DS1104Z @ 56 000 wfms/s real or not..??!

[Rosendorfer]

Hi

Not sure if that have been reported earlier, but while playing with my DS1104Z I have find out that at 50ns timebase with Auto memory settings at Dots display mode this scope is running at 56 000wfms/s (!!!).
This is surprisingly more than advertised 30 000 wfms/s found at Rigol papers.
Not sure if this is true or just some error in my measurements, but in my set-up it is repetitive and constant measurement.

Have not been able to locate some other users wfms/s measurements at DS1104Z or in that sake DS1000Z family.
In next few days I could run more measurements for different scope settings, but first would be nice to confirm that this bit wild 56 000 wfms/s at DS1104Z is real . It seems to be even more that max wfms/s of DS2000 so not sure what to thing about it.
So if anyone could check and confirm this and share about would be great.

At technical side I run sin-gen waves at channel 4 of DS1104Z and measure TriggerOut with old and trusty DS1052.

Rosendorfer

[Rosendorfer]

Hi

Well the point is if this 56 000 wfms/s @ DS1104Z is true number, then this would be really unusual rare case where producer ie. Rigol is giving actually worst parameters at its own marketing papers that is available at real product..!!

I have repeat this with different types of input signals and seems that it is working at 56 000wfms/s with any signal type.
Seems to be the real thing...
Hope someone could confirm this case...

Rosendorfer

[Creep]

IIRC, there was a discussion about this in the forums a while back. From what I can remember, the waveform update speed used to be the specified 30k wfms/s, but rigol managed to improve it tremendously with firmware updates. 50 - 55 k wfms/s was the discussed figure, i think. Still, don't quote me on this as I have not tested it myself (recieving my DS1054Z on wensday I believe).

[rolycat]

I have just tried this on my upgraded DS1054Z and the numbers look even better: 63 000 wfrm/s.

[Rosendorfer]

Hi rolycat

63 000wfms/s on DS1104Z !? ...
How did Yo measure that, I would like to check that on my scope.

Rosendorfer

[Rosendorfer]

Hi guys

I have found Russian (!!) YouTube channel of Maks Krukow where he is doing 2 hours plus review of DS1104Z @  .
And at  about 1:24:30 Maks is very nicely showing 55 000 wfms/s of DS1104Z trigger output with some comments, my Russian is bit rusted to comment at the whole review, but the picture is very clear and definitely worth looking at.
And also just after that he is showing on his DS2072 how exactly trigger output of DS1104Z is looking like and then also 63 000 wfms/s update rate as pointed out by rolycat.

So seems that this Rigol indeed was bit shy in specification of update rate of DS1104Z and 56000 wfms/s or even by some account 63 000 wfms/s is the way the scope is really working in DotMode at 50ns and Auto memory settings.

I have made  more measurements at other Timebase and Memory length settings and put it in table, hope that it is clear.

BUT....!!!!

At this case I would like to point one important thing regarding Vector Display settings at DS1104Z.
PLEASE be aware that the numbers of wfms/s in Vector mode that are point at table are kind of MAX of wfms/s , as scope trigger out and as we like to thing wfms/s is in drastic! way influenced by simple change in Amplitude of signal that is shown on screen or complexity of signal.
Seems like scope display engine have much more work to be done in Vector mode with Linear or sinx/x interpolation and its wfms/s performance can be DRASTICLY! influenced by simple changing the Amplitude of signal. See the pictures of measurements made at 100ns in Vector mode of the same signal where with low amplitude we get 18 000 wfms/s and by just increasing the amplitude only 10 000 wfms/s or possibly even less. 

This is thing that AFAIK never have been mentioned by guys doing similar trigger out aka wfms/s measurements on SD2202 family. Wonder if that could be cleared out.

In Dot display mode on other side wfms/s is just stable and there is no any kind of such changes so You possibly can trust that scope is running with wfms/s as shown in table... and the numbers looks really nice I would say.

Rosendorfer

[Mark_O]

So seems that this Rigol indeed was bit shy in specification of update rate of DS1104Z and 56000 wfms/s or even by some account 63 000 wfms/s is the way the scope is really working in DotMode at 50ns and Auto memory settings.

