Siglent's new product- MSO/SDS2000 Series

[don]

1) zoom does not work in roll mode

Hmm... does Zoom work on any DSO while using Roll Mode? It certainly doesn't work on the Rigol DS2000 series - and I'm not so certain it could.

Roll mode is a completely different kind of acquisition - sampling and doing decimation bit by bit - as opposed to a screen at a time. Zoom mode suggests that the DSO go back to the original sampled data and re-decimate for a different time base. This seems like an impossibility to me in Roll mode due simply to the method of acquisition. But perhaps I'm wrong about this.

It might depend on the price point.  Newer scopes like tek, agilent and lecroy do.  Also, zoom is really just a way of presenting the data.  Some agilents don't have a "zoom" mode (like 6000 series)  but just let you adjust the time base on captured waveform.  This is what the SDS2000 does today.   But the problem with SDS2000 is that it does not adjust the timebase about the center of the screen.  So you see focal point at center of  screen disappear or shift to the left if you are lucky.  The bug here seems to be that you need to align desired focal point on the right edge of screen.  This might be claimed to a "feature" or "works as designed" since there is no trigger and waveform acquisition starts at right side of screen, but it sure is intuitive, looks broken and is inconsistent with other scopes.

[don]

I did some testing on Sequence Acquisition

I can't get I2C decode to work with Sequence acquistion. Scope says seqeunce not allowed if you try to enable
when decode is active.  If you disable decode then enable sequence, you can go back and enable decode...but
it does not work as far as I can tell.  I just get a solid blue line for decoded messages

Also, I can't really get sequence to function correctly.
I'm at 50us/deiv, 700kpts.  Edge triggered and am capturing a full I2C message on screen.

Setup:
Sequence is set for Single (could not get sequence normal trig to fill buffer and stop - I know this is what normal does, but do not understand why it's in sequence mode unless user is supposed to press stop.  This would be fine except it does not work that way on this scope)
Display mode normal
Frame Set shows 1/163
Start=1
End=7
Play button does not do anything as far as I can tell.
The only way I can play through segments is to adjust the start frame.  Then I see waveforms change on screen,
but this does not seem right.

I rearmed again for single, and now single only captures one frame and stops. 

I went back to normal trigger and it won't stop acquiring.  I let it run long enough
 to fill 163 frame buffer.  I pressed stop and it showed no waveform.  Frame set said 1/163.  Scrolling
through frame set did nothing.  Start frame was set to 0 and end frame to 0 and can't be adjusted.

I went back to single trig:
Now it captured 23 frames and stopped automatically.  Playback button does something.  It actually show a pause for a second...it plays maybe three of 21 frames though but indicates it progressed through all as pause changes to play.

Summary:
Scope does not let you decode on sequence acquisition (I thought I read that it did.  Part of reason I bought scope)
Sequence mode essentially unusable at this time.

Does this post need to be in the other thread for Siglent reps to see? 

[marmad]

It might depend on the price point.  Newer scopes like tek, agilent and lecroy do.

I can't find any evidence that the Agilent X-series (2000, 3000, etc) allow Delayed Sweep while running Roll mode. In fact, the way that delayed sweep is based around a trigger position would tend to make me believe it's not a normal feature in Roll mode on DSOs.

Now obviously, you could program the DSO to 'simulate' a delayed sweep in Roll mode when stopped - but that would require a totally separate Zoom routine from that which is used in YT mode.

But the problem with SDS2000 is that it does not adjust the timebase about the center of the screen.  So you see focal point at center of  screen disappear or shift to the left if you are lucky.  The bug here seems to be that you need to align desired focal point on the right edge of screen.  This might be claimed to a "feature" or "works as designed" since there is no trigger and waveform acquisition starts at right side of screen, but it sure is intuitive, looks broken and is inconsistent with other scopes.

The Rigol DS2000 does the same thing - and personally, it feels as if it's the correct way of doing it. What logic would you use to adjust the timebase around the center of the screen? In Roll mode, the 'simulated trigger point' (so to speak) IS, in essence, the right side of the screen.

