Valve Questions

[Synchrodyne]

Indeed, the EF54 had as many cathode connections as UHF triodes had grid connections.

I guess that different designers had different views about the need for multiple cathode connections, particularly where pin count limitations meant there was a tradeoff. Early American B7G miniature VHF pentodes, such as the 6AK5 and 6AG5 had two cathode connections, but the trade-off was that the suppressor was connected internally to the cathode. The 6AS6 dual-control derivative of the 6AK5 swapped one of the cathode connections for a separate suppressor pinout. The 6AG5 was quite widely used in US TV receivers in the earlier days, and the later derivative 6BC5 also had two cathode connections. On the other hand the 6CB6 had one cathode connection and a separate suppressor connection, as did the 6BZ6, which seems to have been used as a TV IF amplifier in the 1960s.

In Europe the EF91 had one cathode connection and a separate suppressor connection, as did the Rimlock EF42 and EF43. In the latter pair, the extra pin was used for the internal screen. Presumably there was some merit in doing it this way. The EF80 and EF85, with noval bases, had two cathode connection as well as a separate pin for the internal screen, and that was carried over to the EF183 and EF184.

Cheers,

[Synchrodyne]

This thread: https://www.vintage-radio.net/forum/...ad.php?t=91537, which refers to a Pye TV receiver using no fewer than 7 PCF80 valves out of a total count of 15 prompted some pondering about the triode pentode as a species. One may subdivide it into two broad groups according to whether the pentode was of the signal type (i.e. “C+F”) or output type (i.e. “C+L”).

The “C+F” type seems to have been entirely a product of the miniature valve era, and very much a television valve. All of those shown in Vade Mecum (1994) are miniatures, and all are noval-based except the Mazda 10C2, which was a Rimlock. (Possibly quite late amongst Rimlock issues.)

The earliest C+F examples as far as I can determine are the American 6X8 and 6U8 of 1951. The 6X8 is mentioned in passing in the RCA 6BQ7 article (see my recent posting #182) as a new triode-pentode. Both the 6X8 and 6U8 were intended for use as VHF TV tuner oscillator-mixers. The 6X8 used a shared cathode, whereas the 6U8 had separate cathodes, which seems to have been the arrangement preferred subsequently. This application of the triode pentode seems to have been something of a paradox, indicated by the upward movement of (American) TV IFs circa 1950 from around 25 to 45.75 MHz, which was then nudging the lower end of the low band (Band I). Evidently the closeness of the lower low band channels to the IF made a pentode mixer a lot easier to use for these channels from the point of view of minimizing Miller feedback, whereas a triode mixer would have required neutralization. With the 25 MHz IF, separation from the lowest signal channels was sufficient that a triode mixer could be used, the 6J6 having been common as a TV oscillator-mixer with that IF. But a pentode mixer was relatively noisy for the high band (Band III) channels, where a quieter triode would still have been satisfactory with the 45 MHz IF. Thus the use of a pentode mixer for the high band channels was satisfactory only when preceded by a low noise, high gain RF amplifier in order to offset its very high noise, which was exactly what the series cascode 6BQ7 circuit provided. I suspect that it was no accident then that the 6BQ7 and 6X8, 6U8 all appeared at about the same time.

Philips released its PCF80 triode pentode TV frequency changer in 1953 at the same time as the PCC84 cascode double triode. I am not sure how well this coincided with movement of CCIR TV IFs from around 23 to 38.9 MHz, but I think that the latter happened within the 1953-1954 period.

Notwithstanding its primary mission, Mullard very quickly advertised that the PCF80 had much wider utility. In the “Valves Tubes & Circuits” series in Wireless World, the PCF80 was announced (along with the PCC84) in #15 (WW 195403), treated in detail as a frequency changer in #16 (WW 195404), with its other applications delineated in #26 (WW 195502) and #27 (WW 195503). In particular #26 discussed applications where the EF80, ECL80 and ECC82 might otherwise have been considered. (That by the way is the first mention that I can find of the ECC82 as a TV receiver valve.) VT&C #26 included a 10x10 matrix showing which triode applications could be combined with which pentode applications. The PCF80 versatility is not so surprising. The EF80 had already been cast as a Band I RF amplifier that was also suitable for use as a mixer, IF amplifier, video amplifier and sync separator. Presumably the pentode section of the PCF80 had been optimized for mixing at 200 MHz, but might not have been a world apart from the EF80.

