[time-nuts] using Fury GPSDO with Rb
df6jb at ulrich-bangert.de
Sat Apr 19 09:31:36 EDT 2008
I have been interested into the tempco of some of my oscillators. To
find out I did the following:
1) I put them into the closed loop of my GPSDO systems for some days or
so. Since I do not use analogue electronics to steer the oscillator but
a DDS, exchanging the oscillator is a snap and does not result in any
parameter resetting, assumed that the loop's time constant is short
enough to keep track with even the most bad of them.
2) Then I recorded the DDS value that was necessary to keep the loop
locked and the temperature. This is pretty much like measuring the EFC
voltage in a anlogue steered system.
3) Using cross correlation techniques I determined the time lag between
temperature changes and the oscillator's reaction to it.
4) Having done that a second cross correlation between time corrected
data sets may be used to determine the tempco of the oscillators.
With this arrangement I received the following results:
HP10811: Tempco 2.2E-11 / K @ 22?C
FTS1200: Tempco 1.0E-11 / K @ 22?C
LPRO101: Tempco 7.2E-13 / K @ 22?C
While it may be said as a rule of thumb that the tempco of a rb standard
is roughly an order of magnitude smaller than that of very good xtal
oscillators one should also keep in mind that temperature is also for an
rb standard the environmental variable with the biggest impact. Thats
why I do not stop to complain that a rb in a GPSDO does by far not suit
us as well as we may expect: Of course the overall variations of the rb
are smaller in size but we cannot take full advantage out of it because
we cannot make the loop's time constant significantly greater than with
the xtal oscillators if the loop is expected to compensate for the the
temperature related frequency changes in a typical flat. Given the fact
that most rbs's stability at short observation times is worse that that
of good xtal oscillators this makes the FTS1200 the best choice in my
GPSDO despite of owning an LPRO101 and an FRK-L.
> -----Ursprungliche Nachricht-----
> Von: time-nuts-bounces at febo.com
> [mailto:time-nuts-bounces at febo.com] Im Auftrag von SAIDJACK at aol.com
> Gesendet: Freitag, 18. April 2008 19:46
> An: time-nuts at febo.com
> Betreff: Re: [time-nuts] using Fury GPSDO with Rb
> Hi Scott,
> that's a great success story!
> What final Loop settings do you use on the Fury with LPRO-101?
> It would be interesting to see if the +-10ns temperature
> related deviations
> are coming from the LPRO or maybe the GPS, DAC etc? What is
> the EFC change
> over these temp changes?
> With the external clock source, it is not easy to use the
> compensation of the Fury, but it could be done - need a
> current sink that is
> sensitive to temperature and uses about 50 - 200mA, that can
> be used as a highly
> accurate temperature sensor for the compensation routines.
> The digital temperature sensor on the Fury has only 1/16 Deg
> C resolution,
> not nearly enough to do accurate temperature compensation.
> In a message dated 4/18/2008 09:21:01 Pacific Daylight Time,
> smace at intt.net
> It's working well with the LPRO-101. It stays typically
> within +-3ns with excursions to +-10ns based on temp changes
> (hvac cycling, etc.) It is very apparent when you plot the
> temp sensor on the fury.
> **************Need a new ride? Check out the largest site for
> U.S. used car
> listings at AOL Autos.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-> bin/mailman/listinfo/time-nuts
> follow the instructions there.
More information about the time-nuts