Author Topic: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems  (Read 389663 times)

0 Members and 1 Guest are viewing this topic.

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #125 on: November 15, 2014, 12:49:38 am »
I can't tell what proportion of people, especially in the YouTube comments, actually understand the distinction (I sure didn't know AC trigger coupling even existed until watching this video; I'm struggling to contrive applications for it), and I see at least a few people who quite clearly don't and think that AC coupling is broken.

For oscilloscopes which support adding a DC offset to the input signal (*) or apply the positioning signal before the trigger pickoff point, AC trigger coupling is used to prevent the trigger level from changing when the vertical offset or position is changed which is annoying.  On some of my analog oscilloscopes, AC trigger coupling is the default and DC trigger coupling is the exception because of this.

This should not be a problem with the Rigols because as far as I know, they do not have a DC offset fuction.  The AC trigger coupling might have just been a checklist feature although it is useful on oscilloscopes which do not support peak-to-peak triggering because you can set the trigger level once and leave it.

(*) A differential comparator input like a 7A13 can add huge amounts of DC offset so for instance ripple can be measured on a 100 volt input using a sensitivity of 10mV/div.  Oscilloscopes which support DC offset are usually not that extreme.
« Last Edit: November 15, 2014, 12:52:21 am by David Hess »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #126 on: November 15, 2014, 01:00:32 am »
AC vs DC triggering... since this seems to largely be a square wave issue, could the trigger be triggering off harmonics (randomly), and that's the cause of the jitter?

I have seen this happen before with oscilloscope triggers but the harmonics are synchronous with the square wave so this would not by itself explain the jitter.  When it happens, the trigger occurs after or sometimes before the fundamental of the square wave but is relatively stable.  It usually indicates a problem with the trigger circuits.

A test I would try on the Rigol is to see if the jitter when AC trigger coupling is used changes when more than one channel is active lowering the sample rate from 1 GS/s to 500MS/s or 250MS/s.
 

Offline eV1Te

  • Regular Contributor
  • *
  • Posts: 181
  • Country: se
  • Your trusted friend in science!
    • richardandersson.net
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #127 on: November 15, 2014, 01:21:18 am »
I have not been able to read all the other posts in this thread, but here are the results from my DS1074Z-S with all options unlocked and the latest firmware from Rigol (04.01.02).

No Jitter on DC with any offset.

But lots of Jitter with AC coupled trigger.

 

Offline alank2

  • Super Contributor
  • ***
  • Posts: 1999
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #128 on: November 15, 2014, 01:25:38 am »
On my ds4k when I start with a 0-5V 20MHz square wave I hit the 50% button and put the trigger line in the center.  When switching to AC triggering that line will now be at the top of the waveform if it is being looked at from an AC point of view even though it is still displayed as DC.  If I adjust the trigger level to 0V it cleans up perfectly.  Could you guys have a trigger level issue on the AC trigger jitter?  Oddly in AC triggering and LF Reject triggering do NOT show a trigger line when adjusting - perhaps this is their way of telling you that the line being visible wouldn't correlate to the on screen waveform.  FOR THOSE OF YOU WITH THE AC TRIGGER JITTER ISSUE, IF YOU CHANGE THE THRESHOLD TO 0V, does it clean up?

Also, on the horizontal jitter issue, my ds4k doesn't get bad then clear up at 5us, 10us, etc., it just has some jitter in it anytime the trigger point is OFF of the display.  You can push the T to the left and while it is still not at the edge, it is rock solid, but push it off a little bit and it adds a slight bit of jitter.  Could it be an issue with the intensity grading code causing this?
 

