VOX Cambridge Reverb ghost

[Randall]

Guys,

Tube VOX on the bench with the complaint of tremolo works fine, but when it is switched off via footswitch, it makes a HFO squeal, and pulls trem plate of V7 down to about 150v. The wire coming off the grid of V7 is about 10" long going to the first .01 cap, and makes different sounds depending which end I probe. I replaced it with a shielded wire, but nothing changed. I verified all the grounds in the circuit. The caps measure good out of circuit, all resistors are close enough.

I foud that if I put my scope probe on the gid, the problem went away, so I clipped in a 270pF cap to ground, and it was good. Until I realized removing the cap, but leaving just a short lead on the grid had the same effect. Then I found probing the scope also quieted it right down. Also saw the noise up into the PS, but it decreased the deeper I went, so I don't think it was coming from there. Also found the touching the shielded cable I installed close to the grid stopped the nose, but not at the other end of the wire.

Eventually I tacked a 47pF cap across the plate resistor R40, thinking it was going to ground, and it stopped the issue. It wasn't until I looked at it again while typing this that I realized it is in fact not going to ground, so I don't know why it is working, but it is.

This is a point to point build that has had most of the caps replaced, new power cord, etc. I don't know where the oscillation was coming from, or why that cap in that position stopped it, but it works fine now, so I am going to leave it. I've already spent half an afternoon on it, and I need a Sunday motorcycle ride.

Maybe someone can tell me what I'm missing. I will wait to button it back up in case someone has an idea.

https://schematicheaven.net/voxamps/...dge_reverb.pdf

[Jon Snell]

I would ensure the changed components are actually what they should be,
Sounds like instability, try eplacing V7.

[Randall]

It worked fine for a long a while after the changed components. Something changed after that fact. They measure close enough in spec. I tried changing V7 as rudimentary. It is obviously instability, why is the question.

[Jon Snell]

Without working on it myself, that is a very good question. If R38 is the correct value;
The reverb/tremolo foot pedal seems the weakest link here but if you place a 100p capacitor across R38 instead and a 22k 1/4W on pin 2 of V7 to the junction of C20 and R35 as a grid stopper, that should cure any instability we see commonly from newer ECC83s/12AX7s.

[pdf64]

Maybe something about the footswitch cable has changed, leading to it having eg higher inductance?

There’s often low pass filtering on the output of a trem oscillator.

[Randall]

I wasn't using the foot switch, I just clipped a lead to chassis. Problem was there in both instances.

"if you place a 100p capacitor across R38 instead and a 22k 1/4W on pin 2 of V7 to the junction of C20 and R35 as a grid stopper, that should cure any instability we see commonly from newer ECC83s/12AX7s."

Can you explain the placement of the 22K grid stopper? There is already 1Meg from pin 2 to ground, how does adding 22K help in this case?

[Helmholtz]

Can you explain the placement of the 22K grid stopper? There is already 1Meg from pin 2 to ground, how does adding 22K help in this case?

A resistor from grid to ground acts as a grid leak.

A grid stopper is wired in series with the grid (not to ground) and constitutes a low pass filter together with the (invisible) grid to ground capacitance.

[Randall]

Right, grid stopper, grid leak. Got it!

[Randall]

I wired it as suggested and it is still working well, but I like this way better than the "it works, but I'm not sure why" method. Thank you gents!