciwrs, the: Learning more about the early days of controlled tempe:ramre coefficient may be di gital computing is difficult. Te xtbooks an important aspect of the circuit. on the subject seem 10 be: very scarce. The only one I ever recall s«ing (unlil As with all componenu, ceramic rC(:c:ntly) was published by IBM in 1986 c.tpacitors are not always what they s«m. and has been out of print sincc: then--it They can � quite useful, but sometimes must have been a sales disaster. were used inappropriately. Knowing that logic in some functions. It was developc:d they are based on special fe rroelecnic How little people care nowadays. The in 1937 and built in 1939. Being a tele ma(erials, with all their pros and cons, big machines of the 1939-1950 pc:riod phone-relay calculator for complc:x num will help you usc them correctly in new were the fo undation of today's PG and bers, the STL- I had no real program designs, and handle them in old equip Internet. All or the basic concc:ptS in such ming capability. Think of it as a huge ment. routine daily use, from pctiphc:rals to HP-35 hand calculator with a teletype programmin to Boolean primitives, were Bibliography (thanks to Wally Chan g machine. originated whc:n transistors werc:n'r even for these): sciencc: fiction. Finally, a video is avail Konrad Zuse developed thc: first pro I. Johanson Dielectrics, Inc., able. Hosled by Cordon Bell of The grammable calculator in Germany. His CompUler Museum, the two-tape sc:t cov Undc:rnanding Chip Capacitors, 1975. idea in 1935 was for a relay system that ers thc: firsl machines 10 use e1ecuicity could perform a sequence of operations 2. F. jona, G. Shirane, Fc:rrodeC1ric and electronics to perform calculations. read off a punched 35mm film snip. Crynals, originally published by Beforc: 1939, this world was all-mc:chani Zuse himself tells some of rhe Story in a Pergamon Press, Oxford and Macmillan, cai, and a "computer" was a young 1981 lecture. Many of these pioneers New Yo rk, 1962, republished by Dover woman being paid 50c per hour to work have died sincc: the lcctures were given, so Publications, 1993. a lever-cranked machine. This changed this video is a piece of history in its own around World War I I. Like lelc:vision, right. 3. Suchiral, Payne, Piaoeleeuic th�r� were many inventors of the pro Measurements on Ceramic Multilayer grammable computer, and they arrived at Electronics appeared in 1939-42. Iowa Capacitors, IEEE Trnns. on the same concepts semi-indc:pendcntly. physicist John Atanasoff, with assistant Component5, Hybrids, and Clifford Berry, built a machinc: for solv It started with telephonc relays. A ManufacturingTec hnology, Vol. CHMT- ing linear equations. It was full of firsts: 8, No. 2, June 1985. pp. 283-288. 1980 lecture by George Stibin of Bell elecuonic AND and OR gates (madc: of Te lephonc Laboratories covers the STL--I, 6C8G (riodes), and an amazing memory the first machine to we binary Boolean madc: of spinning drums full of capaci VACUUM TUBE VALLEY ISSUE 7 • v o E 0 REVIEW IBM AUTOMATIC SEQUENCE CONTROLLED "-�;;-;:;;;:;;;:------;;;;;n noo.... ",.... " •• PReSENTED TO HARVARD UNIVERSITY Photo (()UTt(Sy ()f Tht Co mpuur Muuum rors. Scc this iS5ue's installmenr of THE the nascent computer industry: the The amazing Whirlwind of 1950 is SAVAGE ART (page 28) for morc about EDSAC of Cambridge University, which briefly covered. Built ar MIT to run a the �ABC". led to the development of all British flight simulator, this enormous beast computers; the IAS design; and the could qualifY as the first "supercomput The ASCC, or Harvard Mark I, Eckerc-Mauchly Computer Corporation. er." It pioneered the magnetic core mem became the first really large computing ory, which dominated computing until device. A joint l-Iarvard/lBM project, The British branch of