Hi,
When trying to script for the statistics mode of my MSO5074, I ran into the following observation.
User issue or bug?
The goal is reading the noise level (Vrms) statistics for the different vertical ranges.
For completeness, I am using the following SCPI command which is not listed in the programming manual.
Not sure though if it is hidden for a reason
Setting the moving average window (in display: Measure -> Statistics -> Counter)
:MEASure:STATISTIC:COUNT
range probably [10..100000]
This one exists once for all up to 10 ITEMs
So, I am setting :MEASure:STATISTIC:COUNT to 10.
The issue I am observing is that the reset does not completely reset for larger time base settings e.g. 1 sec / div.
So after a
:MEASURE:STATISTIC:RESET
I am waiting until
:Measure:Statistic:Item? Cnt, VRMS
reaches the value of above COUNT, so I know the window is full.
(and yes, the Cnt comes back as a scientific E-notation float, but should be int. And this is not my main issue)
While waiting, I can see on the display that the first [1..5] values in the statistics box are still from the previous
measurement (from before the reset). The # is somewhat random; mostly 1
So, when next calling
:Measure:Statistic:Item? AVerages, VRMS
the average is wrong.
After every measurement cycle, I am completely erasing the measurement ITEMs like this:
:Measure:Statistic:Display OFF
:Measure:Clear ALL
:Channel' + str(active_channel) +':Display OFF
Is this behaviour known and understood? Any workaround available?
Thanks for taking the time to look into the topic.
And thanks to thm_w. As suggested, I am moving the entry to this topic
-Ulrich
P.S. with regards to
Presumably the workaround would be to read Vrms every trigger, run that ten times, and average the result?
-> Yes, this was my first approach. But, runtime.... I stopped it after three days, and still not complete. So, saw the statistics are much faster.
Next try would be to make statistic from RMS and RMS from the waveform reading. Howerver this is another issue, let me investigate it before I post nonsense.