EL84 vs. UL84 and PL84

[Synchrodyne]

Originally Posted by ricard

EHCC81 etc would be a strange designation, it's only the first letter that signifies heater type. EHCC would be a 6.3V heptode with double triode! (think of ECH81).

I would think the main reason for choosing E as the heater designation even though it can be run at both 6.3V and 12.6V and also in a 300mA (or 150mA) heater chain, would be familiarity. Engineers were familiar with the E series as the standard basic series of valves, and from there on also U and P for series connected heaters. Introducing a new valve at the time as HCC81 would probably have caused many engineers to miss it as few would be using H valves in the rest of the design.

The empirical evidence is that where a given valve simultaneously satisfied more than one heater convention including the “E” series 6.3 volts, then that designation took precedence over the others. There may have been a complete hierarchy involving E, H, P, X, U, but I haven’t checked the evidence beyond the E-takes-precedence case.

Thus the Mullard/Philips World Series of TV valves, most of which were intended for use in 300 mA heater chains, included both E-series and P-series types. Where the heater requirement happened to be 6.3 V, 300 mA, then it was an E-series valve, but where the heater voltage was different, it was a P-series valve. For some designs 6.3 V, 300 mA may have been a target, with valve design constrained accordingly. For example the EF80 had a 6.3 V, 300 mA heater whereas the predecessor EF42 was 6.3 V, 330 mA, and I think might have had a slightly higher slope. Also, the E-series valves in the TV World Series were designed for 170 anode volts or thereabouts, not always the case for standalone E-series valves.

Non-trivial differences between nominally similar E and U series valves goes back at least as far as the Rimlock series, in that the EL41 and UL41 were different. And the later PL82 (in the TV World Series) might have been a UL41 derivative.

Cheers,

[ricard]

Originally Posted by ajs_derby

There was never much DC mains about anyway; certainly not enough to justify so many "AC/DC" sets. It's my personal persuasion that the use of series-wired heaters and dropper resistors was more about a low initial purchase price than about compatibility with DC mains. Otherwise the clock radios of the time, needing AC for the clock motor, would have had power transformers.

I think you are right, although the popularity (or lack thereof ) of DC mains might have varied between countries. Also, most (valve) TV's are of the AC/DC design for exactly the reasons you give, and even more so considering the size of the necessary mains transformer, and the sheer number of valves in a TV tends to justify the series heater arrangement more.

My own experience is clouded by two facts: a) that my father in the late 50's actually lived in a neighborhood which had DC mains (although it was by no means the norm even in those days), necessitating the purchase of an AC/DC radio and battery operated tape recorder. And b): didn't Philips offer many of their sets both in an AC-only and AC/DC version? Thus indicating that there must have been sufficient market demand for AC/DC versions of sets, and not just small sets at that. I don't know if there was a price difference, but assuming the AC/DC version of a set were cheaper, it would have seemed rather pointless to offer an identical set in an AC-only version I would have thought.

[Heatercathodeshort]

The filaments did not go O/C with the 'U' series. They appeared to run into grid current and generally become unstable. Some receiver circuits were better than others. Maybe the printed circuit boards used in cheap receivers were to blame maybe becoming slightly conductive or cabinets too small for good ventilation. The yellow label Mullard manufacture were noted for their short life and inter electrode shorts. Why this is so I have never been able to establish. [That goes for all yellow label valves] The EL84 was indeed a grand valve with it's slight purple glow on it's anode. Quite normal! Mullard said so. They were often replaced because 'a new one might sound better' but it never did. Quite normal! Regards, John.

[kalee20]

Originally Posted by tri-comp

I remember replacing a proportionally higher percentage of U-valves than P & E versions.
They just didn't last for long.

Accepting this as fact, is it an issue with the valves themselves, or just a consequence that U valves shared the cabinet with a hot-running dropper resistor, whereas E valves were generally in a cooler environment?

[Leon Crampin]

For a given "family" of output valves delivering similar power to their loads, the parallel heater types "E" will run at a lower anode current than their series fed "U or P" cousins.

This is simply because in the configuration in which they are wired with a half wave rectifier and no HT transformer, they run at a lower HT voltage. In order to deliver a comparable output to the load, the anode current has to be arranged to be proportionately higher.