the story is the 1970s. ASCC used relays, punched cards for described first by Maurice Wilkes, creator input/output, and plugboards for config of EDSAC in 1948. It was the first elec In J 948 Eckcn and Mauchly started uration control. Grace Hopper, an assis tronic computer in Europe, and was their own company, and made BINAC ram to project leader Howard Aiken, tells intended to test a novel memory device in 1949 for a fire-control system (most how the ASCC was operated. based on an ordinary CRT: the Williams early machines were used by the mili tube. It was not as reliable as desired, so tary). Then came UNIVAC I of 1951, 5SEC was developed by IBM at the Wilkes and his team used mercury delivered 10 the U.S. Census Bureau. prodding of Thomas Watson Senior, who delays. Manchester University had its Ecken-Mauchly sold out 10 Remington demanded a machine to compete with Manchester Computer or British Mark I Rand, resulting in what is now Unisys. ENlAC in 1946. IBM engineers had it at about the same time. F. C. Williams working by 1948. Herb Grosch gives the and Thomas Kilburn were veterans of the Gwen Bell, fo unding president of the Computer Museum, trying to establish leCTure about the SSEC, which was a secret Colossus code-breaking projecl and is hybrid of rdays and tubes and used responsible for the Manchester machine, a branch of the museum in Silicon punched rape loops for on-line storage. mostly made of war-surplus parts. It used Valley, and has encountered some charm the first successful Williams tuhe memo ing apathy. Much of the collection Tape 2 proceeds to electronic-only ry, which we will discuss in a future issue (including pieces of Whirlwind and other machines, Hafting with ENJAC. ofVTV. EDSAC led to Lyons Company bits of history) sits in visible slOrage at Developed at the Moore School of Ltd. devising one of the firsl business Moffett Field, Mountain View, CA. vrv Engineering at the University of comPUterS, LEO. A 1955 promotional readers can help suPPOrt the museum and Pennsylvania by J. Presper Eckert and remember the tubed origins of computing film is shown, with the prototype LEO John Mauchly, ENlAC was all-tube by buying this video. We realize that this being used to run Lyons' main business, eXCept for IBM punch-card machines supplying tea shops with baked goods. has nothing 10 do with audio, bur it's used for some input and Output. Ecken time to free your mind. Besides, so much describes a decimal accumulatOr digit The IAS machine was invented in of this stuff is nearly forgotten ...... and plug-in, containing 28 octal base rubes 1952 at Princeton University by some some of it is very, very weird. Brace your (mostly 6SN7s and 6S]7s). Arthur members of the ENlAC ream. IAS pio self, as we explore a parallel universe in Burks, one of the ENIAC 0 erators, nar neered parallel bus communications, future issues. rates an old newsreel film 0i its use. unlike previous designs. Many institu Mauchly was influenced by Atanasoff's tions built their own work, but EN lAC was quite unlike the lAS machines: AVI ABC. John Von Neumann came to the DAC, ORACLE, EN lAC project in 1945 and thought up MANIAC, ORDVAC, the idea of a control program in the same ILLIAC, SILLlAC, memory as the numcric data. ENIAC did WISEAC, JOHNNIAC have limited. forms of stored programs, et cetera. A 1953 film even with simple conditional branching. about MANIAC shows Mauchly explains, in a 1976 Los Alamos the standard compo- lecture, how they thought up these stored nents of a computer, ., ptograms, which were read off a bank of then and ever since: rotary switches. memory, control, arith metic unit, input and This team moved on to the EDVAC output. The typical in 1951. It had main memory stored in MANIAC register flip mercury delay lines, while a magnetic flop is shown, a 