In Search of the Elusive 1L6
An Obsessive Oddessy
A. Padgett Peterson, P.E.
A few years ago when I began my collection, the 1L6 was available but somewhat pricey - $9.00 when 1R5s could be had for $1.25. By 1997 the AES catalogue did not even list it though they recently had a "special" on JAN 1L6s for $30.00, and while Electron Tubes Enterprises does, it is at $24.00 (if they have any). (AES apparently bought a large number at a Government surplus auction. These are wonderful opportunities for those who can inspect the lots, usually in Idaho or Montana and attend the auction in Virginia or Georgia...)
Always scarce and relatively expensive, the 1L6 has only had one use in its life - American battery operated short-wave receivers such as the Zenith TransOceanic which used it from the 1950 model G-500 to the last B-600 in 1962 (said to be the last tubed battery portable made in the USA).
Other American manufacturers toyed briefly with battery short-wave portables: Hallicrafters, RCA, Philco, Sears,... usually with a knock off of the Zenith circuit using the 1L6 but none had any real staying power compared to the "Royalty of Radio".
More in the mainstream, Zenith even used the 1L6 in certain AM/BC only portables such as the G-503 and H-503 though four tube (lower priced) portables generally used the more commonplace 1R5. They were also used in a range of "lesser" battery shortwave portables such as the 1953 L-507 "Meridien". These generally had only three bands though boasted near continuous coverage.
The 1L6 is really an oddball and has its roots as a cost savings measure from 1932 when SuperHetrodyne circuits as invented by Edwin Armstrong during WWI became "essential". Instead of a simple progressive detector/amplifier as used in Tuned Radio Frequency sets, requiring individual tuning of each circuit, Superhet reduced the tuning necessary to a single circuit with two or (in better designs) three mechanically coupled tuned elements (Antenna, RF, and oscillator. Hence the three coil adjustments necessary for rach band when aligning a TO.
While the Radio Frequency (RF), Intermediate Frequency (IF), and Audio stages could have been lifted from any good earlier design, the center circuit comprised of an oscillator and mixer was brand new.
Superhets operate by taking a radio channel frequency that is desired, and mixing it with a second frequency that is an exact space (usually 455 kc today) away from the desired frequency. When mixed with the original signal, harmonics would be produced, one of which would be exactly on the IF. This could then be amplified by the next stage (IF) without any tuning of that stage. While there were other design advantages, the real selling point was "one dial" tuning.
At first the circuits were separate requiring a tube for the oscillator and another for the mixer - originally two triodes, but later a triode for the oscillator and a "screen-grid" tube or tetrode for the mixer.
By 1932 the first attempts were being made to combine the two elements into a single tube to reduce both manufacturing cost and current requirements. Since the oscillator and mixer were already direct coupled, it made sense to put the coupling inside the same envelope.
Sylvania was always in the forefront of the specialized frequency converters necessary for superhetrodyne circuits with their pioneering type 15, a 2 volt battery tube with separate cathode and filament. Though a tetrode, it proved superior to anything else of the time and even remained popular after RCA introduced the first true pentagrid converter, the 2A7 in 1933. This basic design was produced in many forms (1A6, 1C6, 1A7, 1LA6, 1LC6, 6A8) during the late 1930’s.
These first ‘pentagrid converters" as exemplified by the 1A7 utilized five "grids" with G1 the oscillator grid, G2 the oscillator "plate" and mixer "cathode", and G4 as the mixer grid. G3 and G5 were tied together as "screen grids". This arrangement was very popular despite the low plate resistance and is also the design of first the 1A7, then the 1LA6 and 1LC6, and finally the 1L6.
The "later" type as exemplified by the 6L7 introduced as a metal tube by RCA in 1935 eliminates the oscillator plate/mixer cathode, instead using G2 & G4 as screen grids. G5 now becomes a suppresser grid. While allowing good transconductance with low cathode current, in the words of the Radiotron Designer’s handbook, 4th edition "The lack of a separate oscillator plate results in certain disadvantages in short-wave performance but because a voltage supply is needed for one less electrode than usual, the cost of components required is a minimum.".(section 25.2 pg 999) This is the arrangement used by the 1R5 introduced with the first of the "minature" tubes in 1940.
