EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope

[Martin72]

The fact that I don´t have these problems let me think about a hardware problem.
As I said it before, I had a compensation problem with new probes (on a lecroy scope)- You can´t trim them perfect, with other you could.
And so it seems, in this case, the compensation problem lies unfortunately in the frontend, when any probe can´t be trimmed to normal.
I´m curious about rigol´s answer….And reaction.

[David Hess]

And so it seems, in this case, the compensation problem lies unfortunately in the frontend, when any probe can´t be trimmed to normal.

At least not those probes.  Some probes include more adjustments than just the the LF/MF compensation but this should not be necessary at only 100 MHz or at low frequencies as with the case here.

The acid test is to use a 50 ohm source which requires no compensation; the DSO should display it perfectly.  But if the 50 ohm test and the properly compensated and working probe do not display the same thing, then the DSO is broken.

This is why oscilloscope calibration occurs in two stages starting with a 50 ohm source.  The second stage uses a "normalizer" which simulates a probe although a x10 probe can also be used to calibrate the input capacitance of any input attenuators.

[Martin72]

Helo
New firmware?  MSO5000_00.01.01.04.04
http://www.rigol.com/Support/SoftDownload/3

Here are the changes listed:

https://www.eevblog.com/forum/testgear/hacking-the-rigol-mso5000-series-oscilloscopes/msg2231226/#msg2231226

Today I´ve tested some of them :

- Added 500uV/div in vertical scale.

  Check, OK.

- The waveform can zoom out by drawing a rectangle. If you draw a rectangle
       from the top left to the bottom right, the waveform will zoom in. If you
       draw it from the bottom right to the top left(the opposite direction),
       the waveform will zoom out.

Can´t complain it - Tried to draw a "rectangle", nothing happens for zoom in/out, zooming already existed per spreading sideways with your fingers…

- Added the GND coupling in channel.

Checked, OK.

- Enriched the color options of the LA channels.

Checked, OK.

- The boot time is reduced to less than 1 minute.

Checked, OK ( 59sec... )

- Added the 12-bit high resolution mode.

Checked....ehhh not OK.

Hi-Res is selectable but it happens NOTHING, there´s no difference between Normal Mode and Hi-Res Mode....

[Commodore8888]

Look like some useful updates then.

Wonky hi-res, hmm.

Boot time and color options on digital are the ones I'd get excited over.
Now, does "enriched" mean added 2 or 3? Or a whole bunch. Sort of trivial to just add QT menu options on existing menus....
SSH is gonsky though. Bit bummed there.

I'll be curious to see what Rigol EU says Martin. I'm also in the middle of pinging Rigol NA (but they most likely all share the same "issues" email).

So....wonder if the second set of MSO 7k's the lab ordered a few weeks ago will also show the same issues when they show up. I may be driving a van full of scopes to the UPS store for warranty .........


For reference, a happy DS1054Z, vs a grump MSO5k....
Input is the reference from the DS1kZ.
Tried also at work with a rather fancy Agilent gen and 50Ohm. No ringing...same overshoot. RIP.

Interestingly though, the 50Ohm internal path on the 7k just about eliminated the goofy overshoot.
Still present using a 50Ohm feedthrough however.

[Martin72]

Normally, Rigol EU answers promptly, therefore I expect a reaction tomorrow(when one of them returned into office).

Reduced boot time....I stop the time from power it on until the screen comes up - 59 seconds and a few more, but under 1min...
I don´t have any overshoot-issues here, but your and other ones issues makes me carefully to recommending the 5000/7000 to others.
Until this will be solved.

[Martin72]

Hi,
As I told, rigol anwered me today...very nice.

They already know about and work on it...

So it´s an "official" bug with the overshoots…

Martin

[luma]

Maybe they'll send us a bag of capacitors to resolve it

[Martin72]

 

Wonky hi-res, hmm.

Thought about the "non-functional" hi-res mode…
What Do I expect from - Cleaner filtered waveforms.
What don´t happen if You choose the hi-res mode on the 5000 ?

Right, nothing:



Normal Mode



Hi-Res Mode
But look now, If the memory depth are decreased to it´s minimum of 1k :



Normal Mode



Hi-Res Mode

A-ha....Bandwith limited.
Seems, hi-res does exists but somehow "wrong"....or too less.

Test:  A PWM-Signal with 400Hz sinewave modulation - If you reduce the memdepth in hi-res / enhanced Resolution mode, the scope displays the sinewave because of the limited BW.



PWM signal in normal mode

And now, in hi-res mode:



Tadaa....a sinewave 
The hi-res mode is active on the 5000, but somehow too "weak".
I got a idea and grab the lecroy ws3024 - it has (of course) a hi-res mode called "enhanced resolution" (siglent also).
Same settings but you can choose number of bits adding in 0.5 bit steps.
And this is what I from the lecroy get:



Nearly the same....

