Rigol DS1074Z persistence problem ?

[Big_Al]

Hi All,  just got hold of a Rigol DS1074Z scope and am finding the persistence a pain in the ass.
  I have simply got the scope on Normal Trigger capturing a short burst of waveforms.
Not much to ask of it to end up on the display with the last waveform it triggered on is it ?
  Instead it ends up displaying a merger of some of the previous waveforms.
(picture attached)
I have set the 'persistence' setting to 'min' hoping that this is as good as off but its obviously not ?
  Anyone any idea how to actually turn persistence completely off on this thing, or is this a bug ?
Anyone had the same problem on the 2000 series too I wonder ?
Looks like going back to my TEK tds210 at this rate.
Thanks.
Al.



ps, anyone tried the hacks on this model yet ?

[JDubU]

Use single sweep mode?

[Wim13]

What is the type of signal you want to display..?

here a sample picture of a DS1000z with a burst signal, 5 bursts every 80 mS at 6.3 khz.

[iDevice]

This is not a persistence problem, this inherent to how modern, high frame rate DSO (or DPO in this case) work.

You have about 600micro sec per trace so roughly 1600 acq. frames/sec
All these frames are written in display memory that only refresh the screen like 100 times a sec,
So that gives about 160 acquisition frames superimposed every 10ms.
If your burst of data lasts less than that, well, they all will be stacked on top of each other in display memory.
If you stop the trigger you will display the last display memory with everything superimposed.

So the solution is to either set a time base corresponding to the full length of the burst instead of one data frame and zoom-in to analyze each data frame (you may have to select single sweep if the bursts are repetitive), or set the time base to about one data frame duration but use segmented memory and record as many segments as you have data frames in your burst.
Then you can analyze each data frame by jumping between segments in playback mode.

Remember that you may have also to adjust the holdoff if you fail to sync each time on the start of a new data frame.
Hope this helps.

[Big_Al]

Thanks for the replies guys.
  Seems like a case of using my old tekscope for too long then I think !
Will have to investigate how all this new high speed display stuff works then by the looks of things and try to get used of it.
  Still seems a bit of a pain in the ass though !
The data I was looking at was a simply string of serial characters coming in (19200), a character every 10ms, and about 2 seconds worth of them. I was confused why the last one in wasn't displayed as it 'should' be, and assumed it was the persistence setting.
  I have just found out by accident that if I press the 'CH1' button when the last merged waveform is displayed, it then actually changes the displayed waveform to the actual last triggered raw waveform and removes the merged stuff off it. Oh well, you live and learn !
Cheers,
Al.

[hans]

In this situation, you could probably do various things to "fix" this:

- Use a different trigger. With Nth edge (if you have the hack/trail/option installed) you could trigger on a particular burst, but as you're triggering on data that's not very easy to do (ideal for clock or chip select lines, where you want have a stable trigger on latter part of the burst, especially when the signal is a bit jittery)
- Do a single-shot of the complete burst, and when the scope has stopped zoom in and analyze. You can set up the Rigol to 12MPts (under Acquire) of memory on 1 channel. Most often I leave it auto and let the scope handle it.

- I think (not certain though) the scope would redraw the last trigger if you would zoom in, and the superimposed frames are gone.

With the high intensity setting you're just making each capture more like the others, with little contrast at all. Much like an old TDS210 indeed

[iDevice]

Ok, if you have a serial data frame every 10ms, I guess it's borderline with the display rate.
I don't know what the display rate of the 1000Z is, but if it's like 60hz, it may be low enough to show the phenomenon, but I think it's higher than that with today's TFT lcd's.

Ans if it disappears when you push CH1, then I would go indeed for some bug.
I'll try to reproduce that on my DS4k.

[Rory]

If you still have the trial or have the Official option installed, there is an RS-232 trigger also.

[nctnico]

Ok, if you have a serial data frame every 10ms, I guess it's borderline with the display rate.
I don't know what the display rate of the 1000Z is, but if it's like 60hz, it may be low enough to show the phenomenon, but I think it's higher than that with today's TFT lcd's.

Ans if it disappears when you push CH1, then I would go indeed for some bug.

That is not a bug. Tektronix scopes have been working exactly like that for over 20 years.