[sdiy] SH-101 repair question

[KA4HJH]

>Your instrument was soldered in the factory on a crappy wave soldering
>machine, which left just barely enough solder on the connections to hold
>the parts in place and allow the circuit to work, but certainly not enough
>solder to withstand 10 to 20 years of rocking and rolling.

This is so true. Arcade games suffer heavily from this. Add thermal cycling
to being bounced around and you've got a recipe for failure. I once saw an
angry player lift the front of a pinball three feet off the floor and drop
it--the glass didn't break but it made a HELL of a racket. Now that's never
happened to a vintage synth...yeah, right.

>When are people going to finally start getting it that 90% of electronic
>problems are electromechanical problems?  When things stop working, the
>first thing people suspect is some sort of major chip failure.  After
>that, some kind of mysterious, inexplicable passive component failure.  In
>most cases, both of these are very, very unlikely.

We once had a power supply that would turn off if you whacked it hard. I
must have spend seven hours on the damn thing. Two other people tried and
gave up. I finally found it. They'd bolted a stud rectifier to the chassis
and run the legs up through holes in the board. Since the traces were on
the bottom they put eyelets in the holes so you could solder/desolder from
the top. Despite all appearances at least one of those eyelets wasn't
sticking to the solder--one good jolt would make it come loose again.

>Compared to cold solders, broken wires, dirty contacts, etc. -- actual
>failed components are extremely rare.  Even the mythical electrolytic
>capacitors that are "so old that they're ready to fail".  Naturally, there
>are plenty of anecdotal stories to the effect of "I replaced the capacitor
>and the LFO started working again!"  Guess what, newbies?  It wasn't the
>new capacitor that fixed that LFO, it was the simple act of soldering
>another one in!  You would have most likely gotten the same result by
>simply reflowing solder on the old capacitor.

This can be REALLY hard to see even with magnification. The joint may look
fine.

>One more thing while I'm still perched on this soapbox -- There is
>something to be said about NOT replacing old parts with new ones in an old
>unit (unless they are definitely proven bad).  That is, the old parts have
>a History of Reliability, the new ones don't.  Think about it.

And you've got to be up on the parts in question or you won't know. Those
same power supplies I mention had high-quality, large value tantulums in
them. I never saw one of those fail and we never replaced them. All
electrolytics got shotgunned. The thing is that if you didn't know they
were tantalum you'd think they were some odd-looking electrolytic.

>Anyway, good luck with the repair, Mikael.  If you determine that you
>really, really do have a bad capacitor in there, congratulations.  I will
>still stand by my rant, and by the 90% statistic that I, admittedly,
>pulled from my ass.  Keep in mind, though, that it's a very experienced
>ass.  Wait a minute, that didn't sound right....

Are you available for consultations in your area of expertise?

-- 

Terry Bowman, KA4HJH
"The Mac Doctor"