Car amplifier board short?

[jheatac]

My standard disclaimer - I am very much a novice.

Looking at a 12v audio amplifier. I know no history. Fuses were missing so I assumed it was blowing them. I found a shorted power mosfet and burnt trace leading to it.

I pulled it off the board but now I have another issue. There is something that is changing with a slight twist of the board. I believe I can hear the slight arcing but cannot find where it is.

I have it connected to the bench power supply and limit current. At 14v it draws about 0.65a at idle. With a slight twist of the board on the PSU side, it will draw up to the 5a allowed and drop to about 8v from current limiting.

This only happens with the unit switched on. it will not do it with the SMPS off. That should rule out everything before the mosfets, I would assume.
I have disconnected the power supply from the amp section, so I know it is in the power supply.
When it is drawing excessive current, the bank of mosfets where the blown one was gets hot, the other bank does not.

It almost acts as if the toroidal transformer may have a winding rubbing? Does this happen? I can twist is slightly and create/resolve the problem. I don't see any sign of burnt varnish. The glue is intact.

I have went through and reflowed a lot of the joints, but i'm thinking that a broken joint would open the circuit and not short it.

The scope pictures are normal operation and failed operation. Keep in mind the failed operation is current limited so it is operating at about 8v, not 14v. This is at the output of the power supply and the rectifier. 

Being I still show a waveform on both channels, it would sort of rule out the transformer wouldn't it? If one wind was shorted, I wouldn't expect to see a waveform.

Can anyone point this novice in a direction? Thank you.

[pienari]

Parts missing from the board but nothing important what i can see .
Maybe the transformer short, iv seen few.

[jheatac]

Parts missing from the board but nothing important what i can see .

I pulled the missing components. I originally thought one of the capacitors could have been shorting with a slight twist of the board. Not the case. They all checked fine. Q52 was shorted. I have a replacement coming, but that may have been premature if the transformer is the problem.

[jheatac]

Would it cause any damage if I removed the transformer and powered on?

With my limited knowledge, I'd think it would max out the switching frequency, but with no load, it should cause any damage. Is this correct?

[jheatac]

Okay, so I pulled COP5 (The large board jumper from the mosfet bank that is a man down, and started it up. This was easier than removing the transformer. I could twist and pry and it made no difference.

Would logic suggest that it is the transformer if I disconnect the output of the transformer and I still have a problem but if I disconnect input side of the transformer on that half, I no longer have an issue?

When watching the scope, the disconnected bank did not pulse, but the connected bank did. It appears they turned on, but with no load, they never shut off, therefore no pulse???

[jheatac]

I pulled the transformer and believe it was the issue. While it was out, I checked it with an ohm meter and megohm meter and nothing points to a problem. Megohm meter at 1000v showed no connection between any of the windings, which I believe it maxed out at 5.5Gohm

After putting it back in the board everything works just fine and I can flex the board without an issue.