[ntp:questions] Re: GPS Time source offset

John Pettitt jpp at cloudview.com
Fri Oct 21 20:42:06 UTC 2005


-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

Patrick Klos wrote:
> In article <1129915741.107351.269460 at g47g2000cwa.googlegroups.com>,
> DJ <drnj at freemail.redherring.co.uk> wrote:
> 
>>I have a 4000 GPB ($6000 ?) GPS time source - it works beautifully.
>>I then took a $200 etrex hand-held GPS, set it to NMEA mode, plugged in
>>into a serial port on a Linux machine, I then did
>>ln -fs /dev/ttyS0 /dev/gps0 and configured ntp.conf
>>
>>server 127.127.1.0
>>fudge 127.127.1.0 stratum 10
>>
>>server 127.127.20.0
>>
>>I started ntpd and then my linux platform appears as a stratum 1
>>server. However, it seems to have an approx 300mS offset from the 4000
>>GBP GPS time source. 
> 
> 
> What are the chances that one or the other of your time sources is using
> the wrong edge of the PPS?  300ms sounds like a reasonable "active" time
> for a pulse per second signal.
> 


The Etrex doesn't have PPS so I suspect it's more a case of the NMEA
message not being consistent.    If the offset is a constant number
and the jitter low then it's just a matter of using a fudge to correct
the offset in the NMEA message - however if the jitter is high then
the using is pretty much unusable as a refclock.


John
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (MingW32)

iD8DBQFDWVKeaVyA7PElsKkRA0pKAKDGLamcmybwkr6AYoVQcTzprUTOgwCguBkf
2kmomEbQgUpLZHuX5lU94Xk=
=ru5p
-----END PGP SIGNATURE-----




More information about the questions mailing list