[time-nuts] GPS noise reduction
bruce.griffiths at xtra.co.nz
Sun Apr 6 00:24:35 EDT 2008
Richard H McCorkle wrote:
> The 100M OCXO is undisciplined so your first case should
> be satisfied.
To achieve proper averaging of the TIC measurements the location of the
TIC oscillator output zero crossing has to slowly wander with respect to
the OCXO zero crossings so that the averaged TIC measurements are made
over a wide range of relative phases (ideally over a range of 1 TIC
oscillator period or more). For short averaging times the TIC OCXO may
be too stable.
You will need to measure the relative drift/wander between the TIC
oscillator and the OCXO zero crossings over the averaging period to
ensure that there is sufficient variation.
Either a time interval counter like a 5370 could be used for this.
Failing this an analog phase comparator can be used (first divide the
TIC oscillator down to the same nominal frequency as that of the OCXO).
You also need to characterise/measure the sawtooth corrected PPS jitter.
More information about the time-nuts