|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
12th Oct 2019, 4:04 am | #1 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Why Both EAA91 and EB91?
This thread is a spinoff from the thread Valve Items - Philips/Mullard Rimlock-to-Noval Transition https://www.vintage-radio.net/forum/...d.php?t=159120, wherein the question arose – why were both the EAA91 and EB91 issued, when they appear to be functionally quite similar?
A first pass look at the easily accessible data does not produce a definitive answer. As a starting point, what were probably the final Philips datasheet issues for these valves were dated respectively 1999 June 12 for the EAA91 and 1999 June 14 for the EB91. That for the EB91 is quite brief. No quantitative performance data was provided, simply the note “For further data refer to type EAA91.” From those datasheets, the only discernible difference between the two was a dimensional one. The overall height (top pip to pin ends) of the EAA91 was shown as 44.5 mm maximum whereas that for the EB91 was 54 mm maximum. In that the EB91 referred back to the EAA91, Philips may have viewed the latter as the primary type. An earlier reference is the Philips Australia 1962 Technical Data Book. This provides a tabular listing of a very wide range of valves (including the Innoval types not mentioned in Philips Netherlands sources.) Both the EAA91 and EB91 were listed, and in both cases with the comment “For data and notes refer type 6AL5”. (No dimensional data was provided in this case.) At this stage one could say that in later years at least he EAA91 and EB91 were functionally the same valve, differing only in overall length. Whether that difference was deliberate or happenstance is unknown. Earlier still was a Philips (international) data pocket book, apparently from 1954. This appeared to cover most of the Philips range, including miniatures, Rimlocks and novals, although not all of the Australian Innovals were listed. The apparent date of 1954 seems about right. For example the EF89 (thought to date from c.1954) was included but the EBF89 (thought to date from c.1955) was not. The double diodes listed were the listed the EAA91, EB91 and 6AL5, as well as the EB41 and UB41. Each had its own brief dataset (no physical dimensions) except for the 6AL5, for which it was said “see EAA91”. Furthermore, the EAA91 was the only double diode in the “preferred types” summary at the front of the book. That could have been a change of position by Philips, as the double diode in the TV series, as delineated in Book IIIC, was the EB91. The 6AL5 was the standard double diode in the American miniature series. It had been introduced as a WWII valve by RCA in 1944. It was found that the 6J6 miniature double triode (dating from 1942) was being used in the triode-strapped mode where a high-perveance diode was required for wideband demodulation. Thus RCA developed a high-perveance double diode with performance characteristics very similar to that of the triode-strapped 6J6. The 6AL5 became a part of the initial American B7G-based AC miniature receiving valve range in 1945, one of its recommended applications being as an FM receiver demodulator. The 12AL5 derivative, with 150 mA heater making it suitable for use in AC-DC radio receivers with 150 mA series-string heaters was added in 1947. Then the 3AL5, with 600 mA controlled warm-up time heater was added in 1954 for use in TV receivers with series-string heaters. The overall height of the 6AL5 & co. was 1.75 inches maximum, that is 44.5 mm. That was the same as for the EAA91. From that one might deduce that the EAA91 was simply the European version of the 6AL5, with the EB91 being an electrically similar but physically different (for some reason) variant. But it does appear that the EB91 preceded the EAA91. The EB91 was included in the 1949 Mullard publication “Valves for Industry and Communications”. So it was in place at least by 1949. The same publication also included the EB41 and UB41 Rimlock double diodes. The EB91 had a maximum height of 55 mm, and the EB41 and UB41 57 mm. Whilst the exact release dates for the EB41 and UB41 are unknown, the double diode suitable for FM and TV purposes was certainly within the scope of the original Rimlock receiving valve range as discussed in Philips Technical Review 1946 October. Evidently Philips saw these as its principal post-WWII receiving double diode types. However, in 1949-50, when it decided to adopt the noval base for most of its standard TV receiving valves, it did not develop a noval double diode, but rather took the existing EB91 into its TV range. Whether Philips itself had previously offered the EB91 is unknown; perhaps it was drawn from the Mullard range. One may then ask, why did Mullard (by 1949) offer both the EF41 and EF91, which were very similar in performance characteristics. A possible answer may be found in the Wireless World 1947 November report on the valves shown at that year’s Radiolympia. In respect of Mullard it was said: “Miniature all-glass valves shown by Mullard are mainly for use in industrial equipment. There were a few samples of the valves that will eventually be produced for domestic receivers, with the small B8A pin arrangement.” (The associated picture included both the EF42 (B8A) and EF91 (B7G) high-slope pentodes, another duality that has been covered in the recent thread EF91 to EF80, https://www.vintage-radio.net/forum/...d.php?t=157399.) In 1947 the term “miniature” was mostly used for the B7G types, the Rimlocks being identified as such or as B8A. Thus Mullard probably saw it as necessary that its B7G industrial range should include a double diode, hence the EB91. The competitive situation may also have indicated this. Osram had opted for the B7G base for both industrial and domestic applications, and its double diode was the D77, part of its “77” range. Mazda, whilst a strong proponent of the Rimlock base, had nevertheless offered a small number of B7G valves for industrial applications, including the 6F12 high-slope pentode and the 6D2 double diode. That latter also doubled as its domestic double diode; unlike Philips/Mullard, it did not release a Rimlock version. The D77 and 6D2 both had properties that were more-or-less the same as those of the 6AL5. According to a 1948 May datasheet, the 6D2 had a maximum height of 54 mm, the same as was later specified for the EB91. The D77 had a maximum height of 49 to 55 mm, according to a 1951 October datasheet. Brimar’s offering in this class was, unsurprisingly, the 6AL5 itself. Brimar Manual #7 gave the height of its 6AL5 as 1.875 inches maximum, 47.6 mm, different to that of the original American 6AL5. The EAA91 appears to date from 1951. Perhaps it was intended to be an exact match, both electrically and physically, for the (American) 6AL5. A Mazda datasheet dated 1961 December showed its miniature double diode as being the 6D2/EB91, with 54 mm maximum overall height. A Mazda Belvu datasheet dated 1968 January had what it called the 6AL5/EB91 with a 45.24 mm maximum height A 1956 GEC summary leaflet listed its double diode as the D77/6AL5, no physical dimensions given. Condensing the foregoing, we can say: The 6AL5, EB91 and EAA91, released in that order, were functionally the same valve. So were the Mazda 6D2 and Osram D77 functionally the same as these three. The original American 6AL5 had a maximum height of 44.5 mm, but some European issues were taller. The original EB91 was taller, up to 55 mm, but some later issues were shorter. The EAA91 appears to have conformed to the original 6AL5 physical specification. But that still does not produce a “hard” answer as to why both the EAA91 and EB91 were necessary. One could say that the “short” version might have been needed for some compact equipment that had been designed around its dimensions, and in which the “tall” version would not fit. But that just moves the unanswered question to :why was the (functionally identical) taller version developed in the first place? Cheers, |
12th Oct 2019, 10:06 am | #2 |
Moderator
Join Date: Jun 2003
Location: Oxford, UK
Posts: 27,967
|
Re: Why Both EAA91 and EB91?
The EB91 was by far the commonest version around here. I don't think I've ever seen an EAA91. As you say, it's puzzling that both were produced, especially as the market for the EAA91 appears to have been tiny, whatever it was.
|
12th Oct 2019, 10:11 am | #3 |
Dekatron
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 14,007
|
Re: Why Both EAA91 and EB91?
I wonder if there might have been some more-prosaic production-line-related drivers for these sorts of things, something like "we need to increase our diode-production! We don't want to build another production-line for B8A as the world's not really embraced the idea, and we've got spare capacity on the B7G lines".
|
12th Oct 2019, 10:39 am | #4 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Why Both EAA91 and EB91?
Maybe the taller version allowed use of the same envelope as other B7G valves, which might ease both stock control and production line settings. However, set against this idea were other small envelope valves such as EF95/6AK5 and other 'UHF' valves.
Does the EAA91 have the electrodes set nearer the base, so shorter connections and better high frequency performance? |
12th Oct 2019, 11:16 am | #5 |
Octode
Join Date: Jan 2003
Location: Ware, Herts. UK.
Posts: 1,082
|
Re: Why Both EAA91 and EB91?
The Grundig 5080 that I used to own featured an EAA91 and this was specified as such on the valve lineup printed on the back panel. Being a typical tall, imposing '50s German set there would have been plenty of room for the taller EB91.
John |
12th Oct 2019, 11:46 am | #6 |
Moderator
Join Date: Jun 2003
Location: Oxford, UK
Posts: 27,967
|
Re: Why Both EAA91 and EB91?
I wonder if there was some dispute within Philips regarding the naming of double diodes with separate cathodes. The diodes in the EB/EAA91 are completely separate structures.
The Germans did seem keener on EAA91s than anybody else. |
12th Oct 2019, 5:01 pm | #7 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: Why Both EAA91 and EB91?
I seem to recall that in my own stock of EB91/EAA91 and lookalikes the Russian ones are mostly small envelope versions.
|
12th Oct 2019, 8:16 pm | #8 |
Dekatron
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,398
|
Re: Why Both EAA91 and EB91?
Looking the other way, did Mullard/Philips act as early adopters of the "US format" EB91 with its taller, 6AL5-style envelope, whilst Telefunken et al continued with production/usage of the 0.2A heater EB41 in an environment where Rimlock had been popular and established for some years including in VHF/FM receivers? When Rimlock was finally pushed aside as B9a (particularly) became dominant, it might have seemed sensible to standardise on a B7g double diode, the short and recently introduced EAA91 looking favourable over the established EB91.
