power amp failure, testing methods

[zbyszekskibinski]

1)Does anybody know how to fix power amplifier without soldering out output transistors and testing them with ohmeter? Is there any way to test voltages on the pins of power transistors regardless of them being connected as Darlington or single push-pull? 2) how do you normally check TDA amplifiers by testing voltages? As far as I know, in transistor output stage, if the emiter resistors get hot, and there's 1/2 power supply at the output, this could mean that power transistors are shorted. Sometimes fuses blow up as well. In IC amplifiers, the ICs usually get hot and voltages are wrong while there is no AC signal at the output of the IC chip. Is that right? Do you have any other suggestions, please extend this thread, thanks Paul

[loudthud]

If one or more transistors short on one side, the other side tries like heck to pull the output back to zero due to feedback. The result is usually a blown fuse or the emitter resistor of the shorted transistor opens. So, first check any/all emitter resistors, then you can usually ohm from the output to each rail and the shorted side will be apparent. Don't forget, the driver transistors can also short.

Phase Linear had an interesting tip. Ohm from collector to emitter on each transistor in circuit. The shorted transistor will have the lowest reading and the others will ohm slightly higher by the value of two emitter resistors.

[zbyszekskibinski]

Thanks, this was very helpful. I've always thought that measuring resistance between transistor pins ( when in circuit) may be incorrect unless you solder it out of the circuit, because you could get wrong value in measurements. Would you know what is the rough value of resistance between the output and both rails if the power transistors are good and what it is if they're falty? In case of older amps with coupling (electolitic cap) at the output, is checking ohms between rails and the output a good method as well? I also wonder how to test these TDA ICs, but maybe I should start up a new thread. Thanks anyway.

[loudthud]

There is no typical value. Amps vary and so do meters. But I know it shouldn't be a short. Since there are some large capacitors in the circuit, the reading might show a charging cap, low reading at first that goes higher and higher. Depending on polarity, you might also see one or more diode drops. You are likely to get a different reading on every range. Meters that auto-range may oscillate between ranges. But a shorted transistor typically measures less than one ohm. You may be able to get a vague idea if you have a identical amp to compare against or the working channel of a stereo amp. Repairing a blown power amp is tricky at best. Do the best you can to find the failed components, use a variac or light bulb limiter when you apply power, and if there is a fault you didn't find, keep looking. It's usually best not to have a load connected when you power up the amp. Check for DC on the output before you connect the load.

[Enzo]

To say what loudthud said in different terms, when a transistor is in circuit, yes, there are parallel circuit paths that could confuse your meter, but there is nothing that can make a shorted part read not shorted. If you read a shorted transsistor, pull it out and make sure, but it probably is shorted.

[zbyszekskibinski]

Thank you for your reply. Can you offer some tips how to deal with IC amplfiers as well? I mean, how can you tell that TDA circuit has shorts or it's open. Is measuring voltages on its pinouts enough to diagnose its fault? I suppose it also gets hot when shorted etc. but sometimes external elements fail (like shorted electrolitic capacitors)in the circuit and it might cause IC to malfunction Am I right?

[Enzo]

Usually if the thing blows fuses, then with the power IC disconnected or removed it no longer blows fuses. That right there is enough to get me to replace the IC.

If you find DC on its output and the two power rails to it are OK, then I'd expect it was bad. And unless they are labelled with the same thing (ground, V+, whatever), no two pins should be shorted together.

And lastly, if the voltages are OK to it, and the output is not at DC, touching the input pin should make some sound come out. If not, then it is probably bad.

Oh yes, and check the mute or standby pin for proper level.

And of course soot and charcoal are a bad sign.

[zbyszekskibinski]

Thanks for that. I've also heard about it( what some guys call a 'click test')primarily in tube amps. The way I was checking my amps for sygnal path was just twanging the stings of my guitar and testing (with AC milivolt meter connected via coupling capacitor in series)in different parts of the preamp and power amp at the input and the output of the amp. If there was no signal at the output but there was signal at the input, with correct voltages on pinouts I suspected the IC was bad.
Sometimes I had even more hassle with these IC's. I remember that once I built a kind of stereo amplifier from the kit and in one channel it got very hot and on the other channel it was fine. It didn't seem to pull out to much current from power supply and worked ok as well. I also tested most of the semiconductors in external circuit on the board and decided to replace the IC. Unfortunately what happened was the new IC heated up very easily as well. I don't think there was any idle current adjustment in the circuit cos I would have known about it. I supplied the power IC with bigger radiator as I couldn't find the problem even after I checked everything around...