[time-nuts] Another "atomic" clock question
albertson.chris at gmail.com
Wed Mar 5 22:28:04 EST 2014
On Wed, Mar 5, 2014 at 7:02 PM, Bob Camp <lists at rtty.us> wrote:
> The disconnect is that there is no need at all to correct for the delay with some sort of delay line setup. The sawtooth correction simply sums into the input of your control algorithm for the corrected oscillator. It's nothing more than an adder that sums the TDC output with the sawtooth information. No delay lines, nothing fancy to it.
Thanks this was exactly my plan, but I was wondering if I had missed
something what others were talking about some really complex
solutions. It works out to about one line of C code. Trivial, even
on an 8-bit controller.
Redondo Beach, California
More information about the time-nuts