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

[Teneyes]

btw, I saw waterfall screenshots/pics of DS2xxx DSOs before, so they've been made with this software?

Yes Dave Jones used it in the review of DS1052 vs DS2000,
A great assistant to the Rigol DSOs, fast, easy to demo and report results.
Help Marmad out @paypal

[Radardude]

CH1 and CH2 with no input signal has a very small pulse in the noise. As you can see it is in sync with the trigger out. Is this normal?

[Teneyes]

CH1 and CH2 with no input signal has a very small pulse in the noise.

Hi , welcome to the Forum
Yes we called that the Trigger Hat, and below the trace the 'Ditch'
Playing with the trigger level will pull the random noise peaks out of the noise floor.
BW filtering limits this effect.
I see you were @5nsec, so I'm thinking you have DS2072

We Discussed it back here:
https://www.eevblog.com/forum/testgear/first-impressions-and-review-of-the-rigol-ds2072-ds2000-series-dso/msg178311/#msg178311

Have fun

[Teneyes]

Anyone else has troubles using the rotary encoder right to the menu button? it really is sensitive
everytime I just press it, it registers a rotation
for example: I try to select a different probe ratio. turning is is kinda sluggishly jumping to the next value, sometimes it skips 2 or 3
and when you push it after it's successfully on the correct value it jumps to the next value and selects this before the menu closes
it wouldn't be a bad idea if Rigol would make it possible to ALSO use the up/down buttons below to select something in the menu that's currently open (instead of the menu below it)

Yes I agree, and here is a message I sent to Rigol NA
--------------------------------------------------
Here is some feedback on an annoyance that I have encountered with the DS2072.
When using the Multi-function Knob after rotating the knob to my desired selection, I press the Knob, but
I have selected the incorrect pick, options, or letter, I end up picking the one before or the after!!! .
This has occurred more than 10 times in the 2 days since I received the Rigol DSO
 
Please understand that I have Large Hands (a 250 Lbs. Canadian Mohawk) and as I push the Multi-function Knob
the Knob turns a bit just before the Push switch is engaged/sensed.
I am not Asian with Diminutive Digits that can Discern Delicate Details.
 
I am suggesting the amount of rotation to jump to the next selection be a bit larger, and as such the few more degrees of rotation should not be detrimental to the easy of operating your Product.

Also
When I am Trying to select the letters in a file name from the menu of horizontal letters
then once the letter is Highlight with a Orange Box
when I push the Multi-function button the letter orange box moves.
I have checked the rotation and 6 letters change in 90 degrees rotation
that is 1 letter change in only 15 degrees of rotation ( 24 letters /rev)
NOW the top on the Knob is only 0.375 Inches in diameter
Therefore  a movement of only 0.05 inches ( 1/20 of an inch) will cause the letter to Change
I seriously think this is too sensitive of an Adjustment!!!!!
---------------------------------------------------------------------------------
Here is Rigol's response
    I do. Thank you for the explanation.
     Your measurements are exactly as I experience,
     but I don't seem to have the same issue with the usage.
     My letters are selected as expected.
     I have had this trouble with some of our other scopes....
     and it has to do with how I was pressing the multi-function knob.
     I was actually rotating it a bit when I pressed.
     I will submit that suggestion as an improvement.
     Also, here is the license to reset the trial software options:



To Avoid any rotation of the Multi-function knob during selections,
I have design a new Rigol Selection Tool.
See the Thumbnail picture below.
Feel free to grow your own.

[XaS]

I have similar problems when using the multi function knob. Also, it sometimes skips several entries or in my case causes the cursors to jump over the desired loactions. All in all, the knob is very unprecise in my experience.

FYI, I just found out that there can only be 63 screen captures on a flash drive, after that the scope displays an error that there would be too many files. Interestingly, the 64th picture is then stored as "QuickPrint1" without any file extenstion (since there already is a file "QuickPrint1.png" from before).

XaS

[PA0PBZ]

I am suggesting the amount of rotation to jump to the next selection be a bit larger...

