[time-nuts] Some GPS performance plots
John R. Ackermann N8UR
jra at febo.com
Wed Nov 3 17:00:53 EST 2004
Tom Van Baak wrote:
> Your plots are fine. Presumably you're showing how
> one can use GPS to check the rate of the Rubidium.
> In this case, as in many others, the sawtooth has
> little to do with it.
> You might try a plot or two using the AVG mode in
> Stable32. Tom Clark often used N=300 (5 minutes)
> with his 53131 counters. Not only does this clean
> up the phase plots nicely but it also does a good
> job smoothing the sawtooth without embarking on
> a hardware/software project; it becomes what I call
> a "mouse project" (all you need to do is click on avg).
I normally average at 100 seconds for this kind of stuff, but the
results seem to look smoother when I'm doing the averaging in my data
capture program. I think I recall that Stable32 in "average" mode for
phase data just selects every Nth point rather than averaging the
intermediate values, so the amplitude of the noise doesn't go down in
I wouldn't bother doing the sawtooth correction for this work, but it's
an interesting project to see what the noise looks like with that
applied. I should be able to just add a serial data routine to my
current program to interrogate the receiver for the T-RAIM status which
includes the offset of the *next* 1pps, and apply that to the value from
the counter before logging.
> One other suggestion, it makes an interesting plot
> to show 10 or 20 minutes worth of sawtooth, not
> just 20 seconds. The point here is to observe that
> there is great variability in the "pitch" of the teeth
> of the saw. If you're lucky you can find periods when
> the sawtooth is fine, coarse, positive, negative, and
> in between like a suspension bridge, or upside down.
Good idea. I'll see if I can find some plots like that in the data.
>>Let's see. To do that, you'd have to set the logging program up to
>>record the time interval via the GPIB port, and then get the correction
>>value via the serial port, then apply the correction and log the data.
>>Well, there's my project for this weekend...
>>(Actually, for my real purpose which is measuring the offset of the Rb,
>>it's not worth the effort since I'm averaging over many hours, but it
>>sounds like a good "science fair" project!)
> time-nuts mailing list
> time-nuts at febo.com
More information about the time-nuts