[time-nuts] GPSDO Software & Performance

[Pluess, Tobias]

Dear all

I have been testing my GPSDO already for a couple months and I am also
improving the software from time to time.
I have now uploaded the latest version of my software to Github:

https://github.com/tcpluess/gpsdo

I have now a question about the improvement of this software.
Attached are two pictures of logged data from my GPSDO over more than 600
hours. From the picture "gpsdo-stable" we can see that the GPSDO is fairly
stable under normal conditions!
However in the other plot "gpsdo-outliers" we can see that there are huge
outliers in the phase error. At these times, we had huge thunderstorms here
in Switzerland at my place and I believe that the GPS signal quality was so
bad at these times that the GPS time pulse did not work reliably. (what
other reasons could there be for these 1PPS timing outliers?)
Obviously, the "magic" of a GPSDO control algorithm is in the filtering of
these outliers! so far, my control code (src/cntl.c on Github) does not
seem to filter these outliers good enough - we can see that the DAC voltage
experiences the same kind of ugly spikes as the phase error does.
Apparently, if an outlier occurs, I need to go into "holdover" status - but
based on which criteria should holdover occur, and based on which criteria
could the control program leave the holdover status?

Maybe someone on this list who has also built his own GPSDO has a smart
idea about how to deal with the outliers?

thanks
best
Tobias
HB9FSX



[image: gpsdo-outliers.png]
[image: gpsdo-stable.png]
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gpsdo-stable.png
Type: image/png
Size: 304926 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20210709/2b4ba69b/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gpsdo-outliers.png
Type: image/png
Size: 173499 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20210709/2b4ba69b/attachment-0001.png>