[nctnico]

About 9:
Many Tektronix scopes do this as well. Sometimes this is handy to get a feel about how wide a pulse varies and sometimes it is outright annoying.

I've observed many Tek scopes doing this as well, but I don't think it's an intentional feature but instead an artifact of architecture.   You can take two TDS754D's (different vintages & options) and one will show remnants of last waveform, and one will not.  The newer one does not.  Both high end scopes in the day and both worked this way out of box and useable.

I have seen this behaviour on monochrome Tektronix scopes from the TDS5xx/6xx7xx series but also on the TDS200 series. I dug around a lot in the display section of the TDS5xx/6xx/7xx series and I'm 100% sure it is intentional. In the color versions they opted to show the selected channel as a highlighted trace. On the monochrome version they show the previous acquisition.

Regarding the zoom/changing timebase: I agree it should use the centre of the screen.

[don]

It might depend on the price point.  Newer scopes like tek, agilent and lecroy do.

I can't find any evidence that the Agilent X-series (2000, 3000, etc) allow Delayed Sweep while running Roll mode. In fact, the way that delayed sweep is based around a trigger position would tend to make me believe it's not a normal feature in Roll mode on DSOs.

Now obviously, you could program the DSO to 'simulate' a delayed sweep in Roll mode when stopped - but that would require a totally separate Zoom routine from that which is used in YT mode.

But the problem with SDS2000 is that it does not adjust the timebase about the center of the screen.  So you see focal point at center of  screen disappear or shift to the left if you are lucky.  The bug here seems to be that you need to align desired focal point on the right edge of screen.  This might be claimed to a "feature" or "works as designed" since there is no trigger and waveform acquisition starts at right side of screen, but it sure is intuitive, looks broken and is inconsistent with other scopes.

The Rigol DS2000 does the same thing - and personally, it feels as if it's the correct way of doing it. What logic would you use to adjust the timebase around the center of the screen? In Roll mode, the 'simulated trigger point' (so to speak) IS, in essence, the right side of the screen.

As I said, it might depend on the price point.  I have not used  scopes of agilent DSXO3000 class or below (I suspect agilent will zoom on center though). At work I use higher end scopes and they all zoom on the center of the screen or change time base about the center of the screen. 

Here is the use case. I am capturing in roll mode and see something interesting.  I press stop.  I move the interesting anomaly to the center of the screen.  Depending on the scope, you use zoom or adjust the timebase and see the center of the screen / waveform anomaly expand in place.

With the SDS2000 here is what I have to do. I capture in roll mode, see something, press stop.  I move the interesting anomaly all the way to the right edge of screen.  I change the time base.  Unless I had half the desired focal point off the right edge of screen and half on screen, the focal point will shift to the left as I change time base.  This is because the center point of time base change is the edge of the screen.   For me it's a pain an inconsistent with typical scope behavior. 

SDS2000 will change time base about the center in all other modes except roll.  It's inconsistent at best if this is their desired implementation.

The scope business started with higher end and expands down to lower price point as technology allows.  I consider the norm to be higher end scope behavior and lower end scopes are doing their best to mimic.  Rigol is doing a good job, but they do not set the bench mark of how a scope should perform.
*edit* not trying to offend with Rigol, actually see some nice innovations from them but referring to the more typical usage scenarios (ie zoom on center..)

[don]

About 9:
Many Tektronix scopes do this as well. Sometimes this is handy to get a feel about how wide a pulse varies and sometimes it is outright annoying.

I've observed many Tek scopes doing this as well, but I don't think it's an intentional feature but instead an artifact of architecture.   You can take two TDS754D's (different vintages & options) and one will show remnants of last waveform, and one will not.  The newer one does not.  Both high end scopes in the day and both worked this way out of box and useable.

I have seen this behaviour on monochrome Tektronix scopes from the TDS5xx/6xx7xx series but also on the TDS200 series. I dug around a lot in the display section of the TDS5xx/6xx/7xx series and I'm 100% sure it is intentional. In the color versions they opted to show the selected channel as a highlighted trace. On the monochrome version they show the previous acquisition.