The TV frequency changer application appears to have been dominant for the development of subsequent C+F type triode pentodes. But there were some American types, such as the 6AN8 and 6AZ8, that were designed for multiple applications other than frequency changers. The later European PCF200 was also in this class.

The “C+L” type of triode pentode did exist in the pre-miniature era, an example being the octal 6AD7-G, intended for audio applications. But for the most part, and like the “C+F”, it was mostly a miniature valve with TV applications dominant. The ECL80 was an early example, an original member of the Philips/Mullard TV World Series, and from the start multipurpose in nature. For example, the triode could be used as a field or line oscillator or an audio voltage amplifier, and the pentode as a field or audio output amplifier or a sync separator. Following was a whole series of noval-based C+L triode pentodes of varying primary application and diversity, ECL/PCL82, ECL83, PCL83, ECL/PCL84, ECL/PCL85, ECL86 and PCL86. Then came the decal-based ECL/PCL200. After TV, audio might have been the next most important application, in that a pair of triode pentodes facilitated low valve count in modest output power amplifiers. Conventional radio receiver practice seemed to favour the use of for example EBC81+EL84 or EABC80+EL84 combinations, so was not such a fertile field for the C+L triode pentode, although such would have been found in some push-pull output circuits and in stereo radio receivers.

Some of the American multipurpose triode pentodes are difficult to classify as being “C+F” or “C+L”. I suppose that one might use the ECL80, with its 3.5 W anode dissipation, as an indicator as to where F becomes L, in which case American examples can be found above and below this divide. An interesting mid-1950s example, definitely C+L, was the 15A8 field oscillator/output combination, on an octal base with a topcap. I imagine that its 8.5 W anode dissipation was considered too much for a miniature valve at the time.

Cheers,

[Synchrodyne]

Recently I came across a 1947 article by RCA (*) that provides an interesting perspective on the early development of American miniature valves. Effectively the article is counterpart to the readily available Philips literature on the development of the Rimlock domestic receiving valves.

The conclusion of the article provides a good summary:

“The four miniature tubes introduced in 1939 were designed to fill the need for tubes which could be produced in quantities and at reasonable cost for a compact personal-type receiver. In the intervening years, however, the practical experience with miniatures has shown them to be adaptable for general receiving tube use with an exceptional range of capabilities. The versatility of miniatures holds promise for higher performance and lower costs in electronic equipment of the future.”

It is evident that the original B7G valves, 1T4, 1R5, 1S5, and 1S4, were sized around the needs of battery-operated personal receivers. As well as being much smaller, they were less costly to mass-produce than say GT octal types and had better performance at higher frequencies. Thus they were well-suited for military applications during WWII. Early developments for military purposes were the 6C4 triode (oscillator), 6J6 double triode (mixer-oscillator) and 6AG5 RF pentode, all from 1942. The latter was intended primarily for 30 MHz IF use, but was also used as an RF amplifier up to around 200 MHz. The 6AK6 output pentode (2.75 W) was added in 1943. The 6AL5 double diode arrived in 1944 for use in demodulator service in place of diode-strapped 6J6, along with the 6AQ6 double diode triode.

Following WWII, the 6J6, 6AG5 and 6AL5 also became “television” valves. In its initial post-WWII TV receivers, RCA used the 6J6 (3 of) in its 13-channel tuners, the 6AG5 in its vision IF strip, and the 6AL5 as vision demodulator inter alia.

Presumably because it was an RCA article, only RCA developments appear to be covered. For example there is no mention of the 6AK5 RF pentode, which I think was a Sylvania development during WWII and better than the 6AG5 at higher frequencies.

