[time-nuts] Using a UBlox NEO-6 GPS module for calibrating a PIC microprocessor based timer.

David J Taylor david-taylor at blueyonder.co.uk
Sun Dec 1 13:49:55 EST 2013


From: Attila Kinali

On Sun, 1 Dec 2013 10:57:42 -0000
"David J Taylor" <david-taylor at blueyonder.co.uk> wrote:

> A first approximation would be to assume a Gaussian distribution of the
> error, although there is a systematic component as well due to the system
> clock sampling inside the ublox module itself.  Its likely not to have any
> excursions of the order of 1 microsecond.

This is not exactly true. The accuracy of the PPS is highly dependent
on the position of the antenna, the sky view and reflections around it.
Getting to 1us jitter is not unheard of, even with timing receivers.

It is safe to assume that a TXCO has lower jitter than a GPS receiver
up to a couple seconds. I guestimate that an OCXO is better than GPS
in the range of 10-30s. And if you have a good OCXO, you can beat the
signal received up to a couple 100s [1,2].

You should also watch for oscillations of the GPS "jitter" with a period
of 12h and 24h.

BTW: for a reference on how well the LEA6 modules perform, you can use [3].
Although it's for a LE6-T with fixed position, you can extrapolate from 
this.
Also the histograms should explain you the 99% distribution.

Attila Kinali

[1] "Ultra-stable Oscillator: Limits of GNSS Coherent Integration",
by P. O. Gaggero, D. Borio, 2008
http://plan.geomatics.ucalgary.ca/papers/ion08_ultrastable_pascalg_26sep08.pdf

[2] "Effect of oscillator instability on GNSS signal integraion time",
by P. O. Gaggero, 2008
http://plan.geomatics.ucalgary.ca/papers/msc_thesis_gaggero_feb08.pdf

[3] "GPS-base Timing - Considerations with u-blox 6 GPS receivers",
u-blox application note GPS.G6-X-11007
https://www.u-blox.com/images/downloads/Product_Docs/Timing_AppNote_%28GPS.G6-X-11007%29.pdf

=====================================

Thanks for that, Attilia.  I added the URL for [3].

You are right that I simplified things.  The sawtooth effect will have (over 
a long term) a uniform distribution with a width equal to 
1/<sampling-frequency>.

I appreciate your better knowledge on the "GPS jitter".  Certainly a poor 
antenna location is to be avoided, and the serial output could be used to 
give some indication of the general GPS quality (e.g. by looking at signal 
strength, number of satellites, and precision of fix, in general terms).

Cheers,
David
-- 
SatSignal Software - Quality software written to your requirements
Web: http://www.satsignal.eu
Email: david-taylor at blueyonder.co.uk 



More information about the time-nuts mailing list