Regarding the zoom/changing timebase: I agree it should use the centre of the screen.

That's actually different scenario or feature than I'm trying to flag  The tek option is to determine behavior of selected waveform (ie channel 1 or 2) when overlapping other channels.     I'm talking about remnants of previous capture and new capture of the same channel being combined and displayed.  I can demonstrate on two old TDS754D's.  One will only show the captured waveform from trigger event, the other will show some leftovers from previous capture because it did not fully purge the buffer before recapturing.  Both let you give priority to the selected waveform when overlapping others.  Both work fine...but one is much easier to monitor in real time and the later model scope is the one that purges memory completing before acquiring a new capture.

*edit* By the way, I can't speak to how the old monochrome scopes behaved and was referring to later model color scopes.   Fortunately the norm has improved in this respect for me...I have a hard enough time processing one waveform in real time yet alone trying to deconstruct the combination of two waveforms. 

[marmad]

As I said, it might depend on the price point.  I have not used  scopes of agilent DSXO3000 class or below (I suspect agilent will zoom on center though). At work I use higher end scopes and they all zoom on the center of the screen or change time base about the center of the screen.

Sorry, but you seem to be confused between just changing the timebase while stopped (which is a kind of 'zooming' in/out) - and Delayed Sweep mode (called Zoom mode on many DSOs, including the Siglent). My response was a challenge to your assertion that "Newer scopes like tek, agilent and lecroy...." will do Delayed Sweep while running in Roll mode, which I highly doubt - and which is what I wrote in my original post to which you responded. It had nothing to do with the horizontal reference point of the DSO when changing the timebase when stopped. That's a separate issue entirely.

With the SDS2000 here is what I have to do. I capture in roll mode, see something, press stop.  I move the interesting anomaly all the way to the right edge of screen.  I change the time base.  Unless I had half the desired focal point off the right edge of screen and half on screen, the focal point will shift to the left as I change time base.  This is because the center point of time base change is the edge of the screen.   For me it's a pain an inconsistent with typical scope behavior. 

I understand what you're complaining about very well since I use Roll mode often. Inconsistent with typical scope behavior? It depends on how you define typical behavior. If the horizontal reference point should only ever be the center of the screen, then I suppose it's inconsistent. If, OTOH, you think it should be the acquisition point, it's consistent. On my DSO, I can change the reference point in YT mode to Center, Trigger, or User-defined... but not in Roll mode - so I agree it could be handy to allow setting it there as well.

The scope business started with higher end and expands down to lower price point as technology allows.  I consider the norm to be higher end scope behavior and lower end scopes are doing their best to mimic.  Rigol is doing a good job, but they do not set the bench mark of how a scope should perform.

I never said they set a benchmark - I was just reporting the similar behavior of another DSO.

[nctnico]

Well, let's see what Siglent comes up with. I have pulled the trigger on buying an SDS2204 which I hope to get somewhere next week. Lot's of things to try out 

[marmad]

Well, let's see what Siglent comes up with. I have pulled the trigger on buying an SDS2204 which I hope to get somewhere next week. Lot's of things to try out 

Not to break you balls or anything, but I thought you've been consistently arguing here at EEVblog over the last couple of years that older, used scopes were both cheaper - and could provide the same functionality - as these modern Chinese DSOs. Or did I get it wrong? 

[don]

As I said, it might depend on the price point.  I have not used  scopes of agilent DSXO3000 class or below (I suspect agilent will zoom on center though). At work I use higher end scopes and they all zoom on the center of the screen or change time base about the center of the screen.

Sorry, but you seem to be confused between just changing the timebase while stopped (which is a kind of 'zooming' in/out) - and Delayed Sweep mode (called Zoom mode on many DSOs, including the Siglent). My response was a challenge to your assertion that "Newer scopes like tek, agilent and lecroy...." will do Delayed Sweep while running in Roll mode, which I highly doubt - and which is what I wrote in my original post to which you responded. It had nothing to do with the horizontal reference point of the DSO when changing the timebase when stopped. That's a separate issue entirely.