No matter how large they make the amount of rotation it can always be just on the edge when you press the button, so I think it would make more sense to solve this in the software, like ignoring a turn when it happens less than 100ms before a push or something like that.

And yes, it is annoying, it happens to me too.

[Tasman]

I agree that the control is too sensitive.  The wrong selection problem occurs because the process of releasing the knob after turning to the required setting, then placing the thumb on the knob, then pressing -  always turns the control one way or another.  I find that pushing the knob using  the finger and thumb immediately the setting is correct, is the most accurate method, and I now rarely mis-select.

[Pinkus]

so I think it would make more sense to solve this in the software, like ignoring a turn when it happens less than 100ms before a push or something like that.

Thats exactly the way I would program it.

[NikWing]

yes, I guess the knob/rotary encoder itself is also kinda cheap
but software should fix it
also, from my experiences with other devices, the faster you rotate it, the bigger the steps are to change a value (slow = fine, fast = coarse)
maybe I did something wrong, but I didn't see that on the Rigol yet

but I like the DSO more and more

I'd like some feature to quickly center on peaks of the FFT curve, too ... rotating a lot to move it by hand on a peak can be annoying ^^

(and I had a similar "peak" on a curve as described above, it looked like some "data transfer" but it was the only one on the signal, no matter what time base I selected)

[Rory]

While we're on the subject of the rotary control, FFT and such, maybe someone can tell me how to change the time base scale on the time domain display while in FFT without having to turn off Math mode?  It seems really awkward and I use it a lot since that also sets the total bandwidth of the FFT. 

[Fagear]

Yes, multifunctional knob is very touchy. Rarely it utilizes velocity of rotating. For example in Trigger -> Menu -> Setting -> Holdoff - it does. But most cases - no.
Also it's a pity that trigger level knob doesn't use this feature.
I think that waveform playback knob must act as multifunction knob because it is hardly used at all. And it is bigger and has fixed angles, good for selecting something.

I've discovered some bugs (?) with MATH FFT funtion.
After you turn on FFT (MATH -> Operate -> FFT) horizontal controls move and zoom the FFT graph. It's ok.
But then press CH1 button and use horizontal scale knob. The FFT will scale with CH1 waveform as well and will shift to the side. Playing around with both horizontal knobs (rotating and pushing, while CH1 is selected) will force FFT graph to shift and jump around and sometimes to dissapear. To reposition FFT graph you need to press MATH, set required zoom and push POSITION to center the graph.
Also with FFT turned on it is impossible to turn on "delayed sweep" by pushing SCALE horizontal control knob (with CH1 selected). But this may be a processing power limitation. With other MATH functions it's not a problem, "delayed sweep" turns on and works ok.

Also if you turn on MATH -> Anti-Aliasing while in FFT graph will shift to the side and you have to recenter it by pushing horizontal POSITION knob. After turning Anti-Aliasing off you have to do it again. Strage.
When FFT is active you can not add any measurements, it's ok. But when you select something else in Operate box, measurement menu doesn't change color to MATH. To add some measurements on MATH you must push button of any channel and then press MATH again.

Also I'd prefer to turn off MATH function in way that I can do it with CH1, CH2 and REF. If it is selected and I press its button, it turns off. If not selected it selects and if you press a button again - it turns off. But with MATH the button only selects MATH menu and opens/closes it. It doesn't turn it off, you have to go in MATH menu -> Operate and select OFF with that touchy knob.

My unit: DS2072A, HW: 1.0.2.0.2, SW: 00.02.01.00.03

[Teneyes]

Back in July I reported to Rigol
=======================
I wish to report a Bug!

Using the FFT on the DS2000, the DSO must maintain the Center frequency
selected at the center of the display as you change through span scales.

At the start, an operator  will set the span set widest to see peak,
Then  the operator  will adjust the position so the peak (or point of
interest) is at the center frequency
After that ,the operator  will narrow span and the display must spread the
frequency holding the center frequency at the Center of the Display!!!.
The DS2000 does NOT do this!

