[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: NTP with only GPS/PPS flags PPS as falseticker
From: |
Greg Troxel |
Subject: |
Re: NTP with only GPS/PPS flags PPS as falseticker |
Date: |
Tue, 17 Mar 2020 10:19:24 -0400 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/26.3 (berkeley-unix) |
My $0.02 of advice
Long ago, in 2008, I set up gpsd and ntpd on a mac notebook on a
computer with no internet, and that worked fine, and it provided sync
to about 100 other computers with no internet, 50 of which were
connected by an ad hoc radio network under development. Were we 10 ms
off from true UTC? Maybe, but we didn't care about that last 10 ms.
Set fudge on the non-PPS GPS time (which is based on mesage arrival).
It is typically off a bit, and I dimly remember 100 ms late (on one
computer, running a different OS, with a different receiver -- just
mentioning it that you should not be upset by 20-300 ms of fudge being
needed). This is basically calibrating the delays, and I found that
variance was not that bad (a few ms) once done.
Read about minsane. If you want to sync to just GPS, then you need to
accept a single source. ntpd by default does not do this.
Turn on ntpd debugging and figure out why it's doing what it's doing.
If you have an ntpd problem, ask them :-)
My ntp.conf, for 3 USB GPS mice, NetBSD, base system ntpd, gpsd:
# GR-601W
#fudge 127.127.28.0 time1 0.110 refid GPS
# VFAN (ublox 8)
# u-blox AG - www.u-blox.com (0x1546) u-blox GNSS receiver (0x1a8)
#fudge 127.127.28.0 time1 0.048 refid GPS
# VIK (ublox 7, I think)
fudge 127.127.28.0 time1 0.072 refid GPS
Not that you should take my numbers, but basically if when on the net
the gpsd time is slow, increase, and if fast, decrease, and then you'll
be dialed in where it's pretty good. Again if you care about sub-ms
from this you are out of luck, but getting < 5 ms should be easy, and
1ish probably doable.
Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/16
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/16
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/16
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/16
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/16
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/16
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/16
- Re: NTP with only GPS/PPS flags PPS as falseticker,
Greg Troxel <=
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/18
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/18
- Re: NTP with only GPS/PPS flags PPS as falseticker, Greg Troxel, 2020/03/19
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/19
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/19
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/19
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/19
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/19
- Re: NTP with only GPS/PPS flags PPS as falseticker, Gary E. Miller, 2020/03/19
- Re[2]: NTP with only GPS/PPS flags PPS as falseticker, Nick Burkitt, 2020/03/19