On 9/8/2011 12:27 PM, Jeff Liebermann wrote:
On Thu, 08 Sep 2011 14:12:20 -0500, John
wrote:

On 9/8/2011 1:30 PM, Jim Lux wrote:

And, you can take ESD hits that degrade, but don't destroy, the part.
I've got some LNAs in my desk drawer that appear to work just fine,
except the NF is a bit higher than it used to be.

How do you know it was ESD that caused the degradation?


Oh.. I don't know on the ones in the drawer that's what did it (because
they still work, after a fashion), but on some other ones where we did
failure analysis, you can see the damage on the SEM photos.

We've got whole books full of pictures of this kind of thing.

That's partly why industry has gone from marking just sensitive things
to marking everything ESD, including nuts and bolts. It's just too easy
to get damage without any obvious signs. Consistent practices reduce
the possibility, which increases yield. The concern is the "weakening"
or "partial damage without causing outright failure" kind of scenario,
especially in a hi-rel situation.

There's a lot of interest in figuring out a way to test for this kind of
thing, as you can imagine.

And there's even evidence that the damage can heal (anneal), much like
radiation damage (both are basically deposition of energy in the wrong
place and the wrong time, right), but the mechanics of the process is
poorly understood.


http://www.fainstruments.com/PDF/eosesd93.pdf
http://eed.gsfc.nasa.gov/562/ESD_Failures.htm
http://trs-new.jpl.nasa.gov/dspace/b.../1/00-1338.pdf