[SOLVED] Tektronix 4000 series oscilloscopes - question about measurements and statistics

[tubelectric]

Anyone familiar with these oscilloscopes?

These scopes have functions to measure quantities like RMS value of voltage. When you do that, you get several output values. These are named: "Value", "Mean", "Min", "Max" and "StDev".

(I mean, you get these five output values for each measurement you do. In the case of RMS measurement, for example, you would get all these for RMS measurement - "value" of RMS, mean of RMS, maximum of RMS and so on.)

These names should be self-explaining. But how are these statistics actually calculated? Sometimes the scope shows the exactly same value in the first four fields and zero in the StDev field. Sometimes it seems to perform some actual statistical calculations. How does it decide what to do? And how does it actually do it?

Do these scopes have a waveform/period detection of some sort? If they do, it's apparently not documented anywhere.

I haven't found any options for this in the menus. The "gating" option (whether to measure over full record/screen (visible section)/between cursors) in the measurements menu does not seem to affect this. The manual doesn't tell much either.

 

[KlausST]

Hi,

I assume the scope comes with a user manual. Did you read it?

Sometimes the scope shows the exactly same value in the first four fields and zero in the StDev field.

Makes sense. Do you know what "standard deviation" is? There are more than enough explanations and mathematical formulas in the internet.
My (non programmable) hand held calculator from 1980s is able to calculate StdDev.
Many other calculators and software can do this, too...Excel for example...

Klaus

 

[Dominik Przyborowski]

I was in the lab a very long time ago, so I do not trust to me memories now.
However, the tektronix scopes has various measurement functions (like RMS of waveform - so for 1V sinusoidal it would measure 0.707V of RMS)
and statistical characterization the stability of measurement like Mean value or StDev of what is currently seen in the scope. These statistical characterization has nothing to do with scalar values which would be interesting. You can check this by simple changing acquisition mode. With single mode all this statistics would show higher values than in averaging mode, for which with maximum number of averaged periods StDev number would be close to 0.

And I am pretty sure that all necessary informations about it should be in user or programmer manual of the scope.

 

[asdf44]

I also don't know exactly (though I do have the scope and know the feature you're talking about) however my take is that the 'extra' columns, mean max etc are based on the previous displayed values of the measurement in question.

So as it calculates the selected measurements and display them it keeping a running tally of the min/max/mean/etc of those displayed values.

 

[tubelectric]

Thanks for the answers.

The manual of these oscilloscopes can be found here:
https://www.tek.com/oscilloscope/dpo4054-manual/mso4000-and-dpo4000-series-0

I also don't know exactly (though I do have the scope and know the feature you're talking about) however my take is that the 'extra' columns, mean max etc are based on the previous displayed values of the measurement in question.

So as it calculates the selected measurements and display them it keeping a running tally of the min/max/mean/etc of those displayed values.

This sounds reasonable. I'll try to take a deeper look and also check how the acquistion modes and other settings affect the values.

There's also an option called "Mean & Std Dev Samples" in the statistics menu (123rd page in the manual = page numbered as 101). Value 32 is shown in the manual and I think it's the default. What does "Sample" mean in this context?

If what you suggested is correct (calculate RMS of the current record, store it, capture a new record, calculate RMS, store it... calculate statistics of the stored values...) then I would assume that this number of "samples" means the number of these values to store.

I was in the lab a very long time ago, so I do not trust to me memories now.
However, the tektronix scopes has various measurement functions (like RMS of waveform - so for 1V sinusoidal it would measure 0.707V of RMS)
and statistical characterization the stability of measurement like Mean value or StDev of what is currently seen in the scope. These statistical characterization has nothing to do with scalar values which would be interesting. You can check this by simple changing acquisition mode. With single mode all this statistics would show higher values than in averaging mode, for which with maximum number of averaged periods StDev number would be close to 0.

And I am pretty sure that all necessary informations about it should be in user or programmer manual of the scope.

I'm sorry, I'm not sure if I understood what you said correctly. But anyway, I'm not talking about calculating the mean or standard deviation of the signal seen on the screen. I'm talking about calculating the RMS value (or something else, but I use RMS as an example) of the signal seen on the screen, then calculating the mean, StdDev etc. of this RMS value. The scope should do this automatically. It sometimes does this. But sometimes it shows the exactly same value in the first four fields (value, mean, min, max), and zero in the StdDev field. This means that it has not performed any statistical calculations, or it has only performed the calculations for one sample. The signals I'm measuring are not stable enough to provide zero standard deviation under any imaginable conditions.

 

[tubelectric]

This mystery is at least partially solved...

The statistics of measurements are always calculated, but if you change display options such as horizontal/vertical scaling or position, or use zoom, the scope will reset the statistics. This happens even if acquistion is stopped. This happens when the gating option is set to "screen".

However, if the gating option is set to "full record", changing horizontal options will not reset the statistics. I haven't yet tried the "between cursors" option but I guess it might also affect when the statistics are reset.


In short, the statistics are calculated but changing various options will reset them. Resetting can sometimes be prevented by adjusting the gating option.