Offline softwarejedimaster

  • Contributor
  • Posts: 15
  • Country: us
  • Software Dev. by Day / Electronics Madman by Night
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #129 on: November 15, 2014, 01:31:41 am »
my first post is a bad one! I have the same issue with DS1054Z with jitter on a 15 Mhz sine input, only my jitter starts as early as one micro second and gets progressively worse up to 5 then disappears at 10 etc... - also same problem with AC coupling -  :wtf: Rigol!!!:
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #130 on: November 15, 2014, 02:58:38 am »
So on the trigger AC coupling, what does it mean that it forbids DC signals into the trigger path?

No Jitter on DC coupled triggering on DS2202.

Next, feeding it a 12 MHz square wave (sorry for the bad ground probing) I surely get jitter with triggering set to AC coupling, I've never used it before, not even sure what is supposed to do.

If I change the trigger level to the peak to 2.5 Volts so it triggers on that overshoot due to bad ground probing, it jitters way less.

If I adjust the Acquisition Mem Depth to 56MPoints it dances around the jitter but only showing one wave.

If I change the Acquisition mode to Average, I can get it to clear setting it to 16 but the trigger offset (rising edge) is off (yeah I could change it with CH1 Delay-cal and make it look ok, but I'm not sure in AC coupled trigger where is supposed to trigger at if I have it set to rising edge..

I'm on an old version as well, Software: 00.01.01.00.02 Hardware: 1.0.2.0.0 etc (attached version screenshot)

I'll watch the video to see if I get an insight on what the AC coupled trigger is supposed to be used for :)
« Last Edit: November 15, 2014, 03:02:07 am by miguelvp »
 

Offline LoyalServant

  • Regular Contributor
  • *
  • Posts: 65
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #131 on: November 15, 2014, 03:03:06 am »
I guess I will throw my hat into the ring and say me too.
My DS2072A is quite horrible, and with a sine wave....

 

Offline selkathguy

  • Supporter
  • ****
  • Posts: 88
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #132 on: November 15, 2014, 03:05:54 am »
Throwing my two cents in, just tested it on my DS1074Z after seeing this video.  Signal source is a Rigol DG1032Z.  No jitter problems from what I can see.  I tried to match your conditions the best I could (with the exception of amplitude).  Perhaps different firmware?  I can't imagine the front-end for the two scopes is any different, so it might be some clocking issue.



 

Offline nixfu

  • Supporter
  • ****
  • Posts: 343
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #133 on: November 15, 2014, 03:23:33 am »
Ok... as stated above.

So, from what I gather this is strictly to do with using the AC TRIGGERING feature, where the jitter is occurring, not just from using normal everyday AC coupling of the input.

AC coupling, I personally do lots of times with audio signals/RF signals etc..but I don't think I have ever used any AC triggering settings yet. 


THIS IS IMPORTANT.  I do think maybe people are thinking a regularly used feature such as AC input coupling is not working property. 

If this is not the case, and I think we should make sure we don't get people into a tizzy unnecessarily.


« Last Edit: November 15, 2014, 03:29:56 am by nixfu »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #134 on: November 15, 2014, 03:31:41 am »
but I'm not sure in AC coupled trigger where is supposed to trigger at if I have it set to rising edge.

If the trigger level when DC coupled is midway between the positive and negative peak of a waveform which has a 50% duty cycle like a square, sine, or triangle wave, then the trigger position should not change with an AC coupled trigger.  There should always be some level of AC coupled trigger which produces the same results as a DC coupled trigger.

Not only is the displayed jitter not suppose to be there but the trigger position is wrong as well.  It should be on the fast edge just like when a DC coupled trigger is used.
« Last Edit: November 15, 2014, 04:15:43 am by David Hess »
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #135 on: November 15, 2014, 03:48:00 am »
As for the 5 us jitter my old version on the DS2202 has it, I don't even need to go to 5us just 1us makes it jitter and it doesn't recover at 10us nor ever.
Both test with DC trigger coupling and either DC or AC channel coupling. I guess I will have to upgrade because on Software: 00.01.01.00.02 Hardware: 1.0.2.0.0  it's horrible on my scope.
 

