Hello,

I am trying to measure PPM on my crysal oscillator operating at 32.7688kHz. What method do I use: The specification is +/-50ppm. which is ~1.63Hz.

I have measured PPM using frequency counter Agilent 53131A Frequency Meter. Plotted multiple samples and then took Max minus MIN to be the maximum PPM. But folks are claiming since there is avergaing involved the number is too good and the real instantaneous variation is not captured.

What is bieng proposed is that we should instead measure the cycle to cycle jitter. For this measurement I am using 1Gs/sec scope and the infinite persistent mode is showing  a spread of 132nsSec in few minutes. The folks are claiming thus 132nsec jitter is violating +/-50ppm for a 32.768kHz clock since they are taking the period 30usec*50e-6(50ppm)as 1.5nsec and expecting the jitter to be below 1.5nsec.

I am not agreeing to the jitter measurement.

The other option is to use the spectrum analyzer and measure the peak frequency around 32.768kHz and see how much is the offset and compute PPM and see if we are below the 50ppm limit.

Any suggestions ? Do we do average or no for PPM measurement.

The issue is this clock will be used to time the packets of data for as long as 5mins before transmitting the data.
Note the clock will NOT be used to perform actual RF clocking i.e. no up conversion.


Regards
ic_engr