With the SDS2000 here is what I have to do. I capture in roll mode, see something, press stop.  I move the interesting anomaly all the way to the right edge of screen.  I change the time base.  Unless I had half the desired focal point off the right edge of screen and half on screen, the focal point will shift to the left as I change time base.  This is because the center point of time base change is the edge of the screen.   For me it's a pain an inconsistent with typical scope behavior. 

I understand what you're complaining about very well since I use Roll mode often. Inconsistent with typical scope behavior? It depends on how you define typical behavior. If the horizontal reference point should only ever be the center of the screen, then I suppose it's inconsistent. If, OTOH, you think it should be the acquisition point, it's consistent. On my DSO, I can change the reference point in YT mode to Center, Trigger, or User-defined... but not in Roll mode - so I know it could be handy to allow setting it there as well.

The scope business started with higher end and expands down to lower price point as technology allows.  I consider the norm to be higher end scope behavior and lower end scopes are doing their best to mimic.  Rigol is doing a good job, but they do not set the bench mark of how a scope should perform.

I never said they set a benchmark - I was just reporting the similar behavior of another DSO.

No, I'm not confused - I never said delayed sweep (that was your interjection). You were challenging an assertion I did not make.  No worries.  Ha, we can blame scope vendors for not making scopes  100% intuitive to 100% of population and changing things over the years...

Anyway, I'm only referring to a zoom or timebase change around an event on the captured waveform that is not the trigger or start of acquisition. And only after acquisition is stopped.  It always centers about the right edge on sds2000 in roll mode vs. center. That is inconsistent with the many scopes I have used, maybe not the ones you have used. And of course  there is no configurable trigger with roll mode (other than pressing start) hence no delayed sweep from a trigger.  So true delayed sweep is a tangential topic and not used in roll mode since the trigger is your finger's job.

I credit agilent with using a timebase change post trigger as a zoom. The first time I used an infiniiim (coming from tek tds700 series) i was confused that you could change the timebase after you capture.  I would change timebase and see the sample rate and sample size change.  What?! You just ruined my captured  waveform!! But it didn't. Seemed so simple and intuitive, right?  The bad part? If you press run you now had a new sample rate and memory depth. So you had to set timebase back to what it was before new acquisition.   I liked tek's way of constructing a trigger or sequence of triggers, delayed or not, and then letting you do a true zoom where you leave sample rate and memory alone as you browse the captured waveform.    But today, both agilents and teks now let you zoom or change timebase after capture. You get to decide if you want to change next acquisition (timebase change) or leave as (zoom). Today Siglent is lacking the zoom and does a wonky timebase change in roll mode

[marmad]

No, I'm not confused - I never said delayed sweep (that was your interjection). You were challenging an assertion I did not make.  No worries.  Ha, we can blame scope vendors for not making scopes  100% intuitive to 100% of population and changing things over the years...

OK, I get it    But you did write "1) zoom does not work in roll mode". I assumed that you already KNEW that "zoom" on the Siglent (as well as many other low-cost DSOs) means a delayed sweep mode - thus impossible in roll mode (although, as I mentioned in another post, it could function differently when stopped).

But it's absolutely true that the industry has confused the meaning of the word by using it for different features at different times. BTW, this might be Agilent's fault since they changed the name from Delayed Sweep to Zoom on the X-series, with Chinese makers following suit (though I could be wrong) 

[don]

No, I'm not confused - I never said delayed sweep (that was your interjection). You were challenging an assertion I did not make.  No worries.  Ha, we can blame scope vendors for not making scopes  100% intuitive to 100% of population and changing things over the years...

OK, I get it    But you did write "1) zoom does not work in roll mode". I assumed that you already KNEW that "zoom" on the Siglent (as well as many other low-cost DSOs) means a delayed sweep mode - thus impossible in roll mode (although, as I mentioned in another post, it could function differently when stopped).

