Peavey Triple X - high frequency oscillation

[xtian]

This amp's owner complained of bad noises. I found burned open screen resistor R64 (400R/10W) and replaced. All other components pass inspection. Amp sounds OK at idle now (I have kept the bias at minimum for testing--seeing 470v plates and roughly 15mA each 6L6). Issue is, 10-11KHz oscillation blooms as master volume is turned up past noon. I can hear the pitch, and energy consumption rises quickly from ~80w at idle to ~120w when oscillating.

Oscillation happens ONLY if external speaker is connected; no oscillation if running on resistive load only.

Oscillation does not happen if V2 pulled.

Using scope, I start seeing the oscillation at large amplitude just after V2B (see green circle on my schematic screenshot). Tube swaps don't change anything.

I will try using various snubbing techniques on V2B, but happy to hear anyone's suggestions.

Schematic: https://www.thetubestore.com/lib/the...-Schematic.pdf

[xtian]

I tried touching a 470p ceramic across plate resistors R2, R7, R4 in turn. No change to oscillation.

What's the other obvious approach? Is it snubber cap from plate to cathode?

[The Dude]

You say, "Oscillation happens ONLY if external speaker is connected; no oscillation if running on resistive load only". Then, oscillation is traced back to V2. I don't see an electronic signal path between the two, so I'm wondering if the oscillation is due to mechanical or magnetic coupling. Is this a combo or a head? If combo, try a speaker outside the cabinet and see if oscillation is still there.

[xtian]

This is a head. Speaker cab is a few feet away.

[xtian]

I see no signs the filter caps are failing, but is this a possible vector?

[The Dude]

I'm not saying it is or isn't. There are plenty of possibilities. I will say that I just, 1/2 hour ago, finished up a Vox AC15 that was oscillating at about 24K. A new main B+ filter cap solved the problem. I found it by temporarily clipping in known good caps across the existing ones, while monitoring the oscillation, and found the bad one. FWIW: The bad cap tested perfectly fine out of circuit.

[xtian]

Clipping a fresh 47uF into each node made no change in the oscillation.

[xtian]

Found it! 1n cap from anode to cathode (pins 1,3) of V3A solves the oscillation. 470p didn't do it, but 1000p did.

[tedmich]

glad you solved it but IME these amps work great without adding extra caps IFF all the other components are working!
The Illi-noise (sp?) PS caps spec'd on these are NOT long lived; they will need replacing eventually.