Jinx wrote:
>
> > If Microchip would put numbers about the stability somewhere
> > then they would have to stand by them. They won't do that. It's
> > the same thing as the non-liability for use in 'critical applications'
>
> I appreciate that, but surely the wafers and timing components on
> them fall within a small enough tolerance range to at least give an
> indication of what to expect. They do it for WDT and IntOsc, (but
> stop short of a few more helpful details). As reported to this list,
> parameters of non-xtal timing functions seem to fall within just a
> couple of % over a wide temperature range. What is so special
> about these functions that gives Microchip the willies about putting
> data in the manual ? Some general figures and a disclaimer is
> all that's needed


By maintaining that "secrecy" between the
max-perf specs and the end user it gives them
the ability to release batches of differing
manufacture techniques and then test them in
the field, fully paid for.

As they keep their moral obligation by providing
a guaranteed safe spec, everyone wins, they get
to try stuff out and we get PICs that are often
capable of MUCH higher performance than what is
listed in the "safe" spec.

Personally I don't trust written specs in general,
much better to test a few to failure with any
component your product will rely on. As I tested
16c505 for RC drift, and it was always within
1% in my temp and voltage range, i'm happy using
them in a serial app that requires around 5%
freq stability. :o)
-Roman

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.