6J6 drum held long-term data. EN lAC and dual uiode with a neon EDVAC led to three major branches of lamp for display. VACUUM TUB E VALLEY ISSUE 7 • w N T E R I 9 9 7 C E S VTV at the Winter 1997 CES in Las Vegas By Charles KinIeson ©1997 Every year, right after the Christmas Welborne Labs, as they were: discussing holiday blirz., electronics manufaclUrers, the fine pointS of the VAle VV30B ui buyers, dealers, enthusiasts and the press ode. Since this W;1S the first time JOt and show up in Las Vegas and take over the I met fa ce-w-face, he said - "You don't town for a week. At the 1997 CES, it look like: I thought you would.� ! (Old was reported that over 120,000 people him the fe eling was mutual. showed up for the show. No less than five major hotels and the Las Vegas Harvcy Rosenberg was in his full-dress Convention Center were jammed with "Thcrmionic Tcchno-Shaman" regalia fea everything from digital watChes to home turing a 300B headband, a "The Triode theater to high-end vacuum tube audio Fa{her� T- shirt, Scottish ki llS and hiking amplifiers. hoofS. Eric, SIeve and I were dressed in normal street clothes and Harvey said we The high-end speciallY audio event looked like "A bunch of Fuller Brush was held at a new venue this year. the salesrnen.� Needless to say. Harvey Alexis Park Hotel. Most of the major attracted a significant amount of attention players were there and many not-so-major in his salesmen's costume. cussed his spectacular SE833A 100 Watt players as well. As usual, tube equipment Class amplifier. The group fielded quc-s· dominated the specialty audio Scene again Harvcy talked us into an'cnding «The Triode Summit" being held in [he tions aboU! Japanese transformer designs, with at least 60 percent of the rooms the use of transformer coupling and exot using some lYpe of tube equipment. WAVAC suite: in which Nohu Shishido, the famous Japanese audio designer was ic circuit designs. The "Triode Summit" Tubcs arc no longer a fringe product, bUI holding court. The crowd in the room was, wirhoUl :.I. doubt, one of the high are now the dc-facto standard for the hest lights of the show. sound. numbered about 25 or so, bUl included Anhur Loesch, Ron Welborne. SeNt The equipment on di splay was indeed At the Alexis Park, we met Joe Frankland, Frank Garbie, Joe Roberrs and an international extravaganza. We SlOpped Robens, EditOr of Sound Practices others. Nobu did a mini-bio on his back in and challed with Denni! Had ofCary Magazine. Joe was with Harvey �Gizmo" ground in audio and design approach to Audio, discussing the finer points of Rosenberg, plus Ron Wd borne of single-ended amplifiers. He also dis- 300Bs and 8455. Dcnnis k«ps coming up with even more varieties of tuhe amps each year to add to his already impressive stable. Then we met Kevin Hayes of VAC and checked out his beautiful gear including the new push-pull 300B amps. Some of the newer :.I.mps from foreign manufacturers fe atured KR Enterprise tri odes including the VV30B and VV300B. AudioNote presented the new Ankoru paralld-single ended 845 transmitting triade amplifier which used a single 300B as the driver tubc-. The Ankoru also featured dual SR4WGT tube rectifiers! Sonic FrOnliers from Canada displayed an incredible amounr of their tube gt2r, including remOte control tube preamps, VACUUM TUBE VALLEY ISSUE 7 • I w N T E R 1 9 9 7 C E S tube: amplifiers and the Assemblage: D/A facturc:rs displayed this year with an Consequentlr, a number of manufacturers converter. Their products were well impressive array of amps, prcamps and who normally would nOt exhibit at the designed and beautifully executed. Graaf speakers. They also had a �tube city" dis CES, were able to because of more afford of Icaly displayed their delicious-looking play of all the Chinese audio tube types