[time-nuts] Fw: Sawtooth correction with a slower TIC
Tom Van Baak
tvb at LeapSecond.com
Fri Jul 20 23:52:04 EDT 2007
I am very happy to hear you did this already. Impressive.
Is there a prototype I could test here in my lab? I'd be very
interested in measuring it. I hope to finish testing a dozen
or more GPSDO this year; would like to have something
this simple in the mix.
----- Original Message -----
From: "Richard H McCorkle" <mccorkle at ptialaska.net>
> I did most of the software development work using just the
> uncorrected 1PPS into TMR0 and two scaled and summed PWM
> outputs instead of a DAC, and the results were surprizingly
> good. It made a very low cost, small footprint design without
> the external counters and DAC. All the added extras since are
> just part of my continuing education.
> ----- Original Message -----
> From: "Tom Van Baak" <tvb at LeapSecond.com>
>> > My original design used a 50 MHz TIC clock and the TMR0
>> > prescaler as the phase counter giving 20ns single sample resolution
>> > in the prototype system.
>> In this case how about making a super-simple GPSDO
>> and just put the uncorrected 1pps directly into TMRO.
>> The sawtooth jitter of an M12+ is 9 ns rms, which is just
>> the magnitude you need to improve on the 20 ns single
>> shot PIC-based TIC resolution. If you average long enough
>> even hanging bridges will not be a concern.
>> You'd have to do some work to get fancier, but it's hard to
>> imagine a PIC-only GPSDO that would be simpler.
More information about the time-nuts