Usually when one wants to generate an accurate clock, one uses the correct
crystal for the job.  Why are people so stuck on frequency multiples of 1MHz?
Why not use a 3.6864MHz, or 7.3728, or 11.0592MHz crystal or resonator?  They
aren't any more expensive.

The only reason I can see is to make it easier to count cycles.  It's nice to
say, 10 instruction = 10uS.

If you are doing rs-232, these crystals will give you 0% error.

In fact, there are very few situations where 10MHz or 20MHz timing is better
than a timing based on seconds being divided into a base 2 number.

Let's say you use a 18.432MHz crystal.  This frequency factors into 2^14 * 3^2 *
5^3.  You can get practically any usefull timing you want from those factors!
You want a one mS timer?  Set the prescaler to 8, and let the timer run free,
generating an interrupt 9000 times per second.  It is then a simple matter to
set up a jitter free fast counter which gives you your 1mS interrupt.

-Adam

Bob Ammerman wrote:
> What we really need is a way to ensure that Timer 1 interrupts every 5000
> instructions or 1 millisecond.
>
> The problem is that there can be jitter on reaching the interrupt handler
> for the timer. This jitter is caused by 2-cycle instructions, other
> interrupt handlers, and interrupts being disabled at 'task' level.

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