Offline selkathguy

  • Supporter
  • ****
  • Posts: 88
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #136 on: November 15, 2014, 03:51:55 am »
this is strictly to do with using the AC TRIGGERING feature, where the jitter is occurring, not just from using normal everyday AC coupling of the input.

AH yes,  just tested that.  Yup.  Forgive me I admit I did not read all of the posts beforehand.  From how Dave was going on in the video, I thought it was a huge input coupling jitter problem.  I've been using mine for a while, and I've never used AC triggering before.  I'm trying to wrap my head around what exactly AC triggering actually does and how it differs from the other options.  Hmm, to google!

 

Offline i4004

  • Regular Contributor
  • *
  • Posts: 71
  • Country: hr
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #137 on: November 15, 2014, 04:09:16 am »
Ok... as stated above.

So, from what I gather this is strictly to do with using the AC TRIGGERING feature, where the jitter is occurring, not just from using normal everyday AC coupling of the input.

AC coupling, I personally do lots of times with audio signals/RF signals etc..but I don't think I have ever used any AC triggering settings yet. 


THIS IS IMPORTANT.  I do think maybe people are thinking a regularly used feature such as AC input coupling is not working property. 

If this is not the case, and I think we should make sure we don't get people into a tizzy unnecessarily.

correct, only ac triggering throws it off.

on 1054 i saw this:
don't see the 5ms(i used few khz signals, so it's not 5us) jitter on the built-in signal generator (1khz?), but i do see it on a bit higher freq. square wave with A HECK OF A LOT FASTER RISE TIME, in that case i see it well, it's slowly moving the waveform left/right in that jittery way....

like so:


trigger ac coupling issue i don't see, but probably because i tried with just positive-pulses square waves...

i think both issues were not reported earlier because they're not used that much...

when would one need those 2 anyway?
i mean if mikeselectric never used ac trigger somehow i doubt i'll develop the need to use it tomorrow, and for moving the trigger left/right....didn't really used that either...i quite like it centered....

« Last Edit: November 18, 2014, 03:36:34 am by i4004 »
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #138 on: November 15, 2014, 04:16:37 am »
Well on the old version of the DS2202
Software: 00.01.01.00.02
Hardware: 1.0.2.0.0

DC coupled trigger 5 us jitter is present, but so it's at 1 us horizontal positioning.


So I changed it to a lower frequency clock driven by an mcu around 33KHz 0 horizontal offset.


And just changing the horizontal positioning by 3 ms, it's jittery time


I think others are not seeing this on DC coupled trigger mode, so if they did fix it in the DS2000 series and I'm getting it because of my ancient update version then probably they can fix it in the DS1000Z series.

 

Offline TMM

  • Frequent Contributor
  • **
  • Posts: 436
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #139 on: November 15, 2014, 05:07:05 am »
On my ds4k when I start with a 0-5V 20MHz square wave I hit the 50% button and put the trigger line in the center.  When switching to AC triggering that line will now be at the top of the waveform if it is being looked at from an AC point of view even though it is still displayed as DC.  If I adjust the trigger level to 0V it cleans up perfectly.  Could you guys have a trigger level issue on the AC trigger jitter?  Oddly in AC triggering and LF Reject triggering do NOT show a trigger line when adjusting - perhaps this is their way of telling you that the line being visible wouldn't correlate to the on screen waveform.  FOR THOSE OF YOU WITH THE AC TRIGGER JITTER ISSUE, IF YOU CHANGE THE THRESHOLD TO 0V, does it clean up?
Did not help one bit for my DS1kZ as i was already feeding it a signal with zero dc offset and had the trigger already set at 0v.

AC vs DC triggering... since this seems to largely be a square wave issue, could the trigger be triggering off harmonics (randomly), and that's the cause of the jitter?
Unlikely to be an issue unless the AC trigger circuit has a serious bandwidth limitation and the trigger point is set to something other than 0.

