New Rigol DS1054Z oscilloscope

[Sredni]

How do we get twenty functions into two buttons? 

With a bonsUI. 

[ebastler]

But how am I ever going to be satisfied with my Rigol's UI now... it doesn't even include "Whack" triggering !

Right -- instead it has that old-fashioned "force trigger" button...
Much less intuitive, and lacking the haptic feedback. Disapponting! 

[MrW0lf]

At least it has "Russian trigger". In Russia they use it to trigger reception on old CRT TV and many other devices including drunk mates
https://youtu.be/WE9pYUVvr00?t=10m49s

[David Hess]

While it is not usually a problem with low bandwidth oscilloscopes, the relationship between the -3dB bandwidth and transition time depends on the shape of the passband so measuring only the transition time may not be sufficient.

Actually I did it properly also, with ETS and all, real -3dB point is 140MHz, cross checks with pure sine.
https://www.eevblog.com/forum/testgear/picoscope-2000/msg1153233/#msg1153233
Mentioned 200MHz with square on purpose, because Z-box owners often do not have 100MHz+ high fidelity sine sources and will do their bw testing with square. Point being that if classical frontend unhacked 100MHz will do -3dB at 200MHz with square, then 50MHz would probably do -3dB at 100MHz etc...

This does not follow.  I did the bandwidth calculations based on Dave's reverse engineered schematics and posted the results on the forum (good luck finding the post though) for the DS1000Z series and they have no relationship to the maximum bandwidth.  They agreed closely with the specifications.

If you want to know the bandwidth with different settings of the bandwidth limit filters, then measure them.  Measure the transition times and transient response as well.

Unfortunately as you point out, most DS1000Z series users are not equipped to make useful measurements of bandwidth, transition time, or transient response.  Back when this was an issue for me as well, one of my early acquisitions was a sampling oscilloscope so that I could calibrate my sources used for these measurements.

Meaning its not like you cannot measure timing related things @100MHz quite ok with any proper 50MHz scope.

Sure you can.  (1) I do not use my 50 MHz oscilloscope very much except for crushing zombies but often use my 100 MHz oscilloscopes to measure timing (but not transition time) down to 1 nanosecond.

In short: this whole bw hackability thing is bit overrated in this case (especially considering Sinc trickery).  Real point of hacking is IMHO more in getting all other stuff unlocked.

I am aware of the sinc trickery in these oscilloscopes but I suspect it has more to do with Rigol sacrificing accuracy for update rate.

The reason I suggest making both measurements especially with a DS1054Z hacked to 100 MHz is that if the hacked passband is not a single pole rolloff which is likely, (2) then its 3dB bandwidth will not match its transition time based on the 0.35 rule.  Lots of much higher bandwidth oscilloscopes are this way but it is unusual at 100 MHz.

(1)  "A 50MHz oscilloscope cannot track a 5ns rise time pulse, but it can measure a 2ns delay between two such events." - Jim Williams, Linear Technology application note 47, page 20.

(2) I base this off of Rigol's transistor selection and topology of the input amplifier shown in the reverse engineered schematics that Dave provided.  The front end has an unusual arrangement of switched equalization that implies Rigol had problems meeting the 100 MHz bandwidth specification.  I would expect this to result in a peaked response compromising the transient response among other things and some of the published transient response tests do show something unusual going on in the bandwidth hacked oscilloscopes.

[MrW0lf]

This does not follow.

Maybe little mixup. Test was with my new 100MHz Pico, which has "classic" frontend AFAIK.
However dug up my old DS1054Z (hacked) test:



If just look at it and presume Sinc=OFF is actual analog response of DS1000Z then its roughly -5dB at @100MHz, which they compensate with non-standard Sinc producing bit silly amplitude gain @100MHz with Sinc=ON. Input in this test is perfect sine. There is theory around that Sinc=OFF applies some lowpass filter but I do not buy it. Rather Sinc=ON applies artificial bandwidth boost. Actually quite logical thing to do at first glance but easy to lose track with reality of actual input signal...

[frozenfrogz]

It all comes down to what you need it for. Me, I am into tinkering with the lower range of MCUs and most of my troubleshooting involves looking at signals in the kHz range. The four channels of the Rigol come in very handy for these tasks and on-screen decryption of serial protocol, SPI and I2C is a cool feature to have (yes, a Salea logic analyzer might be suitable as well, but that is another story).
For me, this is somewhere between a hobby and my profession as I do not develop electronics for a living, but I am learning a lot from tinkering what I can then incorporate into useful conversation with the actual developing engineers.

For professional use, there is always the range of bigger, better, more reliable hardware you should rely on. There is no free lunch. Cheap tools might get the job done, but I still need to see one example, where this has led to actually saving money. In the maker community niche on the other hand it’s 400€ well spent (and can not be specified as a cheap tool IMO).

