Bad Car Amps Gone Worse...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I hate it when people bring me cheap and nasty amps for repair, but I think this one took the cake... Brand new out of the box connect power and woooof flames. The amp was pasted full of QC and QA stickers, but they didn't notice the one mosfet's gate leg shorted to drain.

I'd like to see anyone elses amps gone BAD.
 
pic1
 

Attachments

  • dsc013351.jpg
    dsc013351.jpg
    44.8 KB · Views: 608
I didn't take a picture, as it happened accidentally, but in one of my Planet Audio amps I, while collecting data to build my own SMPS, touched the PWM chips' MOSFET lines to voltage with my meter probe.

A lot of flames shot out, lots of smoke, and it was quite a sight!

Repairing that myself was a good learning experience.:blush:
 
I Am An Idiot said:
I think these 4 transistors were a close to the upper end of their thermal limit.

This can't be real. Internal transistor bond wires or the own leads would have got fused before so much damage was done. I've repaired plenty of 12V push-pull SMPS with burned MOSFET and damage never went so far. It looks like someone having a boring day found a blowtorch...
 
This one was real. He was joking about the transistors being at their upper limit.

If I'm not mistaken, it acts like an arc welder. The transistor opens and an arc bridges the gap until the gap gets too large. Then the arc extinguishes itself. When conditions are right, it can create a mess like this. This is the worst I've ever seen but it happens to a lesser degree relatively often.
 
I've seen a few like this, think this one an audiobahn that never had full PS output. IIRC the FET/PS was fine.
An externally hosted image should be here but it was not working when we last tested it.


I burned myself on this one, bought and was going to fix. Cleaned it up and had this. It sat there for too long; I was too busy and I bailed on it. Planet VX2200d. Someone got it for next to nothing.
An externally hosted image should be here but it was not working when we last tested it.
 
Perry Babin said:
This one was real. He was joking about the transistors being at their upper limit.

If I'm not mistaken, it acts like an arc welder. The transistor opens and an arc bridges the gap until the gap gets too large. Then the arc extinguishes itself. When conditions are right, it can create a mess like this. This is the worst I've ever seen but it happens to a lesser degree relatively often.


I'm doing 400V-600V high power class-D development, where arcing is a real problem once air gets ionized on narrow gaps on (uncoated prototype) PCB and between transistor legs, and I have never ever managed to get so much destruction. I have got plenty of rotten PCB due to arcing but nothing else.

The following pictures show the effects of *real* arcing on PCB. The copper gets vapourized until the arc extinguishes, but this takes only a fraction of a second. There is no way to produce enough heat to melt a piece of aluminium. The expansion of copper when it gets vapourized and the ionization of air result in a very loud bang but nothing else :D

T_SINE63.jpg


Q_SINE7.jpg


Ironically, no components failed on the first picture, arcing started due to abnormal 3KV spikes on a flyback/coupled inductor.

On the second picture two IGBTs failed resulting in a dead short across rectified 230V mains (with big capacitor bank). Not much destruction either, 15A mains fuse blew immediately.
 
When I first saw a melted aluminum clamp, I was impressed. After the 20 or so I have seen over the years, it was not such a big deal anymore. But when I saw the clamps I had posted earlier, I was impressed again. Trust me there was nobody being bored, and there was no torch involved. Unless of course the customer torched his own amp, which I highly doubt, he almost cried when I showed him what had happened. The following pic is not of the amp I am referring to
 

Attachments

  • cimg2833.jpg
    cimg2833.jpg
    42.3 KB · Views: 343
I Am An Idiot said:
This was a Planet 1400 Looks identical to your damage. I have seen 3 or 4 with this same problem. I think that the capacitors leaked there. I have no idea what caused the fire. Outputs are fine.
Looked close to that, Perry says he has seen them. The outputs were blown (those 2) and board was mostly gone from where they attach over to the three resistor 'hoops'. I didn't have to remove more than 1/8" to 1/4" at most from original hole to get fresh board. Caps tested good and no damage really, but a couple were bulged a little. That is a clone amp I think. It may have arced from under cap to something, but under cap was ok condition. There was no glue on caps there, I've seen others that were glued to seal cap to board there. Hole was maybe a thumb print before, there was a small part of the edge left I cut off, from outputs to edge. It was stinky. There is also a larger diode near the three resistors that was really fried, looked suspicious.
 
They insist in not including features such as under-voltage shutdown with hysteresis, peak-current limiting and a regulated 15V-18V gate drive supply in car amplifiers. That's why these power supplies blow so badly. When one MOSFET fails the PSU does not shut down, current consumption increases dramatically, input voltage drops, the gates end up fed with 8V and all the MOSFET fail.

Also, in most cases, when one output transistor fails, the PSU continues working and everything blows too. Note that most amplifier overcurrent protections only detect excess current in one direction, shutdown is delayed and nothing is done to limit speaker current actively. In contrast, other amplifiers can drive a dead short without problem (I design my class D stuff that way).

They don't include rail fuses between the power supply and the amplifier circuit either, if one channel fails, the amplifier is no longer useable.
 
Eva said:
They insist in not including features such as under-voltage shutdown with hysteresis, peak-current limiting and a regulated 15V-18V gate drive supply in car amplifiers. That's why these power supplies blow so badly. When one MOSFET fails the PSU does not shut down, current consumption increases dramatically, input voltage drops, the gates end up fed with 8V and all the MOSFET fail.

Also, in most cases, when one output transistor fails, the PSU continues working and everything blows too. Note that most amplifier overcurrent protections only detect excess current in one direction, shutdown is delayed and nothing is done to limit speaker current actively. In contrast, other amplifiers can drive a dead short without problem (I design my class D stuff that way).

They don't include rail fuses between the power supply and the amplifier circuit either, if one channel fails, the amplifier is no longer useable.


if they don't design amps that blow catastrophically, they will someday run out of jobs because everyone using their amps don't need to replace them. :clown:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.