Date:         Tue, 12 Jun 2012 10:59:27 -0400
Reply-To:     David Beierl <dbeierl@ATTGLOBAL.NET>
Sender:       Vanagon Mailing List <vanagon@gerry.vanagon.com>
From:         David Beierl <dbeierl@ATTGLOBAL.NET>
Subject:      Re: Vanagon Syndrome Fix: 10 Volt Capacitor OK?
Comments: To: neil n <musomuso@GMAIL.COM>
In-Reply-To:  <CAB2RwfheinSUS2SwJYy3vRe62Rd_5ZH-BLNdB87DVCqQaLARNQ@mail.g
              mail.com>
Content-Type: text/plain; charset="us-ascii"; format=flowed

At 08:52 PM 6/11/2012, neil n wrote:
>The archives mentions using 12 volt or higher rated capacitor, but will a
>10 Volt Tantalum capacitor be ok?

The AFM is supplied with 5 VDC.  50% of rated voltage is a common
rule of thumb, so you're good in my book.  Get your polarity right,
though, or you'll soon find a stream of tiny tantalum droplets
pointing back toward an almost invisible crack in the epoxy.  We used
to find that very useful as it pointed straight toward the culprit,
but of course that was on a nice flat circuit board.

Aluminums were different.  This was before they had rupture lines
stamped into the ends.  We (part of Analogic) shared a factory with
Data Precision; we worked two shifts but they only one.  Every once
in a while some unknown tech would solder a big power supply cap into
a board backwards and head off into the dark and give it power.  A
while later there would be a bang and then a clang as the can
ricocheted off the metal roof.  And the innards would look like a
bird's nest.  This would happen once in a while on the burn-in line
as well, but those had the covers on.  Never saw one blow up on a
tech's bench, but you wouldn't have wanted to get your eye in front
of that can as it took off.  Mostly we weren't leaning over the
boards, but once in a while...

Another thing we found useful was the fingerprint test.  If you
touched all the chips and looked at your finger, the logo that was
burned in to the end of it would tell you what chip was running
hot.  A number of these were ceramics and they didn't look much
different hot. (The product was a high-accuracy 7?-digit front end
for use with remote weighing systems where the load cell might be
half a mile away in an electrically noisy environment.  Rejecting
common-mode noise to find a much smaller differential signal was a
big deal.  Analogic was at the bleeding edge with this
stuff.  Sitting here now I wonder why they didn't have an amplifier
at the load cell powered from the head-end, but no doubt there were
reasons and I may have known them at the time).

Yrs,
d