🚀 go-pugleaf

>From article <1990Sep11.0...@lavaca.uh.edu>, by j...@karazm.math.uh.edu (J. Eric Townsend):
> In article <141...@hp-vcd.HP.COM> jo...@hp-vcd.HP.COM (John Eaton) writes:
>>With some of the early micros it
>>was SOP to clean the leads of all socketed IC's whenever the system started
>>to flake out.
> 
> Was there ever any real field-proof that gold-edge contacts are
> better than <x>-edged contacts?  I remember it used to be a big
> deal, but I haven't heard much about it recently.
...
> Just last week, a friend of mine had a PC's Limited XT go out of
> whack.  We called Dell (the current owner/name of PC's Limited)
> customer support (on a 5 year old machine :-) and one of the
> first things the support tech said to do was clean all of the
> card edges with an eraser.  (He didn't say to clean the eraser
> first, but we did that anyway. :-)

The worst possible metal combination for ICs and their sockets is
Tin/Tin, which is also the cheapest (and therefore often used).  What
happens is the tin oxidises, gaining a thin, tough insulating layer.
Over sufficient time (one or two years ) this layer will extend to the
microscopic contact region that was clean metal caused by the mechanical
insertion of the IC at manufacture.  The breakdown voltage of this layer
is low, but the modern IC draws very little current and may not cause
sufficient voltage across the barrier to break it down.  The result is
that you will never see the supply and ground pins not making contact,
but the data lines on an EPROM will go open.
Replacing either of the metals with gold will stop this problem (I've
been over all our single board CP/M systems and replaced the boot EPROM
socket with gold, and not had a failure since, previously failure was
common, and cured by a thump on the front panel in most cases, or
removal and reinsertion of the EPROM for the serious ones).

Erasers on gold contacts (soft rubber of course) is an approved cleaning
method.  It also will remove tarnish and crud off silver and tin IC pins
(but not tin oxide).


Summary: there is good basis for percussive maintenance of some
electronic equipment!

Allan Duncan	ACSnet	a.duncan@trl.oz
(03) 541 6708	ARPA	a.duncan%...@uunet.uu.net
		UUCP	{uunet,hplabs,ukc}!munnari!trl.oz!a.duncan
Telecom Research Labs, PO Box 249, Clayton, Victoria, 3168, Australia.