Question: Leslie 122 amp output stage instability issues

[tpappano]

Hi All! I'm doing sort of a "face lift" of the 122 amp using modern materials and construction methods. The amp circuit and magnetics remain stock. On firing the new amp up there were terrible symptoms of instability, distortion and low output. Long story made very short, the problem appeared to be caused the .005/.0047 caps from each 6550 plate to ground. Instead of putting out 40 watts of "audio" it was radiating massive amounts of RF at around 50 to 60 MHz. The oscillations were being *caused* by these capacitors, perhaps by forming a resonant circuit with the tube electrodes. Removing them brought forth the performance and sound we all love 8) My question is "are there known issues with these capacitors and if so, typical fixes?"

Thanks!

[geoelectro]

I've never heard of any such problems with those caps. Since you say this is a new build, I would suspect grounds. They terminate to ground and if the ground isn't secure they would connect the two plates to each other. I don't think that in itself would cause oscillations so maybe some other ground reference. If the voltage swings from the outputs were to somehow enter the circuit at an earlier point like the 12AU7 circuit you would have massive oscillations.

Geo

[David Anderson]

I've never heard of issues with these capacitors, but I have heard of issues with people changing the physical layout of a circuit. You're not telling us exactly what you did, but it seems to have resulted in unintended side-effects.

Numerous examples have emerged in recent years of problems arising from printed circuit board layout that doesn't play well with tube circuits. Marshall has been plagued by them.

[muckelroy]

I found this ironic given that I thought those caps were there to prevent oscillations.

[tpappano]

Me too! This is a pcb configuration, I think what is going on is the the pcb layout is providing very low impedance signal paths and the ground system is very low impedance as well. Combine all this with the very low esr of the 'modern' plate (and cathode) bypass caps and I think you wind up with a high frequency series resonant circuit comprised mainly of tube electrode reactances and the plate bypass capacitor. Feedback with the 12AU7 didn't seem to be the issue, it didn't even need to be installed. I'm thinking the caps may need a small resistance added in series with them to act properly as a parasitic suppressor instead of a parasitic enhancer 8)

[David Anderson]

I'd suggest another theory. When a PTP circuit doesn't work properly when transferred to PCB, it's usually because the PCB layout has introduced some unintentional capacitive coupling that was not present in the PTP layout. This can definitely cause oscillation. Having a 6550 grid trace too close to a 6550 plate signal will do this.

The stock grounding layout of the 122 may be slightly crude, but it's not high resistance. A few milliohms one way or another isn't going to be enough to cause oscillation.

I have a Dean Markley CD-120 amp from the 1980s that I got for $50 because I was told no one could fix it. The scope showed me it had pretty severe high frequency oscillation above the range of human hearing. I had to replace the output tube grid wires from PCB to tube socket with shielded cable.

There are potential advantages to the original practice of wiring much of the 6550 circuitry directly to the tube sockets rather than through wires to a board.

[tpappano]

Everything is pretty well separated. Over the next few days I'll be doing more testing, first with the capacitors out, trying to force unstable operation. We shall see 8)