I'm sure Brooks Shera can describe the nuances of his GPS locking circuit far better than I can; but that said, the 24MHz oscillator is not used to directly count the 1PPS signals. It is used over a 30 second measurement interval, yielding a precision of about 1.4nS per count. Also the digital filter designed into the PIC software yields very long time constants once you've moved away from the set-up mode; 2000 to 8000 second range comes to mind. Brooks' QST article shows 1E-11 performance & that was measured when SA was still turned on! This maybe mediocre by comparison to current commercial offerings, but it's not too shabby for something that is easily homebuilt.
Regards, Pete Rawson _______________________________________________ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts