Martin Peach <martinrp@VAX2.CONCORDIA.CA> writes:
> I think the time it takes for the charge to leak out of the memory cells
> would be the limiting factor.

Yes.  EPROMs are usually rated for ten year data retention.  Some chips
are rated more than that; I've seen a few parts claimed by their
manufacturers to have 100 year data retention, but I really don't
believe that.  (Even if they've done accelerated aging tests, there are
some failure modes that such tests don't accurately model.)

In my experience, most EPROMs from 1978 still are holding their data.
However, I have seen a few EPROMs as recent as 1997 that have gone
flaky.  It is not clear in such cases whether the part is defective
(or of poor design), or whether the programming was done improperly.

Anyhow, IMHO it's reasonable to expect a properly programmed PIC to
work for ten years if you operate it within specifications.  I've got
some that are older than that and are still working fine, but they
haven't been in continuous operation.

Probably they will keep running much longer than that, but if you were
to ask me when I was wearing my engineer hat what kind of projected
lifetime you could promise a customer for a product including any sort
of EPROM/EEPROM/Flash technology, I don't think I'd normally be willing
to claim more than five years.

IMHO, it is absolutely f$#%ing amazing that they have managed to make
the capacitor (floating gate) in the memory cell hold a charge for
many years.  The leakage current is incredibly small, many orders
of magnitude less than any other charge storage device I know of.

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu