[time-nuts] Some GPS performance plots

[Poul-Henning Kamp]

In message <4188F835.4010700 at febo.com>, "John R. Ackermann N8UR" writes:
>Poul-Henning Kamp wrote:
>> In message <4188F665.5090502 at febo.com>, "John R. Ackermann N8UR" writes:
>> 
>>>Nothing sexy here, but in response to some serious confusion on the NTP 
>>>mailing list, I put together some plots of the 1pps output from a 
>>>Motorola M12+ Timing Oncore versus an Rb standard.  It may be helpful if 
>>>you want to point people to info about GPS timing capabilities.
>>>
>>>http://www.febo.com/time-freq/gps/performance/index.html
>> 
>> 
>> You're not being fair to the M12+ here.  You should correct the timestamp with the
>> "negative sawtooth correction" field from the serial data stream.
>
>I know, and I should (and will) note to that on the page.  Since I'm 
>capturing the 1pps output from the receiver and feeding it into a 
>counter, I can't apply the sawtooth correction value the way I could if 
>I were processing the data in a PC.

As long as you don't miss out on pulses or serial data it is possible to
correct with post-processing.  At least I have been able to make it work
several times.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.