able ratts. glon black and chrome GM 100 and available including KTGG, KT77 an d GM200 OTI. stereo power amplifiers . KT J 00 types. Steve Parr and I had the same com plaint this year as last. Use larger speakers One of the more in your demo rooms! The use of high innovative amps at the priced shoebox size speakers with six or show was the MONS eight inch drivers JUSt does not give the 300B single-ended tri most dynamic and tealstici sound. Hey ode tube amplifier guys, [Ty using bi�er speakers for maxi using an El34 as the mum effect next tIme. load. The amp uses no OUlpUt transformer, One of the surprises of the show was but has a coupler net the 43 watt Minnesota Audio Labs work for either single MAL6550 push-pull 6550 amps driving or bi-amping. The Von Sehweikert 4.0 s�kefS. This system MONS amps were sounded musical, balanced and JUSt right hooked up to the new to our tired eafS. Probably the most Series 200 Edgarhorns impressive equipmem from an enginCl:ring and u i st dpoint t e WAVA C by Bruce Edgar. q al ty an was h gear. I can't think ofany tube freak who An added t re [0 the: elcclronic fes wouldn't drool of the WAVAC The GM200 is a 200 waH �r ch nnel fea u over any a tivities was �The Altern:nive Hi h-End equipment. hunk usi lG L504s per channel! g ng P Audio ShowM held the Debbie Reynolds Another Italian firm, Audio Devices, at Viva Hotel located near the Las Ve gas. Well, off we go to the May Stercophile dispb.yed their new amplifiC'T using four Convention Cemer. Lots of exhibits fea show in San Francisco which we will cover Svcdan SV572-1 0s. Two of t e SV572s a h turing smaller, but high amp, in vrv#8. were push-pull OUlpl1l and the other [Wo q uality speaker, cable and relate manufacturers. were used as rectifiers. The styling of the d This venue was less expensive for t e r spe kers was, to say Viva amps and h i a exhibitors to display than the reg l r CES. the least, unique. Chinese audio manu- u a VACUUM TUBE VALLEY ISSUE 7 • U N C L E E R C ' S DUMPSTER 417A/;842 audio until the more extreme elementS of G the audiophile world discovered it in the UNCLE ERIC'S 19905. Although its dismnion is reasonably G DUMPSTER low, the 417A15842 really shines in the first 5tage of phono or microphone prc amps. It likes to run :H substantial plate: current, so by keeping the plate volcage [0 The 417A/5842 IOOy or less.it can give the: dC2nest tonal ity possible along with its low noise floor. BY Eric Barbaur © 1997 Even with only IOOv, a 417A gives a dynamic range mOTC than an order of magnitude greater than an equivalent Reading this magazine-;;; makes;;k;;- �you;--- r;::======::;� semiconduclOf. Low noise: tranSislOrs tend t--somethingR.;;;ii,;;-;hi,� of an elite,;;;;;:;i;;; because you are to die when such voltages are applied. aware of the advantages of therm ionic And a low plate resistance makes it possi- vacuum tubes. Yet even today, some very ble to even use the 4!7A as a power tube- smart people JUSt don't get it. A recent -it can handle 24 mA current from its example is the 2/97 issue of Glass Audio, cathode, and has a plate resistance of page 32. Erno Borbely, a well-respected 1800 ohms. engineer and longtime contributor ro Audio Amateur, is trying to ease his way As with any high-gain RF IUbe used into the tube-design game. Yet in this CA for audio, microphonics can be a prob article, fo ot is inserted intO mouth. Erno lem. In low-level applications shock is doing a phono preamp, and startS ofT mounting may be needed. this feature with the blunt statement that value of the resistance as well as the mate A shame that nobody is using this it is ·practically impossible" to obtain rial or COnstruction of the resistive mater tube in high-end preamps today. It's easy noise figures as low as semiconductOrs ial. Generally, more reSiSlanCe means