The Big Black Dial
Meanwhile by 1935 Zenith was beginning to recover and a major selling point was becoming the "Big Black Dial" introduced on the Stratosphere. A BBD needed something to fill it and shortwave bands beckoned with the drama of the "Police" band and the allure of the European cities scattered about the "Shortwave" band (a few "U" and "A" series radios had a fourth band extending the range beyond 20 Mc as did the Strat. but this was more for advertising - like the 160 mph speedometer in an asthmatic 6-cyl ‘68 Firebird I once had - than usefulness). This circuit was the mainstay of Zenith radios throughout the thirties and early forties.
It worked well and Zenith saw no reason to change. The minimal cost savings that could be achieved with the later design was not justified in the "Royalty of Radio" where the need for performance at least to 18 Mc was a major selling point for all but Zenith’s least expensive offerings.
By 1938, Commander Eugene F. McDonald (head of Zenith and why their FM station in Chicago for many years had the call sign WEFM) sensed a need for a portable shortwave receiver to keep in touch while on his yacht and when at his Canadian fishing camp. The new 1.4v 50 ma Loctal tubes had just become available and while true "portables" would not be a part of the Zenith line until 1940, a design team was put to work on the perennial bugaboo of low-voltage shortwave - maintaining oscillation over a very wide range of frequencies.
The first working examples were little more than reworked model 5G401s with a single band to cover all frequencies. MacDonald complained that tuning with the small knob and dial required "micrometer fingers" and insisted that another means be found. This "electronic bandspread" achieved by selecting five fairly narrow shortwave bands (most bands were only about 400 kc wide) focusing on those allocated to commercial stations. Such narrow bands with wide gaps between them was to be a hallmark of all TransOceanics until the very last, the R-7000 (not to be confused with the prior Royal-7000) which added near continuous coverage.
However even though there were six bands instead of the typical three found on a contemporary console, the basic circuit of individual coil sets for each band was the same and the 1LA6 pentagrid converter was of the "early" design and chosen for maximum sensitivity at the higher frequencies.
Unfortunately, that was not enough though and oscillation kept failing so as Zenith had done previously on top-of-the-line units, a separate 1LE3 tube was used as oscillator with the 1LA6 restricted to converter duties.
Even so, the first production 7G605 "Clipper" introduced in late 1941 had its share of problems and at one point owners were advised that "drying in a warm oven" might help if their unit failed to operate as expected. Production continued until halted by WWII in April, 1942 with exactly 35,000 chassis recorded.
When production resumed in 1946, the new 8G005 "TransOceanic" was the flagship with completely new styling but the front end was almost identical to the 7G605 that preceded it. However, the 1LE3 was missing with three tubes now used for push-pull in the audio section replacing the single 3Q5G used before. This was possible because the case size grew by about two inches in both width and height.
Instead all oscillation was now a function of the Pentagrid converter, now a slightly lower rated 1LC6. The RF tower coil assembly was similar to the earlier unit with three coils (Antenna, Detector, Oscillaor). However performance was not as good as evidenced by the shift back to the premium 1LA6 for the TZ1 revision of 1947, the slightly lower tranconductance being offset by the 15% greater plate resistance.
Enter the 1L6
Thus as the 1950 models were in design Zenith had two major concerns for its flagship:
1) The circuit worked well but was "fragile", changes attempted had been detrimental.
2) There was a large manufacturing investment in the mechanical coil tower design.
Even though the new (1940) miniature tubes had proven themselves, in size, power consumption, and longevity, in the matter of the TransOceanic, Zenith was "risk adverse". Examination of the 5G40 (G-500) chassis shows that not only was there room for a 1LA6 /socket, the chassis was actually punched for a loctal/octal sized socket and a filler had to be used for the 1L6 even though it is obvious that all of the other positions were for miniature tubes. Zenith had no intention of risking the circuit design change required for the 1R5, the only listed minature pentagrid converter, or accepting the loss of shortwave sensitivity that would entail.
