UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

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.

Reply
 
Thread Tools
Old 20th Oct 2019, 5:46 pm   #61
Pieter H
Tetrode
 
Join Date: Sep 2017
Location: Waalre, Netherlands
Posts: 59
Default Re: EF91 to EF80

Hi all and especially Synchrodyne,

Due to other activities I only saw this discussion when it has almost been finished. Very interesting and much input to digest.
Let me throw in my few pennies of wisdom, purely focussing on the Philips developments, which I've traced back in my Philips TV and Tuner histories.

As described in that story, the introduction of new valves by Philips post-WW2 went in two waves: in 1947 the Rimlock 40-series was introduced as a major miniaturization step based on the all-glass base. All 1948-49 TV sets were based on the Rimlock valves, with the EF42 as main RF-IF-AM amplifier. Interestingly, Mullard continued the use of the EF50 as main amplifier, switching to the Rimlock series and a design from Eindhoven in 1949.

However, already in the 1946 TV sets 463A and 563A Mullard introduced the EB91, as successor of the EB4, which means that the 90-series was the first in which Philips implemented its all-glass base. This would at the same time be the only 90-series valve used in any Philips TV. The only exception is a DAF96 that was used in 1952 in a French (TF2323) and German set (TD2314) in a protection circuit for the MW6-2 rear projection picture tube. [And to be fair and complete: an EF91 could turn up in UK TVs were the original EB91 sync separator was replaced by an EF91 plus EA50 service solution, see e.g. the 383A in RM.org]. So I fully support your conclusion that Philips never promoted the 90series in its consumer application. However, the EB91 apparently hit a sweet spot, because it was used for 10 years, even alongside the first germanium diodes.
The French Philips/Radiola sets from 1956 were the last to use it, as AGC detector and protection diode. Your interpretation that the EF91 was mainly a local Mullard exercise driven by the UK market is probably correct.

The Noval series was introduced in 1948. My interpretation is that this base came from the US (RCA I suppose) and was quickly adopted by Philips to give it access to a much larger market. Most Rimlock types were almost overnight copied into Noval types, with all TV sets from 1950 onwards 100% on Noval valves (except the EB91 of course). And with the EF80 as core pentode amplifier covering the RF, IF, AM and FM functions, replacing the EF/UF41 and 42 as well as the EAF/UAF42.

At the time of introduction the EF80 was essentially the best consumer pentode available within Philips, and was thus also used as first RF amplifier in the first generation tuner modules. At the same time a pentode is obviously not the best solution for an RF pre-amplifier (more noise than a triode) although it survived surprisingly long in that role until replaced by the PCC84 in 1953. The French Philips designers were the first to make the switch to triode RF inputs, using the ECC81, although it is questionable whether that gave an overall better performance, the ECC81 being essentially a low frequency valve.

Throughout the 1950s the EF80 remained the undisputed backbone of all Philips TVs, being used as video and sound IF amplifier. In 1954 it started to be replaced in that last role by the PCF80, while the EF184 replaced it completely as IF amplifier from 1960.

I hope it helps to clarify the story.

Cheers, Pieter
Pieter H is offline   Reply With Quote
Old 22nd Oct 2019, 3:23 am   #62
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,448
Default Re: EF91 to EF80

Thanks Pieter, that does help a lot, and puts the whole issue into perspective. One may see that prior to the arrival of the EF80, Mullard was travelling a different pathway to the rest of Philips, one that involved offering both the EF42 and the EF91 for TV receiver applications rather than the EF42 alone. Once the EF80 was available, its superiority for the TV job made it the natural successor to both. Not only that, but the UK valve industry generally came on board as it were by offering clones of the EF80, something that had not happened with the EF42. You could say that in the EF80 era, Mullard crossed over from the EF91 branch line to the EF80 main line.

ECC81 was the European designation for the American 12AT7, which had been introduced by GE in 1947 as a TV and FM mixer-oscillator, with improved performance as compared with the established 6J6, in particular lower microphony. Most data sheets claim that the 12AT7/ECC81 was good to 300 MHz as an amplifier, with some makers claiming higher than this. In US practice, it was used as an FM mixer-oscillator through the 1950. However, its TV role fell away when triode pentodes such as the 6X8 and 6U8 were introduced late in 1951 as being better suited to the new 45.75 MHz standard IF. (The first cascode valves, 6BQ7 and 6BK7, were introduced at about the same time.)

