Author Topic: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems  (Read 409963 times)

0 Members and 1 Guest are viewing this topic.

Offline eV1Te

  • Regular Contributor
  • *
  • Posts: 179
  • Country: se
  • Your trusted friend in science!
    • richardandersson.net
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #375 on: November 20, 2014, 10:33:27 pm »
BTW, as far as these bugs go - while they are certainly annoying and need to be fixed by Rigol - if I was a new owner of a DS1000Z series, I'd be more pissed about the fact that High Resolution mode doesn't work correctly (certainly not to 12 bits) than I would about these bugs.

I think the Hi-res mode on the D1000z series averages many measurements for each point (as it is supposed to do), but the resulting values are only 8 bits. So you get less noise due to the averaging, but you never get any higher resolution. (At least not on my DS1074z)
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2200
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #376 on: November 20, 2014, 10:40:32 pm »
No, it's really not that interesting if you understand how the DSO works...

Completely missed the point of my message. I've been saying the whole time, and in my video, that the behaviour of the DS2202 is not a bug, and in fact, superior to a standard analog CRO. Because an analog CRO waits for the sweep to complete before even starting the holdoff, whereas the holdoff on the DS2202 is correct (modulo some fixed error). This is demonstrated in my video.

OK, so that sets the scene: The DS2202 has either edge trigger, and it works beautifully. We're in agreement so far. The other thing I showed is that the triggering behaviour of the DS2202 is not affected in any way by the selected timebase. This is wonderful behaviour. Unless the demonstration I showed in the video was a magical frequency that just happened to behave consistently.

Then, wlanfox comes along and demonstrates that the MSO2202A behaves somewhat differently to the DS2202 -- the trigger behaviour is affected by the selected timebase, and adding one period's worth of holdoff (which should have no effect) has a huge effect. I claim this is very very strange and very very interesting, and I'd like to understand how this is happening (timebase affecting the triggering behaviour kind of makes sense, but the one period's holdoff having no effect makes no sense in any universe.)

To clarify: there is absolutely NO BUG involving the DS2000's Either Edge triggering - it works just fine.

Ya, never claimed otherwise.

...at least on the DS2202, because the EXT input is not available for triggering modes other than "Edge".
This is also wrong. It's available as a source to the "Pulse" trigger as well.

I just tried, and on my firmware, if you select "Pulse" trigger, the channel select turns into a toggle button (as opposed to a scrollable list), pressing it just flicks between CH1 and CH2?  :-//
« Last Edit: November 20, 2014, 10:45:15 pm by rs20 »
 

Offline i4004

  • Regular Contributor
  • *
  • Posts: 71
  • Country: hr
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #377 on: November 20, 2014, 10:50:52 pm »


scrolling i ment in a way the bug was originally reported ie

do you use it like that?
ie i'm asking do you capture it and inspect, or keep it "live" and scroll like in the video above...
« Last Edit: November 20, 2014, 10:53:51 pm by i4004 »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 32201
  • Country: au
    • EEVblog
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #378 on: November 20, 2014, 11:06:04 pm »
if I was a new owner of a DS1000Z series, I'd be more pissed about the fact that High Resolution mode doesn't work correctly (certainly not to 12 bits) than I would about these bugs.

What's that issue?
Sorry, I can't up to date with all the posts.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #379 on: November 21, 2014, 12:08:20 am »
I think the Hi-res mode on the D1000z series averages many measurements for each point (as it is supposed to do), but the resulting values are only 8 bits. So you get less noise due to the averaging, but you never get any higher resolution. (At least not on my DS1074z)

This is normal behavior in High Resolution mode since the DSO has to downsample for the 8-bit display memory as the final step anyway (although it would be handy if it saved the 12-bit values for offloading). But the problem I'm talking about is averaging to 12 bits to begin with - which the DS1000Z is not doing, as far as I've been able to discover so far.

Completely missed the point of my message.

Well, no, I didn't - but I'm guessing you didn't watch my video and so missed the point of my message.

Quote
The other thing I showed is that the triggering behaviour of the DS2202 is not affected in any way by the selected timebase.

Wrong. Watch my video - I change timebases and affect the triggering behavior clear as day on my DS2202.

