Mystery amplifier

[trobbins]

The schematic in post #79 appears to be missing a link from the V2 heater to the first filter cap C2 pos. That link may be direct, or via a CT of the heater winding (although I don't see any spare terminals for that).

Does V1 screen (pin 4) also connect to a TC anode, to act as a triode connection pass regulator?

[Kokotoni Will]

Originally Posted by trobbins

The schematic in post #79 appears to be missing a link from the V2 heater to the first filter cap C2 pos. That link may be direct, or via a CT of the heater winding (although I don't see any spare terminals for that).

Does V1 screen (pin 4) also connect to a TC anode, to act as a triode connection pass regulator?

Yep, right on both counts. Thanks for pointing those out. Here's the corrected schematic.

[trobbins]

Interesting regulated supply schematic indeed.

I can't identify it as a direct copy of anything commercial or magazine related that I have info on.

The bottom diff pair is somewhat normal with filtered reference on one side, and variable setting on the other, although the variable setting includes a regulator to drop the output voltage signal/level. The single sided output is via a cascode configuration that is then passed to a cathode follower stage to control the pass regulator's grid.

[Richardgr]

A useful tool for schematics is LTSpice. It is not necessary to master the power of it to simulate circuit operation, but once you master its way of working for adding components, rotating, linking, setting values, etc, then you have the advantage that all the connection points are 'glued' so it is much easier to turn a physical layout into a much more understandable logical one.

The schematic above is hard to read because the heater circuitry is included. It might be easier to have two drawings; one for the heaters and the other showing the remaining part of the schematic.

[Kokotoni Will]

I've gone over it with a biro, and uploaded a higher resolution photo here:

https://harnessontheleftnape.blogspot.com/2025/10/mystery-power-supply-schematic.html

[trobbins]

The only ps schem with a cascode I have is page 210 of this link https://www.worldradiohistory.com/UK/Radio-Constructor/60s/RC-1961-10.pdf.

Often a sketch of a 'schematic' is fine and can be done fairly quickly, with the sketch showing a 'typical' layout of parts. And then a photo is often easier than a scanner, as used in #85. I'd anticipate that many find a schematic much easier to interpret, and to then use for queries, as well as marking on diagnostic voltages.

Reviewing commercial and magazine articles on valve regulated supplies can be a good way to identify how such supplies were designed, and how some aspects like noise and ripple and regulation were improved on (especially in the high end commercial supplies). Examples include Advance PP1, APT 506, Fluke 407d, Heathkit PS3 and SP2717, HP-710B and 712B, Kepco 400 and 615, Philips PE 1531, Solartron SRS153, Short Wave May 1962, AudioXpress 2008, RTV&H March 1960.

[Kokotoni Will]

Originally Posted by Richardgr

A useful tool for schematics is LTSpice.

I'll download it for my Windows XP laptop later. I'll draw another couple of schematics later, with and without heater connections, as you suggested.

[Kokotoni Will]

Originally Posted by trobbins

The only ps schem with a cascode I have is page 210 of this link https://www.worldradiohistory.com/UK/Radio-Constructor/60s/RC-1961-10.pdf.

Thanks for the links / suggestions. Can you think of a plausible explanation as to why my power supply uses a cascode?

[trobbins]

Many aspects of design need to be considered and worked through. Without a design guide, or a very similar schematic, then reverse engineering may light up the way - but that is harder to work through without basic operating levels, and a good handle on the intended load and what aspects of the dc supply are most important.

A basic schematic and some deduced operating voltage and current levels (ie. from the amplifier output stage) are perhaps an easier starting point.

[Maarten]

Originally Posted by Kokotoni Will

The two small valves circled in yellow are proving hard to identify. They are marked something like R78, R73 or R7B, and have 7 pins. I've attached a photo of one of them.

R7B is most likely the factory/date line of the small print and would point to a valve made in february 1957 by Mullard Mitcham. They were one of the factories that produced mostly 7 pin miniature valves. R7 as a type code isn't plausible as that would point to a GZ32.

Maybe you can try to see if there's another line of small print.

[PJL]

Do we know if the PSU unit was used to power this amplifier? Using a regulated supply for the HT for a PP pentode hi-fi amplifier would seem a bit pointless. It would make sense for all but the output valves and particularly a pre-amplifier or tuner.

I would recommend we focus on getting a circuit for the amplifier unit first. What are the valve numbers of those under the screenng cans? As someone has already mentioned, screens on an EL84 or similar would cause the valve to overheat and thermally runaway.

[PJL]

Found your post #34 where the output valves are identified as CV2179/A2134. This suggests the output will be around 10W/channel and the HT should be around 250V but this can be confirmed by finding the cathode resistor value.

[Diabolical Artificer]

Originally Posted by Kokotoni Will

Thanks for the links / suggestions. Can you think of a plausible explanation as to why my power supply uses a cascode?

More gain & input sensitivity, IE a bit more control in a closed loop, very simply put that is.

Andy.

[Radio Wrangler]

Sometimes, the driver stage in a regulated power supply needs to run at higher anode-cathode voltage than suits common low power valves. Rather than the expense of using something fancy, or wasting a high power high voltage valve on the job it is sometimes done by sharing the voltage across two valves in cascode formation. Of course, this now would mean high voltage from cathode to heater on the upper valve, unless a dedicated, floating heater winding is used for it.

Cascodes are usually used where their better bandwidth than a simple common cathode stage, so the higher voltage trick gets forgotten about. They're a useful circuit configuration to have in your design toolbox and a configuration which also works with bipolars, fets etc.

David

[Marconi_MPT4]

After looking at the regulated PSU load step response, whoever designed this circuit knew exactly what they were doing and reflects the outer system build quality.

My simulation subbed EL509 for 12E1 and 82V zener diodes for 85A2. A constant current load of 100mA was stepped up to 200mA which is a fairly severe test. Voltage output dipped from a nominal 240V by approximately 0.2V sine shaped for about 3ms. No sign of oscillation which is a surprise in itself as many of these circuits tend to ring or oscillate uncontrollably. I can post results if anyone is interested.

Rich

[Radio Wrangler]

Now, that is very impressive. I wonder if that cascode is for bandwidth in the stage which is doing all the heavy lifting, bandwidth-wise.

David

[Marconi_MPT4]

There appears to be two methods of keeping a wider bandwidth in the feedback loop where the cascode stage is coupled via a cathode follower to the 12E1 grid.

As 12E1 is strapped as a triode there will be significant miller capacitance present.

Rich

[Radio Wrangler]

Originally Posted by Marconi_MPT4

As 12E1 is strapped as a triode there will be significant miller capacitance present.

If the 12E1, strapped as a triode is working as an emitter follower, then there won't be any gain to Miller-up the capacitance, so the bandwidth will be quite secent.

If the 12E1 strapped as a triode is part of cascode structure, the cascode structure blocks capacitive feedback from the anode of the top valve to the control grid of the bottom valve which similarly helps.

David

[Marconi_MPT4]

Originally Posted by Radio Wrangler

If the 12E1, strapped as a triode is working as an emitter follower, then there won't be any gain to Miller-up the capacitance, so the bandwidth will be quite secent.

David

Yes, my mistake, thanks for correcting this point. That's me trying to do several things at once.
Rich

[Kokotoni Will]

Originally Posted by PJL

This suggests the output will be around 10W/channel and the HT should be around 250V but this can be confirmed by finding the cathode resistor value.

The CV2179 valves have 470 ohm cathode resistors.

Originally Posted by Marconi_MPT4

I can post results if anyone is interested.

Yes please.