But it's absolutely true that the industry has confused the meaning of the word by using it for different features at different times. BTW, this might be Agilent's fault since they changed the name from Delayed Sweep to Zoom on the X-series, with Chinese makers following suit (though I could be wrong) 

I don't think sds2000 does a delayed sweep. Maybe Siglent calls it that, but you can't say trigger, wait 1s, and then capture. You always capture the trigger. So I'd say siglent does not have a delayed timebase at all regardless of what they call. I'd go back to early agilent infiium vs x series for combining zoom / "delayed sweep". But definitely blame agilent.  Dual time bases let you delay or count events and then capture. Today's "delayed sweeps" on some scopes just let you move trigger position around and not do a true trigger, wait, capture.

[marmad]

I don't think sds2000 does a delayed sweep. Maybe Siglent calls it that, but you can't say trigger, wait 1s, and then capture. You always capture the trigger. So I'd say siglent does not have a delayed timebase at all regardless of what they call. I'd go back to early agilent infiium vs x series for combining zoom / "delayed sweep". But definitely blame agilent.  Dual time bases let you delay or count events and then capture. Today's "delayed sweeps" on some scopes just let you move trigger position around and not do a true trigger, wait, capture.

I feel we may be having a divergence of terms again.  A delayed sweep provides a detailed look at a small selected portion of the main timebase. The main timebase normally serves as a controllable delay, after which the delayed timebase starts. DSOs allow waveforms to be displayed in this way, without needing to offer a true delayed timebase as such (as was needed on analog oscilloscopes to do the same thing). So yes, the Siglent (as well as Rigol, Agilent X-series, many many other DSOs, etc) certainly does do the equivalent of a delayed sweep. That's exactly what's happening when you use "Zoom" in YT mode while the DSO is running.

While you can argue that a DSO's delayed sweep can't do everything delayed sweep on an analog scope (or with dual timebases) can do, you can make the same argument for any number of features.

[nctnico]

Well, let's see what Siglent comes up with. I have pulled the trigger on buying an SDS2204 which I hope to get somewhere next week. Lot's of things to try out 

Not to break you balls or anything, but I thought you've been consistently arguing here at EEVblog over the last couple of years that older, used scopes were both cheaper - and could provide the same functionality - as these modern Chinese DSOs. Or did I get it wrong? 

You didn't remember wrong. It's just that the Chinese have moved forward rapidly. I also argued that there was a huge gap between the low end 2 channel Chinese scopes and the higher end 4 channel scopes from the 'A-brands'. The DS1000Z series from Rigol and even more so the SDS2000 series from Siglent finally fill that gap. If you compare the Siglent SDS2000 series with scopes offering similar functionality from Rigol, R&S (Hameg), Keysight (Agilent) or Tektronix you pay 2 to 3 times more. Spending €2500 would also buy me a 10 year old second hand MSO from Ebay but that is likely not to have protocol decoding, LAN connectivity, USB, etc.

[marmad]

You didn't remember wrong. It's just that the Chinese have moved forward rapidly. I also argued that there was a huge gap between the low end 2 channel Chinese scopes and the higher end 4 channel scopes from the 'A-brands'. The DS1000Z series from Rigol and even more so the SDS2000 series from Siglent finally fill that gap. If you compare the Siglent SDS2000 series with scopes offering similar functionality from Rigol, R&S (Hameg), Keysight (Agilent) or Tektronix you pay 2 to 3 times more. Spending €2500 would also buy me a 10 year old second hand MSO from Ebay but that is likely not to have protocol decoding, LAN connectivity, USB, etc.

Good points - thanks for clarifying! 

[don]

I don't think sds2000 does a delayed sweep. Maybe Siglent calls it that, but you can't say trigger, wait 1s, and then capture. You always capture the trigger. So I'd say siglent does not have a delayed timebase at all regardless of what they call. I'd go back to early agilent infiium vs x series for combining zoom / "delayed sweep". But definitely blame agilent.  Dual time bases let you delay or count events and then capture. Today's "delayed sweeps" on some scopes just let you move trigger position around and not do a true trigger, wait, capture.