On my DS1k it occurs equally as bad with sine, square or triangle signals, which indicates that it is not dependent on the slope of the rising edge or the frequency domain content.

At this stage, i think people are on the right track with the cause being a spread spectrum internal clock. The fact that it doesn't happen when people use the 'S' models internal sig gen as a source is pretty solid proof.

Throwing my two cents in, just tested it on my DS1074Z after seeing this video.  Signal source is a Rigol DG1032Z.  No jitter problems from what I can see.  I tried to match your conditions the best I could (with the exception of amplitude).  Perhaps different firmware?  I can't imagine the front-end for the two scopes is any different, so it might be some clocking issue.


Trigger offset should be 5us instead of 5ns
« Last Edit: November 15, 2014, 05:11:37 am by TMM »
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #140 on: November 15, 2014, 05:41:06 am »
Can someone please explain when do we need to prevent DC signals from triggering?

Also when would you use that feature when you need AC triggering?

Thanks.
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2173
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #141 on: November 15, 2014, 05:48:04 am »
Can someone please explain when do we need to prevent DC signals from triggering?

Also when would you use that feature when you need AC triggering?

Thanks.

Your first question doesn't make any sense to me, and I asked the second a while back -- David gave a good answer (although it's not really a situation where you would use it, it's more a "this seems like a historical artefact"):

For oscilloscopes which support adding a DC offset to the input signal (*) or apply the positioning signal before the trigger pickoff point, AC trigger coupling is used to prevent the trigger level from changing when the vertical offset or position is changed which is annoying.  On some of my analog oscilloscopes, AC trigger coupling is the default and DC trigger coupling is the exception because of this.

This should not be a problem with the Rigols because as far as I know, they do not have a DC offset fuction.  The AC trigger coupling might have just been a checklist feature although it is useful on oscilloscopes which do not support peak-to-peak triggering because you can set the trigger level once and leave it.

(*) A differential comparator input like a 7A13 can add huge amounts of DC offset so for instance ripple can be measured on a 100 volt input using a sensitivity of 10mV/div.  Oscilloscopes which support DC offset are usually not that extreme.

However, I could make up some contrived situation where you have some small clock signal overlaid on a slowly varying ("DC" as far as the AC triggering coupling cap is concerned) signal. If you want to see both together on the screen, but trigger on that clock signal, then AC triggering would be useful and you'd be annoyed if you had a DSxxxx right now. Could still work around the problem by sending the signal to both channels and AC coupling one of them, and triggering of that channel.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #142 on: November 15, 2014, 05:51:31 am »
Ok... as stated above.

So, from what I gather this is strictly to do with using the AC TRIGGERING feature, where the jitter is occurring, not just from using normal everyday AC coupling of the input.

AC coupling, I personally do lots of times with audio signals/RF signals etc..but I don't think I have ever used any AC triggering settings yet.

correct, only ac triggering throws it off.

i think both issues were not reported earlier because they're not used that much...

when would one need those 2 anyway?
i mean if mikeselectric never used ac trigger somehow i doubt i'll develop the need to use it tomorrow, and for moving the trigger left/right....didn't really used that either...i quite like it centered....

I had to double check but the DSOs that I use the most, a 2230 and 2232, do not even have AC coupled triggering and I have never missed it but they have peak-to-peak automatic triggering which is practically the same thing.  All of my other DSOs and analog oscilloscopes have it though and my 7000 series mainframes would be more difficult to use without it.

Unlikely to be an issue unless the AC trigger circuit has a serious bandwidth limitation and the trigger point is set to something other than 0.

On my DS1k it occurs equally as bad with sine, square or triangle signals, which indicates that it is not dependent on the slope of the rising edge or the frequency domain content.

At this stage, i think people are on the right track with the cause being a spread spectrum internal clock. The fact that it doesn't happen when people use the 'S' models internal sig gen as a source is pretty solid proof.

There are two different problems Dave identified in his video.  One is the AC coupled triggering issue which I doubt is related to frequency modulation of the sampling clock if that is what is happening.  The other is the jitter which appears between the trigger point and 5us later but not 10us later which spread spectrum clocking could cause.
 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #143 on: November 15, 2014, 05:57:55 am »
Thanks rs20,

The first question was the same as the 2nd, the built in help system refers to the AC trigger coupling as in ignoring the DC component, probably referring to the DC offset Dave mentioned.

It does make sense now.

So if the clock or AC signal was offset by a varying DC signal, how will the Channel AC coupling deal with that, I guess with some math function averaging of the dc signal subtracted from the signal would come up with the same result, but math being so slow, might not be practical?
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9946
  • Country: us
  • DavidH
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #144 on: November 15, 2014, 06:12:38 am »
For oscilloscopes which support adding a DC offset to the input signal (*) or apply the positioning signal before the trigger pickoff point, AC trigger coupling is used to prevent the trigger level from changing when the vertical offset or position is changed which is annoying.  On some of my analog oscilloscopes, AC trigger coupling is the default and DC trigger coupling is the exception because of this.

This should not be a problem with the Rigols because as far as I know, they do not have a DC offset fuction.  The AC trigger coupling might have just been a checklist feature although it is useful on oscilloscopes which do not support peak-to-peak triggering because you can set the trigger level once and leave it.

However, I could make up some contrived situation where you have some small clock signal overlaid on a slowly varying ("DC" as far as the AC triggering coupling cap is concerned) signal. If you want to see both together on the screen, but trigger on that clock signal, then AC triggering would be useful and you'd be annoyed if you had a DSxxxx right now. Could still work around the problem by sending the signal to both channels and AC coupling one of them, and triggering of that channel.

I know of one place where AC coupled triggering comes in very handy.  If your oscilloscope does not support peak-to-peak auto triggering (*) and you are probing different logic levels and want the ground reference to remain fixed so you can see the actual DC levels, then you can set the trigger coupling to AC and the trigger level close to zero and no matter where you probe, the oscilloscope will trigger correctly without adjusting the trigger level.  In this case it acts as a poor version of peak-to-peak triggering.  I use it when dealing with mixed families of logic and mixed signal circuits.

(*) The last oscilloscopes that I know of which supported peak-to-peak automatic triggering are the Tektronix 22xx series analog and DSO models from 1994 and they do not even have AC trigger coupling because it would be superfluous.  Their successors support auto level where if there is no trigger within a timeout period, the trigger level is reset to the midpoint between the signal peaks but this is much slower and they do support AC coupled triggering.
 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2173
  • Country: au
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #145 on: November 15, 2014, 06:25:01 am »
So if the clock or AC signal was offset by a varying DC signal, how will the Channel AC coupling deal with that, I guess with some math function averaging of the dc signal subtracted from the signal would come up with the same result, but math being so slow, might not be practical?

Channel 1/2 AC coupling (as opposed to AC trigger coupling) is achieved using a real-life DC blocking cap in the input stage. So that works just fine; if you want to view only the AC component of a wave, you enable Channel 1/2 AC coupling, and (perhaps confusingly) leave the trigger coupling on "DC" (i.e., a normal, non-DC-blocking connection to the already AC-coupled input signal). That way you see only the AC component on the screen, and you can move the trigger level up and down on the screen and it behaves exactly as you always expected.

Basically, just pretend the AC trigger coupling option doesn't even exist, because I'd guess that half the people on this forum didn't even know the option existed before today (myself included); turns out it's broken anyhoo, so just don't touch it. "DC trigger coupling" is what you've always known and does exactly what you've always needed.

I know of one place where AC coupled triggering comes in very handy.  If your oscilloscope does not support peak-to-peak auto triggering (*) and you are probing different logic levels and want the ground reference to remain fixed so you can see the actual DC levels, then you can set the trigger coupling to AC and the trigger level close to zero and no matter where you probe, the oscilloscope will trigger correctly without adjusting the trigger level.  In this case it acts as a poor version of peak-to-peak triggering.  I use it when dealing with mixed families of logic and mixed signal circuits.

