Jeff,
Running a burn-in test (infant mortality) is meant to find and remove the 
parts that would have failed in the field at first stress. The nature of the 
test will degrade your parts to some degree. This stress (usually heated 
while part is in operation ... ok biased) is determined by you, which brings 
up something else.
Burn-in tests are statistical process. The levels and duration of stress, as 
well as target application (be it solder joints, actual components, entire 
subsystems ... you get the picture) all have to be determined before hand. 
The best thing for you to do is:
1) Check with your reliability dept. (I'm assuming that you have one since 
you've access to an actual burn-in oven) to see what standards apply to your 
design. Someone left suggestions on acquiring MIL standards (they also left a 
brief list of sites to search). MIL standards are great if you want to build 
overkill into your product (which is not a bad thing), however, there are 
specific standards governing each industry that do not require MIL spec 
quality, and hence, cost you less.
2) Look at the specifications yourself. Even if you do have a rel dept., and 
they say they've got everything covered, you'll want to review yourself to 
double-check their assessment. Try these sites as well (keep in mind you'll 
have to pay for any standard you need):
www.itu.org
www.eia.org
www.ansi.org
I tool around these sites from time to time to see what's changed for 
specifications. There are more, which I hope others will point out in 
subsequent mailings. These are three that I have hot linked.
3) Keep on asking questions (like you did on this group). Standards are 
changing all the time and they often get convoluted. Everyone will have an 
opinion about burn-in specs. The trick is digging deep enough to figure out 
what's right and why.
4) This is statistics in action. Everything about it is. Keep that in mind 
when your reviewing a spec that has in it something like "a temperature is to 
applied to the device of 85deg C for no less than 24hrs, while biased at 
-5V." This would be a minimum only. Those folks wrote that with a specific 
tub curve in mind for a typical case. Your design may be a little more robust 
with respect to initial failures but not so robust as to negate the need to 
extend the time and or temp range. You'll need to do some experimenting to 
define the first leg of the tub curve.
Anyway, I rambled enough. I hope this helps (or at least starts a debate).
J
-- 
Author: 
  INET: [EMAIL PROTECTED]

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB CHIPDIR-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).

Reply via email to