[time-nuts] Disciplining Rubidium
James.P.Lux at jpl.nasa.gov
Thu Apr 24 20:04:52 EDT 2008
At 04:04 PM 4/24/2008, Bruce Griffiths wrote:
>Jim Lux wrote:
> > Other uncertainties (multipath, ionosphere, phase center movement
> > with look angle, etc) are of that general magnitude as well, so even
> > if your little receiver could put out a 1pps with 1ns accuracy, the
> > other parts of the system are worse.
>Receiving antenna phase centre movement will only be a few cm not meters.
>Also the phase centre position is highly repeatable and once calibrated
>can easily be corrected for.
>Even the phase centre movement of the SV transmitting antenna (typically
>a helical antenna array) can be modelled.
>Multipath has a relatively small effect on a carrier phase disciplined
>Correction of the ionosphere propagation delay correction can take
>advantage of the difference between the ionosphere's group and phase
>Only a single frequency receiver is required for this.
>However generating an accurately positioned PPS pulse is another matter
>entirely as code phase data is required.
I agree with you..
I think my comment was more that the cost of dealing with all those
other factors (calibrateable or compensateable) is big enough or rare
enough that there's not much commercial market for a inexpensive
receiver that has, say, 1ns, accuracy.
Add that to the all around hassle (as you've described) of trying to
accurately position a pulse to a fraction of a clock coming from an
ASIC that is clocked at around 10-20 MHz.
More information about the time-nuts