[time-nuts] How does sawtooth compensation work?

[Nick Sayer]

I've read Tom's page about sawtooth PPS jitter and I believe I understand where it comes from.    My current GPSDOs ignore the phenomenon. Certainly at the moment, I'm satisfied with that.  The systems gravitate towards PLL time constants that average it all away. 

What I'd like to understand is how sawtooth compensation works with receivers that support it. Is it that I expect an NMEA sentence with a nanosecond offset value that I add to any phase difference observation that I get?

Sent from my iPhone