Dean wrote:
>
.......
>  Good questions you ask Phillip,
>   no extra external components are necacary if I can mearly find out what
>the wake up delay is with an acuracy of ~100ns  (or ~100 clk cycles), unless
>this delay can vary greatly dependent on temperature and voltage or other
>variables, this is all I need to compensate for. Another reason I dont want
>to include another chip is that Im trying to make this circuit as physically
>compact and simple as possible.
>

Yeah, isn't it **always** that way! You probably cannot do it here,
but running a PIC off a 32.768 Khz xtal sucks only about 15 uA, and
any Li battery will operate it for a year or three.

Another scheme - the one I would try if I were doing it - would be
to go with a 28-pin PIC, and use a 32 Khz xtal on Timer1 for the
RTC, and whatever xtal I wanted on the nml oscillator. After wakeup,
interrogate Timer1 to see how long it has been since the wakeup
interrupt occurred. With this, it doesn't matter how long the main
xtal takes to fire up.

Would be interesting to know if the std xtal oscillator fire up
time repeats to within 100 nsec. Let's see, for a 4 Mhz xtal, for
instance, 100 nsec is about 40% of 1 cycle. Who knows???????

- DanM

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]: PIC only [EE]: engineering [OT]: off topic [AD]: advertisements