REVIEW - Rigol DS2072 - First Impressions of the DS2000 series from Rigol

[Teneyes]

It's interesting to see the DSOs interpolation of the individual data points.

@ Sparky
Here are displays showing how the DS2072 interpolates 1 Sample off line ,vectors and dots

Then displays of 2 Samples in a row off line, vectors and dots
Note how 2 samples will make the trace going more offline

These are very unlikely to occur unless a weird event happens, IE a Neutrino hits the ADC
It is more to show the interpolation function

For Sin(x)x See these discussions:
https://www.eevblog.com/forum/chat/sin(x)x-interpolation-and-digital-filters-in-oscilloscopes/#top

https://www.eevblog.com/forum/chat/rigol-ds1000e-series-possible-errorfail-in-sin(x)x-interpolation/
[/quote]

[Sparky]

It's interesting to see the DSOs interpolation of the individual data points.

@ Sparky
Here are displays showing how the DS2072 interpolates 1 Sample off line ,vectors and dots

Then displays of 2 Samples in a row off line, vectors and dots
Note how 2 samples will make the trace going more offline

These are very unlike to occur unless a weird event happens, IE a Neutrino hits the ADC
It is more to show the interpolation function

@Teneyes  --- AWESOME!!  THANKS!! 
That's very exciting to see --- I had no idea DSO was doing truncated sinc interpolation of the data points when choosing "vectors" over "dots".  Until I saw your first result showing the interpolation, I thought DSO just used linear interpolation when choosing vectors!

I have often wondered why is it called "Vectors"?  Clearly it is not "join the dots" mode, so calling it "Lines" or "Linear" or similar would be misleading.  Can anyone tell us, where the name "Vectors" has originated from?  Perhaps they should call it "Sinc"!

Would be nice if there was a way to change the interpolation function (just a few coefficients in memory...) because, for example, it's easy to imagine a scenario in which all data points should be positive, and so linear interpolation might be more appropriate (least some of our data ends up with negative values which would be an artifact of e.g. sinc interpolation).

Thanks again Teneyes!

[EV]

Here are some risetime measurements with my with DS2202. Signals (10 MHz) are from Rigol DG4162 generator. Generators rise time measured with fast oscilloscope is <1.4 ns . Signals are connected to scope with Rigol feed thru 50 ohm terminator.

In first picture signal is from sync output connector. Average rise time is 1.72 ns.
It gives bandwidth 350 / 1.72  = 205 MHz to scope.

In second picture is signal from generators output connector. Its average rise time is 4.1 ns. It is much slower than with signal from sync output.

[Lizerd]

Here you have another one who found this thread and decided to buy a DS2072.
Had a Instek before and the difference is huge.

No finally i gone trough the whole thread here
and wow what an effort 

I did register to be able to comment and contribute 

Where do i find the Firmware updates ?

I have the FW 00.00.01

I cant see that that anyone else has mentioned it, but when i start-up the unit with a USB stick inserted, the unit can not find the momorystick at all. i have to power down and restart without the memorystick, and insert it when the unit is running to be able to use the memorystick.

also another thing that annoys me.
in Diskmanger there is no return arrow like in the other function boxes.
Storage->DiskManager
It's not a big deal , but it is some thing i noticed.

I hope USB-Boot is fixed in the newer FW, i almost always have i memory stick in the scope.

Many thanks for the good reading 

[EV]

Where do i find the Firmware updates ?

You can find it here:
https://www.eevblog.com/forum/projects/software-tips-and-tricks-for-rigol-ds200040006000-ultravision-dsos/15/

It is posted by studio25.

Be careful not to loose trial options! It must be installed when booting the scope. Read instructions here in the forum. You can find the instructions on the first post of this thread.

[Sparky]

For Sin(x)x See these discussions:
https://www.eevblog.com/forum/chat/sin(x)x-interpolation-and-digital-filters-in-oscilloscopes/#top

https://www.eevblog.com/forum/chat/rigol-ds1000e-series-possible-errorfail-in-sin(x)x-interpolation/

Thanks for pointing this out Teneyes! Had a quick look --- interesting stuff!  I will read it closely a bit later

[orbiter]

Hi guys.. Think I've found another small issue with the menu display time (FW 01.00.05)

With trial options still enabled, if you select either 10 or 20 seconds for the menu display time. Only the left hand side menu
will hide itself after the selected time, the right side menu stays put. However if you select either 1,2 or 5 seconds for the display time, both side menu's work
and hide properly. 

