Siglent SDS2000X Plus

[Martin72]

Hi,

it seems to me that the intensity gradng rendering was slightly "broken" in the latest firmware.

Try it a few minutes ago, for me it´s still the same as before, as I´ve bought it.

[maxspb69]

Yes, I have exactly the same on my screenshots.
I guess it seemed to me, so everything is ok.

And on real scope screen  it very much depends on the angle from which you look at the display. And why did they not equip this scope with the IPS display ?!

[Martin72]

It could be going always better...
This Display is better than the one of the rigol MSO5, I´ve owned.
Laugh or not, I´ve searched and found a suitable IPS display for it, with the same dimensions and connectivity.
Would cost 69€ - That sounds cheap, but if rigol would build the scopes with that, they won´t cost the money before.
Same by siglent I guess.

[maxspb69]

I agree, the SDS2000X+ display is not so bad, I just involuntarily compare it with the RTB2000, and this is definitely not worth doing


By the way, if you find a suitable IPS display for SDS2000 - give me a link, please,  I will buy and install it.

[Martin72]

And on real scope screen  it very much depends on the angle from which you look at the display.

Like this, it´s a picture taken from the screen with lowest intensity adjusted - doesn´t looks like.
So videopics could be an illusion too.

[thinkfat]

I agree, the SDS2000X+ display is not so bad, I just involuntarily compare it with the RTB2000, and this is definitely not worth doing


By the way, if you find a suitable IPS display for SDS2000 - give me a link, please,  I will buy and install it.

What type is it now, TN? Might be for a good reason. TN displays are faster than IPS, no?

[maxspb69]

What we see is TFT TN. There is only one reason - the price of this type of display. The scope doesn't need to refresh the screen faster than 60 times per second, all IPS can do this for a long time ago, and many modern IPS are even much faster.

[casterle]

Functional bugs and limitations are problems when those bugs are encountered, but encountering them requires that you use the functionality in question.  The further away from the most common functionality you get, the less likely someone is to run into any problems or limitations there.   But the user interface?   That is used by everyone.

Very well stated. The value of a good UI is often underestimated, with non-critical but irritating fixes delayed by other priorities.

Perhaps those of us with a software background better appreciate this because it was often up to us to design them, so we learned to appreciate the importance of (and had lots of practice creating) good UIs.

[Martin72]

By the way, if you find a suitable IPS display for SDS2000 - give me a link, please,  I will buy and install it.

First we need to know the original display....
In the rigol case, I´ve simply asked them and became an answer with datasheet, so I could search for a suitable Exchange but with IPS.
10" IPS in 1024x600 are very hard to find, in real I´ve found only one.

https://www.banggood.com/de/10_1-Inch-1024x600-720P-65K-IPS-Full-View-LVDS-HD-LCD-Display-Screen-Capacitor-Touch-Board-p-1613920.html?utm_source=googleshopping&utm_medium=cpc_organic&gmcCountry=DE&utm_content=minha&utm_campaign=minha-de-de-pc&currency=EUR&cur_warehouse=CN&createTmp=1&utm_source=googleshopping&utm_medium=cpc_bgs&utm_content=frank&utm_campaign=frank-ssc-de-all-20bf11-1006-11sale&ad_id=471112227280&gclid=CjwKCAiAirb_BRBNEiwALHlnD6_jNQA8kqtaQCAHIJ7L_7qO2FOk18179sBwg7x_T3yCHsDbJPeiCBoClGwQAvD_BwE

[casterle]

Drooling idiot seeks help!

How do I get dot mode to work?

What I've tried:

• Start internal AWG, sending 300KHz 4vp-p Ramp to Channel 3, all other channels inactive
• Run Auto Setup
• Stop
• At this point the 'scope is set to sample at 2 uS/div; far too many points for dot mode, so...
• Set timebase to 500pS, reducing points to 100 total
• Single shot.
• Toggle between Dot and Vector modes
• Sit in slack-jawed wonderment for a moment, wipe drool from chin, start drafting this message

With no visible change between Dot/Vector mode, and having seen this very model run in Dot Mode, I know it works... Nice, nearly identical screenshots below.

