B&D Smart 12v Battery Charger SMPS Repair

[tango17]

Black & Decker VEC1093DBD automotive battery charger runs at 4 to 40A continuous with short bursts to 100A.

While charging at 40A, it blew the switchmode section.

The PWM control chip is a TL494CN that controls four irf640n n-channel mosfets in an H bridge.  The driver circuit for each mosfet is identical, and includes a two resistors, 6R8 and 1K1, an smt driver transistor marked "1P" and a diode (no marking).

The PWM control drives a transformer with 4 output pairs, each connected to the base of its corresponding driver 1P transistor.

The first failure occurred while 40A charging for a few hours. Two of the mosfets on one side of the H bridge blew, taking out the main FWBR, the main fuse, and a cap at the input filter as well as one 1P driver transistor, a diode and a 6R8 resistor in one of the driver circuits.

I repaired with misc. components I had.  I looked up 1P and decided it was a 2n2222.  I had a thru hole 2n2222a which I installed.  The diode I guessed at and stuck something in from another smps.  I had no irf640n but found an old irf640 and used it

After repair, it worked.  I ran it at 4A to charge a small battery overnight.  That worked.  Then I tried a 20A charge.  It worked for a while, and also at 40A for a while (a few minutes) then I left it charging at 20A.  As I walked away, it made a bang and released the magic smoke.

This time it left major black marks where the 2n2222a repair was installed.  That transistor was disintegrated.  It also took out a second 1P (same side of the H bridge) and a couple resistors in the driver circuit.  I believe the two new irf640's may be damaged, too, but they didn't leave as much black as in the first failure - maybe none.

I'd appreciate comments on where to go now.  The substitute diode is fine.  Was the 2n2222a substitution for the 1P smt a bad idea?  How about the irf640 for irf640n sub?

I'm mostly interested in learning a bit more about smps repair.  Would you suspect anything upstream closer to the pwm control, in my mosfet subs or in the driver circuits?

The mosfets are from a no name source.

Would it make sense to make the same repair again and then check out timing with a scope?  Could a cap cause an issue at the tl494? (The charger is prob about 10 years old or more).

Thanks for any comments or suggestions.