[MrW0lf]

It all comes down to what you need it for.

Quite correct, indeed helpful device for mainly digital tinkering. However time to time its described as real 100MHz analog scope and cannot fully agree with that.

[David Hess]

If just look at it and presume Sinc=OFF is actual analog response of DS1000Z then its roughly -5dB at @100MHz, which they compensate with non-standard Sinc producing bit silly amplitude gain @100MHz with Sinc=ON. Input in this test is perfect sine. There is theory around that Sinc=OFF applies some lowpass filter but I do not buy it. Rather Sinc=ON applies artificial bandwidth boost. Actually quite logical thing to do at first glance but easy to lose track with reality of actual input signal...

Are you making the measurement through the Rigol's automatic measurement functions?  That might be an additional problem with Rigol's display record measurements.

Sinc interpolation should have no effect on a 3dB bandwidth measurement but as you point out, Rigol's sinc interpolation is questionable.  It should not affect the bandwidth measurement even if aliasing or undersampling is present; undersampling to produce a low frequency signal is the same as a real low frequency signal except for how the analog frequency response affects it.  (1) Deliberate aliasing on the Rigol might even be the way to go to avoid problems with its weird sinc filter but that is not going to be much help if the results are different than non-aliased usage.

(1) This can be used for calibrating DSO digitizers but the procedure is headache inducing.

[CatalinaWOW]

Another way to judge it is that it is a huge step up from my venerable Heathkit 2 channel scope, with sort of 2 MHz bandwidth.  Costs a lot less than the Heathkit did if you adjust for inflation, and I didn't have to assemble it.  Is it a perfect scope.  No way.  But then I haven't encountered one of those at work or anywhere else, at any price.

[Housedad]

It is not only what you want it for, but also what are your expectations.   For someone that has used and /or owned new or used top end scopes, it may very well be a letdown.  If you are like me, where   the last scope you used and owned was a Tek 535a 30 years ago, the Rigol is a incredible thing to use.  I'm pleased with the scope.  Of course I see better scopes out there, but the Rigol was affordable to me now.    In the future, I will probably outgrow it and want something better, but it will fill my needs learning and most things fairly well.

[ebastler]

If just look at it and presume Sinc=OFF is actual analog response of DS1000Z then its roughly -5dB at @100MHz, which they compensate with non-standard Sinc producing bit silly amplitude gain @100MHz with Sinc=ON. Input in this test is perfect sine. There is theory around that Sinc=OFF applies some lowpass filter but I do not buy it. Rather Sinc=ON applies artificial bandwidth boost. Actually quite logical thing to do at first glance but easy to lose track with reality of actual input signal...

I don't understand what problem you see here? You have chosen to sample at only 2.5x the signal frequency (at 100 MHz), so of course the sin(x)/x correction becomes relevant. If you switch it off, I assume the scope reconstructs the signal by simple "connect the dots" interpolation of the samples. That, of course, will reduce the apparent amplitude since the peaks are not properly reconstructed. This has nothing to do with "some lowpass filter", but only with the crude interpolation that is used (by definition) in the absence of sin(x)/x interpolation. Once you enable sin(x)/x interpolation, the signal and its amplitude get properly reconstructed, as expected.

If you want to assess the bandwidth of the scope's analog front end, I would suggest enabling a single channel only, to sample at 1 GHz. That will minimize the effects of interpolation.

[JPortici]

I don't understand what problem you see here? You have chosen to sample at only 2.5x the signal frequency (at 100 MHz), so of course the sin(x)/x correction becomes relevant. If you switch it off, I assume the scope reconstructs the signal by simple "connect the dots" interpolation of the samples.

except it doesn't

[ebastler]

I don't understand what problem you see here? You have chosen to sample at only 2.5x the signal frequency (at 100 MHz), so of course the sin(x)/x correction becomes relevant. If you switch it off, I assume the scope reconstructs the signal by simple "connect the dots" interpolation of the samples.

except it doesn't

Could you elaborate on that, please? My comment was intended to be constructive; yours does not help me so far. Thanks for explaining.

[rf-loop]

Rigol Z box "Sin(x)/x" works totally wrong. THis is demonstrated many times. Starting with "famous" toy DS1052E.  They have always done it wrong and they do not listen anybody because in Beijing they think that Beijing man is always right and others wrong. 

Rigol flush out real sample points. They violate all basics how real Sinc function need work.

Ten years they knock they head to great wall and never listen anybody for repair it.

In T&M instrument this is unbelievable error. Made for nice image instead of try follow data. Perhaps they are "green" and thay have adopted "selected  facts" policy.

I will not anymore repeat my tests for show this error again and again.

In case when sinc do not draw thru real raw sample points it is joke. I understand if this is art drawing machine but they call this test and measurement instrument. They do not respect real ADC samples - only true data.   Least they need correct menu text. Instead of Sin(x)/x there need read Smooth.

