Mystery amplifier

[PJL]

Update circuit diagram.

[Diabolical Artificer]

Upon powering up the PSU, what would normally happen if this valve had failed?

Depends but if you power the amp on with it upside down so you can see the resistors etc whilst monitoring HT, using your senses, you'll be fine for a quick DC conditions check. Valves are hardy beasts & don't tend to fail SC like transistors.

The X could mean it's really good or something else though more likely to be duff, Andy.

[PJL]

Next steps are:

1. Repeat the resistance checks for the other channel.
2. Perform a similar resistance check on the power supply unit.
3. Order and fit new parts.
4. Get the power supply running. You don't need a Variac as you have paper in oil capacitors which, unlike electrolytics that can explode when you first apply power, do not need to be 'reformed'.
5. Make up connecting lead.
6. Use the power supply to reform the capacitors on the power amplifier unit (just needs series resistance on the HT line).
7. Use the 6.3V supply to work out the turns ratio of the output transformer. Strap secondaries to suit your speakers.
8. Test amplifiers individually.
9. Do you have a CD player?

[Richardgr]

I think a variac buys a lot of security, even if it is just a comfort blanket. I like to slowly power up to about 20%, no valves present, and just verify the different secondaries are performing as expected. Then replace the rectifier, power up to about 50%, when the rectifier will start conducting, and check around the OPTs and other places B+ and other power rails should be present.
Then all valves replaced, and slowly up to max volts, listening for anything odd, smelling for anything getting too hot, hopefully no plume of smoke (although I have experienced a nuclear bomb mushroom cloud in miniature at this point).
Since you have the regulator circuit you have to be extra careful because the screen supplies are critical to the health of the output valves, if that is the need for that supply. Maybe that need to be checked in isolation at around 75%.
Do you have some sacrificial speakers to ensure the output stage is loaded when you start up?

My variac was £50 second hand, and when I am ready to part with it it will be worth £50.

Another alternative is a dim bulb tester, and have a selection of bulbs so the proportion of power to the bulb and amp can be changed, but that solution feels a bit 'binary' when you have those irreplaceable output transformers to consider.

[Diabolical Artificer]

I'd second using a variac or at least a LL after making changes. We are all human & make mistakes, powering up slowly & at about 200v AC mains allows the valves to work & any shorts etc to become apparent without bangs.

A dummy load or scrap speaker as load again just in case, bobs your Uncle. We all have different ways of working though.

Andy.

[PJL]

Post 119 already shows the PSU has been powered up using a lamp limiter. A lamp limiter will limit the current but a Variac just changes the input voltage and in this case the regulator will compensate for this. I would always reform the electrolytics by applying HT voltage through a high value resistor with no valves in before putting back the valves. I would then apply full power whilst monitoring the HT voltage but in this case it is a regulated supply so we need to monitor the regulator to get an idea of the current drawer..

[Kokotoni Will]

PJL, thanks for compiling that list / amending the tag strip - that's very helpful. Would you suggest replacing cathode resistors R17 and R22, both 470 ohms, with 330 ohm resistors, to suit the valves' specification?

There are two paper-in-oil caps in the PSU: C4 and C5. Should these be replaced?

After completely removing the tag strip, it appears that the top of R15 is not connected. I have corrected another error in the attached photo - the bottom of R15 does not loop back on itself. Would this circuit work? If not, what alterations are needed?

I have powered up the PSU, using the lamp limiter, with all valves in, and taken the following measurements. The lamp remained at a constant, dimmish level.

V1 (CV378):
2 (H): 296V
8 (H, K): 296V

V2 (12-E1):
2 (H): 1V
4 (G2): 298V
5 (G1): 134V
7 (H): 1V
8 (K): 139V
Cap (A): 296V

V3 (ECC 83):
1 (A2): 57V
2 (G2): 22-33V (adjusted using variable resistor VR1)
3/8 (K2/K1): 33V
4/5 (H2/H1): 0V
6 (A1): 134V
7 (G1): 31V
9 (H centre tap): 0v

V4 (ECC 83):
1/7 (A2/G1): 22V
2 (G2): 51V
3 (K2): 57V
4/5 (H2/H1): 0V
6 (A1): 297V
8 (K1): 136V
9 (H centre tap): 0V

V5 (85A2):
1/5 (A/A): 136V
2/4/7 (K/K/K): 48V

V6 (85A2):
1/5 (A/A): 81V
2/4/7 (K/K/K): 0V

I do have a CD player. I bought a 15ohm speaker at the same auction house last week for only £6 (see attached), would that make a fitting sacrifice? I also have some 56ohm 10 watt resistors.

