Bill,

We called them 'cold flows' at UNIVAC.

They were nasty to find. The faults usually occured at random. For
troubleshooting this problem an effective 'low tech' method was used with
great success. We used to cycle the 'diagnostics' on the computer and
'massage' the back plane with a wooden spoon. If any movement was on the
'shorted' pin -- the program would 'die' or come to an error stop.

The backplane on our computers were usually wirewrapped with three levels of
wire on the pin. Level 3 wraps were on the top of the post, level 1 wraps on
the bottom. If the 'cold flow' happened on a level 1 wrap -- it was not
uncommon to remove a dozen [or so] wires to get to the 'bad' one.

Speaking of unconventional 'tools', one of the best tools for wirewrap
troubleshooting was using a small  'darning hook'. We would buy them at the
sewing supplies store. The small darning hook was ideal for snagging buried
wires in a backplane. To this day, I still carry one in my tool case.

Dave Gunderson
-----Original Message-----
From: William Chops Westfield <billw@CISCO.COM>
To: PICLIST@MITVMA.MIT.EDU <PICLIST@MITVMA.MIT.EDU>
Date: Wednesday, January 20, 1999 11:04 AM
Subject: Re: [OT] Wire wrapping questions


>By the way: number one "mysterious problem" in old mainframes with
>wirewrapped backplanes?  After 5+ years of operation, some of the wrapped
>wires that were positioned against other pins would eventually have the
>insulation displace enough to cause an inadvertent connection to the post!
>
>BillW