Show even one evidence it works right.

More fun, user can not even turn off this bullshit. (exept if 4 channel simultaneously in use - afaik)
How about dots mode. Do you see alone and only ADC produced real raw sample dots or...

[tautech]

Could you elaborate on that, please? My comment was intended to be constructive; yours does not help me so far. Thanks for explaining.

I see you've found rf-loop's work and for those that have not seen it already:
https://www.eevblog.com/forum/testgear/rigol-ds1074z-weird-signal-level-problem/

It just follows on from other work this long sever member did many years ago, I think in his first post on the forum.
https://www.eevblog.com/forum/chat/rigol-ds1000e-series-possible-errorfail-in-sin(x)x-interpolation/

[JPortici]

I don't understand what problem you see here? You have chosen to sample at only 2.5x the signal frequency (at 100 MHz), so of course the sin(x)/x correction becomes relevant. If you switch it off, I assume the scope reconstructs the signal by simple "connect the dots" interpolation of the samples.

except it doesn't

Could you elaborate on that, please? My comment was intended to be constructive; yours does not help me so far. Thanks for explaining.

yes, sorry for that. the "sinc interpolation" button is enabled only if you have more than two channels active, otherwise it says "ON" and is greyed out.
pressing it doesn't do anything anyway, as the sinc interpolation is not turned off: you can't have linear interpolation on this scope.
then rf-loop and others will explain in more detail how it's flawed but basically not only you can't turn it off (other than switching from vectors to dots), the reconstructed line doesn't pass through the samples

[Howardlong]

Yep. If it wasn't hackable we'd all be buying the $400 GW-Instek.

GWI 1054B has surprisingly high non-hacked bandwidth. Same with my Pico. Very sharp edge square will show -3dB only at 200MHz, on 100MHz unhacked scope (Many Rigol BW claims are made using square also).

Perhaps I've misunderstood what you're trying to say, but that's not how bandwidth is traditionally measured.

Either the 3dB point of a sine wave is used, or the rise time of a rectangular wave is used to derive it.

While it is not usually a problem with low bandwidth oscilloscopes, the relationship between the -3dB bandwidth and transition time depends on the shape of the passband so measuring only the transition time may not be sufficient.

Indeed, that is why I stated "is used to derive it", to be deliberately vague. Now I could have gone on about Gaussian and brick wall filters, and I did originally write a clause to that effect, but felt that, in view of the level of the discussion, it was adding unnecessary confusion so I removed it. There is an Agilent document describing it 5988-8008EN.pdf

[MrW0lf]

Once you enable sin(x)/x interpolation, the signal and its amplitude get properly reconstructed, as expected.

Observe, normal 100MHz scope with exactly same sampling rate options 1-0.5-0.25GSa/s + 20GSa/s ETS:
https://www.eevblog.com/forum/testgear/picoscope-2000/msg1153800/#msg1153800
Relative amplitude error @100MHz below 0.5dB across all sampling rates. Difference @250MSa/s between Sinc ON|OFF is below 0.025dB or 0.3%. Which can only mean that in both cases it does calculation based off actual sample points and Sinc is just for CGI show (like it should be).

However theres another matter... I suddenly remembered that Rigol has quite long autocal procedure which is done with open BNC connectors. I seem to remember people have gotten less-than-perfect cal results. Now with "russian trigger" feature on the table. What if you cannot fire machine guns and maybe even stroke a cat on table where DS1000Z is doing autocal? Would do experiment but do not own unit any longer.
When I did autocal back in the day I instinctively did leave it alone in peace & quiet and cal turned out pretty ok.
Edit: In other thread there is opinion that this is not issue, because hopefully all cal related is done with internal low impedance sources. So dunno, many other gear has "open" part in non full-auto cal procedure 

[ebastler]

Thanks rf-loop and tautech! I have read up on rf-loop's analysis now, and agree that the DS1054Z implementation seems very wrong indeed.

• Fudging the individual dots in sin(x)/x mode is obviously not what one would expect (are they doing the correct sin(x)/x interpolation behind the scenes, and then displaying the data points from the interpolation, instead of the raw sample data?!)
• Showing smooth curves in vector mode when sin(x)/x is OFF is also disconcerting. No idea how those are derived.
• The actual sin(x)/x interpolation might well be correct, as far as I have seen -- or is there evidence to the contrary?

Strangely enough, Rigol seems to get it right on the DS2000 series:
https://www.eevblog.com/forum/testgear/new-rigol-ds1054z-oscilloscope/msg519934/#msg519934

[Howardlong]

It is not only what you want it for, but also what are your expectations.   For someone that has used and /or owned new or used top end scopes, it may very well be a letdown.  If you are like me, where   the last scope you used and owned was a Tek 535a 30 years ago, the Rigol is a incredible thing to use.  I'm pleased with the scope.  Of course I see better scopes out there, but the Rigol was affordable to me now.    In the future, I will probably outgrow it and want something better, but it will fill my needs learning and most things fairly well.

