It could have been. We have only limited information about Philips’ intentions, so it may well have explored other possibilities that never came to commercial fruition, including novalization of the EF91.
What we do know or can reasonably deduce about Philips’ novalization programme is as follows:
By 1949, and perhaps during 1948, Philips had decided that it would adopt the noval base for mains-type domestic receiving valves. In detail:
1. Future radio and audio mains valve issues would be on the noval base, except for large output valves and large rectifiers. The first radio novals that were new types, and not reworks, (EBF80 and UBF80) were issued in 1949.
2. The planned standard range of TV valves would be on the noval base (with one or two exceptions).
3. For the Australian market – and perhaps only for this market – a small number of existing Rimlock valves would be reissued as novals under the Innoval name. The Rimlocks had not been released in Australia, so were effectively bypassed there. The Australian release also included new noval issues, such as the EBF80.
4. In the European case, the existing radio and audio Rimlocks would stay in place until either replaced by improved novals (e.g. ECH42 by ECH81, EF41 by EF89, EL41 by EL84), or in a minority of cases, noval versions of existing Rimlocks (e.g. EF40 by EF86, EBC41 by EBC81). In some of the latter cases this involved novals that had already been released in Australia, but had been held back in Europe. This process was completed by c.1955. Late changes were UL41 replaced by EL84, and EL42 replaced by EL85 (which was not quite the same as its predecessor).
5. The new TV valve range would replace the handful of “interim” TV Rimlocks, such as EF42 by EF80, UL44 by PL81, UL46 by PL82 and PL83.
6. The B7G base would be retained for battery portable valves, including new issues.
7. The Rimlock battery home radio series, DK40, etc., was not replaced by a similar noval series. Presumably the B7G battery series would in future be used for this application.
8. The B7G base would be retained for industrial valves.
If the EF91 was viewed as being primarily an industrial valve, then #8 suggests that Philips would not have had a strong reason for developing a noval version. For domestic receiving applications, it appears to have seen the EF80 as being equal to or better than the EF91 (and EF42) along most applied performance vectors, and good enough along those where it was not quite equal. At least that seems to have been the main thrust of this concurrent thread:
https://www.vintage-radio.net/forum/...d.php?t=157399.
Nonetheless, given that development work often involves at least a preliminary look at ideas outside the scope of the initial commercial brief, someone in Philips might have looked at the EF91 as a candidate for novalization – e.g. with the same basic characteristics but with a second cathode pinout and an internal screen.
That notion that Philips had a distinct B7G industrial range is at least supported by indirect evidence. The French Cibot-Radio Miniwatt catalogue of 1956 February, available at:
https://frank.pocnet.net/other/Philips/doc-cibot.pdf, listed several categories and subcategories of valves. These included the following:
Série Miniature – 9 broches (Noval)
Série Miniature – 7 broches pour postes “batterie”
Série Miniature – 7 broches pour postes “secteur”
Tubes Série “Rimlock”
As far as I know, “secteur” referred to the industrial sector.
Also I have on-hand a Mullard publication entitled “Valves for Industry and Communications”, Part One, second edition, 1949. This did include the domestic battery miniatures and Rimlocks, but in addition had several 6.3-volt B7G-based valves that did not usually appear in the domestic receiving lists, such as Philips Book IIIA. Included were the EF91, EF92, EAC91, EC91, ECC91, EB91, EL91 and EY91. Most of these were industry standards rather than Philips originals.
Whilst there was a specific industrial B7G series, where noval valves were required for industrial applications, these were simply drawn from the main noval series.
Re #2 above, it was noted that one or two of the TV valves were not on the noval base. One of those was the EB91 double diode, already in the industrial range. But for domestic applications, Philips had also introduced the Rimlock EB41 and UB41 double diodes. Presumably this was done in order to have a complete set of Rimlocks, no exceptions, even though the EB41 was quite similar to the EB91. For the TV case, apparently sanity prevailed, and rather than develop a noval double diode (say an “EB81”), Philips simply plucked the EB91 from the industrial range and made it part of the TV range. At about the same time the EB91 displaced the EB41 in the radio range, so that was an unusual case where a Rimlock was displaced by a B7G type rather than by a noval. (The UB41 was also eventually replaced by a B7G valve – UB91 or UAA91 – I am not sure which.)
Subsequently, there were a few other B7G interlopers added to the mostly noval radio/TV valve range. These included the EF97 remote cutoff and EF98 sharp cutoff car radio 12-volt HT pentodes. Possibly the smaller size of the B7G valves was seen as an advantage for what were size-constrained units. The frame-grid triodes, such as the EC95, developed for use as VHF TV RF amplifiers also had the B7G base. Perhaps this was advantageous to performance. Then there was the EH90 heptode, used as an FM sound locked oscillator demodulator. This was simply the American 6CS6 gated amplifier with a European number, so it could be that Philips decided to retain the original design on its B7G base rather than rework it as a noval.
Cheers,