Bob Ammerman wrote...

>Olin opined:
>> But if you're talking about a dozen chips that were purchased all at the
>> same time or in a couple of lots, I wouldn't draw ***ANY*** conclusions
>> whatsoever from what you measure.  Absolutely none.
>
>Of course, if the dozen chips are all you'll ever need for the application,
>and you carefully characterize those (and perhaps a few spares), then you're
>all set.

That's me you quoted, not Olin.

What you say is certainly true: if all you're interested in is what that
specific group of a dozen chips will do, then there's nothing wrong with
characterizing them and designing around the data you obtain.

The error, though--and I've seen instances in which that error turned
out to be horribly expensive--is in assuming that those results have ANY
value at all in predicting what future units will do, especially when
someone is starting out with hopes for "better than spec" performance.

Dave D.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads