I'll add that the GPS chip I mentioned previously has a sleep mode, wherein
it goes to a low-power state and wakes up and takes a reading every 2
hours.  Toggling a pin spits out one reading, then goes back to sleep.
Additionally, it has a hibernate mode that just keeps the static RAM alive,
and claims around 20 uA consumption.  Generally the receiver can "warm
start" or "hot start" if it has enough stored state that it knows which
satellites to look at (and probably their code phase info) so that taking a
reading doesn't take much time at all.  Even if this is overkill for your
application, it may be of use in the future for others reading this thread.

I see opinions that solar tracking doesn't need that much accuracy, and
also that they do.  If anyone could throw out some numbers, that might help
us put it in perspective.  I guess if optics to "focus" the sunlight are in
use, the problem is then magnified.  I've also read comments that the power
you gain slewing the panels around can be spent away running the motors to
do it.

Seems most RTC chips could come close to the needed accuracy, with only the
occasional hint required.  Then again, some ingenious person might come up
with a way to tell the time from the....sun!  I mean, after all, you're
pointing sensors at it :)
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.