RE: [chrony-dev] pps source is still marked as synchronised when no absolute clock sources are available |
[ Thread Index |
Date Index
| More chrony.tuxfamily.org/chrony-dev Archives
]
- To: "Bill Unruh" <unruh@xxxxxxxxxxxxxx>, <chrony-dev@xxxxxxxxxxxxxxxxxxxx>
- Subject: RE: [chrony-dev] pps source is still marked as synchronised when no absolute clock sources are available
- From: "Hattink, Tjalling (FINT)" <T.Hattink@xxxxxxxx>
- Date: Thu, 21 Jan 2010 10:36:15 +0100
- Thread-index: AcqZ7ab0UXfzeovuRVCh1MIAFE/PWQAjlf7Q
- Thread-topic: [chrony-dev] pps source is still marked as synchronised when no absolute clock sources are available
>> I agree with you that chrony will be able to keep time between the
pps
>> pulses. But in my situation I cannot blindly trust the incoming PPS
>> signal being the correct time. When the GPS card has no lock it keeps
>> generating
>
>This is a hardware problem, external to chrony. If your gps card is
>providing PPS without those PPS haveing anything to do with accurate
>time, that is an external problem to chrony, or any time discipline.
>Chrony has to trust its sources in the absense of evidence to the
>contrary. You are trying to solve a hardware problem-- an unreliable
>gps card, by chrony or ntp, and that is not their jobs.
>
>If you are really going to rely on the time put in three gps cards of
>sources ( the Garmin 18x is only about $60. Buy 5, and use them if you
>are really going to rely on the time for that server).
>
It is not really the GPS card being unreliable. When the card does not
have a GPS lock it just cannot provide an accurate clock. The clock
status is also given in the NMEA messages so the SHM interface knows
that the time is unreliable, but PPS is a simple electrical signal from
which you cannot see if it is reliable or not.
Btw, I have to mention that the system is operational on a moving vessel
so GPS signals are sometimes lost because the antenna is obscured by
bridges/trees/buildings.
>
>> PPS pulses from its own oscillator, but it will be free running and
>> drifting fast as no GPS corrections come in. So somehow I have to
tell
>> chrony not to trust the PPS source anymore when GPS lock is gone. I
>> used to do this
>
>That should be a job for the software. for example gpsd uses both the
>NMEA time and the PPS. It should be its job to decide if that gps
>receiver has gone crazy, for example because its nmea time has stopped
>working. That should not be a job for chrony or ntpd.
>
I can alter my SHM daemon to also correct the timestamps using the PPS
interface and remove the PPS references from chrony. So chrony will only
use the SHM interface for synchronisation. This would move my PPS
problem to the SHM daemon.
--
Tjalling Hattink
---
To unsubscribe email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "unsubscribe" in the subject.
For help email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "help" in the subject.
Trouble? Email listmaster@xxxxxxxxxxxxxxxxxxxx.