Electronic – Why time period of GPS PPS signal is not accurately equal to 1 second. (7 to 10 mirco sec less)

gps

I've multiple nxp K60 controller running at different ppm crystals. All have a GPS modules, i want to adjust their clocks using pps signal from gps.
Whenever K60 controller will start it will adjust their clock by taking reference from pps signal.

But the pps signal time period is not equal to 1 sec. First pps time period is 0.99999270 (7.3 usec less) and after first pulse it goes more less for some time.All positive edge of pps signal on different module is alligned within 1 usec. The time period for raising edge of pps is 0.8usec.

enter image description here

Best Answer

This is definitively your computer clock that is inaccurate as you are using the Salae Logic Analyzer software.

7ppm is within the expected precision bounds for your computer clock's crystal.

You can expect the GPS clock of a commercial receiver to have an accuracy that is better than 1us - and I think that this is short and long term jitter accuracy. The (absolute) precision actually depends on your receiver and meteorologic conditions.

More or less every 18 months, UTC time is adjusted so that a leaps seconds must be taken into account to convert the GPS clock to UTC time. That is due to earth turning slower and slower. This the precision of a second as these leap seconds only synchronise our calendar with the rotation of the earth.