I will address several messages, one at a time, here.

The part, incidentally, is a high-speed analog mux- the NLAST4599.

>From Dave VanHorn:

> Having been in similar situations with other vendors, are you sure you
> are within the parts ratings?

Fairly sure- run down of the application follows:
It's an analog mux, two channel, switch the positive rail of a 2:3 voltage
divider from the output of a 0-5V DAC to floating.  VCC is also 5V, but
the selector input is driven by 3.3V logic.  The switching rate is
non-existant; perhaps two pulses of 250 us or so long every several
hundred milliseconds.  It's a high speed mux, so no real worries there.
The output of the divider is buffered by an op-amp, which has NOT been
dying on any of these (10, now, with more examples in the offing) boards.

Spehro observed:
> Stuff happens, and because of liability concerns you may have
> trouble getting them to admit it.

Agreed; a couple of years ago we had a problem with the wirebonds in
Sony CCD modules on this same product.  It took a long time to get them
to admit to it and even longer to get them to stop shipping us bad product.
They just kept adding lot numbers to the "bad" list after they shipped them.

Bob Blick:

>Any chance they are counterfeit parts?

Absolutely!  It wouldn't be the first time we got counterfeit parts
from this CM.
This is a topic already on the radar to be pursued.

>From Russell:

> Examples of the general situation (almost) always help.
> Sometimes people may spot an inobvious fault cause if you
> are free to divulge enough information.

See above.  I don't think I should go further into detail than that; it's not
MY IP to give away.  And, at the risk of dipping into hubris, we have enough
engineering expertise here that a flaw should be found easily enough.

> ON Semi are usually regarded as a quality supplier.

That's part of what I was looking for- a general statement of confidence in ON.
I'd prefer someone to say "I just saw an article about a huge counterfeiting
ring specializing in ONSemi parts" or "You know, we've had a strangely high
failure rate of ON components lately, too".

Google, FWIW, has very little to say about this.

Forrest asked:

>Have you asked Onsemi about why these parts are failing?

Not yet- this started yesterday.  Some of that, at least, will be handled by
our supplier- since they are a CM, we buy WORKING boards from them.
If the boards AREN'T working, that's their problem, and they need to sort it
with their suppliers.  Of course, without being privy to the ins and outs of
our design, they need our help, too, but a lot of the supplier comms goes
to them.

Mike H.
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist