Why did my amp start on fire?

[soundNERD]

Now that I think about it they were barely touching the heatsinks, so could they really get hot enough to start on fire in about 5 seconds? Wouldn't the self protection kick in?


also, is it normal for the 7812/7912 regs powering the 134 and 833 to get warm? do they need to be heatsunk?

edit: Finally, I tried it, and when turned loud it doesn't blow up, but instead starts making a distorting/popping sound. I have a 470K resistor setting the gain of the crossover and a 10K setting each amp. Is the 470K too high and the 10K too low so the crossover is distorting, overdriving the 134s and then overdriving the 541s? Should I put something more like 50K on the 833 and 150K on each amp?

Or is that distortion/popping mean something is wrong with the amp? I know its not amplifier distortion though because the wave on my scope isn't going flat. Just breaking apart at parts of it.

Thanks!

[johnnyx]

I would test each part seperately. First the crossover, making sure that it does what is intended. Then the power amp with a resistor load. This way you can reduce the number of possibilities.
Have you checked for HF oscillations/instability? I've had oscillations start at one point of the waveform, just the fourth quadrant was affected. I cured it by earthing the chassis. I thought that was very strange. Maybe your problem is similar.

[soundNERD]

about the popping, it was just the chips got too hot and started doing that. I'e had that problem with 3886 amps before. letting it cool fixed it.

Everything works perfect now, hopefully when i crank it nothing blows up!!

but really, is it normal for regulators to get warm?

[Upupa Epops]

Whatabout to use for new thing variac ? Do you heard about this ?

[soundNERD]

I'm sorry, what should I use a variac for?

OK, the first actual listening test was very good! I just can't get over the volume without clipping. I turned it VERY loud, with the entire room rattling, and my scope showed NO distortion at all. Also my eyes showed no smoke or fire . I had to absolutely crank the thing way beyond where it would be usually, and with a running bass line going to get the thing to even begin to distort. Plus, I have a cheap rockford fosgate 12" car sub as the driver. I'm sure it doesn't have too good of efficiancy. If this thing really works as good as I hope it does when connected to a surround reciever, then it may be worth it to buy a good driver with high effifiency. Then my room would really shake.

Does anybody know how these chips handle 2ohm loads? I want to get a 4ohm speaker if I buy another for increased power. Would it take a 4ohm in bridge? I would probably be pushing 300W with that wouldn't I?

[soundNERD]

OK this is really aggrivating me.

The piece of s**t blew up again. This time, the sound stopped comming out unless I turned the volume VERY loud, at which point i would hear a slight sound. So, I tried each channel seperately. The one amp I tried, nothing happened. The other, it made a really loud thumping sound, with the speaker moving a lot, then popped, cracked in half, then the other followed.

So, I've just completely given up on it.

[Upupa Epops]

In this case logical questions are : What is DC voltage on rails ? How is value of fuse ? Is fuse slow or fast ? Do you are sure, that is this reprobox OK ?

[carlosfm]

Let me see if I've got it right, soundNERD...
Are you using the two OPA541 chips in bridge mode and trying to drive a 4 ohm woofer?

[soundNERD]

The voltage is +/-33.5VDC (trafo is 26.2-0-26.2). The 541s can take up to +/-40VDC. I also have a PSU board with four 3300uf 35V and one .1uf cap on each rail. On the amp board are two more of the .1uf caps.

I have no fuse on it.

No, my speaker is an 8ohm, 400W 12" speaker.

I was asking if a properly built amp using these chips can power a 4ohm load in bridge, in case I ever get a different speaker.

Do I need a current limiting resistor on the chips? Right now I have the three pins connected to the speaker, the current limit pin and the two output pins.

whats a reprobox?

[kilowattski]

The problem is that you are asking the opamp to do way more than it is capable of. Your power supply is probably 32 volts per rail and if you are cranking it up all the way you are probably putting about 56 volts across your 4 ohm load. If you do the current calculation you are asking each opamp to deliver 7 amps when they are rated for only 5. You are simply blowing them up because you are exceeding their SOA and internal power dissipation capability. Its not the components fault.