Philips Book IIIC included the ECC81, and showed various circuits including push-pull RF amplifiers, shunt cascode RF amplifier and mixer-oscillator. Lower noise at Band III frequencies was claimed as compared with the EF80. At least according to RCA’s work, the shunt cascode was less amenable to multichannel applications, which is why it developed the series cascode circuit, which it originally named “driven grounded grid”. On the other hand, the shunt cascode was probably fine for single-channel front ends.

The ECC81 did have something of a split personality though. It was often listed as an AF valve, along with the ECC82 and ECC83. Mullard advertised it both ways:

Click image for larger version

Name:	WW 195310 p.03 ECC Industrial.jpg
Views:	20
Size:	77.3 KB
ID:	192225 Click image for larger version

Name:	WW 195312 p.70 Mullard ECC Audio.jpg
Views:	17
Size:	61.6 KB
ID:	192226

Returning to the EF80 and EF91, this table (from Fisher; VHF TV Tuners) provides some comparative data and I think shows that the EF80 was pretty good for its time. The 6CB6, dating from 1950, was RCA’s best TV pentode of the time, having been developed specially for 40 MHz IF strips.

Click image for larger version

Name:	Fisher p.34.jpg
Views:	14
Size:	69.5 KB
ID:	192224


Cheers,
Synchrodyne is offline   Reply With Quote
Old 22nd Oct 2019, 4:22 am   #63
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,448
Default Re: EF91 to EF80

Both the EF91 and the EF80 had American (RMA, RTMA) registrations and American designations.

The EF91 was registered by Cossor as the 6AM6, with registration date 1949 January 28.

Prior to this, did Cossor have its own designation, perhaps in the VPT series, for this valve?

Once the 6AM6 name was available, Brimar switched to this from its own 8D3.

(Brimar had a mixed naming system. For American-origin valve types, it naturally used American designations. For its own designs, it sometime used its own designation style, such as 8D3, 9D6 and 20D4, and sometimes American-style designations, such as 12AH8, 6BW6 and 6BW7. I think that it might also have used Pro-Electron designations as primary names to some extent in the 1950s.)


The EF80 was registered by Philips as the 6BX6, with registration date 1951 February 15.

It would appear that at the time, Philips was routinely undertaking American registration of its new noval valves. In part, this was a requirement for the Australian market, which generally used American designations, as noted in the “Innoval” thread, at: https://www.vintage-radio.net/forum/...d.php?t=159120.


Cheers,
Synchrodyne is offline   Reply With Quote
Old 24th Oct 2019, 7:52 am   #64
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,448
Default Re: EF91 to EF80

The first page of the attached WW 1954 June article provides some information pertinent to the EF91 and EF80 discussion.

It was argued that as a Band I RF amplifier, valves like the EF80 were better than the earlier generation such as the EF91 because they did not suffer as much performance fall off at the upper end of Band I.

Also, that they were better IF amplifiers because they had sufficient input resistance not to be the sole source of tuned circuit damping.

It would appear that for TV receivers, the case for the EF80 and its ilk over the EF91 & co. was clear-cut, whereas with FM and communications applications it was a closer call and so probably situational.


Cheers,
Attached Thumbnails
Click image for larger version

Name:	WW 195406 p.272.jpg
Views:	22
Size:	72.4 KB
ID:	192369  
Attached Files
File Type: pdf Valves for Bands III, IV and V WW 195406.pdf (1.10 MB, 7 views)
Synchrodyne is offline   Reply With Quote
Old 24th Oct 2019, 10:46 am   #65
Radio Wrangler
Dekatron
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 12,583
Default Re: EF91 to EF80

Regarding the 7pin/9pin issue. I never did see that there was much point in sticking to the same format throughout a set. The valves were hardly going to be swapped around were they?

(I've just done your 6000mile service and rotated all the valves in the radio for even wear, sir.)

There will be a small saving in production over using two different types of holder, but that might be exceeded by the saving on 7 pins.

I suspect it was Philips telling the market what they ought to do.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 24th Oct 2019, 12:51 pm   #66
kalee20
Dekatron
 
Join Date: Feb 2007
Location: Lynton, N. Devon, UK.
Posts: 5,133
Default Re: EF91 to EF80

Quote:
Originally Posted by Radio Wrangler View Post
Regarding the 7pin/9pin issue. I never did see that there was much point in sticking to the same format throughout a set. The valves were hardly going to be swapped around were they?
Good point - in fact there's an argument to say that, within a piece of equipment, the valves should all be on different bases so that misfitting is not possible, by design. Same as using different connectors, or different keying configurations, so that it is not possible to plug the wrong cable into a connector.