The DSO moves the selected Peak way off the display,
and Making the 'Pissed Off' User constantly  needing to adjust the Center
Frequency.
++++++++++++++++++++++++++++++++

This is Fixed Now 

[Fagear]

Something strange with trigger level.

Attach CH1 probe to internal 1 kHz generator and push AUTO. Timebase sets to 200 us and vertical scale to 1 V/div (1:10 probe). Trigger is auto-set to 1.50 V.
Set scale to 500 mV/div and set CH1 position to 1 div from bottom (POS: -1.5 V). Trigger is in the middle of the screen (same 1.50 V) and waveform is 6 divisions amplitude.
Set trigger level to 500 mV (2 divisions from bottom).
Now move CH1 waveform with POSITION knob to the bottom until trigger level dissapears from the screen (it will point that it is somewhere at the bottom). Continue moving waveform down until it is barely visible. Than return it back to 1 div from the bottom. Trigger level will be offset to the top somewhere (it was at 2.54 V in my case).
This also happens with SCALE knob. Readjust trigger level to 500 mV (damn non-velocity sensitive knob!). Than adjust CH1 scale to 20 mV and then set it back to 500 mV. Once again, trigger level will be shifted, this time - down. To 160 mV in this case.

So while trigger level stay on the screen, SCALE and POSITION of CH1 do not affect its level at all. But when trigger level dissapears from the screen it level starts to change. I think is strange. In this case trigger level icon pointing "up" or "down" is pretty useless, because its previous setting is corrupted. It must stay on dialed level or clip withing screen with no "up" or "down" icon.

[marmad]

I think that waveform playback knob must act as multifunction knob because it is hardly used at all. And it is bigger and has fixed angles, good for selecting something.

It does. Note the menu symbol for Navigation Knob from the DS2000 User Manual:

[Rory]

But then press CH1 button and use horizontal scale knob. The FFT will scale with CH1 waveform as well and will shift to the side. Playing around with both horizontal knobs (rotating and pushing, while CH1 is selected) will force FFT graph to shift and jump around and sometimes to dissapear. To reposition FFT graph you need to press MATH, set required zoom and push POSITION to center the graph.

  @ Rory
 Yes the CH1 and Math buttons change the function of the <Scale> Knob.  But as Fagear states  , ackward or bugging

Thanks Fagear and Teneyes. Exactly what I needed to know. I got tired of turning off Math to make the timebase changes then turning FFT back on. Sometimes need to see envelopes in the time domain at slower clockrates than the right rate for the FFT. 

[marmad]

@Marmad:
The % of blind time (between 200ns and 2ns) in the following table is it real or effective?

Effective blind time. You can see that the active acquisition time is listed as 14x the time base (i.e. the display window - not the 'real' acquisition time).

[Carrington]

@Marmad:
The % of blind time (between 200ns and 2ns) in the following table is it real or effective?

Effective blind time. You can see that the active acquisition time is listed as 14x the time base (i.e. the display window - not the 'real' acquisition time).

Then the real blind time (%) between 2ns and 200ns would be: 82.46, 88.65, 92.80, 63.06, 81.36, 86.70 and 91.46.
So we can say than for these time bases (2ns-200ns) is always better to use 14K?
If so, would be interesting an automatic mode that selects the most appropriate memory to get the lowest blind time?
What do you think?

[marmad]

Then the real blind time (%) between 2ns and 200ns would be: 82.46, 88.65, 92.80, 63.06, 81.36, 86.70 and 91.46.
So we can say than for these time bases (2ns-200ns) is always better to use 14K?

It depends what you're doing. For most Normal uses, the effective blind time is, in a sense, the 'real' blind time - because YOU will be blind to it (i.e. it won't appear on the display) even if the DSO is not technically blind to it. So if a glitch happens in the part of acquisition memory that is off-screen, you'll never see it. But when capturing waveforms for later examination (whether just one - as in Single Shot mode - or many - as in Segments), then the 'real' vs 'effective' blind time actually makes a difference.

