[ntp:questions] First attempt GPSD/PPS ->NTP time server

Unruh unruh-spam at physics.ubc.ca
Fri Jan 25 17:41:52 UTC 2008

Nero Imhard <nim at pipe.nl> writes:

>David J Taylor schreef:

>> The fact that you have a 0.2s offset suggests you are synching to the
>> trailing edge of the PPS signal and not the leading edge.

>The NMEA output would be my prime suspect. It is not surprising to have
>an offset there. The time in an NMEA sentence doesn't tell you what time
>it is "now", but instead it specifies the point in time that the rest of
>the data (e.g. position) refers to. Also, NMEA sentences lack proper
>time markers. It's always tricky business to get clock synchronization
>from NMEA alone. I would say it is impossible.

>The Garmin GPS 18 LVC manual explicitly mentions that emitted NMEA
>sentence describe the when and where of the *previous* rising edge of
>the PPS signal. It does not specify how long after that edge it will
>give you the NMEA, or even that this time will be constant.

It cannot tell you since that depends completely on how you have set it
up-- ie how many sentences it delivers. As mentioned if it is only the one
sentence, then 180msec is not a bad estimate at 4800bd. Ie, the unit just
sends out the sentences as soon as the second has occured.


More information about the questions mailing list