I feel we may be having a divergence of terms again.  A delayed sweep provides a detailed look at a small selected portion of the main timebase. The main timebase normally serves as a controllable delay, after which the delayed timebase starts. DSOs allow waveforms to be displayed in this way, without needing to offer a true delayed timebase as such (as was needed on analog oscilloscopes to do the same thing). So yes, the Siglent (as well as Rigol, Agilent X-series, many many other DSOs, etc) certainly does do a delayed sweep. That's exactly what's happening when you use "Zoom" in YT mode while the DSO is running.

Actually...just looked at scope and it can do a delayed trigger . Meaning trigger, wait 10s, capture. But this is separate from zoom mode as you suggest.  Zoom on sds2000 is more of a true zoom of what you see on the screen. When you exit zoom all settings remain. In fact, zoom setting have no impact on the acquisition at all (true zoom in my mind). For example, I can move the trigger 10s to the left with horizontal knob. Scope triggers, waits 10s, then captures. Now if I press zoom button, I can expand / contract the waveform that is on screen. No way to see part of waveform that was triggered on (correct though). When I exit zoom, I still see "delayed 10s" and all was the same settings before I pressed zoom.  This is consistent with other scopes that have dedicated zoom mode.

Interestingly enough, this is another example where there used to be a dedicated function that was redefined over time and now most scopes do a superset of old and new behavior.  Last discussion was on timebase, this is horizontal knob. There used to be a delayed trigger setting and horizontal position only referred to on-screen (tek's old way).  Tek had a separate trigger position not tied to horizontsl knob.  Agilent let the horizontal knob  also control the trigger position.  Today both agilent and tek (depending on scope)  let you do both. The sds2000 (need to circle back to topic!) only let's you do a delayed trigger via horizontal position knob.

[tautech]

Does this post need to be in the other thread for Siglent reps to see?

Thanks don, they will read it, have a think if any posts need editing/correcting or additions(point to a subsequent post).
All good stuff, thanks all, made for a good first read of the morning. 

[Hydrawerk]

1) zoom does not work in roll mode

Hmm... does Zoom work on any DSO while using Roll Mode? It certainly doesn't work on the Rigol DS2000 series - and I'm not so certain it could.

Roll mode is a completely different kind of acquisition - sampling and doing decimation bit by bit - as opposed to a screen at a time. Zoom mode suggests that the DSO go back to the original sampled data and re-decimate for a different time base. This seems like an impossibility to me in Roll mode due simply to the method of acquisition. But perhaps I'm wrong about this.

Zoom does not work on  my Agilent DSOX2002A while using Roll Mode. This is probably very common on many scopes.

[Wuerstchenhund]

Hmm... does Zoom work on any DSO while using Roll Mode? It certainly doesn't work on the Rigol DS2000 series - and I'm not so certain it could.

Roll mode is a completely different kind of acquisition - sampling and doing decimation bit by bit - as opposed to a screen at a time. Zoom mode suggests that the DSO go back to the original sampled data and re-decimate for a different time base. This seems like an impossibility to me in Roll mode due simply to the method of acquisition. But perhaps I'm wrong about this.

It might depend on the price point.  Newer scopes like tek, agilent and lecroy do. 

At least in regard to LeCroy you're right, they can indeed zoom when in roll mode. I just tried on my LeCroy's, and lo and behold, no problem zooming in when the scope is in roll mode.

The first two screenshots are from a WaveRunner 2 LT264M, the second two from a WaveRunner 64Xi (the signal is a 3Hz square wave).

I don't have an Agilent or Tek scope at hand at the moment so I can't check but if I remember right their scopes can't zoom in roll mode.

[don]

1) zoom does not work in roll mode

Hmm... does Zoom work on any DSO while using Roll Mode? It certainly doesn't work on the Rigol DS2000 series - and I'm not so certain it could.

Roll mode is a completely different kind of acquisition - sampling and doing decimation bit by bit - as opposed to a screen at a time. Zoom mode suggests that the DSO go back to the original sampled data and re-decimate for a different time base. This seems like an impossibility to me in Roll mode due simply to the method of acquisition. But perhaps I'm wrong about this.

