unexpectedly developed: The fuses in the transformer
secondary started blowing. That's very bad. It means
something has developed a major short circuit and is
drawing far too much current through the transformer
secondary. Thinking it was a fluke or I'd touched
something wrong, I replaced the fuses and fired it up
again with all my probes removed. It lasted two or three
seconds before popping again. Not good at all.
If the excessive current was drawn through the
resistors in the low voltage section of the power supply,
they'd have smoked immediately, but those resistors were
all fine — this meant the failure was in the output
transistors which are directly connected right after the
transformer and bridge rectifier. I pulled them for
The transistors themselves all tested good — somehow
— but the interesting part was their insulators. The
collector of the 2SC1030 in this is tied to the metal case,
mounted to a grounded heatsink. An insulator failure
would short the collector (and supply rail) to ground,
which is no good. As luck would have it, one of the mica
insulators looked slightly cracked and had a carbonized
trace on it. That'd be the problem, although it picked a
weird time to show up.
I replaced all four mica insulators, applied new
thermal grease, and for good measure I replaced the
original 2SC1030 with all new ON Semiconductor
MJ15003G transistors. These are drop-in substitutes, but
like the capacitors are very de-rated in this application and
should be reliable for a long time to come.
Back On Track
After I popped in new fuses again, the next power-up
was stable and I was back where I'd started: one channel
50 July 2015
New ONSemi MJ15003G output transistors.
Close-up of the unmodified final amplifier board.
Schematic snips ©1978 Harman/Kardon Incorporated, referenced from the factory technical manual.