[time-nuts] TIC resolution impact on GPSDO's performance
Tom Van Baak
tvb at leapsecond.com
Sun Dec 24 18:03:11 EST 2006
Thanks for your note on TIC resolution and the very
nice plot you attached. Here are some comments:
1. Could you clarify for us what points in your plot
are real measured data and what points, if any, are
2. Have you looked into why your Comparison.pdf
plot doesn't have the same look as SigmaTauBoth
Do both represent the same thing (M12/sawtooth
vs. M12/corrected) or did I misunderstand one of
them. Because they do seem to have a different
The red and black lines are parallel while black
and yellow appear to merge (like chopsticks).
But the horizon point (where the lines meet) for the
SigmaTauBoth plot is around tau 1 day while the
horizon point for Comparison.pdf, as best I can tell,
is around tau 10^8 seconds (about 1000 days).
3. For the sake of contrast, and to make your plot
even better, would you be able to add another OCXO
For example, perhaps add one that performs closer
to the 10811 spec (e.g. drift rate of 5e-10/day) and
one that is a lesser grade, an Ovenaire or CTS class
of OCXO or TCXO with short-term stability in the -10's
and drift rate in the -9's. I can provide the data if you
Perhaps other readers can suggest their own OCXO
4. I'm wondering if you could conclude something
interesting if your plot separated the OCXO issue,
from the GPS engine issue, from the TIC issue
rather than lumping the GPS and TIC together into
For example, in addition to different examples of
OCXO (low, medium, and high-performance) and
different GPS engines (e.g., Oncore VP, M12 with
sawtooth, M12/corrected), do you think you could
you also add different TIC resolutions?
For example, HP 53131 (500 ps) vs. 53132 (150 ps)
vs. 5334 (2 ns) vs. 5370 and SRS 620 (25 ps). You
can also add Shera (42?ns) and Jackson (8?ns).
Perhaps other readers can suggest different ones.
I'll have a few more comments later.
More information about the time-nuts