JCM 800 Output Transformer

[JoeK]

Have a JCM 800 2203 (100 watt 4 EL34) and just finished some work on it. New filter caps, bias supply caps and set of matched JJ EL34. When I setup the amps bias on idle it has 25 milliamps on each tube with 530VDC on plates and 525 on the screens. When I run a test signal I get one half of the OT (brown lead) doing 80ma on each tube and the other half (red) doing 50ma on each tube. If I swap the tubes the current draw stays with the OT leads. If I measure the resistance across the plates (tubes out) to OT CT I get 14.5 ohms Brown and 16 ohms Red.
I could swap bias supplys to each half of the OT which will prove the last two resistors in the bias supply (one for each half but they measure 220k each so OK) and the bias is good at idle -44vdc on each side, under load it becomes -41vdc Brown -43vdc Red.
Is it possible the OT has an imbalance or fault that is causing this issue? Thanks in Advance Joe

http://www.drtube.com/schematics/marshall/jcm800pw.gif

[Jazz P Bass]

Most, if not all output transformers do have a different reading from red to brown & red to blue.

[g1]

Your output section is matched when idling but unmatched with signal. Swapping tubes showed that the problem is with the circuit, not the tubes.
You need to check the AC voltage of the signal on the output tube grids when the mismatch occurs. Chances are you are not getting the same drive signal from the 2 sides of the phase splitter.

[JoeK]

Your output section is matched when idling but unmatched with signal. Swapping tubes showed that the problem is with the circuit, not the tubes.
You need to check the AC voltage of the signal on the output tube grids when the mismatch occurs. Chances are you are not getting the same drive signal from the 2 sides of the phase splitter.

hmmm didn't check that out yet Shall report back.

[JoeK]

Ok I've got some measurements. The phase inverter is delivering more AC signal to one side of the power tubes than the other, about 20-30% more. The B+ is 396 at the PI plate resistors Pins (82k) 1:252vdc 2:23vdc 3,8:39vdc (100k) 6:215vdc 7:27vdc The 100k side is delivering more signal and I realize the PI won't be balanced exactly but the voltages are off the mark when I look at http://mhuss.com/MyJCM/JCM800_2204.gif

[J M Fahey]

Ok I've got some measurements. The phase inverter is delivering more AC signal to one side of the power tubes than the other, about 20-30% more. The B+ is 396 at the PI plate resistors Pins (82k) 1:252vdc 2:23vdc 3,8:39vdc (100k) 6:215vdc 7:27vdc The 100k side is delivering more signal and I realize the PI won't be balanced exactly but the voltages are off the mark when I look at http://mhuss.com/MyJCM/JCM800_2204.gif

1) to begin with, don't worry about that.
Guitar amps have been running like that since the Dawn of Humanity.
I'm *certain* that it sounds killer as is.

2) that said, you can adjust if you wish.
Usually the long tailed PI used there is *not* symmetrical by itself , so to compensate a higher load resistor (100K) is used on the supposedly weak side, and an 82K in the strong one.
*BUT*, enter modern crappy tube production, often both halves of same tube do not match , this might be the case.

So:

a) put another PI tube, it might even out, or

b) replace 82 K with 100K, should do, or

3) leave it as is and be happy.

Pick one of the above

As of voltages not matching published ones, +/-20% is within tolerance.

[Amp Kat]

1) to begin with, don't worry about that.
Guitar amps have been running like that since the Dawn of Humanity.
I'm *certain* that it sounds killer as is.

2) that said, you can adjust if you wish.
Usually the long tailed PI used there is *not* symmetrical by itself , so to compensate a higher load resistor (100K) is used on the supposedly weak side, and an 82K in the strong one.
*BUT*, enter modern crappy tube production, often both halves of same tube do not match , this might be the case.

So:

a) put another PI tube, it might even out, or

b) replace 82 K with 100K, should do, or

3) leave it as is and be happy.

Pick one of the above

As of voltages not matching published ones, +/-20% is within tolerance.

^^^Can't agree any more with what JM is saying. That is very normal operation and probably sounds better like it is !!

[Tom Phillips]

I agree with JM too. If you make the kind of detailed measurements that you described you discover all manner of things that are "wrong" based on classic audio amp design theory. If you were to "fix" them all (major redesign required for some "faults") then you would end up with a clean HiFi amp which would "reproduce" recorded music well but would not "produce" a good guitar sound. In your current case the non symmetrical drive that you have observed is most likely contributing to some nice harmonic content in your amp's output signal.
Cheers,
Tom

[JoeK]

Hi I decided I might test what happens if I changed the 82k/100k phase inverter plate loads to 100k/100k and when measuring I found that the original 100k plate resistor was measuring 180k! So now have a much more balanced output signal to the power tubes. Gone over the other resistors to check what's right or not and found a few but I guess the amp is over 30 years old so it's not unusual as a welsh singer once said Cheers and thanks Joe.