SVT4 PRO one side out

[rf7]

I've got this amp on the bench, and Power Amp B works fine, but Power Amp A has no output. The preamp tests fine. There is nothing visibly burned on the PCB that I can see.

I signal traced it to the shared 5532 opamp IC1. The output for Amp A (pin 1) is sitting close to the +15V rail, and has no signal on it. I actually lifted the feedback from the output circuit and also lifted the connection to the following Q102 thru Q105 circuit, but that just moved the output to the negative rail.

I have socketed and replaced that opamp, but nothing changed.

Any ideas?

Thanks in advance!

[Enzo]

Never forget that the solid state amp is massively fed back on itself, it is a loop. The op amp controls what voltage that whole Q102-105 group sits at, just as they contol what the output stage does. The op amp then sees a sample of the output and in its role as a differential amp, it "corrects" what goes through the amp to make the output agree with the input. If you disconnect that string of transistors, then the output can;t move and no feedback exists (which you also disconnected) so the poor thing has nothing to correct with. It is like trying to stretch a rubber band from one end.

With the full circuit connected, what conditions do you find? Does it blow fuses? You already said no signal passes, right? This amp has a speaker relay, so does that kick on? Look on the output bus - the line all those 0.47 ohm 5W resistors connect to. Is your signal THERE? If the amp works but the relay won;t come on, or has bad contacts, then no output occurs. Oh wait, nothing coming put the 5532, means nothing coming out the speaker, sorry. Still, if the op amp has 15v on the out pin, I would expect the output bus to have DC as well. Does it? If the op amp cannot control the output of the amp, then it will slap over to one rail trying to "correct" this.

The power transistors on the output stage are source followers. Just like emitter followers or even cathode followers, those sources conneted to the output bus "follow" their gates. The gates are driven by A_HI and A_LO drive signals from page 1. Those are also TP9 TP10. They are about 7vDC apart, but they track each other with the signal. If they move more positive, so does the output - it follows. Back to page 1. Q106 107 maintain that 7v spread between the drives. They are your bias. And that brings us back to Q103, 104, who take the signal from the op amp and control the drives with it. They depend upon TP6,TP7 being there. You got +5 and -5 at those? Without those 5v references at their bases, those two transistors have nothing to push against.

If your op amp is over against +15, it leads me to wonder if it isw not able to move the output. If the output sits a couple volts negative, say, then the op amp will move to positive to try and move it back to zero, where it ought to rest. But if it cannot control the output, it will try harder and harder to push that voltage back until it reaches its power supply voltage limit. And while it is doing all that, it has no ability to make music too.

[rf7]

Wow, there's +80Vdc on the output bus, so I've got shorted output devices. Thanks Enzo!

[Enzo]

And that is what your op amp is trying to correct.


You may have shorted output devices, OR they may be doing what they are told. Look at those drive signals, TP9,10. Are they also at 80v? If they are, then the output will follow them there.

STretch a rubber band in front of your face. Now let go one end. The rubber band all snaps over to your other fingers. That is what happens when a power supply rail is missing, everything tends to swing over to the other supply. On page 1, look at that whole totem pole of Q102-105 stretched between +80 and -80. (I assume your rails are 80v) Also note they have a separate pair of 80v rails, not the same ones as the output transistors use. Now what if the -80 disappears from Q105? Or if Q105 goes open? Then like that rubber band, the whole thing will snap up to +80. And the output will follow. SO Q105 might be at issue, or maybe the TP23 power rail on page 3. But check that out.

You may indeed just simply have a shorted output device, but don't assume so. Those other things are not unlikley either.

[hodgy]

When you find that your MOSFETS are done for, also look at the .47 ohm resistors (the big white guys), and your 47 ohm resistors that are on the Gate of the MOSFET. Those all seem to go out together, in my experience. There's also a good chance that you lost one or more of the following: Q202, Q205, D210, D213.
Good luck! These can be time-consuming beasts.
I almost forgot: look at the 1.5k resistors also.

[rf7]

So 9 out of 10 of the output devices are shorted, and 7 of the 5W 0.47 ohm resistors are open, along with most of the 47 ohm resistors too. I have not checked the driver transistor stuff yet, but I will.

[rf7]

The customer decided to not repair Amp A, since Amp B still works.

Thanks for the help guys!

[Enzo]

Yep, that is always an option. I'd remove ALL the bad channel outputs, even the one that seemed OK.

[rf7]

Wow, really? He's been running it like this for months. You'd think that something would blow, with shorted outputs in it, though. Guess I gotta open it up again. I know now not to unhook all those connectors to get that PCB out. That's a good thing.

[lowell]

just finished fixing one of these BEASTS. PA B was fried.. and I mean fried! Burnt pcb and exploded components. These are a bitch to take apart and reassemble. If you want, for ease, you could snip out those ballast resistors in the pa. You should be able to get at em without taking that monster board with the monster heatsink all apart. There's also a possibility that the fried MOSFETS are leaking the rail voltage back to the biasing circuit, which I believe has some top-of-pcb wire jumpers running to those MPS transistors... and those wires could be snipped too. It's kinda butchering the thing, but he doesn't wanna pay. All this said I guess it's almost as easy or just as to clip all MOSFET leads and just leave em in.