Rigol DS1000Z series (ds1054z, ds1074z, ds1104z and -s models) Bugs/Wish List

[Fungus]

Doesn't the Rigol position control shift the ground level?  Position controls usually operate as offset controls with the position signal added before the digitizer.  The difference is whether the signal is added before or after the attenuators.

Screenshot attached how it works on Rigol.

So... it's well within specification? 

[JPortici]

As I said, it honestly never occurred to me to look at a rise time without zooming in on the wave. Isn't it 100% natural to zoom in and look at the rising edge when you measure it? Make sure there's nothing weird there? 

If Picoscope owners aren't doing that then it says a lot about their user interface.

i understand that you either don't get it or you need to be fed like a good troll

if you have acquired at the correct sample rate AND your math is calculated from memory, not from display, zooming doesn't do shit to make the measurement more precise/true. you can zoom in to check if the waveform is nice and clean but beside that there is no need for correct calculation

(as for your beloved fiesta, it's like it advertised/it's written that it does 0-100 in 10 seconds and if you're alone it's fine, but if you are in two it takes 15 seconds because the engine can't keep up with a bit more load)

[Fungus]

i understand that you either don't get it or you need to be fed like a good troll

if you have acquired at the correct sample rate AND your math is calculated from memory, not from display, zooming doesn't do shit to make the measurement more precise/true. you can zoom in to check if the waveform is nice and clean but beside that there is no need for correct calculation

I understand it perfectly, I just don't see it as a big deal.

(the fact that nobody has ever complained about it here makes me suspect I'm not alone, either...)

The only way this "debate" would be interesting is if we examined a whole load of other oscilloscopes to see which ones do this and which ones don't. At what price range do they start to do it 'properly'? etc.

[JPortici]

PC: Remind us again the price of a 4-channel, 100Mhz Picoscope. I forgot.

If we talk non-hacked, diff is not that big. If we talk hacked, low-end VHF and only analog work, maybe yea (that's by I bought Z-box). If we talk lower freq then Picos just wipe the floor with Z for lower price.

https://www.picotech.com/oscilloscope/2000/picoscope-2000-overview
100 MHz, 4 channels - $1235
50 MHz, 4 channels - $659

Is that the pricing we are talking about? Then I don't get your "wipe the floor for lower price" comment, even looking at a stock DS1000Z in comparison. And if I compare a hacked DS1054Z to a $1235 Picoscope then - "maybe yea", as you put it, there might be just a tiny price advantage indeed...

if i wanted to be picky i'd say that you also have a signal generator w/ AWG (albeit a bit crappy). add that to the cost
also more protocols decoding that any rigol combined, ets, eye diagram, reference waveform, segmented memory, advanced math. And if you're a programmer and you have the need, you can add other stuff via the SDK.

[Fungus]

if i wanted to be picky i'd say that you also have a signal generator w/ AWG (albeit a bit crappy). add that to the cost
also more protocols decoding that any rigol combined, ets, eye diagram, reference waveform, segmented memory, advanced math.

Sure! At two to three times the price you should demand more from your equipment.

And if you're a programmer and you have the need, you can add other stuff via the SDK.

You can also connect your DS1054Z up to Matlab (or whatever).

[nctnico]

As I said, it honestly never occurred to me to look at a rise time without zooming in on the wave. Isn't it 100% natural to zoom in and look at the rising edge when you measure it? Make sure there's nothing weird there? 

if you have acquired at the correct sample rate AND your math is calculated from memory, not from display, zooming doesn't do shit to make the measurement more precise/true. you can zoom in to check if the waveform is nice and clean but beside that there is no need for correct calculation

This varies from brand to brand. Tektronix for example uses the memory for math but Keysight uses the screen data. On a Keysight scope you'll get different results at various zoom levels. Whether this is good or bad is debatable and depends much on the kind of measurement. When it comes to risetime you may want to select a specific edge, when it comes to RMS or some average you may want to calculate it over as many cycles as possible.

[metrologist]

This varies from brand to brand. Tektronix for example uses the memory for math but Keysight uses the screen data. On a Keysight scope you'll get different results at various zoom levels. Whether this is good or bad is debatable and depends much on the kind of measurement.

So this statement basically nullifies all of MrWolf's arguments? Whether it is wrong or not?