Actually I think it works both ways. The Rigol (MSO1074Z-S) is my field scope/swiss army knife, it works well enough for almost all field work I do. At the bench I have a selection of scopes well beyond the capabilities of the Rigol, but none of them are as compact, convenient and feature packed for the field. Yes the Rigol has its limitations, and I've found a few bugs myself, but I've found similar unwanted features on top tier TE too.

I'm pretty sure most of us are able to know when something doesn't look right, and know to question and validate the results. Whether it's a Rigol or Keysight/Tek/LeCroy/R&S, you should always be in a position to be able to know about what to expect, and use the scope to verify. When it doesn't look right you need to be able to analyse and troubleshoot your findings. Most of the time it will be the DUT itself, but sometimes anomalies will be a problem with your test regime, and occasionally that will be the tool you are using. I am certain most of us are capable of figuring out which it is.

[JPortici]

Instead i think that many of us (hanging in this thread) do not know if something look right or not, i know i certainly didn't when i got my Z. Let's not forget this scope is aimed at hobbyist/students.. it is important to remind that there are flaws that cannot be swept under the rug by the "it's only 400" broom. The flaws are there, period.
you have to know which are so when you question what is on the screen you can decide if the flaw or the bug is affecting your measurement or not and in which amount.

stupid example, scope Z displays garbage on serial decode while scope P on same settings displays correct data. is scope Z bugged? no. it works differently. Now i know it and i can correct my settings accordingly, unless this way of working is not acceptable for me
and so on..

obviously (or not) this is true for all equipment.

[Fungus]

Instead i think that many of us (hanging in this thread) do not know if something look right or not, i know i certainly didn't when i got my Z. Let's not forget this scope is aimed at hobbyist/students.. it is important to remind that there are flaws that cannot be swept under the rug by the "it's only 400" broom. The flaws are there, period.
you have to know which are so when you question what is on the screen you can decide if the flaw or the bug is affecting your measurement or not and in which amount.

This is true of all test gear, not just the ones you have a personal mania against.

[Assafl]

I'm pretty sure most of us are able to know when something doesn't look right, and know to question and validate the results. Whether it's a Rigol or Keysight/Tek/LeCroy/R&S, you should always be in a position to be able to know about what to expect, and use the scope to verify. When it doesn't look right you need to be able to analyse and troubleshoot your findings. Most of the time it will be the DUT itself, but sometimes anomalies will be a problem with your test regime, and occasionally that will be the tool you are using. I am certain most of us are capable of figuring out which it is.

Well said - but it is always amazing how many people believe a display, especially a digital one.... I think it may be a religion (The LED G-d; or is it a segment g-d?).

In fact, I'd assume a well run university lab (especially an EE101 level) will want an arbitrarily crippled capability scope to test the knowledge of the students. Something for which you can easily build test scenarios (with cheap CMOS devices) where the scope will mislead you. What did you expect to see?; what did you see?; why?

Hypothesis: Perhaps some of the target market for the DS1Z is education?

(They did that to us at the Technion when scopes were all Tektronix and all analog and thus had few limitations, NOT; Alt; chop; triggering, BW, dynamic range etc... poor students... BTW - Doing labs with physicists are even worse: what are you measuring: are you sure that is the drift velocity of the minority carriers??? How are you sure?)

[JPortici]

Instead i think that many of us (hanging in this thread) do not know if something look right or not, i know i certainly didn't when i got my Z. Let's not forget this scope is aimed at hobbyist/students.. it is important to remind that there are flaws that cannot be swept under the rug by the "it's only 400" broom. The flaws are there, period.
you have to know which are so when you question what is on the screen you can decide if the flaw or the bug is affecting your measurement or not and in which amount.

This is true of all test gear, not just the ones you have a personal mania against.

correct. i'll edit previous post to add the obvious

[Assafl]

I am trying to rack my brain looking for an excellent piece of kit I worked with in the past. Just as a playful way to juxtapose the Rigol with "perfection".

I did semiconductor and RF work in the University - so Anritsu, HP, EESOF (prior to becoming HP), all kinds of power meters, Newport, Mini Circuits, Watkins Johnson, etc. Nada. All need tons of workarounds...

I did Naval testing for a long time: Tons of HP, Flukes, Bruel & Kjaer, Krohn-Hite, ITC, Tabor, and many others. Nada - We would amplify and then attenuate because the dynamic range of the inputs sucked etc.....

The only thing I can think that even comes close to being perfect (and even that suffers from parallax problems) is a 30cm ruler.

Can anyone think of a T&M equipment that isn't highly compromised (as in no bugs and no way to misinterpret the result etc.)?