6L6 PP amp motorboating

[chrish]

Hello, after months of enjying my 6L6 PP amps from this thread 6L6GC AB2 Amp, one of the monoblocks has developed what I think is a motorboating instability. I am getting a 'POP POP POP' soumd at about 4-6 Hz, and the VR tube supplying the MOSFET source followers with -105volts flashes in sinc with the 'POPs'

BTW, cannot duplicate on the bench with a fixed 8R load, only appears in system. Changed all tubes as well with no change.

The amps are monoblock 6L6 PP amps with regulated supply supplying the output and voltage amp/phase splitter. Input is a pair of diff amps with 10M45 CCS in the tails feeding MOSFET source followers to the output tubes. I am guessing that the instability is some kind of feedback through the B+ rail? My thoughts are to try to decouple the two diff amps. The first is already decoupled with a 6.2K 20uF RC filter that should have a cutoff of about 1.3Hz. I have upped the capacitor to 40uF with no change. Thinking of making a sep rail for each input stage and each having its own RC filter, maybe upping to 100uF (as these are the only high voltage caps I have on hand), so 6.2K and 100uF for the first diff amp and maybe 1-2K and 100uF for the second diff amp.

Am I on the right track with this approach, or should I investigate elswhere?

Schematic attached.

Thanks,

Chris

[Miles Prower]

Don't see any gNFB connected, and if that's the case, it wouldn't seem to have anything to do with the early stages. If it had something to do with the UL connection, then I'd expect that changing tubes might have some effect on it.

I'd tend to suspect that one of the MOSFETs is headed south. I would also try disconnecting those capacitors paralleling the VR tubes. That's a great way to make an oscillator.

Also, I wonder about the voltage of that negative rail. 130Vdc doesn't look like very much margin to get the VR105 to strike reliably. Maybe with some hours on it, it's not striking like it used to?

[chrish]

Thanks for the input Miles,

You are correct, there is no global feedback. I did have an issue with enough voltage to strike the VR tubes, so I ended up bucking with one of the unused 5 volt windings from the power transformer to boost the voltage a little. I did screw up a little in my description above, the OA3 +75 volt regulator tube flashes, not the -105 volts reg tube. Of note is the OA3 does drop out of regulation when amp is placed in to heavy clipping (when testing on bench, not in practical use). The amp has been in use for over six months without any problem, and the other monoblock is working fine. Also, no problems when hooked up on the bench with a sine wave input and 8R power resistor load.

Regarding the parallel capacitor to the VR tube, are you talking about the 0.1uF? I am guessing the 220uF power supply capacitors for the +/- supplies are sufficiently decoupled? I might try this first, as disconnecting that 0.1uF capacitor is quite easy, replacing he MOSFET is going to be significantly more involved due to my wiring and heatsinking setup.

Thanks again,

Chris

[chrish]

Further investigation showed that the bias correction to the second grid of the first 6SN7 pair was about -33 volts. The bottom LED of the DC offset adjustment was intermittent, so the bias would swing from some very small +/- value of less than a volt to -33 volts. I am not sure if this was the cause of the problem or a sympton of some other problem. Anyway, I have taken the adjustment out of circuit for the moment and have just connected the second grid of the first 6SN7 to ground via a 1K resistor.

So far so good...

Chris