JTM45 problem

[flatart]

I have a problem with a JTM45 that I've got to repair. When I play hard at load volume I get really an ugly sound, it really seems that I have got a broken speaker.
Obviously it's not a matter of speakers.
Doing some tests with my scope I found something really strange to me:
if I put the scope probe on Pin 5 of one of the KT66 output tubes, the problem magically disappears!
I tried to understand: basically if I put the scope probe on pin 5 it's like putting a high value resistor between pin5 and ground, isn't it?
Anyway I'm not able to fix the problem
HELP ME!

[Phil_S]

What have you tried to fix it? Is bias voltage OK? Have you lifted and checked the grid leak resistors to see if they are toast?

[Structo]

Does it have screen resistors?

[flatart]

In attachment the schematic of the amp.
I tried to move the bias trimmer alla the way up and down but nothing changes: voltage goes from -40V to -52V, so it seems ok.
Another strange thing is that all voltages seem to be ok.
Tomorrow (here it's 1AM and neighbourhood is sleeping) I will try another couple of output tubes cause someone suggest me that it might be microfonic output tubes

[Structo]

How old are the filter caps?

Funny this amp does not have swampers on the power tubes.

You may try 1K5 to 3K resistors between the 220K (R24 & R25) and pin 5 of the power tubes.

[Andy Le Blanc]

double check the tube base and tube socket, it doesnt take much oxidation
on a tube pin to make noise.

[flatart]

little mp3 sample of the problem

[billyz]

Could it be Parasitic ocsillation ? Fender had to put small values caps on their later cbs amps to cure these ills. Right on pin 5 , a cap to ground. Your meter will also have a small capacitance. Might need to to redress the wiring. Is this a homebrew amp?

If it's an old marshall , they were notorious for arcing on/in the tube sockets . sometimes you can clean them but most of the time I had to change them out.

[flatart]

it's an homebrew amp I made about 2 years ago. I have used it for more than one year without any problem, now this strange issue.
Sockets are ok, they are not so old.
Today I tried also with a pair of new 6L6, but still the same problem.
I think I will resolder all output tube pins and also bias circuit.

[billyz]

Did you use cloth wire ? Could have absorbed moisture or something else. Same goes for if you used the fender style tag board, it too can absorb moisture or oily contact cleaners. Try the 2000pf cap to ground, if it goes away, at least you know what might be the cause.

[dansamp]

May I suggest,
check your solder joints from the PI caps going to your power tubes
C12 and C13
it seems to me it sounds like a loose solder joint type noise
are the power tubes biasing properly ??

[Structo]

Listened to your clip.

I can see why that would bug you to no end.

Try the suggestions here and report on each thing tried so we aren't redundant with the help offered.

[jelle]

Quick test...happened to me too once...


What happens when you disconnect the NFB wire from the speaker jack?

[Clyde]

Did you try some 5.6k grid stoppers on pin 5 of the outputs? I'm surprised that it still had the problem with 6L6's, suggesting this might not be the problem but KT66's spec a higher value than 6L6's in this department-1.5k is pretty standard for 6L6's. Hope this helps.

[flatart]

I re-soldered almost all the circuit: output tubes pins, all bias circuit, all pots and all solderings that may appear not so good, but nothing changes.
So next try I will put a grid stopper on pins 5 (I will go for 5k resistors).
I will also try to disconnect feedback (never tried yet).

Another strange thing I noticed is that also when I put scope-probe on pins 1 and 6 of PI the problem seems to disappear or at least is quite quite less evident.

So another question for you: do you think that filter caps could be the origin of this problem, even if I read correct values on plates or cathodes of tubes (both output and preamp) ?

Thanks everybody here for your suggestions
I'll keep you posted