Having read this, it seems that most "normal" possibilities have been
eliminated.

It's not known though if you are mounting the drives or letting them
hang around, but if you are actually mounting them in the case you
might actually have a ground fault.

Now nothing has changed inside the case and the problem is gone, but
what about outside the case ? Grounds can be funny things at times.
What's more, there may have been some sort of ground fault that
eventually got blown out. The problem with that theory is that a
ground fault would almost for sure be in the PS, at least in an ATX
case right ? I mean the wires to the power switch are not on the hot
side.

I have seen some strange ground fault problems. Try to think back, if
the drives were not mounted, were they in electrical contact with the
case ?

And of course don't forget what might have been plugged in. Hell under
certain conditions a ground fault in the monitor might cause something
like this. Not to coin a phrase LOL - but think outside the box. I
think that somewhere down the line you might wind up blowing another
one. What would be interesting to know is if there are any significant
differences in the architecture of the power distributoin between IDE
and SATA drives. How many grounds and where ?

Whenever a problem seems to cure itself I get butterflies in the
ulcer. (I don't have an ulcer, that is a PutneySwopeism)

J