And now, again the lecroy:



Bingo, a nice sinewave....
What have I done ? Right, the full resolution taken, in this case +3bit, the lecroy works then with 8+3bits  - this is the hi-res mode.

So what does it means...
It seems, the rigol add max. 1.5bits only - and NOT 4bits as mentioned in the changes of the firmware ("12bit resolution").
I guess, that´s the reason for seeing nearly nothing changing in hi-res mode.

Martin

Edit : Another example, a pfc-current via 1ohm resistor, quiet noisy….

MSO 5000, Normal Mode:



Hi-Res Mode:



Same Signal, but now the siglent sds :

Normal Mode:




Hi-Res Mode (+3bits) :



THIS is what I expect from a hi-res mode….

[Commodore8888]

Hi,
As I told, rigol anwered me today...very nice.

They already know about and work on it...

So it´s an "official" bug with the overshoots…

Martin

Here's hoping it's just a few tweaks in #defines.

(The pessimist in me thinks it still means hauling a van load of scopes to UPS, and dying in a fire to shipping fees  )

[mabl]

Hi guys,

I recently received my 5074. While trying to compensate the 10x probes, I'm unable to get the perfect square shape of the signal. On channels 1,2,4 I always get some overcompensation. On channel 3 it looks perfect. I've attached pictures to illustrate. Is it normal or should I return it, while still can?

Thanks!

Interestingly, I observed the same overshoots after upgrading to 01.01.04.04 and doing a calibration. Before everything look very nice. (It also appears in the non-hacked firmware)

[Martin72]

Meanwhile I don´t have any Problems in this case, from the beginning till the upgrades ( to beta, to official).

[Fungus]

Wonky hi-res, hmm.

Thought about the "non-functional" hi-res mode…
What Do I expect from - Cleaner filtered waveforms.
What don´t happen if You choose the hi-res mode on the 5000 ?

Right, nothing:

What happens if you set memory depth to "auto"? Both of your choices might be outside the range that can be used for this.

[Martin72]

Interesting objection, test it out on monday - Scope is on work.
Although I did tests under various mem depth conditions already, but I´ll test it under auto again.
Nevertheless, I want the hi-res option selectable by 0.5 bits steps, as the lecroy and siglent models got.

[nctnico]

Interesting objection, test it out on monday - Scope is on work.
Although I did tests under various mem depth conditions already, but I´ll test it under auto again.
Nevertheless, I want the hi-res option selectable by 0.5 bits steps, as the lecroy and siglent models got.

This basically is the same as input signal filtering but with a variable cut-off frequency. I'd prefer the cut-off frequency to be constant over memory depth and samplerate. Do keep in mind that the signal will need a certain amount of noise for hires to actually work and not show fantasy signals. Hires isn't a replacement for having more bits from the ADC.

[David Hess]

Interesting objection, test it out on monday - Scope is on work.
Although I did tests under various mem depth conditions already, but I´ll test it under auto again.
Nevertheless, I want the hi-res option selectable by 0.5 bits steps, as the lecroy and siglent models got.

This basically is the same as input signal filtering but with a variable cut-off frequency. I'd prefer the cut-off frequency to be constant over memory depth and samplerate. Do keep in mind that the signal will need a certain amount of noise for hires to actually work and not show fantasy signals. Hires isn't a replacement for having more bits from the ADC.

There is almost always plenty of noise even after the vertical attenuation stages.  The only exceptions I have seen are in some old DSOs which used flash converters where the combined ADC and vertical signal chain noise was less than 1 LSB; on these if you adjust the position/offset just right, you can get a straight line with no or very few LSB transitions which might make you think the DSO is broken.  Feed these DSOs with a clean triangle wave and average to 16 bits and you get a nice picture of the ADC's differential non-linearity as a stair-step.

Do not forget that high resolution mode on earlier Rigol DSOs did not actually return higher resolution; the sample resolution remained 8 bits where the user might expect 16 bit samples.  Could this explain the discrepancy and limitation here?

This might be easy enough to check; is the record length halved when high resolution mode is used?  What about in average mode?

In high resolution mode, old DSOs just boxcar averaged the input samples to produce a higher resolution acquisition sample at whatever the oversampling ratio was which is relatively easy in limited hardware and produces consistent results.  Constant bandwidth would require at least a different FIR filter for every oversampling ratio which could get very messy at high ratios.

[Martin72]

Hires isn't a replacement for having more bits from the ADC.

I knew it already, but other brands could handle it as it allows to show noisy signals crisper, actually the rigol couldn´t do it.

This might be easy enough to check; is the record length halved when high resolution mode is used?  What about in average mode?

Will be checked on monday.