What happened was that Sylvania (Zenith’s major supplier) evidently made Zenith an offer they could not refuse - a 1LA6 in a miniature tube form factor, the 1L6 and so the G-500 introduced in 1949 became an all-miniature-tube shortwave portable, the first of its kind, and "long-life" batteries were a selling point. (Later units advertised 200 hours of play from a single A/B battery of 9 and 90v).
In fact examination of the Sylvania data sheet for the 1L6 indicates that the curves were not released until early 1949 - a very short leadin for a radio due to begin production around August.
In fact from the late 1949 (1950 model) G-500 to the last B-600 produced in 1963 - the last tube portable made in America, the basic coil tower/oscillator/converter circuit was really never changed though the mid-1951 H-500 replaced the narrow 49m band with two "marine" bands offering continuous coverage from 2mc-8mc (the A-600 of 1957 - a 1958 model - extended this to 9 mc).
Further, Zenith was essentially alone in offering a premium shortwave portable (several others, notably Hallicrafters & RCA tried briefly in 1952-1955 to crack the market- usually with identical tube lineups to the Zenith including the 1L6) but America in the ‘50s was near-isolationist and people were turning from radio to TV anyway.
Consequently, the 1L6 was never a high volume item and always expensive (even in 1966 it was the most expensive miniature tube in the Allied catalogue) since it had only a single use - premium (e.g. low volume) portable shortwave receivers.
Further, being a pentagrid converter (heptode) it was also more complex than an ordinary pentode or tetrode amplifier to manufacture. The 1R5 overcame this through sheer volume. The 1L6 did not.
As a result supplies of 1L6s began to dry up as early as late 1994. Asking prices were generally still in the $10 range but supply was becoming erratic. By late 1995, prices had hit $15, early 1996 $20, and 1997 is in the $24-$30 range. Surprisingly Antique Electronic Supply unearthed a number of JAN (military specification and generally a bit better than commercial) 1L6s made by Sylvania in 1968 in early 1997 which they sold for a comparatively reasonable $30. And some people complained.
First, the one found in nearly all substitution books, the 1U6. This is an identical companion to the 1L6 except for one thing, it is a 25 ma tube instead of a 50 ma.
This means that it will test identically to a 1L6 in a tube tester but when placed in a radio, nothing will happen since it has double the filament resistance and the TransOceanic uses a series filament string. To use a 1U6 in place of a 1L6 a 56 ohm resistor must be added in parallel with the filament so that the other tubes in the series string will receive proper current. (Note: some sets may have already been so converted. If in doubt, check).
However this is unlikely to be a problem since the supply of 1U6s is as short as 1L6s (maybe shorter). One caveat however: I have a number of National tubes brightly marked as 1L6s which turned out to be 1U6s (and the original designation was still visible under a strong light). So if it tests good but the set still does not work, check the filament string current (should be 50-60 ma). If in the 25-30 ma range, you may have a 1U6.
The substitution most commonly used in TransOceanics is simply to replace the 1L6 with a 1R5. This is possible despite the seemingly totally different design because while the fifth grid in the 1R5 is tied to the filament internally and to B+ in the circuit, the B+ line contains a 68k resistor which drops the current to a negligable level.
The amazing thing is that the 1R5 actually works - up to a point: In a good T/O the Broadcast Band and up to about 10 mc will have acceptable performance while above 10 Mc sensitivity will drop off noticeably. A really good alignment will make it appear as if it is working well, at least until compared to a really well-tuned T/O with 1L6. I have seen several "weak in the upper bands" T/Os turn out to have good 1R5s.
One enterprising gentleman even offered "1L6 replacements" in the form of 1R5s with pin 5 surgically amputated Unnecessary in a T/O but probably justified the 10X markup in price.
As an alternative, I have been working on using the 1AC6 or as it is known in Europe, the DK92 or military CV-5172. This is a late developed (c.a.1957) tube not often seen in the USA though quite common across the pond and in some Canadian sets. This is a combination tube with a design similar to the 1R5 though with separate screen connections, each grid being handled individually instead of the 3-5 connection of the 1L6 or the 2-4 connection of the 1R5. This enables it to be used with a g1 oscillator grid, g2 oscillator plate-mixer cathode, g3 mixer grid, g4 mixer screen, and g5 supressor with exactly the same socket wiring as the 1L6
On a tube tester, the 1AC6 acts like a 1R5 on steroids with 10% higher transconductance. Absolute ratings are very similar to the 1L6 with 90v maximum screen voltage instead of the 67.5 v maximum of the 1R5 (80v nominal in a T/O). The only limiting factor is a 60v maximum (30v nom.) on G2, the oscillator "plate" while the TO circuit design is for 84v. Is one of the experiments I keep meaning to get around to (see note below). On paper the 1AC6 is far superior to the 1R5.