September, 1945 saw the introduction of a full range of domestic receiving miniatures. Curiously, here RCA gives precedence to the AC/DC (150 mA heater) range, namely 12BA6, 12BE6, 12AT6, 50B5 and 35W4. The corresponding AC (6.3 V) range, 6BA6, 6BE6 and 6AT6 were mentioned on more-or-less and “also ran” basis, although it is noted that the AC range included the 6AU6 sharp cutoff pentode, a valve whose origins were previously unclear to me. The 6AU6, evidently sharp cutoff counterpart to the remote cutoff 6BA6, does not appear to have had a similar counterpart type in the European range of domestic receiving valves. At least I am not aware of any sharp cutoff companions for the EF41 and EF89. Typically the EF80 (counterpart to the EF85) would have been used for applications, such as FM limiters, where US practice favoured the 6AU6.

An interesting comment is that the 6BA6 and 6BE6 were designed for use at the then-new FM band frequencies of 88 to 108 MHz, as well as at lower frequencies, thus facilitating their use in AM-FM receivers. Whilst they might have been marginal in FM service, particularly the 6BE6, the achievement of acceptable performance (by the standards of the time) was no doubt attributable to their miniature nature. Nevertheless, the use of heptode frequency changers for FM was short-lived, and RCA also listed the 6AG5 for FM RF use, where it was presumably a bit quieter than the 6BA6.

RCA used the 6BA6 and 6AU6 in the sound IF (split sound) sections of its immediate post-WWII TV receivers, along with the 6AL5 as discriminator and 6AT6 as audio voltage amplifier. The 6AU6 was also used as the 1st video amplifier.

The 6X4, 6BF6 and 6AQ5 are described as having been released in December, 1945 for car radio applications. Classifying the 6X4 and 6AQ5 as car radio types leads to an inference that the early thinking was perhaps not to use them in AC radio receivers, but to retain the previous octal types (6X5GT and 6V6GT) for those functions, this perhaps on the basis that unlike the car radio case, there was not such a pressing need for miniaturization. Still, the 1947 article listed the 6AQ5 without alternative as the designated output valve for AC FM-AM radio receivers, although no rectifier valve is listed. The 6AQ5 was also listed as the TV sound output valve.
Quite why the 6BF6 double diode 16 µ triode was needed is not immediately apparent, but perhaps it was to align with pre-WWII octal valve practice. Be that as it may, by 1947 RCA was including both the 6BF6 and 6AT6 in its car radio list.

So by late 1945 the B7G miniature valve range included four double diode triodes with similar and somewhat if not significantly overlapping functions, namely the 6AQ6, 12AT6, 6AT6 and 6BF6. The 6AV6 was not mentioned in the RCA article, but I think it appeared during 1947. There was, it seems, leporidic proliferation. The 6AQ6 (70 µ triode)had relatively low heater dissipation (6.3 V, 150 mA) as would have been preferable for military aviation and other applications, but one assumes that it could have been usable in both AC/DC (150 mA heaters) and AC (6.3 V) receivers. But in the former case perhaps a 12.6 V valve was preferred in order to have the desired summation of heater voltages. Perhaps also there was some suspicion that low heater consumption valves were not as good and/or not as robust as conventional (1.9 W) heater consumption valves, and so should not be used outside of applications where they offered direct benefits. (Similarly the later 6BH6 and 6BJ6 low consumption RF pentodes seem to have been rare in domestic equipment.)

As an aside, whilst the 6AT6 and 6AV6 represented something of a final plateau in regular radio receiver double diode triodes, the American species subsequently developed along a different vector, one without European counterpart, for TV (and to a lesser extent FM) applications, including noval-based variants with two and three cathodes.