Not necessarily so much "shy", as having developed those Specs at an earlier date, then found ways to optimize them in new firmware builds later.

At this case I would like to point one important thing regarding Vector Display settings at DS1104Z.
PLEASE be aware that the numbers of wfms/s in Vector mode that are point at table are kind of MAX of wfms/s , as scope trigger out and as we like to thing wfms/s is in drastic! way influenced by simple change in Amplitude of signal that is shown on screen or complexity of signal.

Seems like scope display engine have much more work to be done in Vector mode with Linear or sinx/x interpolation and its wfms/s performance can be DRASTICLY! influenced by simple changing the Amplitude of signal. See the pictures of measurements made at 100ns in Vector mode of the same signal where with low amplitude we get 18 000 wfms/s and by just increasing the amplitude only 10 000 wfms/s or possibly even less. 

This is thing that AFAIK never have been mentioned by guys doing similar trigger out aka wfms/s measurements on SD2202 family. Wonder if that could be cleared out.

That is well worth pointing out, but not surprising in any way.  Nor do I think it needs to be cleared up.  In Vector mode, whether doing Linear or Sin(x)/x, additional points must be generated by the processing engine, which takes time.  And drawing those extra points into the display buffer will also take time.  Both of those are roughly proportional to the waveform amplitude (though other factors may offset or skew that a bit).

In Dot display mode on other side wfms/s is just stable and there is no any kind of such changes so You possibly can trust that scope is running with wfms/s as shown in table... and the numbers looks really nice I would say.

Agreed.  And in Dots mode, since no extra points are calculated, and only the same fixed # of points is always drawn to display memory, the rate will remain fairly stable and consistent, at a particular setting.  That's why the focus has been on that Mode (repeatable results), in addition to the fact that it yields more impressive numbers for advertising. 

[Mark_O]

Oh, and in your chart, Auto-Vector-500ns shows a rate not only 8x faster than dot mode (39800, vs. 5000), but higher than at any other sample rate.  I suspect you have an extra zero on the end.

[Rosendorfer]

Hi

Chart hopefully is corrected now.
And in any case somebody wondered at 1us in 12k scope is "faster" than in Auto mode because in 12k at 1us it changed from 1Gs/s to 500Ms/s. While in Auto it was still at 1Gs/s.

Mark so then should I understand that it is absolutely normal also in DS2000 series that wfms/s in Vector mode is dependant on what is on the screen...?

Can not say that I have read all about DS2000 on this forum but have been reading a loot and do not remember anyone mention that.

Rosendorfer

[Mark_O]

Mark so then should I understand that it is absolutely normal also in DS2000 series that wfms/s in Vector mode is dependant on what is on the screen...?

Yes.  And not just for the DS2000.  This would be true for any DSO, and the remarkable thing would be if there was no dependency.  It could be minimized by eliminating intensity grading, and some scopes do.

[wn1fju]

On my DS1104Z, one channel, no input, 50 ns/div, dot mode, I get about 62,500 waveforms/second as measured on a scope.  But, be careful - if I look over a long time span, I can see the DS1104Z has gaps in the trigger output pulse train.  Thus, when I switch to a counter, I get about 52,500 waveforms/second because the gate time of the counter hits those gaps.  If I turn the counter gate time down enough, I can get just about any reading I want.  Whether the gaps are truly gaps in the processing, or whether the thing really is doing 62,500 consistently, I defer to the experts.  However, I just wanted to point this out because it may explain why people are reporting numbers similar to both cases.

[marmad]

Very comprehensive waveform update rate and comparison tables for both the DS2000A and the DS1000Z were made and posted here in June by EEVblog member Fagear (AKA YouTube poster Maks Krukow).

I personally think it's pointless to compare or advertise waveform update rates when in Dots mode (a practice started by Agilent, I believe). It's deceptive - and rather like comparing how well a car handles on a perfectly flat and empty road. I certainly don't use the DSO that way 98% of the time, and I think the same is true about most users. I only put the DSO is dots mode when I have to see the actual sample points for some reason - otherwise, vectors are always on - and that is the speed you should be interested in.