Response
on CPU's from Jeremy Smallwood
http://www.static-sol.com/
This
is a very interesting question.
My first thought - as you mention storms - is, could the failure be related to
lightning? If there is a close correlation between occurrence of lightning and
failure, then this may be the answer. Lightning could induce large electrical
transients on power or data cables. If such a transient could get onto a cable
running into the cpu board, this could cause failure. Be particularly suspicious
of long cable runs which act as good antennas.
