Marshall 9200 Output Signal Oddity

[AmpFix]

When checking the output signal of a Marshall 9200 power amp, I noticed that the positive portion of the signal was changing in amplitude on a regular interval. The input signal is solid.

Changing the balance pot doesn't really help, although it evens out the symmetry. Bad transistor (the PI circuit is unusual on this amp) or aging caps

The behaviour is the same on both channels. Might be part of the design, but I find it odd nonetheless.

Check it out here:

[The Dude]

Well, a couple of thoughts:

1) If you suspect caps, you could monitor the B+ and see A) if the DC level is changing with the output waveform and B) if there is excessive ripple.
2) That looks like a pretty minimal change. Can you hear it? Is it actually a problem?

[AmpFix]

Its actually odd behaviour. The higher the gain (volume) setting, the more the amplitude varies and speed at which it changes increases. Lower the gain (volume) and the amplitude diff is smaller and the little change that there is toggles more slowly.

The Bias and B+ look solid. I see no ripple there and the ripple wouldn't match the frequency of the fluctuation.

[Mick Bailey]

I also see the negative half modulating, though to a much lesser degree. Is the other channel OK?

What does signal look like coming off the PI?

[Enzo]

I am with Dude here, look at power supply for ripple. Or slow down the scope and see if the dipping seems to correspond with 60Hz or 120Hz. Your test signal will turn into a crammed blur, but the top of the envelope might reveal the ripple pattern.

also agree you need to look at your AC mains to see if it is doing the pulsing. Due to resonances in the power grid, such things occur as surges resonate and bounce around the grid. Look at a power substation sometimes and see huge oil filled capacitors there used to correct phasing issues and such.

[AmpFix]

I am with Dude here, look at power supply for ripple. Or slow down the scope and see if the dipping seems to correspond with 60Hz or 120Hz. Your test signal will turn into a crammed blur, but the top of the envelope might reveal the ripple pattern.

also agree you need to look at your AC mains to see if it is doing the pulsing. Due to resonances in the power grid, such things occur as surges resonate and bounce around the grid. Look at a power substation sometimes and see huge oil filled capacitors there used to correct phasing issues and such.

Have a look at my subsequent post (post #3 in this thread). The pulsing rate varies with the volume setting. The higher the volume, the faster the pulse.
I'm not detecting any ripple on the B+ or Bias supply. I didn't think of checking the AC mains though. I'll definitely do that and report back.

I posted on here to see if anyone had ever seen something like this. When I get to the bottom of this (if I have the patience - the amp seems to work fine), I'll report back.

[AmpFix]

Found it! The fan was causing a slight, recurring drop in DC voltage, which is on the same tap as the bias supply. With the fan disconnected, the output waveform is back to normal.

The fan did seem iffy to start with (sometimes would not start and would vary in speed depending on physical position of the amp) so I'll change it out for a quieter one that also uses less power.