Finally the RCA article notes the then recent introduction of a nine-pin miniature in a larger envelope, namely the 12AU7 double triode, for both industrial and domestic applications. The comment was made that although the larger size was primarily to provide room for the additional pins, it also opened up the possibilities of introducing higher wattage types to supplement the seven-pin miniatures, given that bulb temperatures were a limiting factor. That aligns with Philips’ reasoning behind its choice of bulb size for its Rimlock series. But Philips was thinking in terms of accommodating a 9 W output pentode, whereas RCA had already put a 12 W output beam tetrode into the small B7G envelope. And as far as I know, 12 W was the maximum anode dissipation ever used in a B9A valve. Still, stretching of technology beyond its originally envisaged boundaries and in some cases beyond what might be considered prudent is almost inevitable in the face of perceived commercial needs.

Essentially as predicted by the article, subsequent American miniature valve development seems to have proceeded on the basis of B7G as the default type, with B9A invoked only where more pins and/or a larger bulb were needed. Likely there were some exceptions, but no B9A types come to mind that clearly could also have been executed in B7G form. And most receiving functions were eventually covered by miniature valves.

Cheers,


(*) Miniature Tubes in War and Peace
N.H. Green
RCA Review, Volume VIII, No. 2, June 1947
p.331ff

[turretslug]

From my earliest interest in matters thermionic, it puzzled me that the 6AL5 was such a greedy little bottle in terms of cathode heating. However, its origins as a military technology as opposed to civilian receiver valve makes things a bit more apparent. A diode required to act as a clamp, pulse gate or shaper etc. (radar and other accelerated developments) would need to guarantee higher peak emissivity than an AM envelope detector. Certainly, its internal structure is much more reminiscent of a rectifier than the teensy diodes present in DDT/DDP's of even relatively large IO and pre-IO format.

I think (possibly erroneously) of the 6AK5 as a successor to the EF50 in symbolic as much as actual technical terms, i.e. the mounting dominance of the US over UK technology/manufacturing. A while back, i was given a box of assorted valves that included a sealed carton with familiar JAN label proclaiming that within was a 6AK5 accepted on 6th June 1944- worth nowt, of course, but I salted it away as a little scrap of history. I'm sure that the EF50 counted as a "miniature" valve in 1938 but it must have looked comically large against the B7g offerings from mid-war onwards.

Did the shortcomings of the 6BE6 and 6K8 (discussed in a previous FM front -end thread) inspire a "final push" on the 100MHz multigrid theme in the shape of the 6BA7 or was it extant as an HF mixer anyway and was pushed a bit faster than was ideal for the fundemental topology? I suppose there was the temptation to say, well it works fine at 18MHz, OK at 30MHz, even passably at 45MHz (original US FM BC)- only another octave...

[Radio Wrangler]

The multigrid mixers were a fairly noisy bunch, so they needed generous RF gain ahead of them, and there went the dynamic range at the top end. The E180F did a bit better. Some British naval receivers got good dynamic range out of ECC82 (yup 82 not 81) according to Pat Hawker.

Using a pair of triodes in a cascode mixer wasn't a bad ploy, but the real star for mixers, phase detectors etc etc hasn't been mentioned yet in this thread, unless I missed it. The 7360.

David

[turretslug]

I had made a little mental list of "valves to go" in my last post, I was going to follow 6BA7 with 7360 but forgot.

Honest...

[Synchrodyne]

Hi Turretslug, I think that you are right about the 6AL5. The RCA article notes that it found that the 6J6 was being used diode strapped both to save space and to obtain higher perveance than was available with existing diode types, so the 6AL5 was designed to replicate its characteristics. None of the existing diode types are mentioned, but presumably the 6H6GT was in the frame here. The 6AQ6 was designed for narrow band applications requiring lower perveance.

I was surprised to learn that the 6BA6 and 6BE6 were intended for FM as well as AM use. I rather thought that the 6BE6 had been pressed into such service post facto. I wonder if its putative improved performance at 100 MHz as compared with its octal forbears stemmed largely from the fact that it was a miniature, and not from more fundamental changes. The 6BA7 arrived in 1948, and does appear to have been a last gasp effort at making the pentagrid work at 100 MHz. The data sheet covers both the separately excited and self-excited cases, which suggests that the former might have been preferred where best performance was desired and a separate oscillator valve (6C4 or triode-strapped 6AU6) could be accepted.