Sadly, I'm not smart enough to attach images either, so another dumb question if you please. Once I click 'Choose File' and select the image on my system, what else do I have to do? AFAICT there's no upload button. Docs imply that inserting the markup below works, but preview shows text. If images appear after I actually post this, nevermind!

One final thought: The browser interface made getting these screenshots a piece of cake, and they look good to boot. It was trivial to configure the LAN and it worked immediately. Adding a SCPI command line was brilliant. At first I didn't think I'd use the browser UI, but to grab data it's too convenient to ignore. Good job, Siglent!

Vector Mode Screenshot:

Dot Mode Screenshot:

[Tankj]

Drooling idiot seeks help!

How do I get dot mode to work?

With no visible change between Dot/Vector mode, and having seen this very model run in Dot Mode, I know it works... Nice, nearly identical screenshots below.

Its a bug

My 2000X+ run in to this "dots display failure" thing After a few times of use when I update it to the latest firmware. 
Its probably a bug in the new firmware, Never had this problem on the older version.

recall a setting, or just press the "Default" button on the front panel to set factory default settings, then you should fix this.

I have tried all the operations or settings I have do before, just can't successfully reproduce the bug so far, don't know what caused the problem.

The good thing is, after fix it, it's totally gone.

[casterle]

recall a setting, or just press the "Default" button on the front panel to set factory default settings, then you should fix this.

Thank you!

[casterle]

I have a question about 10-bit mode. It appears this mode is implemented via hardware rather than in software like ERES.

In practical terms, does this mean that, for signals <= 100MHz, it will operate as if it had true 10-bit ADC's?

[Martin72]

Hi,

Please note that Siglent does not advertise this DSO as 10bit

[Zlotnik]

I have a question about 10-bit mode. It appears this mode is implemented via hardware rather than in software like ERES.

In practical terms, does this mean that, for signals <= 100MHz, it will operate as if it had true 10-bit ADC's?

In the comparisons maxspb6 has posted two pages earlier in posts 2276 and 2277 it seems the noise is reduced when activating 10bit mode, which to me seems to indicate a reduced bandwidth due to filtering/averaging before decimation. To me 10bit mode looks like a normal HiRes or ERES mode - would be interesting if someone tested the SDS2k+’s bandwidth in 10bit mode.

Compared to that, the R&S RTB2k shows pretty much the same noise as the SDS2k+, but at native 10bit. Activating the RTB’s HiRes mode reduces noise similarly to the SDS2k+, and actually also reports the reduction in sample rate. Of course, the RTB is quite a bit more expensive as well, so horses for courses.

[Martin72]

would be interesting if someone tested the SDS2k+’s bandwidth in 10bit mode.

Dave did it involuntary...
Apart from this, you can add Eres too and get the 10bit mode + 3bit additional, of course with much more reduced bandwith.

[0xdeadbeef]

I have a question about 10-bit mode. It appears this mode is implemented via hardware rather than in software like ERES.

Note that the SDS5000X (which uses more or less the same SW) has a true ERES (and averaging) mode in HW (i.e. done on the FPGA) while the SDS2000X plus features this 10bit mode in the acquisition menu and offers ERES (and averaging) only as math functions. This makes me think that the "10bit mode" in the SDS2000X plus is probably a simplified ERES implementation. Anyway, it's a digital filter implemented on the FPGA (so in HW) which reduces the effective bandwidth.
Going with the "ERES light" assumption, the manual says that for 2 ERES bits, the -3dB point is about 0.028 times the sample rate, so for the maximum sample rate of 2GSa/s (SDS2000X+), this would result in a maximum bandwidth of 56MHz. For 1.5 ERES bits, the manual states a -3dB point of 0.055 times the sample rate which would result in a maximum bandwidth of 110MHz. This is closer to the "around 100Mhz" limitation stated for 10bit mode.  So I guess either the "10bit mode" is a 1.5bit ERES mode or maybe a slightly simplified implementation. However, it seems implausible the bandwidth limitation is independent of the sampling rate and calling it 10bit mode might be a bit on the optimistic side. Probably it's more like a 9.5bit mode rounded up to 10bit.

