The PL81 TV line output valve.

[kalee20]

I don't know of any other amplifier circuits that do this!

(I have an idea that Philips made a portable radio that used the negative bias from the local oscillator to bias the output stage, thus saving a bit of battery voltage... in this case they must have taken pains to keep the oscillator amplitude constant, when tuning across the band, and between bands).

[Jez1234]

They look like they could be useful as series pass valve in a small regulated supply.

[kalee20]

Possibly - they've got good current capability, but aren't so hot at dissipating lots of watts (well, actually they are, and that's the problem). So best suited as the series pass element in a fixed-voltage regulated supply, rather than a variable one.

[Leon Crampin]

Tektronix designed their valve stabilised power supplies around a low impedance double triode 6AS7 in many cases, with the two sections connected in parallel. This served the requirements of high dissipation and low impedance.

Pentodes and beam tetrodes are more common as high dissipation devices, and universal sets used low impedance output devices to get a reasonable anode current at low HT voltages in order to achive a useful audio power output. The same applied of course to TV sets.

Solartron, in their early 'scopes and power supplies needed low impedance small pentodes as series pass elements, so a "TV" pentode with a 6.3V heater would suit. Enter the A2134 from MOV which is nothing more than a humble N37 with a 6.3V heater. From the Philips stable, we have the EL86 which is a similarly low impedance device, rather similar to the UL84. These devices would all be triode connected in this application.

Strangely, I have not seen many applications of the P/EL81 in linear power supply design - but there must be some.

Leon.

[G6Tanuki]

The PL81 was used as the PA valve in the 1950s KW One Sixty topband transmitter, presumably their ubiquitous availability as a telly valve made it worth Rowley Shears and his designers getting a power transformer specially wound to provide the 21V heater supply.

Back then there were not many small power valves that had the anode as the top cap, which was always a desirable feature for RF amps so you could separate the low level grid side from the high level output side and so avoid the need for neutralising the thing.

[ms660]

Originally Posted by Jez1234

They look like they could be useful as series pass valve in a small regulated supply.

The Marconi TF144H used the 6CJ6 equivalent of the EL81.

Lawrence.

[Robert Gribnau]

Originally Posted by kalee20

It's in here, Richard, and it is No 8, as encyclopaedic Steve Synchrodyne has correctly said!

Curious circuit, using the ECL80 triode circuit as an oscillator to develop the -30V (approx) grid bias for the PL81's without eating away at the total HT voltage as would be the case with cathode bias.

And the 'L section of the ECL80 is the phase inverter, operating at sub-milliamp levels... I'm left thinking why not use a PCF80?

This amplifier was published by Philips in this book from 1952 but with an EF40 instead of the electrically identical EF86 (the EF86 was not introduced yet). See page 69 and further:

https://nvhrbiblio.nl/biblio/boek/el...izen_in_lf.pdf

The PCF80 was introduced in 1953 (and the ECF80 in 1954).

[kalee20]

Ah! That makes sense then - basically an older design, rehashed a bit, but not to the extent of subbing the ECL80.

Thanks Robert!

[Maarten]

Originally Posted by Leon Crampin

Strangely, I have not seen many applications of the P/EL81 in linear power supply design - but there must be some.

I'm pretty sure Philips used them in some of their bench power supplies.

[Ed_Dinning]

Hi Gents, I had a neat little Pye build in unit 250v (settable) and 6.3 V this used the PL81.
Obviously sold on to end users in industry

Ed

[Jez1234]

I built a bench HT and LT regulated PSU about 6 months back and tried to build it exclusively with what's in the junk box for all major components and based on having some good S/H ones to hand used a triode connected 6550 as the pass valve. Seems to work well enough but may get in a 6080.

[trobbins]

Philips sure did some interesting circuitry - the last 100W amp cct in the 1952 book linked in post #47 used one triode in an ECC40 to half-wave rectify a 50V winding for the EL34 fixed bias.

[Richardgr]

The problem with tubes in regulated power supplies boils down to the heater cathode insulation. Having done a bit of research, there were not many options for a useful 400v supply apart from an EL34. I have an old tube PSU and it would be great if I could shoe in something cheap and plentiful like an EL81.

[Ed_Dinning]

Try using a separate heater transformer for the series pass tube(s) with an isolated winding
Heater and cathode should then be linked with about 1meg.

Ed

[Jez1234]

I used a separate winding for the pass valve and then linked heater and cathode with 1M just as Ed says above in my DIY bench psu.

[G6Tanuki]

Yes, a separate heater winding [replace that EZ81 in the original design by a couple of silicon diodes and use the heater-winding so freed up?] is what I did in times past when I wanted to do 'series gate' modulation of a QQV06/40A using an ECL82 as the screen-grid control-valve [the cathode thereof potentially going up to 350V on modulation peaks].

