Path: utzoo!utgpu!watserv1!watmath!att!cbnews!military
Newsgroups: sci.military
Subject: Re: Electronics in Desert Heat
Message-ID: <1990Sep20.022308.15355@cbnews.att.com>
Date: 20 Sep 90 02:23:08 GMT
Sender: military@cbnews.att.com (William B. Thacker)
Organization: AT&T Bell Laboratories
Lines: 39
Approved: military@att.att.com



From: att!utzoo!henry
>From: rinne@mcnc.org (Glenn A. Rinne)
>all components are spec'd to operate at a temperature of 125C (257F).
>I'm assuming that the same  specificatiion applies to equipment.
>Why, then, is equipment failing at the desert temperatures of
>49C (120F)? ...

In general, 125C is the spec for *chip* temperature.  That will be
considerably higher than the temperature at the surface of the chip's
package, which will, in turn, be considerably higher than the
temperature of the air flowing over the package, which will be
warmer than the outside air unless the package is right in front
of the blowers.  And 49C will typically be a shade temperature, not
a vehicle-out-in-sunlight temperature.

Keeping electronics cool is a big problem in general.  The added
constraints of military equipment make it worse.  There will always
be a temptation to cut corners on cooling.  The only way to be sure
the stuff will work in the desert is to try it there.

One particular thing that can cause trouble is design changes after
initial testing.  Often they aren't as thoroughly checked out as
the original; if the stuff keeps on working well in its normal
operational environment, that will be assumed to be good enough.
For example, some proximity fuzes and such that had worked perfectly
well in Vietnam started showing a very high failure rate in peacetime.
The original specs had called for hermetically-sealed chip packages.
Those are expensive and the market is small, so when production rates
rose steeply for Vietnam, there was pressure to cut costs and use
packages that could be had in larger quantities on short notice.
They experimentally switched to commercial non-hermetic packages,
and they worked fine.  Until peace broke out, and the hardware started
sitting on non-airconditioned shelves for years instead of going
straight from the factory to combat.
-- 
TCP/IP: handling tomorrow's loads today| Henry Spencer at U of Toronto Zoology
OSI: handling yesterday's loads someday|  henry@zoo.toronto.edu   utzoo!henry