to find, yet manufaCturers prefer to slap with rubes. He gives the EC86 as an more thermal noise voltage, appearing cheap, nasry Russian 6922s in expensive example of a �lowes[-noise" tube--and it's across the resistOr. This applies to triodes preamps inSlead. So beat the "gurus� of only middling-fair. Apparently, Erno in a similar way--each one has an equiva the world and try it. I guamnlee state-of n�er tried a WE417 A. lent noise resistance, which appears to be the-art noise figures, with nary a transis The 417A was inmxluced in 1948 by connected from grid 10 cathode. The IOr in sight. Western Electric. Intended as a low-noise noise is then amplified and appears in lhe first Ha e for broadband RF preamps, it plale circuit. eventual Iy became something of a stan Triode noise resistance is roughly dard for receiver front ends. Often seen C WE originated it, yet many other companies made it also, usually under the jEDEC designalion 5842. GE, Sylvania, Amperex, RaYlheon and CBS-Hyuon all produced it (Sylvania until 1988). Old stock is very common. Usually found in receiver front ends, it was never used for VACUUM TUBE VALLEY ISSUE 7 • ------�------ v T v UPDATES AND RECTIFICATIONS VTV Back Issues and Subscription Order Form Cash, us Bank Check, AMEX, MC or VISA Accepted VTV#2 _ TheEL34 Reporl.Page 8, hi col. urnn, The EL34 was act\l�lly introduced by Phillip. VTV 1 - Eorly tube cmpl, Dynoco Hiltory, Tube Te5tir>g, Audio Te51 Bench Set-vp ond 12AX7/ECC83 in 1949, initi.lly in England. It became avail.ble History cnd Tests $8 US-SIO Foreign PPO in 1953. for u" in America Is, column, 2nd and VTV 2 - El34/6CA7 History cnd Test., Heothkit WM Seri". Tube Amp., W�"�rn EI"';rk Theoter Sound 3rd 1"''''S'''ph, - Mullanl EL34 -types 1 ,od 11 had 1934, Be.t Guilor Amp, , Tube N.okhing, Magnum Sf Amp Project S 10 US·S 1 3 Foreign PPO welded plates. The Type H1M ullanl. had crimpcJ - pl,le •. Type 1 Mullard EL34. h,d 00 tOP getter VTV 3 WE 300B History and Te sll, SE Tradormer Te ltl, flea Tube Hi Fi, EH $ceM 48 tube rodio, i . ''''0 ., a t r i r ng Type 11. had n ll round get e r ns' and ..v.clnlo$h MI 200 modification, Dynoco Stereo 70 Project $12 US-S I5 Foreign PPO Type 111.had ""rsion. u,ing .ilh", twO ,m,ll getter VTV 4 - 6L6 Hilloryond Tesl$, HK Citation I and 11, Classk Bookshelf Speokers, Designing foru So nd, ring. Or a singie l ..ge tOp getl.. ring. Tube Manool, from tile Golden Era $12 US-SI 5 Foreign PPO 1927-34W est;rn Eleclric!.a [�" Tbm.r Sound VTV 5 - Bendix Red Bank Tube., FM Tune' Shool-out, Vintog6 Hi If in Hong Kong, Gvilor Amp Book Snwn, Po.ge 24. h' and 2nd p.rog"'phs . The Review, Audio Signal Generotors, Uncle Eri<:'s DehJxtI SV811 SE Amp Project, The Birth of theM aron� 108, WE-203 dual 78 rpm pl.y.: r piCtured in Ihe artide Guitar Amp Redi��ation $12 US-51 S Foreign PPD w:l.S .clually USed in an earlier era for in-,hca In 1929, theater. began to show film, that used boIl, rypa of sou"d films. Western EI«tric, Pacem and other produced equipment [ha, could reproduce both ryp.. (>f product. Th.n ks to George McCluney. Se the first In your neighborho to have the official VTV VTV13 - TheQl.w for the Ultjma!e _ Ml "BIG TONE" T..s hirt. 200 Mdmo,h, Page 18, 2nd column, 2nd p. .a graph _ Figure 1 Wa$ not included in this article. It 2 calor heart print avai lable in either was.h.n d-d.. wn schematic of DO\"C: Woltt's mod Blue Spru(e or Desert Sand colors. ifi""'on. (Please Specify when ordering) 100% cotton shirt. Only extra large available. m'i'B ook.hd f Spca kcn P.ge 30, 3rd col Yj Only $14.50. plus $3.00 postage US, umn, 2nd I"'ragraph - The AR-3 'peaker w", rated CA residents add 2 Sales Tax for a minimum of25 ""tUcominuo",. no, 35 $1.' w,,,ts minimum as ...!Cd. Canada and Mexico add $7.00 postage Europe and Asia add $12.00 postage VIV116 • Th. Ei,)""500 , Po.ge 6. 