This will explain why, in general these series fed valves have a shorter life than their parallel fed cousins and also explains their susceptibility to grid current, which is usually caused by gas or by the control grid running too hot and becoming an emitter of electrons itself.

Leon.

[tri-comp]

Originally Posted by Leon Crampin

For a given "family" of output valves delivering similar power to their loads, the parallel heater types "E" will run at a lower anode current than their series fed "U or P" cousins.

This is simply because in the configuration in which they are wired with a half wave rectifier and no HT transformer, they run at a lower HT voltage. In order to deliver a comparable output to the load, the anode current has to be arranged to be proportionately higher.

This will explain why, in general these series fed valves have a shorter life than their parallel fed cousins and also explains their susceptibility to grid current, which is usually caused by gas or by the control grid running too hot and becoming an emitter of electrons itself.

Leon.

This makes very much sense.
It was always a mystery to me why the U-series failed so often but remembering where they were mostly used I think you nailed it.
Thank you.

rgds,

/tri-comp

[tri-comp]

...and then again.

I looked through a few vintage schematics of the kind of radios that used to have the UL's fail.
Like the one attached (Danish Radio made by TO-R) that actually HAS voltages noted on the diagram.
Not a lot of that in the past.
But here you see that the UL84 is run pretty normal in Class-A at around Ik=60mA (Around 11Watt A+G2).
That shouldn't hurt it much with this HT supply.

Maybe how the heaters are stringed has something to do with the problem.
UL84 is put up high as you see.

rgds,

/tri-comp

[WME_bill]

I would endorse "Synchrodyne's" comments about E and U or L numbered valves. As he says, not always the same characteristics He quotes EL41 and UL41 being different. Talking Mullard/Philips.
The same applies for the EL84. I know of no similar valve with different heater rating. The PL84 and UL84 are exactly similar to the low impedance EL86, specifically designed for the lower HT rail voltage and higher current. Perhaps that gives them a longer life under properly designed operating conditions. WME_Bill.

[Synchrodyne]

The following excerpt from Philips Book IIIC (*) speaks to the origins of the P-type valves:

“During recent years the size, cost and weight of television receivers has been considerably reduced. This is to a large extent attributable to the wide¬spread adoption of the transformerless technique, in which the mains trans¬former is dispensed with. The heaters of the valves are then connected in a series chain across the mains supply, the H.T. line voltage being obtained by direct rectification from the mains. For this reason, with the exception of the EQ 80 and the EY 51, the range of valves described here has been designed for series operation of the heaters, the heater current being 0.3 A. The heater wattage of the EY 51 has been kept small (6.3 V, 90 mA), to permit supply from a separate winding on the line-output transformer. The heater current of the EQ 80 valve is 0.2 A at 6.3 V, so that when it is used in a receiver in which the valve heaters are operated in series, the EQ 80 heater must be shunted by a resistor of 63 Ω.

“Some of the valves in the range are suitable for either series or parallel operation of the heaters. With the EF 80 and the ECL 80, for example, the heater voltage is 6.3 V at 0.3 A. Of most of the P-types, such as PL 81 and PL 83, having a heater voltage greater than 6.3 V, equivalent E-types are also available. The type numbers are then EL 81, EL 83 etc.; the heater voltage is 6.3 V, but the heater current is, of course, greater than 0.3 A. The adoption of a heater current of 0.3 A is a deviation from the conven¬tional practice with radio valves, where, for series operation, a heater cur-rent of 0.1 A is used. There are, however, good reasons for this. In the first place, in a television receiver one heater chain would be impossible with 0.1 A heaters, so that several chains would be needed in parallel, each with its own series resistor, and this, of course, would be a more expensive arrange¬ment. With valves having a heater current of 0.3 A, in the majority of cases one single heater chain suffices and consequently this is the cheapest solution. Secondly, the adoption of a heater current of 0.3 A has made it possible to design some of the valves in this range for a heater voltage of 6.3 V, so that these types are suitable for either series and parallel operation.”

Thus it does seem that first cost, not the ability to operate on DC mains was the primary driver for using the transformerless technique in European (and other) TV receivers.