Once the 6BQ7 double triode was available, then for higher quality consumer FM equipment a 6BQ7 cascode RF amplifier followed by a 12AT7 mixer-oscillator was probably a reasonable choice. Maybe a triode pentode (6U8, 6X8) could have been used as a mixer-oscillator, but unlike the TV case, there would not have been the same need for a pentode mixer. With the triode being a bit quieter, even at 100 MHz, then perhaps some judicious backing off of RF stage gain as compared with TV practice would have been possible, thus securing better big signal handling. But then FM practice was more likely 3-gang, whereas TV VHF tuners were usually 4-gang.

From the later 1950s onwards, frame grid valves might have replaced the 6BQ7 and 12AT7, etc. The American series even included the 6JK8, an asymmetric double triode for FM mixer-oscillator in which the mixer was a frame grid but the oscillator was not.

Hi Radio Wrangler, I am not sure I have ever seen beam deflection valves mentioned for consumer FM mixing purposes, although they have cropped up in colour TV and stereo decoding circuits. I am guessing though that the 7360 would be a relatively costly valve as compared with the more familiar possibilities.

Cheers,

[Radio Wrangler]

The 7360 and its relatives were used in some bits of military/aerospace kit, but colour TV decoders were their natural habitat.

As you say they would be a bit too expensive for plain-vanilla consumer radio use, and other devices worked well enough. One thing about the 7360 is that it made a spiffing mixer for HF receivers (EG G3PDM receiver published by RSGB in the 1960s) I often wondered about doing a double-balanced mixer with a pair of them.

The interesting thing for FM is that one could have been used to make a quadrature detector, but I don't know of it being done.

The Foster Seeley discriminator has the capability of being a bit more linear than the ratio detector. but the ratio detector is less critical of the limiting performance of the IF, so the cost of the IF could be cut, and that was the clincher.

David

[G8HQP Dave]

The 7360 and its cousins is good, but not quite as "spiffing" as popular legend has it. Roughly speaking, it has the linearity of a straight RF pentode amplifier with mixer noise which is similar to an RF pentode mixer. Claims that it is amazingly linear and very low noise, which appear from time to time, are simply untrue. It just happens to have the right combination of noise and linearity that you can, just, get away with using it as the first device in an HF receiver. Because it is much more noisy than a pentode RF amp you need very good input circuit design with tight coupling - hence the complaints that it can be fussy about antennas. The linearity improvement comes simply from the fact that no RF gain is used, so the mixer sees much smaller signals.

People not realising the above have tried to use it and ended up having to add an RF amplifier in front of it to get enough sensitivity, thus throwing away the linearity advantage.

[G6Tanuki]

If you want a real oddity look at the 6BN6 "Gated Beam" valve - originally designed for use in oscilloscopes/Radar-displays and as a FM demodulator but it also works well as a SSB product-detector [in a bizarre mathematical way, FM can be thought of as SSB with the carrier reinserted 90 degrees out of phase!]

[Radio Wrangler]

The carrier reinserted 90 degrees out of phase thing is exactly what the quadrature detector does and is named for.

No, the 7360 isn't perfect, but it was an interesting step after all the heptodes etc.

One interesting paper from around the time of the G3PDM design was Rafuse on the ring of Mosfet mixer in 1967. Pity he put a 2 port device in a circuit designed to overcome the problems of using one-port devices, and never wondered about a better topology for 2 port devices. Another interesting stream is Chris Trask's writings on linearising feedback around switching mixers.

David

[G6Tanuki]

Originally Posted by Radio Wrangler

The carrier reinserted 90 degrees out of phase thing is exactly what the quadrature detector does and is named for. David

Thank you! I will freely and openly confess that until you pointed it out, the relevance of 'quadrature' in this context had entirely passed me by. Sometimes you can get bogged-down in the arcane worlds of arcsines and cotangents and second-order differential calculus and not appreciate what's staring you in the face!