However, for me, I confess I'd rather see a nice geometric pattern of similarly sized bottles...
kalee20 is offline   Reply With Quote
Old 24th Oct 2019, 7:48 pm   #67
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 7,739
Default Re: EF91 to EF80

On the EF91 vs EF80 game, yes I can see the benefits that accrued from going to the B9A base (more cathode-connections/better cathode-decoupling) which would become really rather more-important a few years later when we got the high-slope "Frame Grid" EF183/EF184 - hich gave more-than-enough gain at TV-IF frequencies that you could happily cut down the number of valves used in the IF-strip.

In the context of 'consumer vs industrial' range kit, I remember the likes of the 1950s Burndept BE201 transceiver - covering 100-156MHz (AM) which used EF91 in the front-end, and also a number of fixed-tuned military 121.5MHz 'watch' receivers from the same era that used the 6AK5 for the RF-amp and mixer. The EF91s used were generic Mullard ones with the blue coating on the inside of the envelope, whereas the 6AK5s were US-sourced from Tung-Sol or National.
G6Tanuki is offline   Reply With Quote
Old 30th Oct 2019, 7:55 am   #68
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,448
Default Re: EF91 to EF80

The EF80 certainly made full use of the noval (B9A) base. This allowed an RF pentode to have a separate suppressor grid pinout, two cathode pinouts and a separate internal screen pinout.

The Rimlock (B8A) base was more restrictive. The EF42 had one cathode pinout, along with separate suppressor grid and internal screen pinouts. The Mazda 6F13 was the same. The 6F1 was an improved 6F13, with two cathode pinouts. As it retained the Rimlock base, the compromise was to assign the suppressor grid and the internal screen to the same pinout.

Perhaps a useful question here is: did Philips anywhere use the noval base and its associated envelope diameter for valves that could have been accommodated on the B7G base and its associated envelope diameter?

The answer appears to be “maybe”, but if so, it was rarely so. The same question for the Rimlock base would produce a definite “yes” answer, as shown by the EB41/UB41 case.

The valves in the initial noval Radio series and the initial noval TV series either required more than 7 pinouts (e.g. ECH80, EF80, ECL80, ECC81, etc,) or, where they did not require more than 7 pinouts, required the larger envelope diameter associated with the noval base (EL80, PL82, PL83, etc.). The noval envelope diameter was close to that of the Rimlock type, and in the latter case it had been sized to accommodate a 9 W anode dissipation output pentode, which was thought to be beyond the reasonable capability of the B7G envelope diameter. The one valve associated with both series that did not require either more than 7 pinouts or the larger envelope was the double diode, and in this case Philips chose to use the EB91 with B7G base, rather than develop a corresponding noval type.

The FM-AM radio series of the early 1950s, as well as adding necessarily-noval valves such as the ECH81 and EABC80, also included the EC92 VHF triode for FM front end applications, Philips not yet having adopted the ECC85. The EC92 required 6 pinouts (including a separate pinout for its internal screen) and was on the B7G base. On the face of it, this is evidence that Philips was prepared to drop back from the B9A to the B7G where the latter was fit for purpose. Actually, there might have been a simpler explanation. The EC92 was the European version of the American 6AB4, introduced by GE in 1949. as being half of the 1947 12AT7. It was probably easier to clone the 6AB4 than to develop a new valve, possibly noval. Given that the ECC81 was the European version of the 12AT7, then the EC92 was half of an ECC81. As an aside, with the 12AT7/6AB4 (ECC81/EC92) pair, the 12AT7 came first, the opposite of the case with the 12AU7/6C4 (ECC82/EC90), where the 12AU7 came first.

B7G bases were also used for the car radio RF pentodes, EF97 (remote cutoff) and EF98 (sharp cutoff) of the mid-1950s. These appear to have been Philips (or perhaps other European valvemakers) developments, and not clones of American prototypes. In both cases the internal screens shared a pinout with the cathodes. Possibly the B7G base and envelope was chosen here because compactness was a virtue in car radio chassis.

