[time-nuts] Raw GPS signal samples

Magnus Danielson magnus at rubidium.dyndns.org
Wed Sep 4 22:11:37 UTC 2013


On 09/04/2013 11:25 PM, Said Jackson wrote:
> Tom, John,
>
> Wouldn't the software be able to remove any LO instability by simple cross correlation error cancellation by all the carrier phases? After all carrier phase measurements are concerned with relative phase delays, not absolute offsets?
>
> In other words the LO and ADC phase errors should show up on all carriers identically and at the same capture time. Assuming that the stability of the carrier phase is not degraded much by the transmission path I would think the LO errors could be reduced to levels approaching the transmission LO and Carrier stability even if just using a cheapo crystal for sampling?
>
> From what I read using external LO feed into GPS receivers from eg an atomic clock helps re-aquisition time but not locked accuracy or timing stability necessarily?
>
> Certainly most GPS lose lock momentarily if you breathe on them so I could be wrong, but they don't have the benefit of post processing out the LO errors..
You have to recall that you cross-correlate each sat with it's own PRN
sequence and current doppler frequency. Out of that you can get code and
carrier phase observables. You need to work out systematics from this
that would overshadow the noise of the LO. To what degree you would get
a gain after that could be discussed naturally. It might be easier to
just setup two independent LO chains and cross-correlate between them,
that would work.

Cheers,
Magnus



More information about the Time-nuts_lists.febo.com mailing list