19 May
2010
19 May
'10
4:48 p.m.
Tony Duell wrote:
How many people where, who restore and run old
hardware, have never been
caught out by a 'silly fault'? My guess (if you are honest) is exactly
zero :-). I will admit to spending quite a time figuring out why all the
outputs of a PSU were sitting at zero volts ony to discover the fuse in
the mains plug was misisng (not blown, actaully removed). Or the time I
spend a morning tracing waveforms through the video board ina VT101 to
sort out a low-contrast problem, only to find the real fault was that the
CRT screen needed a good clean.
A while ago I was working on a calculator which intermittently would not
respond to several keys (failure seemed to correlate with low ambient
temperature). 5V levels were OK, some involved diagnosis - including
construction of extension connectors - traced it to an IC involved in the
matrix scanning and I replaced the IC.
The next time the temperature was low the problem showed up again. I had
measured the 5V level on the logic board with the power supply. The 'faulty' IC
was on a sub-board, the problem turned out to be a connector feeding the 5V
supply between the two boards: the 5V supply on the sub-board was marginal and
the 'faulty' IC had merely been the first to fail.
I'm still annoyed I messed around with replacing an IC in an otherwise-original
machine, when I all needed to do was clean a contact, esp. when I know better.