Re: [gpsd-users] Garmin 18X-5Hz

[Gary E. Miller]

Yo Robin!

On Thu, 11 Aug 2016 22:12:21 +0200
Robin Schwab <address@hidden> wrote:

> Am 11.08.2016 um 16:07 schrieb Miroslav Lichvar:
> > gpsd supports detection of 5 Hz PPS, but there is a problem that it
> > always sets the fractional part of the timestamp to zero, so with 5
> > Hz PPS the offsets have an error of 0.0, 0.2, 0.4, 0.6 and 0.8
> > seconds.  
> 
> I solved the problem for me by setting the NEMA update rate to
> something clearly below 0.5 s. In my amateur understanding it's then
> always clear to which NEMA sentence the pps pulse belongs: the one
> that is closer.

gpsd only output one PPS time correction per NMEA timestamp.  So the PPS
gets aligned to with the top of the second, the next 4 are ignored.

When we get a good test case it can be fixed better.

 Then you get a good fix although there remains a
> sawtooth-like offset variation.

All GPS have sawtooth effects in the PPS.  It is due to the granularity of
the OSC and PLL in the GPS.  Time-grade GPS output "sawtooth corrections"
so the software can compensate.  gpsd does not handle sawtooth corrections.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

pgpkwHLnBdPxh.pgp
Description: OpenPGP digital signature