[time-nuts] GPSDO using 100Hz

[WarrenS]

Tom Van Baak wrote:
>> This raises some questions on the interpreatation of the M12+T and
>> Jupiter-T receiver specs.
>>
>> Some measurements are required in order to settle the questions once and
>> for all.
>>
>> Can anyone that has either or both of these receivers make the required
>> measurements?
>>     
>
> We know the sawtooth correction message only comes once
> a second, for both receivers, right? That may be a clue.
>
> /tvb
>
>
>   
Tom

The picket fence technique can be used to timestamp the leading edges of
the 100Hz (or 10kHz for the Jupiter -T) outputs.
The 100Hz signal being used to start the TIC whilst a 1kHz signal
derived from a low noise OCXO output is used to STOP the TIC.
Sawtooth error should then be able to be derived from the measurements
after removing phase wraps.
ADEV could then be plotted for Tau ranging from 10ms to say 100s.

With a 10KHz signal the STOP input frequency would need to be around
100KHz or so.
Any known frequency >= 2*(Start frequency) will suffice.

Bruce
****************
Bruce

OR on the 100Hz signal, not the 10Khz,  
can I use a digital scope that can log data time values between say the rise time and a marker, for each sweep.
For the most part, the phase jitter on for the Oncore can be seen easily at 100 Hz output by just waiting until the sawtooth jitter frequency is down to below about 5 Hz.
Something else you said, I did not know was that the 100rd pulse is time stamped by extending it length, Something I've always seen, but did not know the extended time was anything more than random. Next time I have it scoped, I'll check and see if the 1 Hz reported error is for that pulse while in the 100 Hz mode.

Warren