Would you reuse output transistors pulled from a blown output stage, if they test OK? What tests are "enough" (if any) to prove they're safe and effective?
I have a dead Lafayette LR-9090 receiver on the bench. Whatever happened to it hit both channels, which have identical damage -- right down to a burnt trace in the same spot on both sides. This is a head scratcher. There's a lot of damage: the bias spreader STV-4H diodes are open. All driver transistors are shorted or close to it. Surprisingly, all output stage resistors measure OK.
On each channel, both NPN outputs and one PNP output test OK; the other PNP output is cooked. "Test OK" means: all the basic diode junction tests, plus I did a Vceo leak test with 100V across collector to emitter, with the base open, and a 33K resistor in series. No voltage appeared across that resistor for any of the surviving 6 outputs so they all hold back 100V.
Now that MJ21195s and -6s cost >$8 apiece, it'd be nice not to buy all 8 new.
I have a dead Lafayette LR-9090 receiver on the bench. Whatever happened to it hit both channels, which have identical damage -- right down to a burnt trace in the same spot on both sides. This is a head scratcher. There's a lot of damage: the bias spreader STV-4H diodes are open. All driver transistors are shorted or close to it. Surprisingly, all output stage resistors measure OK.
On each channel, both NPN outputs and one PNP output test OK; the other PNP output is cooked. "Test OK" means: all the basic diode junction tests, plus I did a Vceo leak test with 100V across collector to emitter, with the base open, and a 33K resistor in series. No voltage appeared across that resistor for any of the surviving 6 outputs so they all hold back 100V.
Now that MJ21195s and -6s cost >$8 apiece, it'd be nice not to buy all 8 new.
You need parallel output transistors to be from the same batch to get decent matching over temperature.
I've found transistors that survived to pass the double diode test at 2 v, frequently will not pass an Iceo test at 12 v which I do with a car battery charger, ma scale of a DVM, and a 47 k resistor. These failing transistors had .450 v on the diode scale instead of .550 v of new transistors.
I've used survivors as cutoff transistors in a bike battery charger, for example. Not for audio. I tried to use one as a driver, but the Ft was 3 mhz and the Ft of the designated driver part # was 20 mhz.
I've found transistors that survived to pass the double diode test at 2 v, frequently will not pass an Iceo test at 12 v which I do with a car battery charger, ma scale of a DVM, and a 47 k resistor. These failing transistors had .450 v on the diode scale instead of .550 v of new transistors.
I've used survivors as cutoff transistors in a bike battery charger, for example. Not for audio. I tried to use one as a driver, but the Ft was 3 mhz and the Ft of the designated driver part # was 20 mhz.