Thermal runaway with a difference

[Seth Wolpert]

Hello all. I've got a Harman-Kardon Signature series 2.1 5 x 100W power amplifier that powers on and runs for about 30 seconds and then pops the 5-A power fuses on channel 5. I unplugged the transformer lead to that channel, and channels 1 through 3, and the same problem occurred on channel 4, even without speakers connected. There are six output transistors on each channel and none were shorted.

So, my question is what would cause a thermal condition that does not appear to be bias current related? In disassembling the amplifier (there must be 100 screws holding this thing together), I noticed there was heat sink grease between the output transistors and the circuit board, but none between the transistors and the rather massive heat sink itself. It is entirely possible that someone worked on this amplifier before I did, and neglected to apply thermal grease to the heat sink side.

Many thanks for any advice or speculation. I'm new to this forum, but very experienced with electronics.

Seth

 

[chazix]

Welcome to AK!

So, my question is what would cause a thermal condition that does not appear to be bias current related?

Oscillation, maybe?

Cheers,

chazix

 

[Seth Wolpert]

I didn't see any oscillations on the positive or negative DC supplies or the signal inputs on the bases of the output transistors.
It does appear that someone else worked on this amplifier, and may have upset the bias settings. That's my next objective.

 

[triode17]

It might be the driver transistors pushing too much current through the output transistors. If it were bias, that would cause excessive crossover distortion or reaching saturation too fast. To quickly blow a 5a fast fuse, it would take about ten amps. Divided by 3 pairs per channel. would be 1.66A /pair and i doubt you could bias it that high to begin with.

 

[restorer-john]

I would definitely look to the bias circuitry and pot. It sounds like the upper and lower drivers and outputs are getting hard switched on, shorting the +/- supplies and blowing the fuse as it happens with no load to ground.

The output transistors should have thermal compound between them and the heatsink- no doubt about that.

 

[Seth Wolpert]

Thanks, John. I'll have a look at this.

Seth

 

[Mrv8q]

Do you have the matching HK Sig 2.0 PrePro?

 

[Seth Wolpert]

Yes, the amplifier was received with the matching preamp.

Just an update on the repair. I reassembled the main board with heat sink compound, and the thermal runaway
issue has cleared up. Now, channel 4 doesn't pass power-on self test. The other four channels are working.