The postulation that the EB91 was introduced in its taller envelope simply on the grounds that it enabled common envelope processing and handling over a range of B7g types seems reasonable from the production engineering viewpoint, which of course raises the question- what was the point of the EAA91's appearance? |
13th Oct 2019, 12:21 am | #9 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Why Both EAA91 and EB91?
A reverse lookup on RMorg shows generally a higher application count for the EAA91 then the EB91 once the former became available. A cursory inspection suggests that the EAA91 was used by European setmakers, whereas the EB91 was used by UK setmakers.
Here is a postulate, or perhaps a “straw man” as to why things turned out the way they did, at least in terms of what Philips/Mullard did.
If there were any performance differences between the EAA91 and EB91 – and it does seem at least possible that there were - these have not been mentioned in any of the datasheets and literature that I have seen. Cheers, |
13th Oct 2019, 12:49 pm | #10 |
Dekatron
Join Date: Dec 2007
Location: Haarlem, Netherlands
Posts: 4,203
|
Re: Why Both EAA91 and EB91?
Are you sure those aren't the dates that Frank Philipse (a private person unrelated to the company), compiled those sheets?
|
13th Oct 2019, 1:12 pm | #11 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Why Both EAA91 and EB91?
I'd never even been aware of the EAA91 until this thread. EB91s have been breeding in my junk box for almost 60 years, however.
THe EAA91 number seems a bit silly when there's long history of suffix letters in valve designations to cover different bulb shapes. Does the EAA91 do the bright heater flash on turn on in series string heater systems that I associate with the EB91? David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
13th Oct 2019, 3:44 pm | #12 |
Heptode
Join Date: Dec 2014
Location: Lincoln, Lincolnshire, UK.
Posts: 583
|
Re: Why Both EAA91 and EB91?
I have some EB91 and EAA91. I regard them both as equivalent. My EAA91's are branded RFT whereas my EB91's are RST.
I just had a look at some curves I plotted a while ago which show similar characteristics. |
13th Oct 2019, 8:55 pm | #13 | |
Dekatron
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 7,088
|
Re: Why Both EAA91 and EB91?
Quote:
Hopefully nobody has found evidence of an EAAAC80... |
|
13th Oct 2019, 9:32 pm | #14 | |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Why Both EAA91 and EB91?
Quote:
In the usual Philips way, each page has its own data. I have taken the issue (or really reissue) date to be that of the final page. The actual data in the EAA91 case dates from 1954, as does the limited data provided for the EB91. The “refer to the EAA91” statement also dates from 1954 June 06, so a reasonable inference is that by mid-1954, if not earlier, Philips viewed the EAA91 as being the parent type and the EB91 as a derivative. That aligns with the presentation in the 1954 Pocket Book. The EB41 data dates from 1952. Cheers, |
|
13th Oct 2019, 9:54 pm | #15 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Why Both EAA91 and EB91?
Here is the section on the EB91 from Philips Technical Library Book IIIC, “Data and Circuits of Television Receiving Valves”. The English edition was published in 1953, although the book appears to have been based upon the (predominantly noval) range of TV valves introduced in 1950. Anyway, it is apparent that at that time, the EB91 was Philips’ preferred double diode type.
EB91 Philips Book IIIC 1953.pdf The EB41 had been covered in Book IIIA, “Data and Circuits of Radio Receiver and Amplifying Valves; (Second supplement; Valves developed during the period 1949/50”. The English edition was published in 1952. EB41 Philips Book IIIA 1952.pdf As may be seen, the EB41 was also said to be suitable for FM and TV applications. The Philips preferred double diode progression was thus: EB41 (from c.1946) EB91 (from 1950) EAA91 (from 1954, if not earlier). Mullard apparently stayed with the EB91, not moving on to the EAA91. Cheers, |
13th Oct 2019, 10:01 pm | #16 |
Dekatron
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 14,007
|
Re: Why Both EAA91 and EB91?
Thinking further on this: the B8-valves are only ever going to have been produced by a few European companies (so can best be disregarded in the greater scope of things).
Is there any evidence of world-market manufacturers [ITT/Brimar/Sylvania/RCA/Tung-Sol/GE/Hytron/Raytheon being examples] ever producing an EAA91? Or there being military CV- or VT-numbers issued for it? |
14th Oct 2019, 3:03 am | #17 | ||
Dekatron
Join Date: Dec 2007
Location: Haarlem, Netherlands
Posts: 4,203
|
Re: Why Both EAA91 and EB91?