I appreciate the discussion and opportunity to learn more about how to use my instrument to get the most out of it. 

[Fungus]

This varies from brand to brand. Tektronix for example uses the memory for math but Keysight uses the screen data. On a Keysight scope you'll get different results at various zoom levels.

So this statement basically nullifies all of MrWolf's arguments? Whether it is wrong or not?

Basically, yes.

People like to point the finger at the DS1054Z but if you take a sample size greater than 'one' you usually find that the DS1054Z isn't as black as they're trying to paint it.

That's why I said above: "The only way this 'debate' would be interesting is if we examined a whole load of other oscilloscopes to see which ones do this and which ones don't. At what price range do they start to do it 'properly'? etc."

[metrologist]

Yeah, but if that's the way Keysight does it...is it really wrong or worth concern? I'm sure they know more about it than I do.

I'd be happy comparing the top (or bottom?) 3 or 4.

[David Hess]

I mentioned several times that people are free to pay more money and get this problem "fixed". Going on an on about it is like complaining that Ford Fiestas can't do 200mph. Technically true, but not going to make headline news or scandalize anybody.

But I expect the Ford Fiesta to operate as a car doing car things within its specifications.  I do not expect the manufacturer of my car or DSO to lie or mislead about its capabilities.  (1) Incidently, some cars like the Fiat Chrysler Jeep Grand Cherokee which killed Anton Yelchin fail this test because of poor human factors engineering; they are defective by design.

To give a single example (I could have picked from several), the Rigol does not support RMS measurements and it apparently never did even when it was working.  I do not know what the automatic RMS measurement is measuring, but it is not the RMS value of the signal.  To verify this, apply some Gaussian noise and measure the RMS value; it will be wrong.  Now I know *why* the RMS measurement is broken but I only recently discovered it after becoming increasingly suspicious; the processing necessary to produce the display record corrupts the RMS value and measurements are made on the display record.

And it is not like digital RMS measurements are difficult; they are trivial.  Calculate the standard deviation and you get the RMS value.  The icing on the cake is that an *analog* oscilloscope can make this measurement to high accuracy.  I expect my Ford Fiesta to carry my ESI 250DA impedance bridge as cargo just as well as other much older vehicles that I have used.

As I said, it honestly never occurred to me to look at a rise time without zooming in on the wave. Isn't it 100% natural to zoom in and look at the rising edge when you measure it? Make sure there's nothing weird there? 

It is natural to do this for the very reason you identify; to know exactly what is being measured gives confidence in the measurement.  Oscilloscopes are particularly useful tools in this respect.

On the other hand if the measurement changes when it should not, that undermines confidence.  The DSOs I have used made accurate transition time and other measurements over a broad range of time/div and volt/div settings; if a weird result was produced, it was always because the signal was weird and not because of the instrument.  They even tailored the number of significant digits reported to reflect the measurement precision.

Doesn't the Rigol position control shift the ground level?  Position controls usually operate as offset controls with the position signal added before the digitizer.  The difference is whether the signal is added before or after the attenuators.

Screenshot attached how it works on Rigol.

That does not really show me how the Rigol position controls work.

Position controls work over a specified number of graticule divisions.  When the volt/div setting changes, the range of the position control in volts/div changes so the trace does not shift.  If the position control operates by injecting a signal directly before the digitizer instead of only adjusting the display, then it serves as a very coarse offset control increasing the input range of the DSO.

Offset controls inject a signal earlier in the signal chain before some or all of the attenuation and gain stages.  This increases the input range of the digitizer enormously which is important in some applications.  In an extreme case like the Tektronix 7A13 differential comparator, the offset control can operate over a range of 20,000 divisions (it really operates like the display is 20,000 divisions tall) while the position control still operates only over 12 divisions of an 8 division display.

I think people would be complaining if the Rigol's position control did not inject an analog signal before the digitizer and only operated through the display because the limited input range would be a major problem.  With a x10 probe yielding 50V/div, measuring the DC output of the input filter of an off-line switching power supply (340 volts DC) requires 7 divisions which is just within range of an 8 division oscilloscope (2) *if* the position control works as a fixed offset control.  If the position control only changes how the signal is displayed, then the DSO would require 14 vertical divisions.
 