--G6Tanuki

[Synchrodyne]

Originally Posted by Radio Wrangler

The interesting thing for FM is that one could have been used to make a quadrature detector, but I don't know of it being done.

Would the 7360 more easily have allowed better linearity than the other forms of valve-based quadrature demodulators?

Tibbs & Johnstone show curves for the nonode (EQ80) and gated beam valve (6BN6) cases, and neither is all that linear as compared with the Foster-Seeley discriminator. Both the EQ80 and 6BN6 appear to be of the quasi-synchronous quadrature form. Tibbs & Johnstone do not cover the locked-oscillator quadrature case, which might be thought of as being fully synchronous, but I suspect that this would not have been as linear the Foster-Seeley, either. Early American practice used a heptode (e.g. the Philco FM1000 valve), but later dual control pentodes, such as the 6DT6, seem to be favoured for locked-oscillator TV sound demodulators. In Europe the EH90 heptode was used, although this had originated as the American 6CS6 which was developed for noise-gated sync separator applications. (As was the 6BY6, which found additional employment as a colour TV synchronous demodulator, as an alternative to the 6AS6 dual-control pentode.) Curiously, at the same time that Mullard was promoting the EH90 for TV FM sound demodulation, it was offering the ECH84 as a noise-gated sync separator. But then Philips had previously used the ECH83 (car radio triode heptode) for the same purpose.

Whilst the quadrature FM demodulator was something of a secondary type during the valve era, used mostly I think to save an audio voltage amplifier valve in TV receivers (although given the availability of ECL types in Europe at least, even that gain was questionable) it became the primary type in the integrated circuit era, being an early consumer IC type, and best represented by the iconic RCA CA3089 and its ilk. Available IC gain and bandwidth and the use of double-tuned tank circuits evidently allowed impeccable linearity.

Originally Posted by Radio Wrangler

The Foster Seeley discriminator has the capability of being a bit more linear than the ratio detector. but the ratio detector is less critical of the limiting performance of the IF, so the cost of the IF could be cut, and that was the clincher.

The poorer linearity of the ratio detector as compared with the Foster-Seeley discriminator seems a little odd if one follows its conventional description as a variant of the latter. On the other hand, if it is viewed as being of the sampling type, and so I think following the quadrature form (as discussed in this thread: https://www.vintage-radio.net/forum/...ad.php?t=91712) then perhaps it is not so surprising. But with regard to the conventional description, Tibbs and Johnstone give a pointer with the comment: “In the ratio detector design, it is not possible to adjust the coupling factor to obtain best linearity as in the Foster-Seeley circuit, these parameters being fixed by considerations of a.m. rejection.” Improved linearity could be obtained by using a wide bandwidth, and this approach seems to have been adopted in the solid state era before the IC quadrature demodulator took over.

As well as eliminating the need for a limiter valve, the ratio detector made IF design a bit easier in that it required a lower level IF input, say 100 mV or lower as compared with 1 V or so for a limiter. Assuming a two-stage IF, I imagine that there was probably more room to have the second stage as a combined amplifier/limiter, providing high level limiting.

Cheers,

[Synchrodyne]

The current thread “Which Valve Came First”, https://www.vintage-radio.net/forum/...707#post614707, essentially about the EL84, also recycled the pentode vs. beam tetrode issue, discussed in detail upthread.

Not yet mentioned though as far as I can see is any comparison between the KT77 (beam tetrode) and EL34 (pentode). Given that the KT77 was intended by GEC to be an alternative to the EL34, here would appear to be a case where beam tetrode pentode met head on, with no allowances needing to be made for significantly different slopes, as say with any 6V6 family vs. EL84 comparisons.

Was the KT77 known or shown to exhibit any advantages over the EL34, at least outside of individual hearing preferences? Or perhaps did the design brief of matching the EL34 more-or-less force it to be a clone in terms of outcomes, although not in terms of construction.