Quote
Then, wlanfox comes along and demonstrates that the MSO2202A behaves somewhat differently to the DS2202 -- the trigger behaviour is affected by the selected timebase, and adding one period's worth of holdoff (which should have no effect) has a huge effect. I claim this is very very strange and very very interesting, and I'd like to understand how this is happening (timebase affecting the triggering behaviour kind of makes sense, but the one period's holdoff having no effect makes no sense in any universe.)

Again, watch the video; this is easily explained. EACH timebase has an acquisition cycle time - which includes the acquisition time plus the blind time - which is a frequency equivalent to the fastest possible waveforms per second. Frequency multiples falling just outside (or inside, depending on your point of view) of this cycle time will change the triggering behavior.

Quote
I just tried, and on my firmware, if you select "Pulse" trigger, the channel select turns into a toggle button (as opposed to a scrollable list), pressing it just flicks between CH1 and CH2?  :-//

You must be running old firmware.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #380 on: November 21, 2014, 12:17:40 am »
if I was a new owner of a DS1000Z series, I'd be more pissed about the fact that High Resolution mode doesn't work correctly (certainly not to 12 bits) than I would about these bugs.

What's that issue?
Sorry, I can't up to date with all the posts.

I've yet to discover a timebase/memory setting where the DS1000Z is actually doing 12-bit (256 samples) averaging. I'm not saying for certain that it doesn't happen, but it certainly isn't doing it at anywhere close to the same timebases as the DS2000 - and if it IS actually doing it, it's at very slow timebases.

This "vagueness" in the DS1000Z High Resolution mode is compounded by the way that Rigol defines it in the DS1000Z literature compared to the DS2000:

DS2000 User Manual/Datasheet:
High Resolution
12 bits of resolution when >= 5us/div @ 1 GSa/s (or >= 10us/div @ 500 MSa/s).

DS1000Z User Manual/Datasheet:
High Resolution:
The highest resolution is 12 bit

It would be nice to hear from Rigol the specs of this acquisition mode on the DS1000Z - and when it's actually averaging to 12 bits.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 32201
  • Country: au
    • EEVblog
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #381 on: November 21, 2014, 12:26:05 am »
I've yet to discover a timebase/memory setting where the DS1000Z is actually doing 12-bit (256 samples) averaging. I'm not saying for certain that it doesn't happen, but it certainly isn't doing it at anywhere close to the same timebases as the DS2000 - and if it IS actually doing it, it's at very slow timebases.
DS1000Z User Manual/Datasheet:
High Resolution:
The highest resolution is 12 bit
It would be nice to hear from Rigol the specs of this acquisition mode on the DS1000Z - and when it's actually averaging to 12 bits.

OK, but is this of any importance on a $399 entry level scope?
As long as it does some high resolution boxcar averaging, job done IMO.
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2200
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #382 on: November 21, 2014, 12:28:48 am »
Wrong. Watch my video - I change timebases and affect the triggering behavior clear as day on my DS2202.

Could you please link in this video? I've looked through the past month of your posts and nothing jumped out at me. I vaguely suspect we're talking about different things (waveform updates per second has virtually nothing* to do with what wlanfox and I are discussing [how dual-slope triggering looks], but that's pure idle speculation on my part until I can actually watch your video). A proper  understanding of the behaviour of the DS2202 triggering system must explain what's seen in both of our videos; I look forward to figuring this out.

EDIT: * Oh, unless the very short timebases I was playing with all have exactly the same "acquisition time" as you put it. That's another model that explains what I saw in my video. OK, that's easily tested by varying signal frequency. I feel I probably stand corrected after all.
« Last Edit: November 21, 2014, 12:34:10 am by rs20 »
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #383 on: November 21, 2014, 12:32:43 am »
OK, but is this of any importance on a $399 entry level scope?
As long as it does some high resolution boxcar averaging, job done IMO.

I would think the importance you attach to failures and bugs in features is proportional to the importance those features are to the work you're doing.
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 19713
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #384 on: November 21, 2014, 12:39:09 am »
OK, but is this of any importance on a $399 entry level scope?
As long as it does some high resolution boxcar averaging, job done IMO.

I would think the importance you attach to failures and bugs in features is proportional to the importance those features are to the work you're doing.
Agreed.
ANY function of a DSO that does not work as expected needs fixing and continued pressure on the manufacturer until it has.
If we collectively continue to accept these shortcomings God help us in the long term.
Avid Rabid Hobbyist
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #385 on: November 21, 2014, 12:42:54 am »
Could you please link in this video?

