Ampeg SVT 450 power amp bias

[Nyqusit5]

Hi all,

I have an SVT 450 in for repair. All o/p transistors have bmlown as well as their ballast resistors and the 47Ohm at their gates. D18 (13V Zener) has also been replaced. The amp outputs a sine wave now but I can't set the bias according to the schematic the pot is ok): it only goes from 4.5 to 5.6V accross the 47Ohm/5w resistors. Q3 was also burnt and I have replaced Q5, Q6 and Q8 so far.

Also, with no load, the output appears "doubled" on the scope, as if it was modulated by a very high frequency (a bit like with a class D amp). i doubt this is normal. With a load, there is no such modulation.

Any ideas as to what could still be at fault here?


ampeg SVT450.pdf

[Enzo]

As to your doubled waveform, try adjusting the sweep rate on your scope. Without a photo, your descriptions sounds like the scope is just losing sync.

Perhaps you could be more specific, and use part numbers. I don't see any 47 ohm 5 watt resistors, did you mean 0.47 ohm, as in R37,39,41, etc? And you surely do not want 4-5v across any of those, the schematic calls for 25MILLIvolts. 5v there would represent 10 amps flowing.

On the left is R77, do you have the 11 volts across it? (the 5.5v pos and neg at each end?)

[Nyqusit5]

Sorry, I guess I wrote my message too quickly.
Yes, I meant R37,39,41, etc .47Ohm resistors. And I also meant 0.005V or 5mV accross them.

Here are pics of the output: the first one shows the output wave WITH a load, the second WITHOUT a load:



Thanks for your help!

[Enzo]

OK< the amp is unstable under load. What you have is a high frequency oscillation coming from the amp, and then the amp moves it around with the test signal. Even your unloaded waveform has a crossover hitch. I guess one could debate which is modulating which, but your high frequency should resolve if you speed the sweep rate way up. The sine wave will stretch way out and the high freq band in the middle will resolve into a little sine wave of its own.

I am still wondering about R77 voltages, and the gates of the output FETs should have a similar spacing. Your bias trim should affect that voltage spread, does it?

You have the four 13v zeners, are you sure only one of them was damaged? The current that burnt out D18 had to flow through D17 at the same time. I think if one of them were bad, I'd replace all four - they cost just a few cents each. FOur of them for under a dollar.

[Nyqusit5]

I replaced R77 even though it tested fine. I also replaced the zeners but there is no change to the amp's behaviour.
I also should point out that R24 was replaced as it had overheated. I am waiting for 2n3440 so I can replace Q13: it tests fine but may be a problem?

[Enzo]

Note that I did not ask if R77 was OK, I asked if the voltage across it matched the 11 volts the schematic calls for.

[Nyqusit5]

I am decidedly not focused enough. Sorry about that, too much stuff on my hands at the moment. I'll check that.

[Nyqusit5]

There is 10.49V accross R77.

[Enzo]

OK, close enough there. So is the +5 and -5 getting to the gate pins of the outputs? And does the bias trim change that?

[Nyqusit5]

Well I have 1.7 to 1.97, and -1.91 to 1.67 on the gates depending on the bias setting.

[MarkusBass]

I think that Q4 and Q13 are the most suspected and I wonder why you haven't check voltages on them yet. Can you provide B-E and E-C voltages?