AMP11 blown up. Any ideas?

[MickB]

No joy yet.

None of the caps are shorting. Without the STA fitted, the resistance from the output pads of the STA to the speaker terminals measured about 13k. I fitted the new STA but there's no sound. It's drawing about 50mA and the STA is getting slightly warm which all seems to be as it should. I jumpered the input pins to check the DC offset on the speaker terminals but it's measuring zero even when I adjust the pot.

Is it likely to be the TC2001? Or something else?

EDIT: I've got a another (working) AMP11 here and it draws about 110mA when it's not playing, so presumably the fact that the damaged one is only drawing 50mA is significant.

[krilli]

Hmmmm. On the working one, is it 110mA when unmuted and silent? If so, then how is muted current draw?

[MickB]

110mA is unmuted, and with a source connected but not playing. It's the same with the input pins jumpered. On the working one I've hardwired the mute off so I'd have to cut the link to test it muted. Not a problem if it's useful but I'd rather not if I don't have to.

[krilli]

Nah, don't cut. I'm no expert at this, just trying to figure out if the broken one is effectively "stuck in mute", you know? The 50mA figure seems like the figure they're supposed to draw muted. V-Bro, what do you say?

If it's "stuck in mute", well, my hunch is that the TC2000 might be the problem. Could be something else though, but we can't exclude the TC2000 yet either I guess?

[MickB]

TC2001 ordered as it seems like the most likely thing. No signs of damage to anything else. Will report back when it's fitted.

[MickB]

Been doing some investigating and according to my logic I think an external short caused by metal debris is unlikely. Of course my logic may not be sound

Legs 5 and 6 of the STA505 are the ones which got vaporised and they are GND2A and GND2B. They are soldered to the same pad so they are at the same potential (ground), so any short must have been between them and something at the supply voltage or whatever. The current was high enough to make both legs vanish instantly and fracture the chip, so whatever provided the other end of the short should show some signs of damage too, but there's nothing visible on the board. Also the voltage source must have been at the IC end of the legs and have been applied to both legs simultaneously as the entire legs have gone all the way into the chip package. The IC has built in protection against over-current and ever-temp in normal operation so I'm assuming that it wasn't a low-current short on the input side which was amplified to high current at the outputs, and surely that wouldn't affect just the grounds anyway. That leaves a failure of the IC which resulted in an internal short.

Does that seem like sound reasoning, or have I missed anything? I admit to a high level of ignorance. It's just curiosity on my part really as I'd like to have a better understanding of all this stuff.

[V-bro]

I have also seen this kind of damage from reversed polarity and then indeed the TC is also toast. Even if the 5v holds up parts usually fry two ways, either the chip becomes solid metal internally and conducts fully or it breaks contact internally and then all applied voltages will seem OK.

After I am back from my business trip I can acces my files of the AMP11 again and investigate further what exactly could have caused this to happen. The piece of metal debris was just a hunch, but I definitely would not exclude the possibility yet. As you already mention the chip is very well protected and when things blow like that it is most often caused by something really unusual or unexpected happening externally...

[MickB]

Oh don't waste any time on it V-bro, it's just curiosity on my part! I'm going to put a cover over the amp anyway to protect it from debris, I should have done that at the beginning. As you said, with a portable device a short can easily happen whether it was the cause this time or not. And this box is particularly susceptible as it's being shaken around a lot and several times has had people fall on top of it and slide along the road while skating with it. It's what it was designed for and I'm very pleased with the way it's stood up to the abuse. This is the first failure after 18 months of heavy use. The case gets scuffed up, but there's no structural damage and as it's solidly braced plywood it just needs some filler and some paint to make it good as new.

[MickB]

I've replaced the TCA2002 but it's not looking good. When I first connected the power the amp was drawing 140mA and everything looked fine. I jumpered the inputs to adjust the DC offset at the speaker outputs and that behaved very erratically. It read about 25mA initially and I was adjusting it down with the pot but then it suddenly shot up to more than 1 V and then dropped down again. After a bit of jumping a round it dropped to zero and I noticed the current draw of the amp was down to 80mA. It's now behaving the same as before I replaced the TCA2002 which suggests to me that the TCA chip has blown again.

I've checked thoroughly for solder bridges and any other defects but everything looks fine. Time to cut my losses and replace the whole amplifier I think.

EDIT: Actually it's not quite the same as before as it now draws 80mA instead of 50mA.

[MickB]

It's happened again! I ended up replacing the entire amplifier and it worked a treat for 2 street skates, then at the start of the 3rd one it went quiet again. I've just had a look and the same 2 legs have been blown off the TP2050 IC, plus another leg on one of the other output ground connections. When I reassembled it last time I hoovered out the box and covered the amp so there was virtually no chance of a short caused by metal debris.

Could this be caused simply by too high an input signal? In which case I need to limit it with diodes.