The closing comment of the excerpt also confirms that some valves were deliberately designed with heater system duality in mind, that is suitable for both 6.3 V parallel systems and 300 mA series systems. Given that such valves are identifiably E-types, that in turn confirms the E-designations took precedence over P-designations.

This thread also crosslinks:

https://www.vintage-radio.net/forum/...ad.php?t=82981

Cheers,


(*) Data and Circuits of Television Receiving Valves
J. Jager
Series of Books on Electronic Valves, Book IIIC, 1953
Philips Technical Library

[ricard]

Originally Posted by Synchrodyne

T
“... Of most of the P-types, such as PL 81 and PL 83, having a heater voltage greater than 6.3 V, equivalent E-types are also available. The type numbers are then EL 81, EL 83 etc...."

Thus it does seem that first cost, not the ability to operate on DC mains was the primary driver for using the transformerless technique in European (and other) TV receivers.

Yes, definitely with TV receivers, and probably largely so with radios. But why did Philips in several cases offer an equivalent AC of a corresponding AC/DC radios?

One possible reason that comes to mind is the ability to operate consistently with different mains voltages. In those days, 220/240V mains was not the norm all over Europe; equipment with mains transformers often had 6 selectable voltages, like 100/110/127/150/200/220/245 volts, which would not have been possible with a transformerless design.

For larger radios, the multitude of inputs and outputs and consequent need for isolation transformers I suppose would have negated the cost savings obtained with the omission of the mains transformer.

The closing comment of the excerpt also confirms that some valves were deliberately designed with heater system duality in mind, that is suitable for both 6.3 V parallel systems and 300 mA series systems. Given that such valves are identifiably E-types, that in turn confirms the E-designations took precedence over P-designations.

It also implies that the original philosophy was to have an equivalence between for instance the E- and P-series, so that it really was just the heater voltage that was different.

[Heatercathodeshort]

50CG6/6CD6. 6BW6/9BW6 6U4G/25U4G are just a few valves in the Brimar range that come to mind that appear to be identical other than heater ratings.
Must admit I always took it for granted that the difference was only the heater spec in the Mullard valves. I wonder why we never employed [?] the EL86? John.

[Synchrodyne]

Originally Posted by ricard

It also implies that the original philosophy was to have an equivalence between for instance the E- and P-series, so that it really was just the heater voltage that was different.

It certainly looks that way. That is, the P-type design constraints (such as 170 anode volts) were applied to the TV valves, and where E-type/P-type duality could not be achieved, then the separate E-type simply replicated the P-type except for heater parameters. And this included the early “L” and “CL” valves in the P-series, unlike the case with the U-types, where the compromise involved with say the UL41 was not visited upon the EL41.

As far as I know, the EL82 arrived somewhat later than the EL81 and EL83. But given that the EL41 was available, there would have been no pressing need for an EL82 until Philips/Mullard decided that there had to be noval successors to all of the Rimlock valves, and this was a few years after the TV world series was announced. The PL82 was I think derived from the UL41 via the UL46.

Thus the PL84 might have broken ranks amongst the P-types in that it (necessarily) followed the UL84 not the EL84. But the EL84/UL84 pair, being primarily radio and audio valves, had logically mirrored the EL41/UL41, with no U-type compromise being forced upon the EL84. I haven’t looked at the later EL/PL and ECL/PCL pairs to see if the original premise held, or whether some of them followed the PL84 model.

The EL86 might be explained simply by the fact that Philips did not anticipate the need for such a valve when the PL84/UL84 was designed, so had to backfit it into the system. Maybe the EL86/PL84 is the only example where two valves that differ only in heater parameters have different sequence numbers?

Originally Posted by ricard

For larger radios, the multitude of inputs and outputs and consequent need for isolation transformers I suppose would have negated the cost savings obtained with the omission of the mains transformer.

Logical, but in the other thread mentioned above there are some examples of reasonably elaborate radios, transformerless with P- or U-type valves. Perhaps in these cases there was a genuine intent to cover the possibility of DC mains, 220 V or thereabouts at least. I am not sure how much use there was much use of audio isolating transformers for AF inputs and outputs on post-WWII “universal” radio and radiogram units. But good idea though it was, I don’t think that it was common in British practice. I know that some B&O TV receivers used audio isolating transformers to provide line level outputs for feeding tape recorders and the like.

Cheers,