This is a very good and genuinely useful idea -- for the first time, now I'm slightly disappointed that the feature is broken on my DS2202. Hopefully it can be fixed. Nevertheless, this still feels like a fairly specialized broken feature, rather than a flaw with the core, everyday functionality.
 

Offline (In)Sanity

  • Regular Contributor
  • *
  • Posts: 229
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #146 on: November 15, 2014, 06:43:59 am »
I'm going to jump on the "It's not the end of the world" bandwagon.   Cause it's not.   For the price I paid for my 300 Mhz scope well..I just can't complain. 

Interesting testing this issue however,  I found it decrease around 14.71 - 14.97 Mhz.

Jeff
 

Offline RigolTechUSA

  • Contributor
  • Posts: 8
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #147 on: November 15, 2014, 07:28:45 am »
**The attached PDF has the scope shot images included and formatted.**

Dave, as always thanks so much to you and the community for always digging deep into instruments. Everyone here always does a great job with all of their testing!

Dave did a great job of documenting 3 different issues across a number of products.
1)   5 us jitter
2)   AC trigger coupling
3)   Rising & Falling trigger
The 5 us jitter issue I have described as:
   Trigger jitter variance displayed at 5 us odd intervals of positive trigger offset. I.E. when the trigger is 5 us, 15 us, 25 us… before the center of the display some jitter appears that doesn’t appear otherwise.

We tested 3 DS1000Z scopes here with different board revisions including the FW and HW used by Dave. Screenshots of the results in the attachment.

Scope 1: shows model and revisions as well as 5 us, 15 us, and 25 us offset.

Scope 2: shows model and revisions as well as 0 us, 5 us, and 25 us offset.
   
Scope 3: shows model and revisions as well as 0 us, 5 us, and 15 us offset.
   
None of the scopes showed the problem at all.

Conclusions:
None of the scopes we have show any change in the jitter based on trigger delay as far as I can tell. All 3 units with different hardware running the latest FW were rock solid. I was actually hoping to find one that was a little off because I expect a proper warm up and self-cal may solve the issue. You can find the self-cal in the utility menu.  Anyone seeing this issue should try a self-cal and if the problem still exists please contact Rigol. If the instrument is out of spec we will absolutely help you with that. We haven’t had any report of this problem to date in the USA and a self-cal is a good first check. Certainly, if Mads from EcProjects scope doesn’t improve when properly calibrated then it is a bad unit that needs repaired, but tests here indicate that it is not systemic and may be fixed with a self-cal. Dave, can you give yours a warm-up and self-cal and let us know if you see any change at 5 us intervals?   

It is also worth noting that if this does continue in some scopes I’d expect a Firmware update solution. Most, if not all, of the DS1000Z triggering sub system is digital. It is handled by FPGAs predominantly. There is no real inherent jitter in those captured signals, so I’d expect a FW solution if there is a verified problem on some units after a self-cal.

AC trigger coupling

Definition: Setting the Edge trigger to AC coupling causes crazy trigger insanity.

We tested this on 4 different scopes (The previous 3 as well as a DS2000A). Using the scope the way Dave does I verified what he sees. That is an issue that we will immediately address. However, the reason no one has noticed it until now is that isn’t really the normal way to look at AC Coupled signals in these scopes. We look at AC coupled signals all the time for a variety of tests without issue. The best reason to do this is often to test a signal with DC offset and keep it centered on the display and not have to get knob cramp by having to always adjust channel and trigger levels as the DC offset changes. The way I do it is to go into the channel menu and set the coupling on the channel to AC and leave the trigger coupling at the default settings. Here is a good example:

The input signal is a 20 MHz square wave (unterminated, like Dave’s) except I’ve added a 5 V offset. So, this signal actually goes from 7.5 V – 2.5 V.
 
