From: Chuck Simmons
Organization: You jest.
X-Mailer: Mozilla 4.61 [en] (X11; U; Linux 2.0.33 i586)
Subject: Re: Temperature cycling of wirebonds.
Date: Sun, 29 Sep 2002 14:26:33 GMT
NNTP-Posting-Date: Sun, 29 Sep 2002 07:26:33 PDT
Ian Stirling wrote:
> I know this is a "how long is a bit of string question".
> I'm wondering about cooled camera chips in vacuum.
> Specifically orbiting ones, thermoelectrically cooled.
> These might see a temperature range of -100C - 50C or so, and
> perhaps over 10K cycles.
> Do many sorts of wirebond fail at this sort of cycling?
First of all, I have seen more wire bond problems in injection molded
parts. The typical image sensor is open cavity. If you look at the bonds
carefully, you will see that they are not pulled tight so there is some
freedom for thermal expansion and contraction.
I don't know about orbiting cameras but cooled ground based cameras are
normally cooled at a controlled rate and then brought back up to ambient
at a controlled rate. A full cycle might take half an hour. I think this
is more for the die bond than the wire bonds. In space, I'm not at all
sure how long the cycling might take but returning to ambient might well
take hours. Ambient is even a little hard to define but probably the
temperature of the surrounding structure is as good as anything. On
second thought, in space, the peltier effect coolers have to be liquid
cooled. This is done with some ground based systems as well to reduce
problems from having a heat source near the optical path. In space, of
course, without convection, there must be some arrangement to remove
heat from peltier effect coolers. (By the same token, the coolers don't
have to work as hard in space as they must for ground based systems.)
In any event, with open cavity parts and proper wire bonding, bonds
should not fail during the useful life of the part which may be more
radiation related than anything.
... The times have been,
That, when the brains were out,
the man would die. ... Macbeth
Chuck Simmons email@example.com