Mystery amplifier

[PJL]

I am not expecting it to work till we have a more powerful lamp limiter bulb (or two!). It is probably the heater supply that is too low with the 40W bulb and that will stop the valves working properly.

I think the builders idea was to make the HT GD connector on the amp the 'star' earth. When you wire up the umbilical you need to take both the 'HT GD' and 'GD' from the PSU Jones socket and wire them both to the 'HT GD' on the amp.

You are probably busy working on your soldering...

[Kokotoni Will]

Originally Posted by PJL

I think the builders idea was to make the HT GD connector on the amp the 'star' earth. When you wire up the umbilical you need to take both the 'HT GD' and 'GD' from the PSU Jones socket and wire them both to the 'HT GD' on the amp.

Thanks for clearing that up.

I didn't mention it before, but when running the PSU from the mains (no lamp limiter) with the 10k load, after a couple of minutes I could smell something either burning or starting to get hot. It wasn't the typical faint "valves warming up" smell which I got when running the PSU with no load, a smell I know from playing guitar through a valve amp. Of course, noticing this, I promptly switched the PSU off. No smoke or any other signs of any issues, and from careful inspection of the PSU, no signs of anything getting warm except the regulator and rectifier (which were just warm), and the 2.5W 10k load resistors, which were very hot.

The same thing happens each time the PSU has been running on mains, with the 10k load, about a minute after HT has reached its operating voltage. It's happened about three times, in total.

I'm guessing the 2.5W load isn't big enough. I'll try running the PSU with a 5W load tomorrow, and see if I still get the smell, and if so, make a note of how much time it takes to appear.

[Kokotoni Will]

Hardly any smell at all with a 5.75W 10k load after two and a half minutes, although load still hot to touch, so I can confirm it's the resistive load and not part of the PSU that was getting hot and making the smell.

[Kokotoni Will]

I tried the PSU with lamp limiter with a 60W bulb, but voltages were only slightly higher than with a 40W bulb. The below measurements were taken with two bulbs in parallel, one 60W and one 40W. The 10K load was used.

HT: 225V

V1 (12-E1):
2 (H): 4.4V AC
4 (G2): 380V
5 (G1): 237V
7 (H): 4.4V AC
8 (K): 236v
Cap (A): 380V

V2 (CV378):
2 (H): 384V
8 (H, K): 384V
Heater: 3.3V AC

V3 (ECC 83):
1/7 (A2/G1): 30V
2 (G2): 94V
3 (K2): 117V
4/5 (H2/H1): 4.4V AC
6 (A1): 391V
8 (K1): 233V
9 (H centre tap): 4.4V AC

V4 (ECC 83):
1 (A2): 117V
2 (G2): 69V
3/8 (K2/K1): 71V
4/5 (H2/H1): 4.4V AC
6 (A1): 231V
7 (G1): 41V
9 (H centre tap): 4.4V AC

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

V6 (85A2):
1/5 (A/A): 231V
2/4/7 (K/K/K): 143V

[PJL]

Are you measuring the AC heater voltages with a probe on each of the valve heater pins?
I suggest you do not measure the ECC83 voltages as the meter load will affect the circuit operation.
Have you completed all of the component replacements on the amp?
The 600V is now down to just 380V which is far safer for the ECC83. Does the pot still change the HT voltage?
Next step will be making up the umbilical chord and reforming the electrolytics on the amplifier unit using part of your 10K 'load' in series with the HT.

[Kokotoni Will]

Originally Posted by PJL

Are you measuring the AC heater voltages with a probe on each of the valve heater pins?
I suggest you do not measure the ECC83 voltages as the meter load will affect the circuit operation.
Have you completed all of the component replacements on the amp?
The 600V is now down to just 380V which is far safer for the ECC83. Does the pot still change the HT voltage?
Next step will be making up the umbilical chord and reforming the electrolytics on the amplifier unit using part of your 10K 'load' in series with the HT.

Yes, I measured the AC heater voltages as you described.
I've fitted all the new parts. Currently in the process of re-soldering the strip, etc.
With the two bulbs in the lamp limiter, the pot allows for adjustment in the range of roughly 197V - 219V.

[PJL]

200V should be enough to complete reforming.
1. Make up the connecting cable
2. Remove all valves from the Amplifier
3. Not sure what your 10K load comprises but you need around 40K to put in series with the amp HT along with your meter also in series and set for DC current. Switch on and monitor the current which should start at 5mA and slowly fall away. Once it has fallen by half you need to halve the series resistance to 20K to maintain the 5mA reform current. Halve again to 10K once it has fallen by half again. Leave like this for a few hours and you should be left with just the electrolytics leakage current (maybe 1mA).

