Go Down

Topic: Amplifiers up in smoke (Read 1 time) previous topic - next topic

pvercello

Jun 14, 2009, 09:34 pm Last Edit: Jun 14, 2009, 09:35 pm by pvercello Reason: 1
So one day while working on a Waveshield project a couple weeks ago, we had a big wind storm here, and the power went out.  When it came back on, I discovered that two amplifiers have been fried, with both the same symptoms.   One in my stereo, and one in a nice studio monitor powered speaker.   They both do the same thing:  blast incredibly loud 60Hz hum to the speakers, regardless of what goes into them (or what the volume's set at).   The powered speaker has a fuse that blows after a short duration of this, while the stereo fries two little transistors which go up in smoke and explode after a short time.  After either of these happen, the noise to the speakers stops; replacing them repeats the process.

I was wondering if anyone's debugged something like this before?   The symptoms are very similar between the two, and other aspects of both keep working (the digital parts of the stereo appear fine: cd player, digital tuner, etc all look like they work; and the indicator lights on the monitor speaker all keep working when that fuse blows).    It seems like a similar component or components blew in both cases (and indeed, the amplifier circuits look similar:  two big audio amplifier chips on huge heatsinks, big rectifier bridge chip, big filter capacitors, toroidal transformer, etc.).    Any ideas where to start?

Thanks for any suggestions anyone might have!

-Paul

retrolefty

#1
Jun 14, 2009, 11:09 pm Last Edit: Jun 14, 2009, 11:12 pm by retrolefty Reason: 1
The bridge rectifiers and/or the main filter caps would be the leading
suspects in my mind.

If you have a DVM you could put it in AC volts function and measure the raw filtered DC voltage across the main filter caps, if you see more then a volt or two of AC voltage you know you probably have main rectifier/filter problems.

Lefty

shelleycat

Google for "audio amplifier C class", and look at the Wikipedia article.  I have no idea exactly what type of amplifier it/they are, and the last time I worked on an (B&O) amplifier was over 27 years ago, however....

The usual way of running the output was as shown in the wikipedia diagram, two big transistors as the final drive, one npn, one pnp.  They may be darlington pairs.  Nowadays, they might be MOSFETs.

If one of these blows short circuit, you get power output into your speakers.    Furthermore, it is common to have a time delay when the amplifier is turned on, to avoid a sudden power surge - if the power went out for a short period of time, perhaps this did not reset and so you did get a transient that burnt out the output stages.

If the drive circuits blow, you could get both the + and - side turned on together - that is not good.

So,
1   Check the power is OK - no ripple on the DC.  Don't short it!
(A possible tester is an crystal earphone with a 100K resistor in series with it)  A toroidal transformer sounds like a high current DC supply.

2  Check the drive part is working - locate the drive audio signal and see if that is intact (earphone again)

3.  look at the output devices and see if they have a serial number, then search for that.  That might show you what you are dealing with.  

4  See if any of the output devices are running hot.  If one if hot and the other is not (assuming two devices per channel here), then the hot one is probably blown.

5  Disconnect the output devices, and test them (of course, you have to know what they are, but a simple analogue device can be tested fairly easily - remember to put resistors in circuit when testing!)

3.  If its just an output device, any similar power transistor will probably do.  The quality comes from the feedback.

WARNING:  If you do not understand power supplies and where the mains input is, do not attempt live testing.

pvercello

#3
Jun 15, 2009, 04:37 am Last Edit: Jun 15, 2009, 04:50 am by pvercello Reason: 1
These are excellent suggestions; it's much appreciated.   Yeah, I've also got an oscilloscope I can put on it, in addition to the voltmeter with AC, etc.   For the studio monitor, I've got one good one still, so I can compare along the signal chain between the two.    

While I've had a few EE classes, I don't really have much practical knowledge like "what parts are mostly likely to fry from a big transient?"

FWIW, the amplifier chip in the speakers is one of these:
http://www.national.com/ds/LM/LM3886.pdf   (Says ESD Susceptibility = 3000V, not sure if that's relevant).
I can't quite read the part number in the stereo, but it's got the same form factor as the speaker's amp chip.

They both use the same rectifier chip.

I'll give these suggestions a go.

Cheers,
Paul

Go Up