There's a link to the video in the first message to you I posted. And I think the graphic I posted shows fairly clearly how a frequency (or multiple of it) that has it's opposite edge falling within the blind time will cause the DSO to lock to the same edge over and over again (very bottom image labelled Either Edge Trigger). This, of course, changes with the timebase - since the blind time changes with the timebase.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 32201
  • Country: au
    • EEVblog
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #386 on: November 21, 2014, 12:57:54 am »
I would think the importance you attach to failures and bugs in features is proportional to the importance those features are to the work you're doing.

Sure, but if you buy a $399 scope and expect that to get the utmost in performance from the hires mod when all they can spec is "The highest resolution is 12 bit" (which I read as "upto 12 bits, maybe, at one single sweet spot when the moon is full, kinda sorta"), then you are probably expecting too much?
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #387 on: November 21, 2014, 01:22:35 am »
Sure, but if you buy a $399 scope and expect that to get the utmost in performance from the hires mod when all they can spec is "The highest resolution is 12 bit" (which I read as "upto 12 bits, maybe, at one single sweet spot when the moon is full, kinda sorta"), then you are probably expecting too much?

Absolutely; no doubt about that - although their reluctance to actually specify the parameters makes me wonder if it really does manage to do 12-bits at all - even when the moon is full  :)  - or whether that's their attempt to make it seem as if the feature is closer in performance to the other UltraVision models than it really is. I've tested timebases down to about 50 or 100ms and haven't seen any 12-bit averaging. But as you say, it does some averaging, so it's fine for most people - unless they really expected to be getting 12-bit equivalent resolution regularly (and really need that much filtering for their work).
 

Offline TMM

  • Frequent Contributor
  • **
  • Posts: 437
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #388 on: November 21, 2014, 07:34:50 am »
OK, but is this of any importance on a $399 entry level scope?
As long as it does some high resolution boxcar averaging, job done IMO.
The point is that the high resolution mode doesn't seem to work as you would expect. The DS1K doesn't seem to be adding the 1LSB of Gaussian noise needed in order to achieve higher than 8bit resolution. All it seems to achieve is a slightly less noisy 8bit signal, negating any benefit of having >256px vertical display resolution. If you add 1LSB of noise externally, it does achieve >8bit. edit: maybe not.

Compare it to this MDO3000 which has a beautifully working hires mode (jump to 4:50):


I'm well aware of the price difference between an MDO3000 and a DS1000Z, but Rigol are being cheeky to claim that box car averaging alone is 'high resolution mode'.
« Last Edit: November 21, 2014, 05:17:30 pm by TMM »
 

Offline thn788

  • Contributor
  • Posts: 30
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #389 on: November 21, 2014, 12:58:44 pm »
For those of us, who are not "well aware of the price difference between an MDO3000 and a DS1000Z": the Rigol DS1054Z costs $399, the Tek MDO3104, which is shown in the linked video, costs $13,900.  ;)
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #390 on: November 21, 2014, 01:31:04 pm »
The point is that the high resolution mode doesn't seem to work as you would expect. The DS1K doesn't seem to be adding the 1LSB of Gaussian noise needed in order to achieve higher than 8bit resolution. All it seems to achieve is a slightly less noisy 8bit signal, negating any benefit of having >256px vertical display resolution....
...but Rigol are being cheeky to claim that box car averaging alone is 'high resolution mode'.
I think you're confused between how you expect High Res mode should work, how it's actually supposed to work, and how well the DS1000Z achieves it's results.

Rigol is neither the source of claims about averaging - nor the company that first named it High Res (I believe that honor goes to Tektronix - see image below from 2003 Tek paper on acquisition modes).

There are papers detailing the effects of successive sample averaging (also called boxcar filtering or moving average filtering), such as:
http://www.daysalive.com/share/EffectsofAveragingtoRejectUnwanted Signals.pdf
...but the main points are that it removes noise by decreasing a DSO’s bandwidth (it applies an LPF function with a -3dB point approximated by 0.433*sample rate/averaged samples), achieving 1 bit of effective resolution improvement for each factor of four of averaged samples (e.g. 4 bits improvement = 256 averaged samples).

