[ntp:questions] Garmin GPS18x LVC 5m - Centos 5.5 - Kernel 2.6.18 - gpsd

[David Lord]

David J Taylor wrote:
> "Q" <.. at ..> wrote in message 
> news:4d18cb5a$0$12172$fa0fcedb at news.zen.co.uk...
>> I wonder if someone would be kind enough to help me out a little here. 
>> I have done lots of googleing and found nothing that helps me out with 
>> the issues I'm seeing.
>>
>> I am trying to reduce the PPS offset & delay (Which seem quite high 
>> for a local device) and stop the NMEA driver from swinging and 
>> becoming a false ticker.
>>
>> Any ideas please would be great. This is only a personal project, but 
>> I was expecting slightly better results! Below are debug outputs, 
>> configs and some other 'useful' stuff if you could take a look over 
>> them and let me know what I could be doing wrong, or would otherwise 
>> benefit from changing.
>>
>> Cheers!
>>
>> GPS;
>> GPS18x - NMEA outputting GPRMC to serial port.
>> GPS18x - PPS outputting to Pin 1 of serial port.
> 
> For what it's worth, I have been using a GPS18x LVC for quite some time, 
> and seen no problems.  As it happens, this is on a Windows-based 
> stratum-0 clock.  Then a few months back, time keeping became erratic.  
> I saw there was a new firmware out, so I flashed the puck with that 
> firmware and all was well.  For a few more weeks, at least.  Then the 
> same happened again and, as it happened, there was another new firmware 
> out, so I flashed it again.  This time it didn't cure the problem.  More 
> careful observation showed that the NMEA output was being seen as up to 
> a second out compared to the PPS output.  Like you, I have only the one 
> sentence being used, but I was still at 4800 baud (which should easily 
> allow one sentence per second, and had for about a year).
> 
> The situation now is:
> 
> - I don't know why this GPS18x LVC is behaving like this, and why its 
> behaviour has changed.
> 
> - as this system has good links to other reliable servers, I commented 
> out the NMEA driver, and left in just the ATOM driver:
> 
> Old
> ______________________________________
> server    127.127.22.1    minpoll 4
> server    127.127.20.1    minpoll 4    prefer  mode 33
> __________________________________________
> 
> 
> New
> __________________________________________
> server    127.127.22.1    minpoll 4
> __________________________________________
> 
> 
> and marked one of the reliable servers as "prefer".  I wonder if anyone 
> else has seem similar problems with the GPS18x LVC?

I think it's dependent on the ntpd version as to best
settings. I was using "fudge 127.127.20.0 time1 0.650"
and checking stats to make sure NMEA offset remained
low. That was with ntpd 4.2.4p6-o.

Other versions of ntpd handle fudge factors in different
(better?) ways. I'm moving to ntpd 4.2.7p98 and fairly
sure method for setting up is different.


> As to PPS offset, I can't measure it accurately here but it appears to 
> be under a millisecond, and perhaps under 100 microseconds.

On NetBSD I've had periods when rms offset as from
peer_summary, has remained below  10us for many days.
Mean offset is usually near 0.000ms. I don't have
reliable reception so would  expect lower offset with
a higher aerial position.


David