Unintended Low Freq Oscillation

[tubeswell]

Okay I decided to alter my recent 5F6A build to try and get a better reverb balance. The schematic I came up with is attached.

It is now a one channel bassman. 1/2 the 12AY7 goes to the vol control and thence to a 270k mixer resistor and into the DCpair. The other 1/2 12AY7 is tapped off the top of the vol pot and used for a gain stage in front of the parallel 12AT7 reverb transformer driver. Then there are 2 cascaded halves of a 12AX7 for the recovery stages with a kind of 5F2A level and tone control after the 2nd stage, which blends back into the other 270K mixing resistor in front of the DC pair. The rest of the amp is as I built it before.

But when I switch it on and turn up the vol I get a funny thumping at about 180 and 240 thumps per minute. (Sounds like an iroqouis chopper)

With the PI pulled there is no noise, except for a little static to be expected with no PI tube. The trem controls don't affect it. I pulled the trem tube just to make sure it wasn't that, and it wasn't. The trem footswitch had no effect on it either.

Twiddling with the level and tone control seems to affect it. The vol control cuts it completely.

So have I created some sort of inadvertent LFO? What would be my immediate test to cure it (short of completely undoing everything I've done)?

Edit - Correct schematic in next post (sorry)

[tubeswell]

Actually I had the wrong schematic posted. Here is the circuit that is oscillating. Could it be to do with the shared cathode resistors in the first two stages and in the two reverb recovery stages maybe?

(Edit - Layout also attached)

[loudthud]

Tack another 22uF across the filter cap for those first stages. Does it lower the frequency of the thumps? I suspect the power supply is part of the feedback path. You need to decouple one or more of those stages from the others. I'd try an independant RC from the PI filter node just for the DC pair.

[tubeswell]

Hi loudthud

I posted the wrong schematic earlier. The version that is oscillating gas shared cathode resistors in subsequent stages. Would that be the culprit?

[loudthud]

Hi loudthud

I posted the wrong schematic earlier. The version that is oscillating gas shared cathode resistors in subsequent stages. Would that be the culprit?

That's certainly possible. Tack 1000uF across that 250uF.

[tubeswell]

Thanks I try that. Would alternating the shared cathode resistor arrangement also work maybe? (i.e.; V1a shard with recovery 1a, and V1b shared with recovery 1b)?

[loudthud]

I'm not a fan of shared cathode resistors but it can't hurt to try. This amp is way too complicated to predict how it will things will affect each other. Don't forget you can just leave one cathode unbypassed with a small penality in gain and not use much board space.

[melvin]

You may have to separate the cathodes of the 12AY7. Layout could also be a factor.

[melvin]

what about when you pull a reverb tube?

[tubeswell]

Well it didn't oscillate before I changed it, so it was something I did. I'll try separating the cathodes first I think. With the V1 tube pulled it still oscillated when I turned the recovery level up, so it is most likely in there. I'll let you know.

Also I wasn't sure about the best way to add a tone control, so I opted for the 5F2A type. (I previously had a 6G15 tone control when the reverb circuit was more like a 6G15, but I wasn't sure how that would go into this circuit.) Any opinions about better types of 1-tone controls in this circuit?

[tubeswell]

Okay I had some spare time today so I re-did the cathodes in the first stages and in the reverb circuit (by going to separate cathode resistors for each stage) and it cured the problem. (So it must've been the shared cathode resistors).

Then after I got it going properly I found the gain too much in the driver and recovery side, so I did a bit of tweaking to add more attenuation between the stages there, I've got the reverb sounding fairly strong and (for the most-part) clear now. (I just might go to a 15k/33k split for the split-load resistor in the driver circuit)

Edit - put the split load 15k/33k in the stage before the reverb driver. Also reduced the 2nd coupling cap in the recovery stage to .01uF and bypassed the first recovery cathode resistor with 2.2uF. Sounds really good now (if I may say so) - got the bassman vibe as well as verb and trem which only enhance it. The reverb controls have a good dynamic range. I was playin' away this avo and I just couldn't put my gat down. Very happy for now. :-)

Revised schematic attached FWIW