The DS2000 does a good job of implementing 256 sample averaging in real-time (functioning almost identically to the Agilent X-Series), but the DS1000Z does not - I'm guessing because the engine is not fast enough to support it (except perhaps at very slow timebases). The difference between the two DSOs is quite apparent, so the problem isn't with the technique but the implementation on the DS1000Z.
« Last Edit: November 21, 2014, 01:33:10 pm by marmad »
 

Offline jkrichards

  • Contributor
  • Posts: 36
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #391 on: November 21, 2014, 02:13:05 pm »
I received a response from Rigol this am in reference to the ticket I created early this week with them about these issues.  It seems they may be in denial that most all DS1000Z series have this issue or maybe they just want verification.  I received this scope last Friday.
Here is the message I received.  I left the names out.

"We are aware of a potential issue with the DS1000Z series.

Have you experienced the issue or is it in response to information that you found?

I have attached our preliminary research based on Dave Jones' EEVBlog tests.
Can you take a look at the document, perform the self calibration after warm-up, and let me know if the issue persists?

We will surely cover the return, even if it is past 30 days."
 
Here is the pdf document and a screenshot of my DS1054Z.
« Last Edit: November 21, 2014, 02:24:18 pm by jkrichards »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 12107
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #392 on: November 21, 2014, 02:30:37 pm »
Rigol is neither the source of claims about averaging - nor the company that first named it High Res (I believe that honor goes to Tektronix - see image below from 2003 Tek paper on acquisition modes).

Tektronix referred to high resolution mode at least as far back as 1992 when it was included in the TDS400 series of DSOs.  Distinct from this, they used the terms "smoothing" for DSP filtering and "averaging" for multiple acquisitions years earlier.

The scan I included is from their 1993 catalog.  Their November 1991-1992 Product Catalog Supplement released in 1992 mentions high resolution mode and these oscilloscopes as well.

At least in the TDS400 series, their high resolution mode only produced results up to 15 bits because of word size limitations.  I assume from the earlier discussion that the DS2000A series works similarly and the DS1000Z series is limited to 8 bit results making its high resolution mode more of a noise reduction mode.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 4461
  • Country: ca
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #393 on: November 21, 2014, 02:45:23 pm »
I received a response from Rigol this am in reference to the ticket

We've seen that shit before, they sent you the same thing that has been in circulation here for a while.
Facebook-free life and Rigol-free shack.
 

Offline jkrichards

  • Contributor
  • Posts: 36
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #394 on: November 21, 2014, 03:16:45 pm »
2nd email from them this AM in response to the questions document.

"Ok. Excellent. Thank you very much for the video.

We haven't been able to reproduce the issue on units that we have here.

Would you be willing to be without the scope for a week or so?

We would like to ask if you could ship the scope to us (we will send you a FedEx label.. just need you to box it up and ship it) for further testing?

We will test it with beta firmware revisions and then ship it back once we have a solid fix in place.

WRT the 30 day return.. we are extending it until we get the fix from Engineering.. so, no risk from that standpoint"


Are you kidding me!  I have a hard time believing they have no scope to test with.  They want to use my personal scope for testing???  Seems like a mickey mouse operation.  Lets see... I purchase a new scope so they can do R&D.
I am very disappointed in the way this whole thing has been handled.  This is the last Rigol product I will purchase.

Rigol, Send me one of the working scopes you have. and I will gladly send you this broken one for R&D.  :wtf:
« Last Edit: November 21, 2014, 03:28:22 pm by jkrichards »
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #395 on: November 21, 2014, 03:45:18 pm »
Tektronix referred to high resolution mode at least as far back as 1992 when it was included in the TDS400 series of DSOs.  Distinct from this, they used the terms "smoothing" for DSP filtering and "averaging" for multiple acquisitions years earlier.

The scan I included is from their 1993 catalog.  Their November 1991-1992 Product Catalog Supplement released in 1992 mentions high resolution mode and these oscilloscopes as well.

Thanks for the added info; I assumed it was earlier than 2003, but had no idea when the feature was first developed.

Quote
At least in the TDS400 series, their high resolution mode only produced results up to 15 bits because of word size limitations.  I assume from the earlier discussion that the DS2000A series works similarly and the DS1000Z series is limited to 8 bit results making its high resolution mode more of a noise reduction mode.

There seems to be a bit of confusion about how DSOs display the final result of the averaging. You'll notice in your Tek scan that it says:

"Vertical Resolution - 8 bits (256 levels over 10.24 vertical divisions)"

and then later:

"Hi-Res - Vertical resolution improvement and noise reduction on low frequency signals..."