3rd col Cash, Check or Money Orde r umn, hI paragraph - Fi,her 400 r«(iven with the Credit Cards OK Gold." Syncrode from-end did D.l!luse Nuy.i rors as ""led. They used,6HA5 and a 6AM tuoo. VacuumTubeValley 6550artjcle. Page 23 - MPD ended produc 1095 E. Duane Ave .. 5uite 106 ,ion of 6550As in 1993. Sunnyva.le, CA 94086 USA VACUUM TUBE VALLEY ISSUE 7 • Design In the "Real McCoy" 3 10 280 Wall Tu be Output Tra nsformers Power & Filamenl Sleel Cages Tra nsformers Chokes (open & enclosed) Sleel & Aluminum Chassis Q[laggic Q[omponcntg trom: We've been in the transformer & enclosure business for .... HAMMOND over 70 years. In fact, most of the components shown £TI MANUE4CTURINGN here have been in continuous production since the 1950's! Trust your design with our experience .. ..,www .hammondmfg.com 4700 Genesee St., Cheektowaga, NY 14225 Phone (716) 631-5700 Fax (716) 631-1 156 394 Edinburgh Rd. North, Guelph, Ontario N1 H 1 E5 (519) 822-2960 Fax (519) 822-0715 Tube Lore A REFERENCE FOR USERS AND COLLECTORS For serious users, a new look into tubes from 1920 d -L;;:-) to the present: history of esign trends, "where us flejELECTRA-PRINT ed" information, tech data, equivalencies. "user's guide,� even auction prices. See review by Eric aUALlTY HANDWOUND S.E. OUTPUT TRANSFORMERS Barbour in the Winter 1997 VTV. Contains 186 pag • AUDIO Primaries from 1.8K to 10K es, 8-1(2" x J 1" softbound. Orderable from: • Exclusive wideband design Antique Electronic Supply, (602) 820-5411 · 16, 8, 4, or 2 ohm secondary 6221 S. Maple, Tempe, AZ 8.'i283 • Gold plating Antique Radio Classified, (508) 371-0512 CUSTOM·BUILT POWERTRANSFORMERS/CHOKES PO Box2, Csrlisle, MA 01741 • Wound to your requirements Museum of Radio & Technology BookslOrc • Heavy·duty design for Class A use 4204 Thorn Apple Ln., Charleston, WV 25313 • DC filament supply windings W7FG Vintage Manuals, 807-6146 (BOO) • HV chokes to 15H, 300ma 3300 W d aysi e Dr., Bartlesvil1e, OK 74006 • Low voltage, high current filament chokes Or order directly from the author: Component pricing and specifications cata/og available Ludwell Sibley 44·Y E. MDin SI. ELECTRA-PRINT AUDIO Flemington, NJ 08822 41 17 Roxanne Drive Las Ve gas, NV 89108 Price if ordered direct, postpaid: 519.95 in US or Canada, 702-396-4909 FAX 702-396-4910 $24.95ppd. elsewhere. Discolmls available to groups on orden on quantitypurclrllJts . • It's Like Music to Yo ur Ears•.. Amperex Bugle Boy audio tubes are back and they're better than ever! When it comes to selecting premium audio tubes for your amplifier, accept no less than the best, Amperex Bugle Boy! Amperex Bugle Boy tubes are burned in under full load with all voltages applied for 24 hours before being tested for noise a microphonics. The aging racks used are from the original Amperex production line built in Holland decades ago. Complete test results are supplied with each tube, including plate current, screen current, grid one current, filament current, and total harmonic distortion (THO). The THD measurement is performed using one of the finest spectrum analyzers available - the Hewlett Packard 3561 A. Matching in pairs, quartets 11 octets is available for all tube types. Special test requirements? Not a problem. For a slight additional charge we will test and match to your specified operating conditions. Ty pes available: 6DJ8/ECC88, 12AU7A1ECC82, 12AX7A1ECC83 and many more! So if you haven't dealt with a source capable of meeting your demands, call us at 1-800-TU BE USA or call 630-208-2200 today for information on additional products and the worldwide locations nearest to you. Yo u'l1 1ike what you hear! J..Richa rdson W. Electronics, Ltd. Audio Produd Division IS09002 Registered a division of Richardson Electronics, Ltd. 40W267 Keslinger Road P.O. Box 393 You