[time-nuts] Checking the Frequency of a Rubidium Oscillator

Magnus Danielson magnus at rubidium.dyndns.org
Wed Nov 12 07:33:34 UTC 2008


Poul-Henning Kamp skrev:
> In message <491A210B.30401 at rubidium.dyndns.org>, Magnus Danielson writes:
> 
>> Once doppler bin and phase has been achieved for each PRN, [...]
> 
> Just a footnote to say that as soon as you start receiving ephemerides
> from the first sat, the search-space can be significantly reduced
> if you care to do the, rather longhaired, trignometric math.

True, but breaking into the code phase for each sat is nowdays fairly 
quick, and after setting up the receive channels the rest is much more 
parallelized. It is simply just quicker to do the code phase break in 
and start tracking than receiving the ephemerides data from the first 
sat. The time it takes for a calender to be received is fairly long.

>> A sat based receiver must handle higher doppler offsets due to its 
>> higher speed, [...]
> 
> While this is true for any non-geo-stationary satellite, it may not
> be true for the project the initial poster talked about.
> 
> As I remember it, he said that the mission would be in an earth-following
> orbit, ie: in the same orbit as the earth around the sun, but
> trailing it by some distance.

That was never clear in my mind. Ah well, if so then that part would not 
need any specific modifications, not that they are particular hard.

However, which ever orbit we are discussing, the doppler aspect needs to 
be studied.

> Given that the distance in GPS terms is "vast" and furthermore that
> the GPS orbits have a pretty steep angle relative to the earths
> orbital path, I would expect the doppler offsets to be much smaller
> than here on earth.

Another aspect to remember is that there is usually a earth bound 
assumption used to bootstrap the position calculation. This would 
naturally need to be adapted. Fortunately it can be adapted and tested 
very easily if needed.

> Obviously, getting a position fix will suck with the worst
> DOP seen to date, but a frequency fix should not be out
> of the question.
> 
> Obviously, the situation on the way to the final orbit is entirely
> different, and there I would expect doppler to be totally out
> of the lower end of the window.

Actually, you can expect both ends of the doppler spectrum. As long as 
you are below the GPS sats, you will also see high positive dopplers as 
you goes towards them and negative from those you are leaving behind. As 
you go past them, all will show up on the negative side. However, those 
closest to you will not be looking at you any more due to directivity of 
the antenna array.

> Remember to figure out the relevant relativistic corrections.

There are several relativistic corrections that needs to be considered. 
Also, while the sat is in transit to its final orbit one can expect 
these to be with a higher dynamic than a circular orbit. An elliptic 
orbit would always need orbit-based relativistic correction for that 
extra correctness.

The intended orbit and transit-orbit is certainly of great importance 
for a number of key processing requirements. It is however not extremely 
hard to handle it.

Cheers,
Magnus



More information about the time-nuts mailing list