[time-nuts] Best GPS 1PPS Accuracy

[SAIDJACK at aol.com]

In a message dated 12/14/2006 12:00:36 Pacific Standard Time,  
tvb at leapsecond.com writes:

>  Hello Said,
>
> from my own experience in detecting and removal of  outliers i can asure
> you that it is a challenging and ambitious task  in statistical math. I
> fear you expect too much from a receiver's  TRAIM firmware to compute the
> necessary statistic math on a pulse to  pulse base. Nevertheless you can
> do it on your own if the  microcontroller that compares the LO to the
> GPS's 1pps has enough RAM  and processing power. 

I second Ulrich's comments. A certain amount  of
sanity checking on your end is worthwhile. You can
get GPS 1PPS  glitches, OCXO glitches, and TIC
glitches. No need to decide which is  which; just do
outlier detection and you solve them all.

The beauty  of a GPSDO is that you're only changing
the DAC every few minutes so  there's plenty of time
to look carefully at the quality of your 1 Hz  samples
before you commit the DAC update.

It would be a more  difficult problem if you needed to
update the DAC in  realtime.

/tvb


Hi guys,
 
"enough RAM" is a keyword here! Doing outlier detection using  i.e.  median 
filters etc seems prudent...
 
Part of my question was if anyone had any experience with setting the TRAIM  
limits really tight, to say +-20 or +-50ns and let TRAIM do the work. I rather 
 have no pulses than bad pulses.
 
Thanks,
Said