Callan Multibus 68k computer - the evil capacitor

It began with plugging 3 cards into the old Callan multibus chassis. I checked the voltage levels and the 12 volt was down at 0.9 volts.

I misinterpreted this as a power supply problem. I had just had to repair the -12 volt side of the supply I was using, and it seemed logical to assume that the +12 volt side also had a problem and was not delivering current.

After spending a day "repairing" the big linear power supply and giving up because the various components I was testing all were good, I swapped to a big switching supply, only to have similar problems.

I was seriously puzzled, but began checking things more carefully. Could I have run into 2 power supplies in a row, both with +12 volt problems? Possibly, but not likely. The light came on when I measured with a meter from +12 to ground on the chassis with cards in it and measured 0.4 ohms -- A dead short. From this point it was a matter of pulling different boards until I narrowed it down to a Micro Memory 512K ram board.

This board only uses +12 to power a little circuit with a 555 that generates -5 volts for certain (4116) memory chips. On this board with 4164 chips, the circuit is entirely unnecessary and in fact the 555 and related components are not even present. But the bypass capacitor is. The cure is to remove and discard it.

Apparently my switching supply was damaged by the short circuit, and it was necessary to swap to yet a third power supply after performing the "repair" (removal of the shorted capacitor).

I guess such things happen just as a matter of age. The board has chips with 1982 date codes (so it is 40 years old!) It probably has not been powered up in 25 to 30 years.

The lesson is to not make assumptions, but test things more carefully. I can put the original linear supply back together and it should be just fine for some other use.