Cheers,

[Radio Wrangler]

I suspect that technical matters of performance and construction were running in second place to commercial matters. Valve manufacturers were using patents to lock the opposition out of each other's technologies. What they were offering equipment manufacturers were watts-per-pound (Sterling....) and the waters were muddied by ties between equipment and valve makers.

These large valves were going into performance amplifiers for musicians, public address amps, and auditorium amps for cinemas etc. They got used in pairs and quads. The next size up was a big jump to RF transmitting devices like the 813 and upwards, as well as some purpose-made modulator valves in the kilowatt league.

Quad and Leak were interested in overall distortion and the tradeoff with the amount of feedback needed. Musicians either just wanted a lot of power (Get enough power and the glory will look after itself) or just a lot of power allied with a lot of hard distortion.

The EL34 certainly could deliver the goods in a hard amplifier for guitarists. I knew Matt Mathias responsible for the Matamp family (also sold in appropriately coloured rexine under the 'Orange' brand name) He used push pull arrangements of parallelled EL34s.

For more domestic power levels, the EL84 pretty much cleaned up in Europe. The EL84 didn't seem a particularly long-lived bottle and produced alarming discolouration. Given the number in use and the shortish life, I wonder just how many were made? Philips/Mullard must have done well out of it.

Commercial interests often overrule the technical (Beta/VHS/2000?) so I think it may be an error to try to work out who was doing what to whom with what and for how much purely on technical matters.

David

[Synchrodyne]

This item came up in the thread: https://www.vintage-radio.net/forum/...519#post627519, and I thought it would be worthwhile to bring it across to here rather than to wander off-topic in that thread.

Originally Posted by Leon Crampin

The dual control of variable mu pentodes is an interesting point. The EF50 is fully characterised for this type of operation and probably quite a few post-war set designers would have cut their teeth on designs using this excellent device during the war.

I wonder if Mazda were a bit more straightforward in describing this mode of operation with types such as the 6F18, which is very like the EF89. Pre-war, Mazda characterised types such as the AC/SP1 for noise suppression with g3 control. I'll check the post-war literature.

I’d say yes, given that Mazda had offered the 6F33, allegedly the only European purpose-designed dual-control pentode; see: http://www.radiomuseum.org/forum/dua..._pentodes.html.

In fact Mazda seemed to be quite alert in the pentode department. Its 6F1, similar to the EF42, had the two cathode connections often used for VHF valves whereas the EF42 had one; Philips/Mullard did not introduce the second cathode connection in this valve class until the EF80, which I think was essentially a somewhat trimmed EF42, as required to move the heater current down to 300 mA from 330 mA.

The 6F15, similar to the EF41, had a separate suppressor grid pinout whereas the EF41 had the suppressor internally connected to the cathode, as did the derived EAF41. Philips/Mullard corrected that with the EAF42 and EBF80, but never amended the EF41 itself, instead waiting until its EF89 successor.

And then the 6F18 went one-up on the EF89 by having a second cathode pinout, per VHF pentode practice. Presumably the potential benefit of this at HF is recorded somewhere in the Mazda literature.

I think too that Mazda was the first with a remote cutoff VHF TV mixer pentode in a triode-pentode combination, namely the 6C17.

Cheers,

[Synchrodyne]

Also from the thread: https://www.vintage-radio.net/forum/...69#post627569:

Originally Posted by Radio Wrangler

RCA produced a linear power valve and called it a 'Beam Tetrode" and would sue anyone transgressing on their territory. Philips had some sort of "Power Pentode" which they didn't seem to talk about, just sell, and MO-V developed their Kinkless Tetrode and said it was developed to dodge the patents of RCA and Philips, and do a better job as well.

They all relied primarily on aligned grid wires and all the firms could have lost fortunes to our learned friends in one huge legalfest of mutually-assured destruction.

That RCA development is interesting, because the conventional wisdom has it that MO-V developed the beam tetrode, but was unsure whether it could achieve the precision level required for its manufacture, so passed it over to RCA who could do the job. (I think that RCA had access to the Marconi-EMI work on television, as well, so there seemed to be an open door there.)