[funk1980]

I got my DS2072 delivered today. This forum/thread and Marmad have been a great help in the deciding what to get. I've played with a for a couple of hours and it has been been awesome so far. Solid feel, low noise fan, clear menus. Comming from an 1984 analogue scope, I'm amazed by all the features, but was able to utilize them with relative ease. Very intuitive. Ultravision is cool as well. Feels like a CRT .
Hopefully I can be of some help to this thread.

[funk1980]

I've found a bug. Very minor, but a bug nontheless. I'm using firmware 00.00.01.00.05:
When using the menu timeout feature (menus retract after a set amount of seconds of inactivity), pressing any of the menu buttons or flipping to a next page resets the timer. It doens't do this for the left-side menus. No matter what button you press, the left menu always retracts after the set time from initial activation. The only way to 'keep it open' is to press a button the instant it retracts.

I can confirm the X-Y swap is fixed in 00.00.01.00.05 (still in the list in the first post).

[Teneyes]

I've found a bug. Very minor, but a bug nontheless. I'm using firmware 00.00.01.00.05:
When using the menu timeout feature (menus retract after a set amount of seconds of inactivity), pressing any of the menu buttons or flipping to a next page resets the timer.

DS2072 FW1.00.02
Related to this is the very annoyance/bug that once the menu retract time is set, the selection of letters in a file name does NOT reset the timer so it is hard to label files especially if switching to Capital letters in the short time.
I'm not sure if FW 1.00.05 still has this.   (we continue  to do Beta testing for Rigol)

[marmad]

I've found a bug. Very minor, but a bug nontheless. I'm using firmware 00.00.01.00.05:
When using the menu timeout feature (menus retract after a set amount of seconds of inactivity), pressing any of the menu buttons or flipping to a next page resets the timer. It doens't do this for the left-side menus. No matter what button you press, the left menu always retracts after the set time from initial activation. The only way to 'keep it open' is to press a button the instant it retracts.

I gave up trying to use the menu timeout feature because I didn't think it was well-implemented. The animation often isn't smooth (or is interrupted by external events) and I assume this is because the DSO is giving redraw priority (rightly so) to waveform display. But why have animation at all (especially if it isn't smooth)? If I use a menu timeout I just want the menus to go away - from fully out to fully in instantly would be fine with me.

I can confirm the X-Y swap is fixed in 00.00.01.00.05 (still in the list in the first post).

As far as I know, the X-Y swap bug has not been fixed in any version of firmware - except the engineering version which Dave had briefly. It's not fixed in my FW 01.00.05.

[Lizerd]

You are not reporting the full Firmware revision

Aaah thats right,
I have FW 00.00.01.00.02

Storage->DiskManager
Yes that is annoying but the return is there on the second page of the menu,

I totally missed that, think i checked the second page, but noo i had not
Thanks

[orbiter]

So i did got back the trail options, and for test i did an auto cal to get rid of the trail options.  ( version ..005 )
That worked, all the trail options are gone after the self calibration.

I will test now to see if i can get it back again, i want to discover a solid reset.

I have found that following a re-cal the trials will normally return after a couple of cold restarts.

[orbiter]

I've found a bug. Very minor, but a bug nontheless. I'm using firmware 00.00.01.00.05:
When using the menu timeout feature (menus retract after a set amount of seconds of inactivity), pressing any of the menu buttons or flipping to a next page resets the timer. It doens't do this for the left-side menus. No matter what button you press, the left menu always retracts after the set time from initial activation. The only way to 'keep it open' is to press a button the instant it retracts.

I can confirm the X-Y swap is fixed in 00.00.01.00.05 (still in the list in the first post).

Hey.. I'm claiming that find Two posts earlier

[Lizerd]

And my trail options are back again

Ooo how did you do that ?
Did it return by itself or by cold starts as orbiter said ?

[funk1980]

I can confirm the X-Y swap is fixed in 00.00.01.00.05 (still in the list in the first post).

As far as I know, the X-Y swap bug has not been fixed in any version of firmware - except the engineering version which Dave had briefly. It's not fixed in my FW 01.00.05.