can also visit our Classic Components LaFox, IL 60147-0393 site on the Internet at: (630) 208-2200 http://www.tubes-r-us.com • Va cuum Tu be Stereo Line-Preamplifie r Designed ami Made ill USA • 7-50,OOOHz within IdB • Low Noise & Distortion • Contains no Feedback • Accepts five Stereo Inputs • Stereo Recorder Outputs • Gold plated RCA • 5 OOVA Switched AC Outlets • Designed & Made in USA • Superb Quality and Sound • Operates on IOO/1l5f230VAC • With or without Tone controls Tech nical Specifications ModelPA-l Model PA- 2 Line Inputs (five) CD, Tuner, Tape, Video, Aux CD, Tuner, Tape, Video, Aux Frequcncy Response IdB from IS-2S,OOOHz I dB from 7-S0,OOOHz Vo ltage Gain +6dB maximum +6dB maximum Vo lume Control Rangc 80dS continuously variable 80dB continuously variable Tone Control Sass and Treble variable None Tone Control Range From + J OdS boost 10 -ISdS CUI Not applicable Channel Separation -43dB minimum al 1 kHz -SOdB minimum at 1kHz Hum and Noise (-80dB) .2 millivolt maximum (-80dS) .2 millivolt maximum Input Impedance 2S0K Ohm all inputs 2S0K Ohm all inputs Output Source I K Ohm cathode fo llower 1 K Ohm cathode fo llower Tube Compliment 4 each S814A low noise dual triode 4 ea 5814A low noise triode Front Panel Controls AC, Selecl., Vol ., Bal., Bass, Treble AC, Selector, Vo lume, Balance Rear Panel Connectors S stereo inputs, PIA output, S stereo inputs, PIA out, Recorder Ollt, Recorder out, AC line, 2 ea !\Ux. AC out AC line, 2 ca Awe AC out AC Power Required 100ill5/230VAC 50/60Hz 12VA IOOill5/230VAC 50/60Hz 12VA Size W x 0 x H 15 x 7 x 3.S" (38 x 18 x 9 cm) 15 x 7 x 3.5" (38 x 18 x 9 cm) Weight (net) 6 pounds (2.7 kg) 5.5 pounds (2.5 kg) Enclosure Aluminum black powder paint Aluminum black powder paint Front Panel (machined) 3 Choices: black, gold, natural 3 Choices: black, gold, natura! List Price (assembled) $995.00 (5795.00 Kit) 5899.00 (5699.00 Kit) "Our world-wide dealer network is now being fo rmed, apply now!" http://www. amcmura.eomlsoundl An6717tiqu NE 181"e Sound Street, Seattle USA WA 98155 USA, Phone 206-481-8866, Fax 206-485-3836 T��Lf)VAl\tm ••• VA CUU," TUUI:� ff"()m the tieart ()f I:uf"()pe% Introducng Ihm totally new I'3cuum tubes from Ir:SL()V�: thc KT88s, These [uht, are being handmade in limited quantities exclusively for Pema the E34Lsi and [he E83CCS. These new vacuum mix desi s arethe rodu t of Labs and Groove Tubes in the classic Europe n trndirion. This legend ry years of reso:arch, dl':Vdopment, and hardwork by our USAgn epngi nttCring tube faclory has manufactured many OH." tubesa under the brand namesa staffof GI'OOVl: Tubesand Ptnr.t Labs in a join! \�nture withbased t1� staffof the such a.s;TddUnken, Siemens, Amprn. Mullard, and Brimar, to name a few. IleW!yTcsb rc :-If:!!.l-- HEAVY GAUGE SUPPORT RODS ' . --. I PREMIUM "RUBY REo" MICA SPACERS • - """ .. GOLD WOUND CONTROL GRIDS · ----It- . --r-- "K� H�T DI�IAAn':G�NGS · �" WELDED TO PLAn STRUCI1JRE _-t_ I SPECIAL CARBONIZED. �IC."l ••__ PLATE MATERt>\!. �_ KT&&AllOY )0 "111 PUtt diuip,ation E}iLJ 30"'tll Pht� dissipation • ALUMINUM BANDED BASE \X'1TH TAPERED PINS .....,.,.,.• ANODIZED BASE \VIm GOLDRED Gu..ss mE � VINTAGE BROWN BAKElJTE ' / BASE WITH TAPERED PINS Our new ·1·I:SL()VAl{,.m KT88s OUT new Tf:SI_()'V4I{,.m E34Ls The first legi timate European KT88 made in This totally new Pentod design upgr-ades and over 30 years. This new bc:am p Out new p�mium THLf)VAI\m E83CCs prc:amp robe: with Grc:ala ention isgiven 10 t ma e i and bly with 30 special fr:tme grids, wound wim gold wire, help to reduce: micro hand weldedtt contaCtS in this singlerals dual !riodc!Thisassem yea!over sound ph nic imeraaions, lower noise, and improve both transient and ing new robe replaces the 12AX7/7025 in all applic:ttions. frequencyo response. rru� pi Enter the KR Enterprise