(1) Of course car manufactures do this all the time.  That is why I hope to have nothing to do with GMC ever again.

(2) I am well acquainted with this measurement because with 1 or 2 very old exceptions, all of my analog and digital oscilloscopes have 8 vertical divisions and just barely make this measurement using only their position controls.  I would be surprised if any DSO trying to stay out of the toy catagory cannot do this.  Some of my oscilloscopes can make this measurement to very high precision and accuracy using their offset controls; a 7A13 can make a 340 volt DC measurement at 100mV/div.

[Fungus]

Offset controls inject a signal earlier in the signal chain before some or all of the attenuation and gain stages.

You can see the Rigol doing this when you run the calibration. The traces move up and down the screen looking for 'zero'.

[tautech]

Then you could spend all your time complaining it was too slow, right? 

Well I'm just learning about my new DSO. Seems indeed it's some new kind of DSO, I will name it SCREEN SAMPLING DSO, yeah    Sadly in the manual does not teach how to account for differences compared to normal DSO operation. You see I've used only normal ones so far and when buying this one was looking only at analog bw, memory sample rate and channel count. But there's another and unique param - "screen sampling rate" - very important for risetime auto-calc!

So now I must create helper formulas and tables to navigate around it's limitations. Would much prefer for it just to be slow. Slow computes well with measurement instrument. Readings off by 4 orders of magnitude do not (in the situation where normal DSO wouldnt be even stretched).

We don't know that with 100% certainty but that's what we observe and it makes sense to do it that way.

Let's observe some raw data then. I adore raw data, it makes s*it float
Reports attached below.

Short summary:
Fed 10Vpp, 32768Hz 50% duty square signal into Rigol DS1000Z and my old PicoScope 2205 MSO (25MHz, 2ch+16ch, 2013 out of prod model).
In case of normal DSO it can be seen readings reflect the nature of signal rather well. Lowest sampling rate observed at the 5ms timebase was 0.9804MS/s.
However in the case of "screen sampling" DSO it can be seen that despite hardware superiority it's readings are up to 4 orders of magnitude off. Despite running at massive 125MS/s at 5ms timebase in reality it's "screen sampling rate" for risetime calculation is only 0.005MS/s. Does a bit better with period calculation, only by about order of magnitude off.
So clearly it cannot be used anything like normal DSO. But then manual should reflect it. Customer is not to be expected reading forums for weeks when buying equipment.

rf-loop has duplicated your risetime tests with a X series Siglent for comparison here:
https://www.eevblog.com/forum/testgear/siglent-sds1000x-series-oscilloscopes/msg1091032/#msg1091032

[Fungus]

rf-loop has duplicated your risetime tests with a X series Siglent for comparison here:
https://www.eevblog.com/forum/testgear/siglent-sds1000x-series-oscilloscopes/msg1091032/#msg1091032

So... Rigol does it, Siglent does it, Keysight does it. It seems that this is the way DSOs work. 

(ntnico's sig is correct when it says, "There are small lies, big lies and then there is what is on the screen of your oscilloscope.")

PS: Other 'scope owners are still invited to participate in this. Just connect the probe to the test signal and zoom in/out. Does the displayed "rise time" change?

Anybody here own an R&S? Where are the a GW-Instek owners when you need them...?

[rf-loop]

So... Rigol does it, Siglent does it,...

Yes, and difference is only small "nitpick" class difference.  1 vs 100

[Fungus]

So... Rigol does it, Siglent does it,...

Yes, and difference is only small "nitpick" class difference.  1 vs 100

Is the number correct? No.

(and honestly, if it's going to be incorrect then less correct is better - it stands out more...)

[rf-loop]

and honestly, if it's going to be incorrect then less correct is better - it stands out more...

 

Best of day. But it is good: old peoples say that "laughing makes you live longer".

[MrWolf]

OMG, then I've been using technology of the gods thinking it's normal human tech! It's like seeing forest miles away and having capability to count all the needles in an instant... 
Originally I bought Picos for doing work on experimental electric motor. There it proved extremely useful to monitor both slow (mechanical) and fast (electrical) processes at the same time to a high accuracy. I had no idea that this capability is beyond state of the art in DSOs... But yolo, enuff cockfighting 

Since "Wolfs test" seems to characterize scope capabilities rather well (in fact much better than datasheets) and there is some public interest. What if I do separate thread with clear instructions how to do test and ready-to-download unfilled tables?