It doesn't say:

"Vertical Resolution - 8 bits (except when in Hi-Res mode...etc.)"

The resolution or scale of the display doesn't change automatically somehow in Hi-Res mode. As far as I'm aware, all DSOs that do Hi-Res downsample the final results to the 8-bit display (or display memory). They have to - they don't even have the display resolution to manage 9-bits, let alone 12 - otherwise the scale would suddenly shift dramatically. But the DS2000A is definitely averaging 256 samples (4 effective bits) before downsampling for the display, while the DS1000Z appears to be doing mainly just 2 or maybe 3 bits (16 or 64 samples).

I'm not sure about Tek's implementation, but where Rigol's implementation on the DS2000A differs from, for example, Agilent's X-series (unfortunately), is that Rigol doesn't save the results of the averaging in sample memory for downloading to an external device as 12-bit values.
« Last Edit: November 21, 2014, 03:56:30 pm by marmad »
 

Offline TMM

  • Frequent Contributor
  • **
  • Posts: 437
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #396 on: November 21, 2014, 04:06:56 pm »
The point is that the high resolution mode doesn't seem to work as you would expect. The DS1K doesn't seem to be adding the 1LSB of Gaussian noise needed in order to achieve higher than 8bit resolution. All it seems to achieve is a slightly less noisy 8bit signal, negating any benefit of having >256px vertical display resolution....
...but Rigol are being cheeky to claim that box car averaging alone is 'high resolution mode'.
I think you're confused between how you expect High Res mode should work, how it's actually supposed to work, and how well the DS1000Z achieves it's results.

Rigol is neither the source of claims about averaging - nor the company that first named it High Res (I believe that honor goes to Tektronix - see image below from 2003 Tek paper on acquisition modes).

There are papers detailing the effects of successive sample averaging (also called boxcar filtering or moving average filtering), such as:
http://www.daysalive.com/share/EffectsofAveragingtoRejectUnwanted Signals.pdf
...but the main points are that it removes noise by decreasing a DSO’s bandwidth (it applies an LPF function with a -3dB point approximated by 0.433*sample rate/averaged samples), achieving 1 bit of effective resolution improvement for each factor of four of averaged samples (e.g. 4 bits improvement = 256 averaged samples).

The DS2000 does a good job of implementing 256 sample averaging in real-time (functioning almost identically to the Agilent X-Series), but the DS1000Z does not - I'm guessing because the engine is not fast enough to support it (except perhaps at very slow timebases). The difference between the two DSOs is quite apparent, so the problem isn't with the technique but the implementation on the DS1000Z.
I know how 'high-res'/oversampling works. The issue is that the DS1000Z is doing something wrong because there is an apparent lack of increased resolution when the high res mode is turned on.

You can only achieve increased resolution if there exists 1LSB (or more) of Gaussian noise on the signal. If the signal is not sufficiently 'noisy' within the window of the averaging filter, it won't work properly. For example, you feed in a DC signal that falls somewhere between quantization levels '232' and '233' of the 8-bit ADC, but the ADC outputs '233' consecutively since the signal is closer to 233 than 232. You apply box car averaging to a long string of 233s and you still get 233, the result is no increased resolution.
Now if you add 1LSB of analog noise before the ADC, the ADC will output out a series of samples which alternate between 232 and 233, with a distribution that when averaged will give a value that is in between 232 and 233, thus increasing the resolution beyond the number of quantization levels of the DAC.

If the DS1000Z isn't adding 1LSB of noise internally, then it is obviously not going to guarantee increased resolution in every circumstance. I haven't been successful in getting significantly increased resolution out of my DS1000Z, even when attempting to add external noise there are still obvious 8bit quantization levels present (it jumps up/down by 2 pixels in places instead of 1). I'm not 100% sure why this is. I suspect the result might be rounded back to 8bit somewhere before it is rendered on the screen. The high res mode is obviously doing some averaging, it just doesn't seem to increase the resolution. The intensity grading also effectively does some averaging (but not in a way that would increase resolution), which makes it even harder to observe what is going on.

On some vertical ranges there is even less than 8bits of resolution displayed on the screen. Every time you hear a relay click, you are changing the real analog range being fed into the ADC, so certain ranges have 8bit and some are effectively 7 or maybe even 6 bit. It wouldn't be such an issue if the high res mode worked properly!