VV300B contest. Win one of three pairs of KR Enterprise VV300B tubes valued at over $600.00 to be given away on July 1, 1997. Send us your name, address and a photo of your 3008 amplifier. No purchase necessary. For rules and regulations call or lax Welborne Labs or visit our website at: http://www.welbornelabs.com 11 you can't wail until July I" , contact us now for details on our guaranteed in home audition offer. Welborne Labs p. a. Box 260198 Littleton, CO 80126 USA Vo ice: 303.470.6585 Fax: 303.791.5783 E·Mall: [email protected] OEM and dealer inquiries invited. The W Valves are warranted for one year or 2000 hours. Listen to the VV Va lves today and experience a whole new horizon of musical pleasure. Ask top engineersGO and AHEAD music lovers worldwide what capacitor they choose for their uNlmate designs. 11 0 V L l, N-tlO-- "After auditioning MU0 wide varietySIC of Apcops, / ®find these film· and-foil capaCitors to be the most musical performers and by 0 large margin. The MusiCop specializes In textura/ llQuldlty (lock of grain) and dellcocy of treble nuances.· Dick Olshet Reviewer and Designer; November 1990Sanodhl Acoustics wtlit9pq:>er This is the sanw amplifu:r built in the ""Audio From Tubes- clus laught TheCOMP noIlIottmost AREcopocIor revet:*'og . FI'nNOW101 cnd � Acwood by lohn lithe Randall Museum in San Francisco. Availabk construcIion. 0pIirilsd laYOCUUTl /hespoOOirsand I\.tleeIecIrorics. either completely assembled or as a partial kit for the serious wnslructor. SiIver-plaled leads. Precision modein the VduesUSA. from01 to10).lF. • Scen-o,linsk ...... • Svetlana Audio Tubes ..... � .. Svetlalla's (SV)6L6GC showed Ill' ..and were easily ,he best-sounding lIew 6L6s we've heard.. " Power Tube Hme Te sl Guitar Pla)'cr /l.13gazine November. 1996 .i.. "The Svellan(l (SV)6550C beat the P(lIItS offmy treasured TU llg-Sol coke bOIf/es. bass was fighler and more defined. The 7img-Sols had (/ little too mllch bass, they were /00 chesty alld IIIbby by comparison" Dan SchmaHc Publisher. Valve Magazine "To our ears. the best sOllndillg EL34 of the Ill> grollp was Svellalla's Gold Top £L34. � Po... er Tu be Taste Tesl Guitar Player Magazine November, 1996 Visit Qur ... ebsile or collfor IIdis/riburor Ilear Joul .. SveELECTRONtlana DEVICES Headquarters: 8200 S. Memorial Parkway • Huntsville, AL 35802 • phone (205) 882-1344 • lax (205) 880-8077 Marketing & Engineering: 3000 Alpine Road . Portola Valley, CA 94028 • phone (415) 233-0429 • lax (415) 233-0439 www.svetlana.c::om WE HAVE ENOUGH COMPONENTS T" . O BUILD A SMALL CITY TUBES: PARTS: Over 3509 audio, receivin , transmitting and industrial types Resistors. tube sockets, potent!ometers chassis boxes and in stock. ini:luding many foreigng and early types. We offer aluminum enclosures, knobs, dial belt, lamps, diodes. speakerl, the complete line of Svetlana audio, amateur wire. phonograph needles and cartridges and much more. radio and industrial power tubes. LITERATURE: TRANSF"ORMEAS: Extensive selection of literature and books and tubes, Hard to find power, filament and output hi·fi equipment. circuits diagrams. communication transformers as well as filter chokes gear and antique radios. Some items for tube equipment. We not avai lable elsewhere! feature HAMMOND. One Electron, MagneTek and Thordarson as SUPPLIES: well as many new old stock Grill cloth, transformers. cabinet CAPAC ITORS: restoration supplies, High voltage electrolytic and mylar capacitors, multi· batteries, chemicals, section capacitors and tools, test meters, more for tube circuits. gi fts and kits. Svetlana "'0'"0" ""0" HAMMOND MANUFACTURING. G) one electronn, CALL OR FAX FOR OUR FREE 40 PA GE CATALOG --- ANTIQUE ELECTRONIC SUPPLyrn-� Umited PMtnership 6221 S. MAPLE AVE .•TEMPE,AZ 85283 USA (602) 820-S4 1 1 • FAX (602) 820·4643