Ed:

> The Microchip literature on the PERIOD makes sense, it's their DUTY CYCLE
> calculation that seems to forget the factor of 4 between the input clock
> and the instruction clock.  Ignoring their equation and working it out
> based on a straight match between TMR2 and the duty cycle value (following
> what you said) is much easier, and makes sense to boot.

The formula in the databook is correct, the 10-bit time base counts
oscilator cycles (if the prescaler is set to 1.) What you're saying
is right IF you ignore the first two bits of the time base and the
duty-cycle register. This is what Microchip calls
"standard-resolution mode." I'd recommend that you read more
carefully what I sent you before.

I don't think the docs have any errors, what I find misleading is the
stuff about this "standard-res mode." They make it sound as if it was
a special feature of the PWM instead of just a recommended practice
for certain applications.

> Thanks,

You're welcome. Regards,

Andres Djordjalian
adjordj@aleph.fi.uba.ar