Re: [gpsd-dev] Direct GPS support in ntpd makes me nervous

[Hal Murray]

address@hidden said:
> A Dallas Semiconductor I2C RTC chip<http://www.maximintegrated.com/datasheet/
> index.mvp/id/4627>can be tuned and synced so its "PPS" syncs with the GPS
> and it would be far better a time source (2PPM!) than any raw GPS not
> designed for time (SkyTraq has a 10nS GPS chip designed for just this) 

The problem with 2PPM is that it drifts farther from the truth the longer you 
wait. 1 PPM will be off by a whole second after a million seconds.  That's 
11.57 says.

There are many GPS devices that have a time mode.  Normally, it takes 3 or 4 
satellites.  4 gives you 4 equations to solve for 4 unknowns: X, Y, Z, and T. 
 (You can get away with only 3 satellites if you assume you are on the 
surface of the Earth.)

If you know where you are, you can get the time with only one satellite.  If 
you have more, you can do a better job of balancing the errors.


-- 
These are my opinions.  I hate spam.