[Martin72]

Note that the SDS5000X (which uses more or less the same SW) has a true ERES (and averaging) mode in HW (i.e. done on the FPGA) while the SDS2000X plus features this 10bit mode in the acquisition menu and offers ERES (and averaging) only as math functions.

"Only" ? 

[0xdeadbeef]

In the sense that the SDS5000X supports these as acquisition modes on all channels and without (major?) impact on refresh rate like the big brothers from LeCroy. It's important to note though that even the Wavesurfer 3000Z supports ERES "only" as math function.

[tautech]

In the sense that the SDS5000X supports these as acquisition modes on all channels and without (major?) impact on refresh rate like the big brothers from LeCroy. It's important to note though that even the Wavesurfer 3000Z supports ERES "only" as math function.

There were changes for SDS2kX Plus for ERES and Averaging as these both are now treated as Math functions instead of Acquisition modes.
From Dec 2019:

ERES is not an acquisition mode anymore, but a math function (LeCroy send their regards!).

Average and ERES are now implemented as math functions.

[Tankj]

I have tried all the operations or settings I have do before, just can't successfully reproduce the bug so far, don't know what caused the problem.

The good thing is, after fix it, it's totally gone.

Update: it's simply caused by the "Zone trigger".

Try this: set display type to dots, timebase under 2ns. Open Zone trigger menu, toggle between on and off, you'll see the magic.

I think this may for better trigger judgment with the zone trigger function, not a big deal anyway.

[casterle]

Update: it's simply caused by the "Zone trigger".

I don't think it was related to zone triggering in my case as I haven't tried that feature yet.

[Performa01]

No need for speculations about 10 bit mode and ERES - it's all there, even a frequency response graph (replies 32, 57, 58 and 62):

https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2786952/#msg2786952
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2840282/#msg2840282
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2840298/#msg2840298
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2857298/#msg2857298

In short: 10 bit mode is hardware accelerated, with no impact on the waveform update rate. It equals 1 bit ERES, thus provides 10 bits of resolution and a theoretical increase of 1 bit ENOB.

ERES 1 bit is certainly fit for 100 MHz bandwidth.

We can even add an ERES 3 bit math function, for a total resolution of 16 bits and 4 bits theoretical ENOB enhancement. Keep in mind that we still have the linearity of an 8 bit ADC. so it heavily depends on the application whether this makes sense (at least the filter effect should).

[Tankj]

I don't think it was related to zone triggering in my case as I haven't tried that feature yet.

Have you fix it yet? It doesn't really needs you to USE the zone triggering, the scope will forced to Vector mode once the zone trigger menu is "on", even NO trigger zones are in use, and no zones show on the display.

If you can try (or successfully) reproduce the problem by other processes, it should certainly considered as a bug.

[0xdeadbeef]

In short: 10 bit mode is hardware accelerated, with no impact on the waveform update rate. It equals 1 bit ERES, thus provides 10 bits of resolution and a theoretical increase of 1 bit ENOB.
We can even add an ERES 3 bit math function, for a total resolution of 16 bits and 4 bits theoretical ENOB enhancement. Keep in mind that we still have the linearity of an 8 bit ADC. so it heavily depends on the application whether this makes sense (at least the filter effect should).

My main point was that the bandwidth limitation is a function of the sample rate (and the filter length resulting in the number of ERES bits). Directly counting the ERES bits as additional ENOB bits as done in the Siglent manual might be a bit optimistic. Note that LeCroy doesn't do that in the WS3000Z manual/data sheet. So LeCroy defines 3 ERES bits on an 8bit ADC as equivalent to an 11bit ADC - which would mean that a 10bit mode equivalent mode would need 2 ERES bits according to LeCroy's definition. We agree though that the "10bit mode" of the SDS2000X+ can't be actually a 2bit ERES mode. Whether it's 1bit or 1.5bit doesn't really matter so much.