Power Caps installed Backwards - Setton RS-660

[Ken770]

Repairing a Setton RS-660 that had new 15000 uf power supply caps installed wrong. The 2 negatives were strapped together and the two positives went to + 53vdc Pink wire and -53vdc white wire. This damaged one expensive cap. I reinstalled the original caps. I connected the receiver to a variac and slowly brought up the voltage 50 vac and the the left heat sinks to overheat almost immediately. The voltage at the caps was +/- 13vdc. The right amp board remaind cold. I disconnected the white wire at pin 3 of the left amp board and the negative and positive rail voltage came up to -/+ 35vdc. Let me meantion the customer had done the power switch triac mod but it was incorrect and blew the 7 amp fuse. So luckly it never got full 120 VAC or it would be a mess. I will assume the diodes on the power supply board are good.
I have recapped the amp, security and preamp boards previosly. I removed the left amp board and toublingshooting the components on the board. Found Q8 to be flakey and will replace Q7 everything else test ok.
One question. THe power transistors collector case have a high resistance to ground. Shouldn't there be zero resistance to ground? When I clean power transistors and reinstall i always be sure there is zero case resistance to ground. BTW, all output transistors test good.

I'm going to connect the amp board outside the unit to further test the board and redo the left heat sink Any ideas on why the Left channel is over heating?
I have read the lenghty post about the Laffetty 9090 / RS-660 and other posts. No help for this situation.

Ken

 

[ohm470k]

Per the schematic, there is +/-52.4V at the collector. Definitely, should not be connected to GND.

 

[gadget73]

if its connected to the rest of the circuit it may read some resistance through the caps or whatever. If its just mounted to the heat sink with no other connection I would expect open circuit.

heat is generally high current. Could be a problem with the bias circuit, could be oscillation. Either will show up as high voltage across the emitter resistor(s) but oscillation will show up on the scope

 

[Ken770]

Gadget73 I replaced the bias adjustment pot and thought the bias was to high. I set the pot to zero. Measured it to confirm.
The 2sd381 is on the replacement list. I will replace those. Also, on the security board I found R5 overheated and reading higher resistance then 4.7k.
I dont have an o-scope. But will definitely look at the emitter voltages.
Need to figure out how to live test these boards outside the unit. I suppose connecting +/- and gnd to the amp boards is all i need to test it.

 

[Hyperion]

Need to figure out how to live test these boards outside the unit. I suppose connecting +/- and gnd to the amp boards is all i need to test it.

If the object of this exercise is to chase down that collector to ground issue, you shouldn't need to do that. Have the unit powered off, and just remove the connections to the output power transistors, then measure the collector to ground resistance - it should be open-circuit.

If it is general Output board functionality, I would remove and test the semiconductors to verify them replacing them where needed.

 

[Doug G.]

Grounding collectors? Ixnay.

Doug

 

[Ken770]

I have the board removed with heat sink and transistors connected. The collectors are all open to ground.
There are a couple of transistors on the bad list that need replacing. I will do your recommendation and pull the semiconductors and test. Some have corroded leads or appears to be corrosion.