Zoom does not work on  my Agilent DSOX2002A while using Roll Mode. This is probably very common on many scopes.

Good to know - I wonder if it works on Dsox3000. I know it works on higher end Agilents (ie 6000 / 8000 /9000)  and Tek scopes (dpo7000, 4000) series. 

Can you at least change the timebase after capture to mimic zoom?

[marmad]

At least in regard to LeCroy you're right, they can indeed zoom when in roll mode. I just tried on my LeCroy's, and lo and behold, no problem zooming in when the scope is in roll mode.

Nice, I've never seen it before.

So it appears they're doing a magnification of the samples from the center of the screen while new acquisition samples appear at the right? Can you change the window of magnification to a different horizontal reference point, as opposed to the center?

[Wuerstchenhund]

So it appears they're doing a magnification of the samples from the center of the screen while new acquisition samples appear at the right? Can you change the window of magnification to a different horizontal reference point, as opposed to the center?

I can move the zoom cursor at any part of the signal, it's not limited to the center. It works exactly as it would do when running in normal (non-roll) mode.

There's also no noticable slow down of the scope or the screen update, not even on the old WaveRunner 2.

I wouldn't be surprised if even the old LC and 9300 Series scopes can do that.

[don]

I just discovered something really odd on SDS2074.  I'm in roll mode at 100ms / div and 14Mpts.  If I stop acquisition and just start changing time base the sample size decreases.  As I transition from 100us to 50us, the sample size decreases from 1.4Mpts to 700kpts.  If you monitor in the center of the screen at this transtion, you'll see the wavefrom change as if there is a void in the waveform at the center of the screen and the scope is populating the void with data from elsewhere in the waveform.  As you scroll the right you'll see the data changes at this discontinuity at the center of the screen. If you scroll left,  the message "Horizontal position at limit!" pops up if you (recall that for some reason the SDS2000 changes timebase on the right side of screen instead of the center of screen when in roll mode). 

Two pictures posted show how waveform at center of screen changes (look at CH4)

So, something very odd going on here.  Might be related to the odd behavior I saw with roll mode if sample size was 7Mpts or or 14Mpts and you decreased the time base down far enough from this post:

https://www.eevblog.com/forum/testgear/siglent's-new-product-msosds2000-series/msg493459/#msg493459

[don]

Regarding Sequence mode testing I did:

 https://www.eevblog.com/forum/testgear/siglent's-new-product-msosds2000-series/msg493694/#msg493694

So, looks like the Sequence menu under acquisition just lets you enable sequence, select the maximum number of frames (frame set) and lets you select a 19 frame window to play back.  The play back does not seem to work.  If you want to manually scroll through the frames you captured, you need to go to the Utility menu (?) and select History, and then you can manually scroll through the frames you captured.  Not intuitive to have to exit the sequence menu and hop into the utility to see what you captured, but it works.  Decode does not work in sequence mode. 

Can anyone confirm that decode is supposed to work with segmented memory? I thought I read that but can't find it.

I also noticed that I can't enable decode if I have 14Mpts set and sample rate is <20MSa/s (maybe 10..don't recall).  At smaller sample size I can enable decode and much less than 20 or 10MSa/s. 

Summary of issues here:
1) Decode does not work in sequence mode.  (is it supposed to?)
2)  can't enable decode if I have 14Mpts set and sample rate is <20MSa/s (maybe 10MSa/s..don't recall). 
3) playback in sequence mode does not seem to work

[marmad]

I can move the zoom cursor at any part of the signal, it's not limited to the center. It works exactly as it would do when running in normal (non-roll) mode.

There's also no noticable slow down of the scope or the screen update, not even on the old WaveRunner 2.

I wouldn't be surprised if even the old LC and 9300 Series scopes can do that.

It seems a handy thing to have. Clearly it works (from a firmware standpoint) differently than the normal 'delayed sweep' type of zoom with triggered acquisitions, so would require different routines to be written. I suspect that's the reason it doesn't exist on lower cost DSOs - which is too bad because I can imagine certain situations where it could be very useful.