I am told that it is quite common in Europe which did not succomb to the solid state revolution as quickly as the US and that the early 1997 price is around $4.00.
With further experimentation and having collected a number of different examples of 1AC6/DK-92 tubes, I can now report that there are two quite different examples of this class of tube, so different almost to be two different tubes. The first I have designated the "Standard" type. This example is charactorized by a plate (outer shell) which is solid and extends to the outer edge of the mica wafers. The second, designated "Steroidal" has a noticably smaller diameter plate which has two large vertical slots, one on each side.
When tested on a Hickock 533 Dymanic Mutual Conductance tester, the Steroidal type registers about 50% higher than the Standard type and alignment is much closer to the 1L6 settings than the other. Further, when plugged into a B-600 with AVC disabled, the volume from the speaker is much louder (about a 45 degree rotation of the volume control)
Consequently, in testing, both types had to be tried. In the single component experiment, different values had to be used though in the three component version, this was not necessary.
Note that the listed substitute for the 1AC6, the 1AB6 is a 25 ma tube like the 1U6 and so should have the same caveat. I have never seen a 1AB6 though several Telefunken models used it.
Since the TransOceanic circuit was designed for the 1LA6 loctal tube and it is evident that it was a consideration for the G-500 to use one with the rest of of the minature tubes so there is plenty of room. Thus with a suitable loctal tube to minature socket adaptor, a 1LA6 should be able to operate quite well in a later TransOceanic though a tube shield might be advisable.
This is one I originally thought would be more difficult than option 6 so naturally someone went and did it 8*). A solid state replacement for the 1L6.
Bob Pierfelice has developed just such a solid state 1L6 replacement, with his permission I have mirrored his article and schematic as it appeared in the MARC Chronicle web page. For the shopper, I have added an electronic parts list for the project.
The final possibility is to return to the same mechanism used by the first TransOceanis, the 7G605 - a separate oscillator. Since the oscillator never needs to be exposed to B+, an under-chassis solid state solution might restore weak 1L6s to operation or allow full frequency use of the 1R5.
So the 1L6 being a complex and low demand tube, the few remaining tube (e.g.Svetlana) manufacturers are unlikely to be interested, if they were, the price would likely be over $50. Further, the supply of 1L6s (barring discovery of a cache in an Army warehouse somewhere) is just about exhausted. Undoubtedly some people have acquired collections and are waiting for the market to peak however, it is likely that the market having reached the $30 range, either a substitute or a solid state answer will occur first, 1v tubes do not glow anyway.
If enough people want an answer, and are willing to pay for it, there will be one but not when good working T/Os are still often found for under $100.
Orlando, Florida, January, 1997
Note: when I first wrote these words in December, 1996 & January, 1997
little did I know that within two months both option 3 and 5 (appropriate, what ?) would have been completed, both making use of most readily available materials.
A long time ago I discovered that there seems to be a common characteristic of tinkerers: given a problem, a hundred different people will try a hundred different solutions and a couple of them will even work. The big difference is that even just a generation ago, few would have ever known...
Orlando, Florida, March, 1997
General Electric "Essential Charactoristics", 1968
RCA RC-19, 1959
RCA "Radiotron Designer’s Handbook - Fourth Edition", 1953
Combs & Bryant "Zenith TransOceanic - the Royalty of Radio", 1995
Stokes "70 Years of Tubes and Valves", second edition, 1992
SAMs "Radios of the Baby Boom Era" volume 6, 1991
Terman "Radio Engineering", 1947
Ghirardi "Radio Physics Course", second edition, 1937
Zenith "The Zenith Story", in house publication, 1955
Return to top of page
Go to Trans-Oceanic Home Page