[time-nuts] Cesium C Field Set?
cfharris at erols.com
Fri Feb 11 08:27:36 EST 2005
As I understand it, the wander is in the GPS satellites, and
comes from two sources. The first is doppler shift. The GPS
satellites are in motion. The second is from the DOD's downgrading
of the GPS system for civilian users. The way they downgrade
the system is by adding a random error to the output that varies
in such a way as to be on a very long term average, correct.
The way the differential GPS guys get around this error is to
have a receiver at a known location, and use it to broadcast the
error in its resolved position to a remote receiver. The remote
receiver subtracts out the error determined by the fixed receiver,
and the result is precision that is good into the centimeter range.
If you compare your Cs standard to a single GPS satellite, you will
see the frequency of your Cs to be initially low, then it will rise
until it is "correct" and then it will continue to rise until it is
above the "correct" frequency. When you let your receiver average
out a constellation of satellites, you are going to see a pattern that
is bumpy on the short term, but on average correct.
All the above are why I still use Loran C.
Tom Van Baak wrote:
> Interesting plot. You'll have fun getting to the bottom of it.
> I would say the 100 ns wandering you see is not GPS.
> If that were the case it would be like saying your one
> surplus FTS cesium is better than all the USNO clocks.
> Now, there is wander in GPS due to a host of factors
> but it's more like 10 ns for the plain receivers guys like
> us use. Unlikely you'd see that with a FTS4060.
> If you have a rubidium my recommendation is to run
> a comparison between the Rb and Cs in parallel. Then
> you can narrow down where those spikes are coming
> I would turn ION corrections on. Just this week I'm doing
> a parallel run between two M12's; one with and one
> without corrections. I'll let you know what I find.
More information about the time-nuts