[time-nuts] Orbiting crystals

[Lux, James P]

On 6/28/09 4:35 PM, "Magnus Danielson" <magnus at rubidium.dyndns.org> wrote:

> iovane at inwind.it wrote:
>> James Lux wrote:
>> 
>>> Ah.. That's easy.. Pretty much every satellite in orbit derives its transmit
>>> frequency from a crystal oscillator, either a TCXO or a OCXO.  So all you
>>> have to do is pick a LEO satellite that is easy to receive and measure the
>>> transmitted frequency, and that is at a frequency where other effects won't
>>> dominate (ionospheric uncertainties, for instance, probably rule out VHF and
>>> UHF downlinks).
>> 
>> Hi James,
>> 
>> that's easy for you at Nasa! You have rotary antennas, good receivers and so
>> on...
>> I thought to GPS sats because there are cheap receivers all around, and hoped
>> the 1 PPS would tell something, so freeing me from considering doppler,
>> relativity etc.
>> 
>> Anyway, please let me know if you find any easily detectable signals that
>> doesn't
>> require a tracking antenna.
> 
> GPS will not help you at the PPS level, but you can avoid the tracking
> antenna anyway. The doppler-stuff and alot of other things is very well
> described. It is not hopeless, but not a very easy feather out of the
> hat thing.

Basically, it's a sort of tedious whittling away.  Once you have the basic
capture receiver working, it's just modeling stuff and writing countless
Matlab to remove this, that, and the other thing.  Check out the paper by
Peter Yoho for the analysis.  My paper (with Jon Adams) describes the
hardware.