400a with hot emitter resistor

j beede

Member
2007-12-24 5:04 am
NorCal
After re-biasing I was randomly probing easily accessible points with my IR thermometer, I was surprised by the variation in temperature across the 0.68 Ohm PNP emitter resistors. 20° difference between the coolest and warmest. I lifted a few of the resistors, they all measured ~1.1 Ohms on my DMM's 200 Ohm scale. Close enough.

I found 600mV dc across one of the 0.68 emitter resistors on one of the PNPs whose collector is at the rail. This is the hot one. I pulled that PNP and it passes a basic back-to-back diode check. Its forward resistance measured a little low (13 Ohms v 17 for the others).

With the amp idling, nothing connected, what voltage would you expect to see at the collector of FT317? FT417? These are the intermediate nodes in the output NPN and PNP stacks.

Is it safe to lift the hot 0.68 emitter resistor to temporarily open that dc path and see what it does to the voltage at the intermediate node?

...j
 
Even matched devices will vary somewhat. As a consequence, the current will vary from device to device and the resistors will heat differently. As to how to relate that to the actual measured temperature, I couldn't say. It would depend on the design of the resistor.
One thing that comes to mind is that those resistors may be carbon composition resistors. A lot of '70s era equipment used them. The problem with composition resistors is that they don't age well. The resistance varies, usually rising. You might measure the voltage across them and then measure the actual resistance of each resistor to see if they've changed enough to be a problem.

Grey
 

j beede

Member
2007-12-24 5:04 am
NorCal
Yes, this amp uses 10% carbon emitter resistors--crude. Cleaning the boards after soldering does not appear to have been standard practice in those days as the joints have plenty of hard brown flux "decoration".

I have done a bit more debug... There is a 600mV across the "hot" resistor with the amp idling. 600mV and .68 Ohms sounds like almost an amp of emitter current at idle for this one PNP. This design uses complemetary series stacked NPN and PNP for output drivers. The intermediate node in the stacks should be around half the rail at idle?
...j