What exactly is effecting or limiting the counters on the scope? I think the discrepancies between the different counters are a little funny. I realize the Measure section has a much faster sample rate so it's going to see more variation, but is that it?
Why are different 10MHz sources pretending to be the same on the main counters?
The attached screenshot is using the Reference Out from the SDG2122X. Of course, plugging the AWG's Ref Out to its Counter gave a perfect 10.000,000MHz result lol.
Hi Josh!
There is only one counter (hardware counter). That one works on channel that is trigger source and uses trigger threshold as a point on waveform where it counts. If you open Analysis -> Counter, that is that same counter, with more data (and stats) displayed. It is a simple gated counter, nothing fancy.
All time related measurements in Measure window are derived differently. Frequency measurement there is simply an 1/t calculation from signal period measurement from actual sampled waveform. Actual thresholds where the signal pickoff is are governed by measurement thresholds. Make note that these measurements benefit from statistics that will even out all noise processes over thousand of measurements and be able to extract very decent accuracy by very different method.
Those two different methods practically guarantee there will be some difference in numbers.
But what are the differences.?
If we compare signal period of 10,00001e6 and 10,0000226e6 :
(1/10,00001e6)-(1/10,0000226e6)= 125,999589241054315091e-15
1,256 picoseconds.
That is pretty good in my book...
To be honest, i would never pretend that if you need very accurate frequency measurement (like 10 or 12 digits) that scope's built in counter would be instrument to do it. But for most of the general purpose work (like checking if µC clock is 10MHz, where you use 20ppm quartz anyways) it is more than sufficient.
PS: Why do I even bother? Performa already answered....