Given the relative cheapness and availability of the likes of the PL81, coupled with their robustness and high-HT-tolerance I am surprised that more people haven't caught on to using them in applications apart from their prime TV applications.

A PL/EL81 makes a good 20-Watt-PEP HF linear amplifier; a friend used one to boost the output of a Plesey SL600-series-IC transmitter back in the 70s (I used a 250-0-250V receiver-type mains transformer, bridge rectified.... yes a PL81 can happily handle 700V on the anode).

[Jez1234]

I used an ECL82 in my old regulated bench supply actually. Triode as error amp and pentode as pass valve with an 85A2 as the reference. This lash up did me well for many years! I reused the transformer for something else that seemed important at the time and when I decided to reintroduce it to the rest of the PSU I couldn't find the rest of it! I actually found it about a month ago. In the mean time I decided I wanted a really good bench PSU for valve stuff and it had to cost nowt so I designed and built the latest one completely around things in the junk box. Elstone transformer, Svetlana 6550 pass valve, ECC81 error amp (both sections in parallel then cascoded with a mosfet under it for really high gain) and ahem.. quite a bit of solid state stuff including the voltage reference. It can be set in 1V increments between 50 and 350V in 3 ranges, drifts only around 2mV in 3 hours and the 6.3V heater output is also fully regulated. The heater supply ended up being the one that had me pulling my hair out as it needed the lowest forward drop Schottky's known to man, stupidly large smoothing capacitor and the lowest drop out regulator I could design and build. got there in the end and it will do 3A 6.3V but man was it a sod to get it stable at all output current levels!

[Ed_Dinning]

Hi Jez, how about a copy of the circuit?

Ed

[Jez1234]

Hmm.. If a few people are genuinely interested in building one I'll publish the design here I don't want to go to the trouble of neatening up the circuit diagram, converting it to a format like PDF etc etc just so numerous people can give it "ooh I'd have made R34 a larger value if I were you"

BTW having just checked my notes it is 100V - 350V not 50V - 350V as I said above. It could easily be modified to go down to 50V and another range added at either end depending on the mains transformer available, pass valve/s used and user requirements. The way I did it the maths and parts available meant ranges of exactly 100V were apt and with useful 25V overlap so the easiest thing (with suitable mains TX) would be to add a range and make ranges 50 - 150V, 125 - 225V, 200 - 300V and 275 - 375V. One could even add a 350 - 450V range. Usual trades off's and warnings about pass valve dissipation and desirability of switched transformer secondaries etc etc as available output range becomes more extreme. I find the range of 100 - 350V covers the vast majority of applications I have for it. YMMV.

It's reasonably complicated as far as valve bench supplies go and not particularly cheap to replicate. The output voltage adjustment is actually continuously variable and uses a 10 turn precision 50K pot with a turns counter vernier drive (hence readout from this is in 1V steps) so there's a probable £30 - 40 just for those parts....

Valves are used "at the business end" partly as it seemed fitting it should have some nice glowy glass things on top but mainly for their electrical robustness against short circuits, inductive spikes etc etc. However, it also uses 3 TL431's, 5 mosfets (4 of them small TO92 types) and lots of Zener diodes so traditionalists may not approve!

I built with an eye on it looking traditional on the outside though, hence mines on an aluminium chassis with valves, smoothing caps, mains transformer etc bolted to it (and people could replicate that concept as they saw fit), but it having decidedly modern performance in terms of output impedance, stability and reset-ability of output voltage, low hum and noise on both HT and LT outputs etc. It has a ground lift switch for the LT output. You can eg select the 100 - 200V range then dial in 5 turns of the pot on the vernier and the output will be 150V +/- around 10mV. Add 5 on the vernier window and it will be 155V to the same accuracy

[Jez1234]

Here's a few pics of it. Remember that it was built entirely with what was to hand and cost me nowt so yes it could have been laid out better, looked neater etc...if I wasn't reusing a chassis and trying to reuse all the holes already in it etc! The extra unused B7G base both covers a hole and at the time when I started mounting valve sockets etc I still hadn't made up my mind whether or not to use an 85A2 as the reference. The floral background only adds to it I think I hadn't noticed the "Do you want to compress and reprocess photo to fit bandwidth?" (or something like that) check box before so I ticked it... and got flowers! There's a mini Bulgin socket (keeping in olde world style, along with the neon indicator light) and another fuse around the back. The switch is a ground lift for the LT output. There is no mains on - off switch as I considered having to plug and unplug it at the mains a worthwhile safety precaution with so much touchable HV on an open chassis.