AnsweredAssumed Answered

Time-interval, single-period measurement with 1pps gate start

Question asked by arcticsearcher on Mar 31, 2016
We have a 10MHz TCXO system clock which I would ultimately like to discipline with our 1pps GPS reference. Counters are a new instrument to me, so as an exercise Id like to "benchmark" the 10MHz stability over days/weeks/months. We purchased a 53220A with the TCXO intending to use the 1pps to tigger or gate measurements, minimizing participation of the counter's internal time base. However, I am confused with these results..

*Measurement 1 - Time interval, single period
gate - advanced - ext BNC (1pps )
trig - internal

Result 1 - 100.070 ns, 50ps std dev (500 readings) : 9.99300 MHz

*Measurement 2 - Freq/Period - reciprocal
gate - chan1 (1pps) - 0mV
trig - internal

Result 2 - 100.000096 ns, 60fs std dev (500 readings) : 9.99999040 MHz 

*Measurement 3 - Freq
gate - chan1 (1pps) - 0mV
trig - internal

Result 3 - 9.999994MHz, 6Hz std dev (500 readings) 

These numbers seem radically different from each other and I am not confident that the internal time base uncertainty is excluded from either measurement. Im currently working through application notes AN200 and AN200-3. 

What other details/concepts are needed to better set up this 10MHz drift measurement? I appreciate any nudges in the right direction.