Edit :

Have a look at the pics:

https://www.eevblog.com/forum/blog/new-rigol-scope/msg2233980/#msg2233980

Record length won´t change or will not displayed.

[RedCali]

Hi All,

had anyone tried the "Color Grade" mode - setting is found in the display menu.
Just saw the scope on the fair in Nuremberg - and the Color Grad mode was buggy as well - just displayed a gray colored line if switch on the colr grad-  just works only once :-)
Thanks for testing!

[David Hess]

Hires isn't a replacement for having more bits from the ADC.

I knew it already, but other brands could handle it as it allows to show noisy signals crisper, actually the rigol couldn´t do it.

It is a great feature.  I first encountered it on the Tektronix TDS400 series of DSOs.

This might be easy enough to check; is the record length halved when high resolution mode is used?  What about in average mode?

Will be checked on monday.

Edit :

Have a look at the pics:

https://www.eevblog.com/forum/blog/new-rigol-scope/msg2233980/#msg2233980

Record length won´t change or will not displayed.

That probably means only 8 bit samples are returned from the digitizer in any mode so it is the same crippled high resolution and average modes as before in the Rigol DS1000Z series.  I doubt Rigol would sacrifice half of their marketed record length with the accompanying loss in speed or include twice as much memory and bandwidth like Tektronix did.

[mabl]

Interestingly, I observed the same overshoots after upgrading to 01.01.04.04 and doing a calibration. Before everything look very nice. (It also appears in the non-hacked firmware)

I copied my backuped calibration files back (all *.hex in /rigol/data) and the spikes are gone. So it is definitely something in the calibration routine.

EDIT: After using the old calibration files, I cannot reproduce the state with the extra peak anymore.  Strange. But at least it is okay now.

EDIT2:
Starting from the default calibration files in /rigol/default, I get the problematic calibration. So it looks like the current calibration effects the new calibration.
I believe my original calibration was overwritten during my upgrade, which then resulted in the problematic autocalibration.

EDIT3:
The problematic calibration is lfcal.hex. Just replacing that file gives perfectly shaped squares again.

[Fungus]

EDIT3:
The problematic calibration is lfcal.hex. Just replacing that file gives perfectly shaped squares again.

Sounds to me like they're driving one of the front-end opamps harder than before and creating a problem with overload/recovery.

It's good to know it can be fixed and it's not a hardware problem.

OTOH it needs fixing ASAP.

[Martin72]

Told them(rigol eu).
What I can´t complain in general:
Why do I haven´t these issues ?
Bought it, everything was fine - where others with the same firmware have the problem.
Then the new firmware is avaible, I did the upgrade - And still have no problems with overshoots.

[nctnico]

It may be that not every unit is affected in the same way. After all self-adjustment is to counteract component variations.

[2N3055]

EDIT3:
The problematic calibration is lfcal.hex. Just replacing that file gives perfectly shaped squares again.

Sounds to me like they're driving one of the front-end opamps harder than before and creating a problem with overload/recovery.

It's good to know it can be fixed and it's not a hardware problem.

OTOH it needs fixing ASAP.

Not necessarily. As i mentioned before, they might have some kind of signal equalisation that is tunable in software.
There are all kinds of bandwidth tuning, step response tuning, delay tuning etc. that can be done in front end and in pure digital domain.
It is a known concept.

Take a look at :
http://cdn.teledynelecroy.com/files/whitepapers/dsp_in_oscilloscopes.pdf
and
https://download.tek.com/document/55W_17589_2.pdf

And nctnico is right.  Self cal and tunable front end is there to achieve calibrated results despite component variations.
It is obvious that platform is fully capable of very advanced techniques. But they will not have as much experience doing this as those doing it for years.
So it might take them some time to get it right... Or not so much time.. We'll see..
It took R&S 2 years to make their new platform fairly stable and feature rich. They are still missing search over basic serial protocols, and they have no search whatsoever over massive segmented memory that makes it pretty much useless, except for few tasks.

Rigol is still in a very early production phase where they still get the benefit of the doubt, and a pretty good chance they will make it work decently.

[Martin72]

I am excited about what rigol will do with the not really workable hi-res mode….
It even don´t work proper on the DS1000Z models and they are from 2015.

[offmar]

Told them(rigol eu).
What I can´t complain in general:
Why do I haven´t these issues ?
Bought it, everything was fine - where others with the same firmware have the problem.
Then the new firmware is avaible, I did the upgrade - And still have no problems with overshoots.

I will check later to be sure, but I suspect that auto-cal does nothing... I have the overshoots on 3 channels and nothing changes when I use auto-cal. I even did it while I had input signals fed to all channels - the result didn’t change, everything looked as before, and I suppose that in that scenario the ‘scope should have lost calibration.