By changing to AC Coupled in the Channel menu it centers around 0 V as expected.
 
Now notice the trigger line almost right at 0 V. The trigger system is being fed the AC coupled signal from the channel allowing you to trigger with the signal as it is shown on the screen. This is the common and normal way to use AC Coupling on a channel. As you can see this works great! AC coupling just the trigger and not the channel is pretty confusing because it either does nothing if the signal is centered around 0 V or you would then have to trigger at a level close to 0 V at a different point on the display than where the signal appears. So, you could no longer set the trigger level visually. Because of that confusion I’d never use the coupling in the trigger menu and would always use the coupling in the channel. That’s not to say that it shouldn’t work, but I can’t think of any reason why I’d want to use that setting since any signal I put into it I want to have coupled AC on the display as well. So, I think there is a much better way to use the scope, but it is still something we will fast track a firmware version to improve.

Problem 3: Rising/Falling Trigger issue
Checking this on our DS2000A here I believe what Dave sees is a combination of trigger holdoff and the AC coupling issue above. When I measure on an AC coupled channel leaving the Trigger settings alone except for setting it to Rising and Falling it looks great on our scope.
 
With rising and falling you have to be careful with these repeating square waves. Especially, it turns out at 20 MHz. A 20 MHz square wave takes 50 ns and the default trigger holdoff on the DS2000 is 100 ns so you are always catching one side of the wave because of the holdoff. If you really want to catch both sides of the wave make sure your trigger holdoff doesn’t force you to repeat triggers on one side of the wave consistently.

In summary, great detective work, as always, by Dave and Rigol users worldwide! Keep using your scopes to capture all types of waves including AC coupled waves. We are allocating resources to make sure even the corner cases we have found here exceed expectations. As the instruments are clearly capable of triggering and measuring all signals within their domain whether DC coupled, AC coupled, rising and falling, or with a specified trigger offset I don’t see any hardware issues across any of the released versions. Improvement in guiding users on how to make measurements and when to use settings is and needs to be a continuous process.

I can’t think of any use cases on this scope where you would want to AC couple the triggering of a DC coupled channel. It doesn’t work for me. If anyone is willing to share insight with us please follow this link to help us improve the interface.

http://beyondmeasure.rigoltech.com/acton/form/1579/0019:d-0001/1/index.htm

Here is a summary of the questions after the link in case you want a sneak peak:
•   How often do you use AC coupling on your scope?
•   When you use AC coupling on a channel, why do you do it?
•   When you trigger on a AC coupled signal how often do you want to view that signal DC coupled on the display? If you do, please explain the use case.
•   If you do this, how do you want the trigger level shown on the display since it is now decoupled from the signal visualization on the display?

Lastly, if anyone has questions or concerns they want to discuss with us, we are always here. Contact your local Rigol office or distribution partner. Our USA number is 877-4-RIGOL-1. Every DS1000Z bought in the USA is under our 3 year warranty program and Rigol quality is important to us all over the world.

Thanks,

Chris Armstrong
Director of Product Marketing & SW Applications
Rigol Technologies USA
 

Offline pickle9000

  • Super Contributor
  • ***
  • Posts: 2132
  • Country: ca
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #148 on: November 15, 2014, 08:00:20 am »
Two thumbs up for RigolTechUSA takes guts to post.

What I heard:

If the product has a problem it can be fixed let us fix it, we want to keep you as a customer. 

 

Offline miguelvp

  • Super Contributor
  • ***
  • Posts: 5549
  • Country: us
Re: EEVblog #683 - Rigol DS1000Z & DS2000 Oscilloscope Jitter Problems
« Reply #149 on: November 15, 2014, 08:04:01 am »
Also I want to note that Dave's test where done with the channels AC coupled, not DC coupled for the 5 us tests as your pictures show.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf