Turning on power amplifiers after repair.

[Diver300]

Well I've had a few occasions where I have repaired a big power amplifier, then when I've turned on the power, there have been bangs and flashes, and more magic smoke.

A lot of these power amplifiers have multiple power supplies, which often have to be floating, so it is quite difficult to come up with several power supplies to run the amplifier from instead of its mains transformers.

Has anyone got any recommendations for how to limit the damage if it turns out that something has been missed?

 

[canadaelk]

On many power amps you can safely remove the output transistors and drive , at low level, the load using the output drivers. Than measure voltages to make sure they are within specs. E

 

[tunedwolf]

A couple of big NTC's on the supply rails will soak up the overload gently, unfortunately semiconductors are often damaged long before that happens. Polyfuses and thermal breakers are just not fast enough either. That kind of leaves you with a purpose designed adjustable current limiter. I would suggest a light bulb in series with the supply rails. If the bulb lights up brightly, then you still have a short somewhere...Not bulletproof, but better the lamp dissipates the heat rather than the semiconductors

rgds

 

[OlPhart]

Howdy, I'm a Big Fan of power resistors ~5-10 ohm in line with the output power feeds along with dummy ~10 ohm loads. Power up & pay attention to overcurrents & waveforms. Good Hunting.. this isn't trivial..... <<<)))

 

[BeeBop]

I would suggest a light bulb in series with the supply rails. If the bulb lights up brightly, then you still have a short somewhere...Not bulletproof, but better the lamp dissipates the heat rather than the semiconductors

rgds

I agree with this; it works fairly well!

 

[lemofus]

I would suggest a light bulb in series with the supply rails. If the bulb lights up brightly, then you still have a short somewhere...Not bulletproof, but better the lamp dissipates the heat rather than the semiconductors

I'm now also a big fan of this method; was reluctant at first, messing with wires carrying live current, but now that it's setup it's great; very elegant way of buffering and showing a clear sign of something being very wrong in the circuit!

 

[KeepItSimpleStupid]

Generally a variable transformer with a current meter works. Also keeping an eye on the bias voltage between the bases of the outputs or the bias regulator is always good too.

It's also better to have a signal applied, but no load and to have a scope on the output. Attach your scope prior to the protection relay.

 

[Tann-44]

Generally a variable transformer with a current meter works. Also keeping an eye on the bias voltage between the bases of the outputs or the bias regulator is always good too.

It's also better to have a signal applied, but no load and to have a scope on the output. Attach your scope prior to the protection relay.

Would agree there...

 

[Diver300]

There are two problems with variable transformers. One is that the size needed to power some of the amplifiers would cost a lot. The other is that some amplifiers would not turn on the outputs until the voltage is good. If there is then a short, having a 180 V supply, and having all the capacitors charged to 75% of their maximum means that the power transistors will still blow.

 

[KeepItSimpleStupid]

Diver:

All amps can have issues with troubleshooting especially if they are part of a preamp. You almost always have to look at the signal before the protect relay and insert it at the correct spot.

It's not necessary to have the variable transformer rated for full output. Only those during real performance testing that require the line voltage to be set to some value exactly.

I've rigged my homemade variable transformer/isolation transformer differently. The autotransfomer is plugged into an external 1000 W isolation transformer. I picked it up for $10 many years ago.

The autotransformer is a 3 Amp unit. The power switch is rigged to be direct, off or isolated.

It does have an analog ammeter. A voltmeter would be nice. I'd prefer a digital voltmeter and an analog ammeter though. A big pilot lamp would also be useful, but I don't have.

My outputs are a 3 prong standard NEMA out and a binding post.

If I did it again, I would add stuff to make leakage measurements easier or make it built-in. I might add dual voltage capability.

I would consider an electronic fuse as more important, though.

 

[RichTheDude]

TuneWolf's suggestions are good to do on the cheap, I think it's known as the dim bulb method. Start with low power bulbs and work your way up, I would add a analogue ammeter inline too. If the bulb suddenly gets bright, you know your pulling a lot of current.

I got a variac second hand quite cheap (and an isolation transformer) they are worth their money

 

[4pyros]

I was using 300 watt light bulbs for testing power amps. Once you get it fired up at low levels then hook it up straight.

 

[unclejed613]

a dim-bulb tester will save a lot of burned components. another device i use quite regularly is replacing the output transistors with 1500 ohm/2W resistors (from B to E), which allows you to check everything with power and signal (but no load). doing this closes the feedback loop (assuming the output emitter resistors are ok) and you can troubleshoot the amp with an oscope, as well as check the bias circuit operation.