Looking towards industrial applications, one may consider the EC80 and EC81, both of 1949 and both novals. The EC80 was a UHF grounded grid triode amplifier or mixer, for use up to 600 MHz. It had four grid pinouts, so required 8 pinouts in total.

The EC81 was a UHF triode oscillator for use up to 1500 MHz, and constitutes the one “maybe” in this simple analysis. It had one grid pinout, so 5 in total. It probably could have been fitted on the B7G base. Still, the B9A base probably allowed wider spacing between the cathode, grid and anode pins, and so lower interelectrode capacitances. Anode dissipation was 3 W, I think well within B7G envelope capacity, but perhaps the larger B9A envelope conferred smaller temperature variations and so better stability. So, there might have been technical reasons for using the B9A base and envelope. Otherwise, commonality with the EC80 was probably a reason. (Both valves were described in Philips Technical Review 1949 September.)


Cheers,
Synchrodyne is offline   Reply With Quote
Old 30th Oct 2019, 10:47 am   #69
G8HQP Dave
Dekatron
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,858
Default Re: EF91 to EF80

Quote:
As an aside, with the 12AT7/6AB4 (ECC81/EC92) pair, the 12AT7 came first, the opposite of the case with the 12AU7/6C4 (ECC82/EC90), where the 12AU7 came first.
Did you mean to say that the 6C4 came first?
G8HQP Dave is offline   Reply With Quote
Old 31st Oct 2019, 3:22 am   #70
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,448
Default Re: EF91 to EF80

Thanks, Dave. Indeed I did mean to say that the 6C4 came first (in 1942).


Cheers,
Synchrodyne is offline   Reply With Quote
Old 1st Nov 2019, 2:03 pm   #71
Maarten
Nonode
 
Join Date: Dec 2007
Location: Haarlem, Netherlands
Posts: 2,596
Default Re: EF91 to EF80

Philips developed a few B7G triodes for TV tuners (PC97, PC900), though that was only in the second quarter of the 1960's. Before that, they only used noval types.
Maarten is offline   Reply With Quote
Old 2nd Nov 2019, 9:49 pm   #72
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,448
Default Re: EF91 to EF80

Some evidence that the B7G base was preferred for industrial applications in the UK was provided by this item in Wireless World (WW) 1950 February. This announced that the Scientific Instrument Manufacturers Association (SIMA) had issued a list of preferred valve types, on the octal base for “standard” valves and on the B7G base for miniature valves.

Click image for larger version

Name:	WW 195002 p.65 SIMA Preferred Valves.jpg
Views:	4
Size:	49.4 KB
ID:	193000

At the time, the noval base was probably too new in the UK to have been considered, although one might expect that some noval types were added to the list later on. I think it is a safe bet that the EF91 was included in the initial list as the preferred high-slope miniature pentode.

In the mid-1950s, the dominance of the noval base for domestic equipment was noted by Baxandall in his WW 1955 January article on a gramophone and microphone preamplifier, thus: “The equipment as described uses Noval-based miniature valves; but certain other valves may be employed if desired, and the slight changes in circuit values then necessary are indicated below Fig. 1. The Noval type appears to be becoming established as the preferred series in British commercial practice, combining excellent electrical characteristics with conveniently small size and satisfactorily robust construction.”

Baxandall chose the EF80 for the output stage of that preamplifier, for which high-slope was desirable because it also served as an R-C feedback type low-pass and high-pass filter stage. Circuit details were also provided for the alternative use of the EF91 and the SP61. Thus one might say that the EF80 was primary here because by the mid-1950s, it had become the “go to” valve for consumer equipment where a high-slope pentode was required, including in applications where the differences between the EF91 and EF80 were probably not material.

More generally, in UK practice (and probably in European practice too), one of the VHF high-slope pentodes was the choice for MF and HF applications where a sharp cutoff pentode was required. In part this was because the UK/European valve range did not include HF pentodes that were directly comparable to the American 6AU6 and 6BH6. Both of the latter saw some use in UK equipment, particularly where Brimar was the valve supplier, and elsewhere, with the 6AU6 sometimes under its EF94 designation. But generally, one could day that where the American designers used a 6AU6 for HF applications, the UK designers would have used an EF91 or EF80.


Cheers,
Synchrodyne is offline   Reply With Quote
Reply

Thread Tools



All times are GMT +1. The time now is 9:08 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Copyright ©2002 - 2019, Paul Stenning.