The speaker in the photos is a curious thing - it is fitted in an enclosure, with three transformers. Unknown to me at the time of purchase, it also came with the exact 5A Bulgin mains plug I needed - typical!

[PJL]

Originally Posted by Kokotoni Will

PJL, thanks for compiling that list / amending the tag strip - that's very helpful. Would you suggest replacing cathode resistors R17 and R22, both 470 ohms, with 330 ohm resistors, to suit the valves' specification?

There are two paper-in-oil caps in the PSU: C4 and C5. Should these be replaced?

After completely removing the tag strip, it appears that the top of R15 is not connected. I have corrected another error in the attached photo - the bottom of R15 does not loop back on itself. Would this circuit work? If not, what alterations are needed?

Don't change R17 and R22, the output valves will run a little cooler as they are.

Don't change the paper in oil capacitors as these are regarded as the best!

That speaker is not going to be h-fi but should do the job!

Where does the grid of V3 (CV2179) go? R15 and R20 are known as grid stoppers and in conjunction with the miller capacitance of the valve grid they ensure very high frequencies do not get through.

I will check you voltages for the PSU but first observation is the output is only 139V.

That speaker should do the job but won't be hi-fi! We will work out the taps on the ourput transformer later but you probably won't need to buy 15 ohm speakers, what speakers do you have?

[PJL]

V3 g2 is shown as variable but in your diagram that goes to V4 g2 so I assume you have just marked them wrong in your list?

It all looks reasonable but is stabilising at too low a voltage. This is because the reference voltage V3 g1 is too low as this should be the same as the reference voltage on V5 pin 1 (bar the effect of the DMM load). This is almost certainly because C4 (1uF) is leaking a few uA and its fed via a 1M resistor. Firstly check the 1M resistor (out of circuit) and you will need to replace C4 ideally with a non-electrolytic capacitor. This RC filter will be designed to provide a ramp up start.

PS: I have assumed you are using a modern digital meter with 10M input impedance? You also seem to have V5 and V6 swapped.

[PJL]

Last post today... Looking at the picture I am not sure you have the resistor values correct.
red-yellow-yellow-brown-pink=200K 1% HS but I don't see that value in your circuit drawing?
Same applies to orange-orange-yellow-brown-pink=330K 1% HS.

[Kokotoni Will]

Originally Posted by PJL

Where does the grid of V3 (CV2179) go?

The grid of V3 goes to R16 and C4. One end of R15 is not connected. I've attached a clearer photo.

Originally Posted by PJL

what speakers do you have?

Ekco 5768, 4 ohms:
http://tvhf.blogspot.com/2014/01/ekco-5768-3-way-speakers.html

Originally Posted by PJL

V3 g2 is shown as variable but in your diagram that goes to V4 g2 so I assume you have just marked them wrong in your list?

Yes, I did, sorry. Ditto for V5 and V6. I am using a modern digital meter, DM830B, no brand (see photo).

Will this do as a replacement for PSU C4:
https://www.ebay.co.uk/itm/121904596671

I'll check the PSU resistors out of circuit later today.

[PJL]

I suspect this is an original wiring error, V3 grid should go to R15 as per my circuit drawing.

You need to compile a list of parts before ordering anything, the one you have shown looks like a motor run capacitor and it only needs to be >100V.

PS: Do you have any capacitors? For testing purposes a smaller value would do.

[PJL]

https://www.ebay.co.uk/itm/333198908530? is what I would expect but if you create a list and order from somewhere like cricklewood you can save on postage.

[Kokotoni Will]

I've verified all PSU resistors against their colour codes. I've corrected R4 (330k) and R6 (240k), as per attached photo.

R3 should be 2.2M, but measures 2.7M. R11 should be 16k, but measures 18k. All other resistors are well within tolerance, including R7 (1M).

Cricklewood don't seem to have that capacitor available. A shame, as they have almost everything else I need. Would this one work:

https://cricklewoodelectronics.com/CPH1U0.html

Originally Posted by PJL

PS: Do you have any capacitors? For testing purposes a smaller value would do.

Yes, I have some 1uf 450v electrolytics, among others.