[Kokotoni Will]

I've finished rewiring the amp. I ended up changing out the electrolytic capacitors, so they won't need reforming.

[PJL]

Very neat and in keeping with the rest!
Did you sort out the grid stopper wiring?
Set the pot to the lowest HT.
Have you made the connecting cable? If both LT windings are 6.3V then you need to use the lower resistance one but the difference may only be 0.1 ohms! The alternative is try one winding with half the valves in and measure the heater voltage disconnected from the amp and compare with the value when connected (and settled down). If you then repeat using the other winding, the one with the smallest difference is the one you need.

[Diabolical Artificer]

Cracking workmanship, kudos.

Andy.

[Kokotoni Will]

Originally Posted by PJL

Very neat and in keeping with the rest!
Did you sort out the grid stopper wiring?
Set the pot to the lowest HT.
Have you made the connecting cable? If both LT windings are 6.3V then you need to use the lower resistance one but the difference may only be 0.1 ohms! The alternative is try one winding with half the valves in and measure the heater voltage disconnected from the amp and compare with the value when connected (and settled down). If you then repeat using the other winding, the one with the smallest difference is the one you need.

Yes, I've sorted the grid stoppers. Haven't made the connecting cable yet - I'll get on with that tomorrow.

Both LT windings are 6.3V. I've measured the resistance between the two wires of each winding (between A and B, and between C and D in the photo), but both windings measure 0.00 ohms. Am I doing this right?

For the purposes of the alternative test, is it necessary to connect HT to the amp?

I'm curious as to why it's necessary to use the winding with less resistance.

[PJL]

Your meter is probably auto-zeroing.

You don't need to connect HT for this test. We are simply trying to load the LT winding as we need the one with the highest current capability. I assume one of the windings would have been for a preamp. The total requirements for LT is 6.3V 3.44A so pretty low for a stereo amp.

[Kokotoni Will]

With no load, winding AB gave 6.345v AC, and winding CD gave 6.362v AC (or thereabouts).

When connected to the amp with only the left channel valves in, winding AB gave 6.156v AC, and winding CD gave 6.260v AC. So it looks like I need winding CD for the amp.

[PJL]

Agreed, wire up the cable, insert all valves, put your meter on the HT, connect speakers, and power it up without the lamp limiter. Switch off immediately if the HT falls more than 10%, or there is a burning smell, or anything else out of the ordinary for that matter. Don't leave it on for long as we need to do a few other checks like the cathode voltage on each of the output valves.

[Kokotoni Will]

I've only got one 15 ohm speaker. I've got a pair of 4 ohm speakers, will these do?

Should we measure the output transformer turn ratios first?

[PJL]

4 ohm will be ok. Matching impedance allows you to get the maximum undistorted sound. For valve amps running with no speaker is a greater risk as the amp can create very high voltage peaks that damages the output transformer insulation. For transistor amps its the opposite!

[PJL]

To sort out the output transformer secondary taps you need to use the 6.3V AC heater supply.
1. Remove the valves.
2. Disconnect the HT from the amp.
3. Disconnect existing links from the secondary of one of the output transformers.
Measure the resistance between each of the secondary pins to determine how the windings run.
4. Hook some wire links from the heater connections of a nearby valve to the output transformer primary.
5. Turn on and measure the AC voltage on the primary and each of the windings you previously identified.
6. To work out the phase you need to connect one winding in series with another and measure the AC voltage across the pair. If the voltage is the sum of the individual windings they are in phase. Repeat by disconnecting the link from the 2nd winding and attaching to another to repeat the phase test.
7. You should end up with a primary voltage and a list of the secondary windings with their voltage and phase.

[Kokotoni Will]

I've finished wiring the cable, and powered up the amp with it. PSU HT rose to 214V before tailing off, whereas it reached 240V with no load, so I switched off. Other than that, I didn't notice anything strange, other than one of the amp power valves (V6?) glowing slightly brighter than the others.

[PJL]

Measure the output valve cathode voltages starting with the one that was brighter as it might have 'gone to air'. A look at the datasheet and I would estimate it should be around 14V corresponding to 30mA anode + screen current. This is an approximation and what you are looking for is consistency across all 4 so anything very different switch off.

A2134's are plentiful and affordable on ebay, For example a NOS GEC pair are just £9.73. If only KT66's were the same!

[Richardgr]

Did we shed any light on why all the valves had metal screens? Wouldn't that be a bad thing for a 7 pin output valve with a 13W dissipation?

Also, there was differing wiring for the OPTs on each channel, if this was originally for stereo use. Was there a view on that? I was wondering if it was once a bi-amped mono solution with different speaker impedances.