Quote:
However, for the research at hand, that isn't too relevant. I'm just trying to prevent urban legends from taking off. Besides the dates on the pages, the dates and code numbers on the individual drawings might come in useful to establish the history of those valves. The drawing numbers (7R, 7Z etc.) would have been issued in order, just as the page numbers starting with 93. Only possible exception to that might be when a drawing was revised. 7R xxx xx could have been renumbered to 7R xxx xx.1 .2 .3 etc. I'm not sure whether Philips actually used that convention in this case or just issued a new drawing number. From a quick look at the EAA91 datasheet, I'd say the EB91 was prepared for market in mid 1949 while the EAA91 was prepared in late 1954. This matches your findings. Last edited by Maarten; 14th Oct 2019 at 3:21 am. |
||
14th Oct 2019, 3:29 am | #18 | |
Dekatron
Join Date: Dec 2007
Location: Haarlem, Netherlands
Posts: 4,203
|
Re: Why Both EAA91 and EB91?
Quote:
Also, Philips was a world market manufacturer. It would be interesting to find out what double diode, if any, they marketed (and possibly manufactured) in for example their South American national organisations. Even though their valves were shipped as unlabeled bulk goods from and to their own and 3rd party factories all over the world, some local markets would have seen more local production than others, especially if they had their own preferred types. |
|
14th Oct 2019, 3:52 am | #19 |
Nonode
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,944
|
Re: Why Both EAA91 and EB91?
Thanks for the clarification. That kind of third party “mixing and matching” of and additions to datasheets does raise questions about the provenance of the data, although in this case I don’t think that it changes the basic picture.
The attached datasheets are extracted from Philips Electron Tubes Part 4 1972, and so apart from the extraction exercise, have not been tampered with as it were. Philips Electron Tubes Part 4 197206 EAA91.pdf Philips Electron Tubes Part 4 197206 EB91.pdf The EAA91 was still the shorter version, and appears to have still been current in 1972. The EB91 was listed as “maintenance only”, and was referenced back to the EAA91 for detailed data. More insight into the early 1950s chronology may be had from a couple of Philips valve pocket books. The first of these is Philips Electronenbuizen Zakboekje of 1953. It listed miniature double diodes as EB41, EB91 and UB41, but not the EAA91. The EB91 was shown as the preferred type for 6.3-volt and 300 mA heater applications, with the UB41 the preferred type for 100 mA heaters: The second is the 1954 Pocket Book mentioned upthread. This listed miniature double diodes EAA91, EB41, EB91 and UB41 with the EAA91 now shown as the preferred type for 6.3-volt and 300 mA heaters. (By then there was no preferred type shown for 100 mA heaters.) That puts the change from the EB91 to the EAA91 as the primary miniature double diode as having occurred in the 1953-54 time period. It still does not answer the question “why”. But the previous suggestion remains a good fit. That is, Philips decided that it needed a miniature double diode that was an exact performance and physical replica of the American/International 6AL5, and not simply a performance match. The Australian market requirement may well have been the prompt for that, and once done it was logical that the EAA91, identical to the 6AL5, was made the preferred type, displacing the EB91 from its short reign in that role in most of the Philips world, with the UK being an exception. Cheers, |
14th Oct 2019, 5:17 am | #20 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,901
|
Re: Why Both EAA91 and EB91?
The odd nomenclature suggests two possibilities:
1) EAA was chosen before 'B' was defined in the pro-electron world. This is immediately shot down by the earlier EB41. 2) They already had the EB91, and had created a second version with the same electrical characteristics but a different envelope. Now the numerical part of a pro-electron number (you know what I mean!) is supposed to be chronological in order of issue. So the new thing could have been EB92 without straining the nomenclature system. So the part wasn't a 100% form, fit, and functional replacement and therefore needed a different name. So the question becomes why they wanted so much to stick with the '91' digits that they mucked around with the 'B'? We tend to think in terms of the consumer market. The AM radio designs went for combined bottles with detector diodes included with a triode or pentode. The FM radio designs needed a pair of diodes without common cathode connection. So 6AL5 in the states, and its equivalent over here. The valve was at a high signal level so screening wasn't necessary and so either size envelope could be used. There would be nothing to make the larger part not fit. Same for tellies. But those aren't the only uses for double diodes. They turned up in RF, military and industrial stuff where sometimes exact mechanical size was critical... such as inside RF test probes. Also the height was important where location devices were used to keep valves in their sockets under vibration. Designers of military gear would want multiple sources for their parts. Nato would have gone for 6AL5 but would have disqualified EB91 on fit grounds as any sort of equivalent. This sounds the most likely reason to me for Philips doing another version of their double diode. Money is involved. Maybe Philips had had their consumer radio/TV goggles on when they did the EB91 and used a glass envelope size they already had tooled-up for something else (maybe even a something else that never made it to production) but by doing so they blew its chances as a 6AL5 true equivalent. After a sound kicking from their accountants, they set out to right this missed opportunity. This seems plausible but still leaves the mystery of why they felt so attached to the '91' digits. Pride, perhaps? David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |