From: ethan+@pitt.edu
Subject: IIsi speaker contacts: the last word!

Greetings all-  

After offering some (apparently mistaken) advice on how to clean up
the speaker contacts on the IIsi, I received the following note from
Mike Strange.  It seems that he knows what he is talking about!  He
gave me permission to publish it here.

Ethan Benatan 
ethan+@pitt.edu                          http://www.pitt.edu/~ethan

============================

Mike Strange's advice on cleaning copper contacts
-------------------------------------------------

From: strange@ping.at (Mike Strange)
Subject: Contacts

Regarding your post about your si speaker contacts, I have a little
experience with a similar contact problem.  And the solution that we found
may help you.

A little background:  I used to work as an engineering contractor to the
Navy and we were forever having problems with board contacts going bad in
the computers.  We tried the logical thing and sanded them every time we
could, but as you know, that is only good for about a week.  Much less in
an ocean environment.

Anyway, we studied it a little and found that the sandpapering was actually
accelerating the corrosion process.  One reason was that there were little
tiny particals of sand that remained imbedded in the surface of the
contact, providing a nucleation point for corrosion.  Second was that the
groves left by the paper itself (600 grit) were great nucleation points
also.  We were actually making it worse (A classic example of Deming's
tampering theory).  In addition, after a couple months of sanding, there
isn't much contact left.

The solution was to just remove the corrosion with something less abrasive.
Namely, a pink pencil eraser.  The white ones aren't agressive enough.  An
old Dixon pencil is your best friend in these situations.  We found that
treating with the eraser was almost as good as having a new board.
Treating with the eraser, after they had been sanded didn't really help
that much because the abrasive sanding particals were still in the contact.
The final trick was to apply just a thin coat of Vaselline (I can't spell
it right, you know what I mean.)  When I mean thin, I mean thin.  That gave
the best performance of all.  Hopefully somewhere some of those boards are
still running.

Good luck.

-Mike


Michael Strange
strange@ping.at