> -----Original Message-----
> From: Robert Rolf [mailto:robert.rolf@UALBERTA.CA]=20
> Sent: Sunday, June 15, 2003 1:10 AM
> To: PICLIST@MITVMA.MIT.EDU
> Subject: Re: [PIC]: Using dual-clock chip as a RTC
>=20
>=20
> If you have such a time sensitive loop, disable interrupts,=20
> assuming that you can be COMPLETELY done the loop in less=20
> than a second (or a serial character time). The Timer1=20
> overflow will be 'pending' when you reenable that interrupt,=20
> and will just be serviced late.
>=20

Ah, ok.  I guess I didn't understand exactly how the timers worked.  For
some reason, I was assuming that when they overflowed, then they waited
until the interrupt was serviced before continuing counting.  Now that I
write that down, it doesn't make much sense.  :-)

So you may end up with a sequence like this:

Int1: 0 sec
Int2: 1.001 sec
Int3: 1.999 sec

So the interrupts aren't neccesarily right on 1 sec marks, but the
overall time measurement doesn't drift.  So, assuming you had a perfect
crystal, you would only be off by at most 1 second at any particular
time measurement.

For my application, I don't need a high level of precision, but I do
want my time measurement to not drift over time.

Alex

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