[ntp:questions] Ignore one server except in extreme conditions?

A C agcarver+ntp at acarver.net
Thu Nov 17 06:37:41 UTC 2011

On 11/16/2011 15:24, Harlan Stenn wrote:
> A C wrote:
>>> So are you saying that the -52.888ms is making a significant
>>> contribution to the offset of -0.000001s ?
>> Yes, if I disable the NMEA source with a "noselect" and leave the
>> Internet servers and the PPS clock (22) running, my overall system
>> offset drops and holds at a few tens of microseconds.  If I leave it in,
>> the system offset wanders around.  The magnitude of the wander appears
>> to correlate with the magnitude of the NMEA offset.  For very large NMEA
>> offsets (sometimes exceeding +/- 50ms) the system itself starts to drift
>> away to large ms offsets.  Overall I seem to get better performance
>> without the NMEA driver contributing than with it included hence the
>> desire to make it accessible only when all the other Internet sources
>> fail (but GPS is still working).
> This doesn't make a lot of sense.  NMEA is known to be pretty much
> useless without a PPS. So if your NMEA time data is "off" I gotta wonder
> if your PPS is "off" too.
> Does your GPS report any "health" information?
>> Both PPS and NMEA are coming from the same physical GPS just using two
>> serial ports, one for the PPS and one for the serial data (this split is
>> required due to serial port driver limitations).
> Are you *certain* that your GPS is producing a "locked" PPS signal when
> the NMEA time is wandering?

Yes, I'm certain the PPS signal is not wandering.  That's what this 
particular GPS receiver was designed to do.  It keeps the phase of PPS 
steady but sacrifices the NMEA sentence timing.  Health reports are 
fine, it passes all the diagnostics and is locked on an average of seven 
satellites.  Most of the time it's more like nine or ten.

More information about the questions mailing list