[Carrington]

It depends what you're doing. For most Normal uses, the effective blind time is, in a sense, the 'real' blind time - because YOU will be blind to it (i.e. it won't appear on the display) even if the DSO is not technically blind to it. So if a glitch happens in the part of acquisition memory that is off-screen, you'll never see it. But when capturing waveforms for later examination (whether just one - as in Single Shot mode - or many - as in Segments), then the 'real' vs 'effective' blind time actually makes a difference.

I understand:
"So if a glitch happens in the part of acquisition memory that is off-screen, you'll never see it." But, can the oscilloscope automatically detect it?

[TomThomas]

It depends what you're doing. For most Normal uses, the effective blind time is, in a sense, the 'real' blind time - because YOU will be blind to it (i.e. it won't appear on the display) even if the DSO is not technically blind to it. So if a glitch happens in the part of acquisition memory that is off-screen, you'll never see it. But when capturing waveforms for later examination (whether just one - as in Single Shot mode - or many - as in Segments), then the 'real' vs 'effective' blind time actually makes a difference.

I understand:
"So if a glitch happens in the part of acquisition memory that is off-screen, you'll never see it." But, can the oscilloscope automatically detect it?

If you know how the disturbance Looks like you could trigger on that.
Digital scopes have this kind of blind time. It's the Technologie... I heard a very nice statement some time ago: Measuring wiht a DSO is like driving a car with closed eyes and just open them from time to time for a very short moment.

[JDubU]

Digital scopes have this kind of blind time. It's the Technologie... I heard a very nice statement some time ago: Measuring wiht a DSO is like driving a car with closed eyes and just open them from time to time for a very short moment.

Analog scopes also have a blind time (during beam retrace).

[TomThomas]

Analog scopes also have a blind time (during beam retrace).

yes, right, but the percentage is much, much smaler... DSO have blind times between 99.99% and 90% of the whole acquisition time.

[Fagear]

It does. Note the menu symbol for Navigation Knob from the DS2000 User Manual:

Yes I know. For example it works for trigger holdoff setting.
But there are many many times when Navigation Knob does nothing. When you using any of on-screen-keyboards (MATH advaced function, option key typing...) Navigation Knob does nothing. And you have to use multifunctional knob that does not have fixed angle, so you cannot feel by your hand how much step did you perform. You have to be careful and stare at the screen while turning the knob.

Just open Trigger -> Type list (if you have all advanced trigger options it is better - more options there) and try to move three (or 4, 5, whatever) positions down (or up) fast. You cannot feel anything by your hand and cursor is laggy. So you have to wait, look, correct yourself and so on (or turn multifunction knob very slowly).
As for any list in right menu. Navigation Knob simply does nothing there. But it would be great if it will act as multifunctional knob. Because it has fixed angles and you can feel and hear number of steps while you are turning it, you do not have to watch at the screen and wait.

I'd also prefer right menu up/down buttons to scroll through lists. When there is only menu and no list opened - they will do nothing or will scroll "pages" of the menu as usual. It will be great if they would move cursor through the list (as does function key that opened the list, bit it does it only from top to bottom, remember trigger type list?..). It seems so intuitive to me: vertical list, buttons near have "up" and "down" arrow...

[Carrington]

I understand:
"So if a glitch happens in the part of acquisition memory that is off-screen, you'll never see it." But, can the oscilloscope automatically detect it?

The maskrange, work only for a screen/screen-region, would be great if in a post process, the oscilloscope could capture glitches off-screen (at memory).
Also it could be done with SW on PC.

I have not tried, I wonder if this can be done with the segmented memory... Although I am not clear if it is equivalent in a matter of wfrm/s.
i.e. What is the min. "death time" between interframe?

[JDubU]

But there are many many times when Navigation Knob does nothing.

I wish the Navigation Knob would also do high speed horizontal position in normal mode the same way it does in delayed sweep and record playback modes.
I also wish that the inner knob had a finger indent so that it could be spun continuously (like many professional video playback controllers).  The way it is designed now, I find it to be almost useless and, instead, use the small horizontal position knob which has the same function but can be spun faster and with more control.