Your absolutely right. It was a long day, late at night  . My logic thinking out the window it seems. X-Y is still swapped

[Wim13]

I have also fw ..005,

On ch 1 i get the Y
On ch 2 i get the X

What is wrong with that.., on my anloge scoop it is also this way
ch1 give the vertical and channel 2 the horizontal.

[marmad]

What is wrong with that.., on my anloge scoop it is also this way
ch1 give the vertical and channel 2 the horizontal.

Traditionally, it's written 'XY mode' and, numerically, you think CH1=X  CH2=Y, but it's not ALWAYS done this way - and not a big deal either way in my opinion.

[marmad]

@Teneyes and Sparky:

So, I was curious after seeing Teneyes' posted images as to whether the interpolated curves actually passed correctly through the sample points - so I combined the two PNGs into one - with the original sample points superimposed as white dots:



As you can see, it is perfectly correct.

Perhaps we need to make more checks like this at higher frequencies (approaching BW limit) to see if the DS2000 series suffers the same problems pointed out in the old thread (in which rf-loop pointed out that in the Rigol DS1052E/DS1102E, Rigol's implementation of SIN(X)/X was terribly wrong at frequencies approaching the rated BW)?

[Sparky]

@Teneyes and Sparky:
As you can see, it is perfectly correct.

Perhaps we need to make more checks like this at higher frequencies (approaching BW limit) to see if the DS2000 series suffers the same problems pointed out in the old thread (in which rf-loop pointed out that in the Rigol DS1052E/DS1102E, Rigol's implementation of SIN(X)/X was terribly wrong at frequencies approaching the rated BW)?

Good idea for testing this marmad!  It is good to know Rigol have fixed this problem that was shown in the related DS1000 series thread.  Some people here have 100 MHz function gen's -- may be someone will be able to show dots and vectors and we will see how the interpolation matches up.

BTW, do you know where then name "Vectors" has come from?

[marmad]

BTW, do you know where then name "Vectors" has come from?

No idea. But clearly that choice of menu selection name is, if not semantically incorrect, is at least rather misleading - it certainly isn't doing linear interpolation - it's something involving curves. But perhaps Rigol isn't doing true SIN(X)/X, so that don't want to make the mistake of calling it that again (and they don't mention it in the owners manual) - so they just use 'vectors' as a kind of vague term for some kind of interpolation that they don't want to fully explain. We'll have to ask drieg to ask Rigol at some point if they will say what the actual formula is. For now, I'd just like to be sure that whatever math they do ends up with the interpolated waveform passing through the actual sample points at all frequencies up to BW.

[zibadun]

so they just use 'vectors' as a kind of vague term for some kind of interpolation that they don't want to fully explain.

it looks like they may be using splines interpolation

[Teneyes]

I have taken a display of data from the fastest setting of my DS2072 (5nS/div) with 2 Channels  for 1GSa/s
and saved the waveform, then I manually set the Data Points with Hex File Editor,

I created a spreadsheet of "Idea" sine wave data values, and placed them in the file
I will show you the Data Points and the 'Vector' Display which always goes Through each data point.

Now the 1st two displays show a sine wave made up of 4 points,  now each point is 1nsec, therefore 4 x 1 = 4nSec or 250 MHz

Now the 3rd &4th  displays show a sine wave made up of 6 points,  now each point is 1nsec,therefore 6 x 1 = 6nSec or 166 MHz

Note that with only 6 points, sinewave only 2 points are above center and there are no data points at the peak of the sinewave, giving a smaller  peak-peak display.  I think a result of only 8 bit resolution.

I would like to test displaying waveform data on a 2nS/div display.  I will be sending a PM to DS2202 owner to help and sent me a waveform file "xxx.wfm".
It will be interesting to see if a File saved on DS2202 can be loaded in to a DS2072

If so I will Manually set data for  a small number of point on a Fast scan display

Please note there are data points in the file  before and after the displayed data points such that  affect the display, In these displays I repeated the pattern.

GIVE THANKS to Marmad's RUU software that makes these screen captures very Easy.

[EV]

Can anyone provide a Dots display of a real 400MHz signal into a DS2202 with 2 chans (1GSa/s)?

My generator goes only up to 160 MHz. I send it to you.

[Teneyes]

Hey look at this Noisy Waveform,