[Fungus]

and honestly, if it's going to be incorrect then less correct is better - it stands out more...

Best of day. But it is good: old peoples say that "laughing makes you live longer".

I'm perfectly serious.

(and one day you'll realize I'm right....)

[Fungus]

Since "Wolfs test" seems to characterize scope capabilities rather well

Not really.

It should be somewhere very close to the bottom of the list if you're making a purchasing decision.

[MrWolf]

The only way this "debate" would be interesting is if we examined a whole load of other oscilloscopes to see which ones do this and which ones don't. At what price range do they start to do it 'properly'? etc.

  ...but next thing you say...

It should be somewhere very close to the bottom of the list if you're making a purchasing decision.

Purchasing decision? Are you a salesman or something? 
What if you already purchased something and just want to know it's real specs regarding auto-measurements?
I'll be damn interested what is actual precision of my instruments since in scope business it seems
to be norm to forget "<" sign before measurement value where acutely needed...
Multimeters etc do it in even more logical way. Usually they just start lowering reading precision (count) and
finally display "out of range".

[Fungus]

It should be somewhere very close to the bottom of the list if you're making a purchasing decision.

Purchasing decision? Are you a salesman or something? 

I forgot, you're new here.

The way it works here is:
a) A newbie comes in and asks what 'scope to buy for $400
b) A few people reply "The DS1054Z"
c) All the DS1054Z haters say "Oh suuuure, so long as you don't want to measure rise times (snigger)".
d) I say "Ignore the haters, they're not telling the truth!"
e) I get accused of being a crazy DS1054Z fanboi.
f) Signal to noise ratio tends towards zero.

[ebastler]

Hey all,

This thread originally started as a pretty concise collection of bug reports and improvement suggestions for the DS1000Z scope family.  I think we had hopes that Rigol actually reads it, and takes some cues from it. (And they may actually have done so!)

Yet, this has increasingly morphed into yet another "Rigol is fine" vs. "Rigol sucks" thread. Three full pages so far on the question whether or not it is critical that measurements are derived from on-screen data only... 

Can't we move this type of discussion to separate threads, or to the general DS1054Z rant/question monster thread? Please?

Regards,
Juergen

[Fungus]

What if you already purchased something and just want to know it's real specs regarding auto-measurements?

Objectively characterizing things is good.

Listing only the bad things? Going on endlessly about things with only 0.001% real importance? Not so much.

I'll be damn interested what is actual precision of my instruments since in scope business it seems
to be norm to forget "<" sign before measurement value where acutely needed...

I agree that it could probably display "?zoom" or something.

I've never said the DS1054Z is perfect, just that it's the best all-round value for money under ~$1200.

That's quite something when you only paid $400 for it, IMHO.

[nctnico]

rf-loop has duplicated your risetime tests with a X series Siglent for comparison here:
https://www.eevblog.com/forum/testgear/siglent-sds1000x-series-oscilloscopes/msg1091032/#msg1091032

So... Rigol does it, Siglent does it, Keysight does it. It seems that this is the way DSOs work. 

(ntnico's sig is correct when it says, "There are small lies, big lies and then there is what is on the screen of your oscilloscope.")

PS: Other 'scope owners are still invited to participate in this. Just connect the probe to the test signal and zoom in/out. Does the displayed "rise time" change?

Anybody here own an R&S? Where are the a GW-Instek owners when you need them...?

On my GDS2204E the risetime measurement is pretty much constant unless the timebase gets to >200us/div with 10Mpts record length and a 5MHz square wave (approx. 7ns risetime). At that point the samplerate is still 1Gs/s so it shouldn't matter but it does.

[Fungus]

Anybody here own an R&S? Where are the a GW-Instek owners when you need them...?

On my GDS2204E the risetime measurement is pretty much constant unless the timebase gets to >200us/div with 10Mpts record length and a 5MHz square wave (approx. 7ns risetime).

So.... it isn't constant. Is that what you're saying?

At that point the samplerate is still 1Gs/s so it shouldn't matter but it does.

Yes, that's the point of this test.