s/f princeton fries speakers wtf?

[Alex R]

ok same s/f non-reverb Princeton as on the other thread but different issue. The guy had just fitted a new speaker but it had a buzz. Rubbing voice coil. On the bench at some point (not sure when/how) the thing fried my 150w test speaker, which now has serious voice coil rub. I ran it full output into dummy load for half an hour and observed no DC showing up on the output and just a regular steady AC signal wave.

Could it be the OT? Leaking DC across to the secondaries? The NFB loop maybe somehow? Had a look around but observed nothing in the wiring. I put a little 10K pot in the NFB loop and that's ok. No high-freq transients on the scope that I could see. Come across anything like this before? Any ideas?

[Enzo]

DC? Not likely. That would blow fuses pronto. I'd be looking for HF oscillation. Might only happen under certain circumstances, so it doesn;t do it for you.

[Don Moose]

If it cooks a 150W speaker without DC, wouldn't you expect some serious redplating?

Oh! - DC leakage that only shows up in a HiPot situation? Here's the scenario I have in my unkempt little mind:

Big high-frequency oscillation that's high enough in frequency that, into the speaker's inductance, it's effectively running unloaded - a load impedance so high you get the dimed-into-no-load effect (and damage) even with a load connected.

Plausible?

(I know, make sure there are no horses before looking for zebras.)

[Alex R]

Well I guess it might be the same problem as on the other thread but although that odd-looking bass note distortion is still there I tamed it by putting a 30Hz rolloff in with smaller coupling caps. I think I decided it was a harmonic or something. I am good at persuading myself of these things.

It fried two speakers - one, the original, before I had the amp, and the other, mine, after I made many changes to it. I may have made it worse because my speaker was bigger and tougher and got more comprehensively fried. I put more gain in the amp (it sounds lovely btw, dammit!).

Yep DC leaking would blow fuses, HF might come in bursts under odd conditions. So I guess a 47pf 1000+v across the power amp somewhere, then see if it fries any more speakers?! Where would you put it?

[olddawg]

HF might come in bursts under odd conditions. So I guess a 47pf 1000+v across the power amp somewhere, then see if it fries any more speakers?! Where would you put it?

Log sweep generator, dummy load, amp kranked, scope the output, heat gun or blanket over the amp. You will see the HF oscillation eventually. Isolate with freeze mist.

[Alex R]

well i punished the thing as suggested into the dummy load and looked for HF but found nothing. It is possible that the changes i have made to the amp stopped it, and the damage to my test speaker was done before i put them in - i was sound-testing as i went.

But I would like to put in a HF filter on the power amp somewhere to make sure before i give the amp back... What would you suggest?

[Mars Amp Repair]

Well,
the infamous hf snubber cap that Marshall uses in their TSL is a 22pf @ 1KV that goes between the plate and the screen of one of the output tubes.

The screen acts like a ground to the HF signal, but is at the same potential as the plate allowing the cap to not have to endure a constant 500VDC across it (or at least that is what it seems like to me).

Of course if the oscillation is being created at some other point in the amp & is feeding the power section the cap will be under quite a bit of diress constantly trying to stave off the oscillation. I doubt that is the case as most HF oscillation occurs between the highly amplified output & high gain input.

glen

[Alex R]

- thanks glen, I'll bung one on just to make myself feel better. It's not a permanent feature of the amp, this oscillation, in fact I couldn't see it at all under test.