Oscillating JMP

[Ted]

I've got a 100W Master Volume Marshall JMP here.

Came in with a blown output transformer.

On replacing the tx (and all of the 35-year old power electrolytics), it oscillates at about 35 kHz at some control settings. - Audible as a mushy hiss.
I'm guessing that's what did for the original transformer....

With V1 removed it still oscillates, but weakly. Frequency is stable with changes in tone and presence controls.

All grounds cleaned. Lead dress makes a marginal difference, but not much. NFB sense is correct. All resistor values correct to within 5%.

Tried a set of NOS RCA 6CA7s and some M8137s that I was saving for a rainy day.

Nothing seems to make any difference.

Any suggestions?

[Enzo]

Scope the B+ nodes. Is this spurious signal riding on the B+ anywhere? Loss of decoupling can cause oscillation

[Ted]

Supply rails clean as a whistle.

Screening the wiring back to the Presence pot seems to improve things.

It only breaks into oscillation when heavily overdriven now...

[Chuck H]

Gut shot?

[Ted]

Sorry Chuck - it's gone back now. But it was a mess. Maintained on a shoestring for 35 years I guess...

By coincidence an immaculate 50W JMP came in this morning (this thing looked like it had never even been gigged), so I copied the lead dress.

Fitted a set of JJ E34Ls and all seems to be OK. No nasty noises on overdrive, everything looks fine on the scope.
And loud. Very, very loud.