One wonders then how the RCA and MO-V original work differed. Both would have used electron beams formed by the aligned grids and electrode geometry, the latter also allowing the formation of a low potential point that acted as a virtual cathode, and by its positioning thus as a virtual suppressor grid, so making the valve a virtual pentode. Did perhaps MO-V add the beam-confining plates that allowed the use of non-circular cathodes and grids by keeping the beams in the areas where the geometry was right?

Cheers,

[G8HQP Dave]

My understanding is that MO-V were not avoiding RCA patents, as they had a patent-sharing arrangement. They were both avoiding Philips patents on the pentode. MO-V came up with the idea; RCA worked out how to manufacture it reliably and then developed it further. 'Kinkless tetrode' sounds more like marketing than engineering talk.

[Synchrodyne]

Aha, that aligns with the conventional wisdom. I agree that “kinkless tetrode” does look like a marketing term; it conveys the effect (benefit) rather than the means used to obtain that effect.

Now to the EF85, a specific application of which is the subject of this previously referenced thread: https://www.vintage-radio.net/forum/s...69#post627569. The EF85 seems to have had a curious history. It was introduced circa 1951 in fact if not overtly as the remote cutoff counterpart to the EF80. The latter very definitely started life as a television valve, having application as a Band I RF amplifier and mixer, as an IF amplifier, as a video amplifier and in various other functions such as sync separator.

But I have never come across any early Mullard advertising that refers to the EF85 as a TV valve. It seems to have lived in the dark until it became part of the “new” FM/AM radio receiver range in 1954. For a while thereafter it was described as being primarily an FM or FM/AM radio valve. Possibly this was because Philips was inclined to use the sharp cutoff EF80 with agc in TV applications, so may have had little use for the EF85 as a TV valve, and this attitude informed Mullard’s approach.

At least for FM/AM radio IF strips, and probably for FM-only radio IF strips, it was quietly pushed aside by the EF89, which Mullard presented as an improved EF41, without mentioning the EF85, emphasizing its low Cag and resultant stability in the IF role. That left its main applications as being a TV IF amplifier and to a lesser extent, an FM receiver RF amplifier.

The TV IF application fell away once the EF183 arrived. Here the EF183 (remote cutoff) and EF184 (sharp cutoff) were treated as a pair of TV IF valves, and being announced as the respective successors to the EF85 and EF80 effectively validated the fact that the EF85 was a TV valve.

Much later Philips literature described the EF85 as an RF pentode suitable for use as a wideband amplifier, which description fit both its TV IF application and also the FM RF case, where it was typically used with a broadband input, as in the Chapman models. That description also points to its origin as a TV valve.

The EF80 also found employment in FM receivers, as RF amplifier, mixer, IF amplifier and limiter. Whilst it fitted these, in the IF and limiter cases it may also have been a default choice because there was never, as far as I can determine, a sharp cutoff counterpart to the EF89 (or to the EF41 before it) in the standard European domestic receiving valve series. That was unlike the American case, where the 6AU6 was the sharp cutoff counterpart to the 6BA6; the 6AU6 was the usual choice for FM limiter stages, and indeed for many other applications where a general purpose sharp cutoff HF pentode was required.

Just as the EF80 predecessor was the EF42, the EF85 predecessor seems to have been the EF43. But I do not think that the EF43 was a Philips/Mullard valve; probably it came from the another European valve maker.

Cheers,

[Radio Wrangler]

I'm not convinced that 'Kinkless Tetrode" carries the mark of a marketing decision.

I think they'd have gone for "Low Distortion" as being the end result presented to the final customer. Kinkless would only mean something to someone who already had graphical valve characteristics in his head. The final customer the man on the Clapham omnibus would be deciding between a Quad amplifier, advertised as using the new kinkless tetrodes, or the Quintuplet amplifier advertised as having the new Low Distortion valves.

He'd be familiar with discussions of distortion. The valve manufacturer would have handed a better tool to the amplifier manufacturer's marketeers.

David