Bias drift when volume engaged.

[The Dude]

You can't hear oscillation, or a scope doesn't show oscillation? I ask because often oscillation is not in audio range.

[Chuck H]

This is going to seem like a stupid question. You're not running a signal while testing the bias, are you?

[The Dude]

......per tube at idle no signal /speaker attached/ with a twist of vol knob is odd

Continuing to hunt for a bad ground---

^^^^

[Chuck H]

^^^^

Ah. Thanks. Read through too fast.

[mhuss]

I've had a little experience with these amps. Particularly given the Presence control interaction, I'd highly suspect an UHF oscillation. Do you have a scope to use to check the output? The input grounding is particularly important, daisy-chained to a solid chassis connection near the input jacks.

[g1]

If you hear a hiss that goes away when you turn the volume to zero, that is sound.
Turning up the hiss with the volume control is going to make the idle current increase. It's not really idling anymore, it's amplifying a sound (hiss).
I guess the question is, how much change in current is normal as you increase the hiss level out of the speaker. I would compare it to another amp.
Like m1989jmp mentioned above, idle current should always be set with volumes at zero. If you set it that way, and there is no drift when left at zero, then I think you can safely assume the increase in current is due to the hiss. (unless it's a major difference)

edit: I'm not saying there is not a problem, just that I don't think it's a bias drift problem. A doubling in current seems to indicate something more than hiss going on.

[Ampzone]

If you hear a hiss that goes away when you turn the volume to zero, that is sound.
Turning up the hiss with the volume control is going to make the idle current increase. It's not really idling anymore, it's amplifying a sound (hiss).
I guess the question is, how much change in current is normal as you increase the hiss level out of the speaker. I would compare it to another amp.
Like m1989jmp mentioned above, idle current should always be set with volumes at zero. If you set it that way, and there is no drift when left at zero, then I think you can safely assume the increase in current is due to the hiss. (unless it's a major difference)

edit: I'm not saying there is not a problem, just that I don't think it's a bias drift problem. A doubling in current seems to indicate something more than hiss going on.

Culprit was in presence, the 10kR was faulty, reading less than 1k. Replaced it now and have no drift at all. Vol/Prez pots no longer affect it. Appears solved.

[Helmholtz]

Culprit was in presence, the 10kR was faulty, reading less than 1k. Replaced it now and have no drift at all. Vol/Prez pots no longer affect it. Appears solved.

That supports the HF oscillation hypothesis. I assume the 10k is the feedback resistor from the speaker output. A 1k instead strongly increases loop gain (https://en.wikipedia.org/wiki/Loop_gain) and thus the risk of oscillation.

[Chuck H]

Increasing loop gain in a feedback circuit should only increase the risk of oscillation if it's a positive feedback loop. Increasing loop gain in a negative feedback loop decreases the final output and (usually) reduces distortions and output impedance. It's probable the loop is unstable.

[g1]

After the bad part replacement, was the bias at the lower number (13mA) ?
That would verify oscillation/instability was the cause of the increase in idle current.

[The Dude]

This sounds fishy. Resistors don't lower in value. Was the wrong resistor installed? What was the color code on the "faulty" resistor? If you measured it in circuit, It would measure lower resistance because you're also measuring the parallel 2.2K cathode resistor since one side of the feedback resistor is grounded through the OT. I'm wondering if it was just a connection issue you fixed by resoldering a part.

[Chuck H]

I'm wondering if it was just a connection issue you fixed by resoldering a part.

This^^^^^^^^^^^^^^

EDIT: Wait... If there was a bad connection at the feedback resistor I don't think the presence control would have exacerbated the problem.?. I've seen amps behave almost stable with incorrect feedback loop phase.

[Helmholtz]

Increasing loop gain in a feedback circuit should only increase the risk of oscillation if it's a positive feedback loop.

Well, yes and no.

In most real world amps with a global NFB loop the risk of oscillation increases with loop gain. In fact, without adequate phase compensation any NFB amp will eventually oscillate if loop gain is high enough.
The reason is that unavoidable phase shifts along the feedback path finally turn negative into positive feedback when the frequency is high enough (that's for the "yes").

In a design with positive feedback it suffices that the loop gain is above 1 to start oscillation.