[time-nuts] Newbie question: GPSDO filtering principles

Peter Vince pvince at theiet.org
Tue Apr 24 14:54:44 EDT 2007


Tom Van Baak <tvb at LeapSecond.com> wrote:

>You guessed the step size by manually setting the DAC
>to a couple of extreme values, right?

Yes, although not too extreme.  I first used +/- 100 (hex) steps 
around the currently used on-frequency value, and then +/- 10 (hex) 
steps.  The latter proved a little inconclusive due to the wander of 
the various oscillators, but wasn't a million miles away the figures 
obtained from the earlier larger frequency steps.


>Can you tell from the insides what GPS engine is used?

It's a tiny little Trimble board, but I couldn't see any obvious 
designation on it.


>One other thing you could try, since you have other GPS
>references and counters available is see if you can find
>the 1PPS from inside. This will tell you if the GPS engine
>is working right or not. Did it auto-survey correctly? The
>lat/lon/alt look OK?

Yes, that is buffered and fed out.  Interestingly it has a 
double-sawtooth effect: as well as the ordinary sawtooth discussed on 
the forum at length recently, there is a "tick-tock" effect every 
pair of seconds, with a swing between adjacent readings of about 220 
nanoseconds!

While the antennas don't have a perfectly clear view of the sky (some 
tall buildings and satellite dishes mask the eastern horizon), they 
are high up, and yes, the lat/long/alt seems OK.  I enquired about 
putting them into position-hold mode, but was told they always are 
(presumably after an initial survey).


>Also, another trick to isolate GPSDO problems is to pull
>the antenna and see if the output gets a lot more stable.

It is possible to lock the DAC at a specific value, which I've done, 
and that gave a much more stable output: after a few days it turned a 
corner and drifted off by about 3 parts in 10^13, and about a week 
later it came back flat and level - but I ain't gonna complain about 
that!  I've not tried pulling the antenna out, could be worth a go, 
just so long as the power surge doesn't blow it up!


>By looking at the raw GPS 1PPS, the locked performance,
>and the holdover performance you get a clue if the problem
>is with the GPS engine, the phase comparator system, or
>the LO itself.

The GPS timing has quite a lot of outliers, and I wonder if they are 
not being filtered out?

	Regards,

		Peter



More information about the time-nuts mailing list