EEVblog Electronics Community Forum

Products => Test Equipment => Topic started by: iBeta on March 23, 2017, 12:33:43 pm

Title: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: iBeta on March 23, 2017, 12:33:43 pm
(http://RigolStats.jpg)
The statistics shown in the scope are obviously not correct. I am beginning to regret buying this.
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: alsetalokin4017 on March 23, 2017, 12:41:40 pm
I'd hardly call that a "simple" pulse train, since you have two different periods happening.

With that kind of pulse train, what numbers would you expect to see for the statistics?

I agree that there is something wrong, but could it be because of the signal itself rather than the scope? That is, perhaps your signal is confusing the scope's measurement algorithms.
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: 2N3055 on March 23, 2017, 12:59:17 pm
There is no scope in the world that read minds .. no artificial intelligence either...
So it can't know what do you think it should measure here.

You have two shorter and one wider pulse. Which one do you want to measure? Or is that whole pulse train is one repetitive waveform?
What is the start ? First short one, second short one, the wider one... Which ?

Use cursors, that is the reason they exist. Try enabling auto cursors. They show you what is scope using for the particular measurement.
It will probably make more sense then..
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: iBeta on March 23, 2017, 01:23:30 pm
Quote
I'd hardly call that a "simple" pulse train, since you have two different periods happening.
If not I would of written 'simple square wave'.

Quote
There is no scope in the world that read minds .. no artificial intelligence either...
I am not asking it to read minds, just the oscilloscope data.

Let's take the statistics for the +Width; it has an average, minimum and maximum values. I would expect that the minimum width figures out the shortest time that the signal is at +V.
Looking at the image, that appears to be around 400us on the visible data.

Not rocket science people.

Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: GeorgeOfTheJungle on March 23, 2017, 01:32:23 pm
+Width and -Width are both wrong
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: alsetalokin4017 on March 23, 2017, 02:00:58 pm
If it's any consolation for you, I can also confuse my DS1054z with a similar pulse train. The stats are totally wrong and don't even update after a while, just as in your scopeshot.

I wonder what a high-end scope from, say, Tek or R&S would do when asked for stats on that kind of pulse train.

I don't know any rocket scientists that use Rigol scopes.    :popcorn:
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: PA0PBZ on March 23, 2017, 07:57:47 pm
Tried it on an MSOX3000:
 - it automatically shows the cursors when enabling the measurements
 - It sets the measuring point somewhere in the middle of the screen
 - It does show the correct min and max statistics only if you scroll around in the waveform (see above)

Picture:

(https://www.eevblog.com/forum/testgear/rigol-ds2072a-statistics-for-simple-pulse-train-incorrect/?action=dlattach;attach=301989;image)
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: alm on March 24, 2017, 12:34:41 pm
Not rocket science people.
Calling people trying to help you stupid seems like a really smart way of making people willing to help you.

What do the auto cursors show? I am guessing the measurement functions will only measure around the trigger point (or some place on the screen), and will calculate statistics across subsequent acquisitions, not multiple pulses in the same acquisition. Try restricting the measurement range to the other pulse widths with cursors and see what the measurement gives. To get the statistics you expect, try setting up the triggering so it will trigger on both the short and long pulses continuously, and see what the stats do.

As expected, the user manual is completely useless in its description of measurements and statistics. It includes only textbook definitions of pulse parameters.
Title: Re: Rigol DS2072A Statistics for simple pulse train incorrect.
Post by: 2N3055 on March 24, 2017, 02:17:20 pm
It measures first event after trigger.  Stats are not stats on one screen, but between consecutive screens (trigger events, buffer captures).. That is why I said to use auto cursors, which are showing you what it measures.