Tektronix scopes implement high-res correctly and consistently achieve increased resolution that pretty much matches the display resolution, even for noiseless DC. I'm not sure how/if the DS2000 does it because I don't recall seeing anyone demonstrate it and haven't had the chance to get my hands on one.

You do not need 256 sample averaging to increase the resolution. Even if the 1000Z can only do 16sample wide windows, it should still achieve 12-bit resolution. 256 sample window would achieve 16bit. It really only needs 9bit since the display isn't even 512pixels tall...

That said, complaining that a $400 scope can't do high res properly would be a bit rich, i just think that 'high-res' is an inaccurate description of what it is doing. 'Noise reduction' or 'HF reject' would be a better description.

« Last Edit: November 21, 2014, 05:19:47 pm by TMM »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 12107
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #397 on: November 21, 2014, 04:13:41 pm »
The resolution or scale of the display doesn't change automatically somehow in Hi-Res mode. As far as I'm aware, all DSOs that do Hi-Res downsample the final results to the 8-bit display (or display memory). They have to - they don't even have the display resolution to manage 9-bits, let alone 12. But the DS2000A is definitely averaging 256 samples (4 effective bits) before downsampling for the display, while the DS1000Z appears to be doing just 2 or 3 bits (16 or 64 samples).

Did I say that?

The scale of the display does not change but the waveform record is not 8 bits in the case of the TDS series and similar old Tektronix DSOs and does not even match their digitizer resolution.  I have assumed that the Rigol oscilloscopes do something similar to the old Tektornix DSOs but are they more limited?

Quote
I'm not sure about Tek's implementation, but where Rigol's implementation on the DS2000A differs from, for example, Agilent's X-series (unfortunately), is that Rigol doesn't save the results of the averaging in sample memory for downloading to an external device as 12-bit values.

On the DSOs I am familiar with which operate in a way similar to the TDS420, the waveform record is greater than 8 bits and usually 16 bits.  High resolution mode returns an acquisition record with more than 8 bits of resolution and averaging during post processing also produces a waveform record with more than 8 bits.  I can visually verify this on the Tektronix DSOs I have, including the old ones, by saving and vertically magnifying the display record which will not show 8 bit quantization.  The TDS series and some earlier ones will do this in averaging mode in real time to produce more sensitive vertical scales.  I do not remember if the TDS series will do the same in high resolution mode as I have not used one in 20 years but I think they do.  Automatic measurements work on the higher resolution waveform record as well.

Without high resolution mode, averaging, or smoothing, it is just 8 bit data of course.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 12107
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #398 on: November 21, 2014, 04:17:15 pm »
You can only achieve increased resolution if there exists 1LSB (or more) of Gaussian noise on the signal. If the signal is not sufficiently 'noisy' within the window of the averaging filter, it won't work properly.

It would be an unusual DSO vertical signal chain which does not inherently have sufficient noise to meet this requirement.
 

Offline marmad

  • Super Contributor
  • ***
  • Posts: 2979
  • Country: aq
    • DaysAlive
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #399 on: November 21, 2014, 04:31:23 pm »
I know how 'high-res'/oversampling works. The issue is that the DS1000Z is doing something wrong because there is an apparent lack of increased resolution when the high res mode is turned on.

It doesn't sound like you do.

You can only achieve increased resolution if there exists 1LSB (or more) of Gaussian noise on the signal. If the signal is not sufficiently 'noisy' within the window of the averaging filter, it won't work properly.

According to who? I posted a paper that clearly documents both the mathematical method and effect of successive sample averaging - which, BTW, doesn't mention anywhere "adding 1LSB of noise" - but you've not posted any substantiating material.

Quote
I haven't been successful in getting any increased resolution out of my DS1000Z, even when attempting to add external noise there are still obvious 8bit quantization levels present (it jumps up/down by 2 pixels in places instead of 1). I'm not 100% sure why this is. I suspect the result might be rounded back to 8bit somewhere. The high res mode is obviously doing some averaging, it just doesn't seem to increase the resolution.

I'm not sure what you mean by "increasing the resolution" - it seems you didn't read the paper on how it works. The DS1000Z is definitely increasing the effective resolution - but by only 2 or 3 bits - I can see this clearly on the DS1000Z I have.

Quote
Tektronix scopes implement high-res correctly and consistently achieve increased resolution, even for noiseless DC.

Again, I'm afraid you would need to post actual documentation proving your assertions.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf