Talk:GPS disciplined oscillator

Use of 1PPS for tracking?

Hi there, I think that there was something very fundamental mixed up by the initial author. It was written: "GPSDOs typically phase-align the internal flywheel oscillator ... generate a 1PPS signal from the reference oscillator, then phase comparing this 1PPS signal to the GPS-generated 1PPS signal and using the phase differences to control the local oscillator frequency in small adjustments via the tracking loop. ". No, it's not typical: Never seen such thing in any commercial circuit! As I know it from all schematics of commercial radio base station transceivers: you don't use a "consumer"/"home-grade" GPS receiver having well-known simple 1PPS outputs and phaselock to this!! You use special "timing"-GPS receivers (e.g. see "ublox LEA-M8F", or "Rockwell Jupiter") that do have an additional GPS-locked timing output of *much* higher frequency, e.g. 10khz or typical 10MHz. This "HF"-signal (typically NCO-generated and yes, thus phase-noisy) is then afterwards connected to a PLL-circuit as described by the author - but not to the 1PPS!! (the VCO-PLL gives then a clean signal again, as phase noise of the 10MHz/kHz-NCO-signal is rejected by the PLL's loop filter.) Using 10khz/MHz as loop reference, it gives much higher statistical information than with 1PPS to get same precision and thus magnitudes of shorter wait time for lock!!! Using 1PPS is not industry standard, maybe it's hobby style ;-) 84.148.103.212 (talk) 15:55, 13 April 2014 (UTC)

Thanks for taking the effort to bring up this issue. While I can believe you are correct, there are references, such as this book, that claim the 1PPS signal is the one that is commonly compared. I suppose what we need is a reference that indicates the alternative design is also in use. Thanks, --Mark viking (talk) 19:56, 8 March 2018 (UTC)