[Kokotoni Will]

Here are the resistor values for the right channel, with measured values in brackets:

R23: 1M (1M)
R24: 68K (80K)
R25: 1K (1K)
R26: 470 (458)
R27: 910K (1.07M)
R29: 4.7K (5.9K)
R30: 10K (9.74K)
R31: 10K (9.96K)
R32: 470 (470)
R33: 910K (1.13M)
R34: 4.7K (5.28K)
R35: 33K (39K)
R36: 1M (1.14M)
R37: 39K (44K)
R38: 100K (100K)
R39: 100K (100K)
R40: 150K (185K)
R41: 10K (12.9K)
R42: 120K (120K)
R43: 470K (478K)
R44: 1.8K (1.93K)
R45: 100 (110)

Edit: The updated PSU schematic photo is hard to read at low resolution, so I've updated it here instead:

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

[Kokotoni Will]

For easier comparison, here are the resistor values for both channels, with measured values in brackets.

R1: 1M (L: 1M) (R: 1M)
R2: 68k (L: 80k) (R: 80K)
R3: 1K (L: 1.21K) (R: 1K)
R4: 100 (L: 112) (R: 110)
R5: 1.8K (L: 2K) (R: 1.93K)
R6: 470k (L: 472k) (R: 478K)
R7: 120k (L: 122k) (R: 120K)
R8: 10k (L: 12k) (R: 12.9K)
R9: 150k (L: 175k) (R: 185K)
R10: 100k (L: 101k) (R: 100K)
R11: 100k (L: 99.8k) (R: 100K)
R12: 39k (L: 50k) (R: 44K)
R13: 1M (L: 1.28M) (R: 1.14M)
R14: 33K (L: 40K) (R: 39K)
R15: 4.7K (L: 5.6K) (R: 5.28K)
R16: 910K (L: 1.1M) (R: 1.13M)
R17: 470 (L: 465) (R: 470)
R18: 10K (L: 9.9K) (R: 9.96K)
R19: 10K (L: 9.6K) (R: 9.74K)
R20: 4.7K (L: 5.6K) (R: 5.9K)
R21: 910K (L: 1.23M) (R: 1.07M)
R22: 470 (L:473) (R: 458)

[PJL]

Capacitors to be replaced due to impact of minor leakage are:
Amplifier
C2 0.05u
C3 0.1u
C4 0.1u
Also the same for the right channel

PSU
C4 1u
C5 0.25u

Resistor list is updated in attached.
We are making good progress but you will have plenty of soldering to do!

PS: A modern electrolytic should not leak much so try replacing C4 in the PSU.

[Kokotoni Will]

Thanks again.

Amp C1 is another Visconol Metalpack, like C2-C4. Should this be replaced also?

PSU C3 is electrolytic, should this be reformed before applying power without lamp limiter?

[Kokotoni Will]

I've changed C4 for a new 1uf 450v electrolytic, and powered up the PSU again. For ease of comparison, I've included the old results from before the cap replacement, and typed the new results in brackets.

V1 (12-E1):
2 (H): 1V (0V)
4 (G2): 298V (302V)
5 (G1): 134V (129V)
7 (H): 1V (0V)
8 (K): 139V (127V)
Cap (A): 296V (302V)

V2 (CV378):
2 (H): 296V (303V)
8 (H, K): 296V (303V)

V3 (ECC 83):
1/7 (A2/G1): 22V (21V)
2 (G2): 51V (47V)
3 (K2): 57V (64V)
4/5 (H2/H1): 0V (0V)
6 (A1): 297V (302V)
8 (K1): 136V (126V)
9 (H centre tap): 0V (0V)

V4 (ECC 83):
1 (A2): 57V (64V)
2 (G2): 22-33V (19V)
3/8 (K2/K1): 33V (28V)
4/5 (H2/H1): 0V (0V)
6 (A1): 134V (126V)
7 (G1): 31V (28V)
9 (H centre tap): 0V (0V)

V5 (85A2):
1/5 (A/A): 81V (82V)
2/4/7 (K/K/K): 0V (0V)

V6 (85A2):
1/5 (A/A): 136V (125V)
2/4/7 (K/K/K): 48V (38V)

[PJL]

Yes, I missed that one, C1 0.1u definitely needs replacement. It is fed from a 1M resistor so even a small leakage will upset the balance of the long-tail pair phase splitter.

PSU C3 has already been partially reformed through your lamp limited test. It only got to 139V but should now be OK.