Siglent SDS2000X Plus

[Martin72]

Here is the "instruction" for measuring the max. waveform update rate from siglent...

[rf-loop]

Today I wanted to test the waveform updaterate, using this cute toy: (Picture got lost....wtf)


It performed well, but in sequence mode, you will have set the trigger to normal - And in normal triger mode, the toy couldn´t measure.....

So sequence mode measure will follow when I get an alternative to the toy.

For fast sequence mode it can of course measure like also Siglent simple method advice but if need or want do more perfect there need find perfect limit to real trusted max speed.

Possible and simple way to measure it is Single shot sequence and using enough perfectly known burst signal.

Way I have made other some other models measuremets was this (simplified) and this can do for all time scales not only for 50ns/div what give only UP TO value for sales persons and advertisements. it can do also for all channel combinations and memory limit combinations.

What ever is wanted channe(s)  timescale and memory you want test.
Next step is set amount of segmments. Better is max with used settings.
Then next step is set signal generator for generate Exactly! this amount of pulses or sinewave etc cycles.
Then select pulse frequency inside this burst. Do not start. Of course mode need also be Single shot burst, not repeating.
Set scope for Single sequence mode and start it for waiting trig's.
Then start this signal generator Single burst. If sequence come ready (amount of segments trig'd  and captured)  it have passed test with this  speed what is burst frequency. If there is missing pulses scope stay waiting more until amount of defined count of segments captured. In this case test is failed with this pulse freq.
If not get all, then set more low freq and try again. If pass then try higher freq. Until you find limit. Take small safety margin down and that is it. This is quaranteed max sequence trig speed. Yes it need work and time.

Also

If you select continuous repetitive sequence mode and not watch if every single pulse is captured you get too optimistic UP TO value. In some models difference is more and some other models difference is not so high. Example older SDS2000 have not big difference but SDS1000X-E may have more differnce between these two test methods, because they measure different thing.
Only true important thing is speed where it do not drop out ANY single pulse.
If we go to more deep inside trigger speed we may find some interesting advantages least on some other models. I do not know SDS2000XPlus.

Here just for example some old results from other models what give some imagine what I mean.


Older SDS1000X


and SDS1202X-E

These are guaranteed max speed measured with principle as previously told where is important to not not any dropped out pulses.
But yes these tests takes some time... and cross check and cross check for avoid false claims. 

[shtoz]

Hi guys,

At this time i don't need high R&D capabilities of my Lecroy wavepro oscilloscopes, and i am looking for some smaller oscilloscope.

I have some questions:
1. How good this siglent are in terms of DC offset and saturation? In Lecroy scopes you can offset you signal and "zoom in" up to several times needed piece of you signal without distortion. I mean this https://youtu.be/n_dXvpEV18g?t=155
2. Can someone provide screenshots of sin 200Mhz and 400Mhz with linear interpolation (without sinx/x interpolation) with 2Gs/s and 1Gs/s SR?
3. Can someone provide screenshots of fast rise meander with linear interpolation (without sinx/x interpolation) with 2Gs/s and 1Gs/s SR?

Thanks!

[Vestom]

Got the SDS2104X Plus today 

Here are some quick first impressions:

Positive impressions:
- The scope feels solid and the shell is made from thick plastic. The handle is molded with a soft rubber grip.
- The screen is big, nice and relatively bright and clear. Not iPad clear, but quite OK. The viewing angles are good, except from below. Not an issue for me. Trace drawing is nice and clear.
- The form factor is perfect. I will trade individual vertical knobs for a big screen any day. Maybe in a future model the screen could even be a little bit bigger in the same form factor, if ditching the black bezel.
- The buttons are nice, relatively firm and with a good tactile feedback. Not the mushy kind, where the whole keyboard moves when pressing one button. In my opinion the button LEDs are nice and bright on 50% brightness. The layout is OK and grouped as you would expect. All scopes are different and an experienced scope user will have no trouble finding their way around this one. A couple of buttons could have been saved and put into the menus (See below).
- The encoder knobs are with soft touch and with a good detents on the horizontal/verticals. The least impressive parts are the four detent-less narrow encoder knobs, which are not completely straight and wobbles slightly when turned. They mar the quality feel a bit. Time will tell whether they will last.
- The scope feels pretty responsive. There is immediate audiovisual feedback from touch, so even though the scope may take a little longer to draw menus and make settings effective it does not feel especially laggy to operate. Maybe at some settings and with all 4 channels enabled it sometimes becomes a bit slow - but not unbearably so.
- User interface is nice, discrete and unobtrusive. Small, but easily readable fonts. Looks professional and modern. (Not gaudy and cartoony like some other scopes...). Maybe the top bar could have been slightly slimmer to provide more trace space and the icons left out to reduce the clutter.
- 10 bit mode is pretty good. I will likely run the scope in 10 bit dot mode most of the time - especially since it is limited to 100MHz anyway... (Vector interpolation really shouldn't be the default setting, as it will often fool users to use a scope beyond its true capacity.)
- Noise level is good!
- FFT also seems usable for a scope.

Potential for improvement:
- The probes accompanying the Oscilloscope are 200MHz PP215 1x/10x probes. However the probes have no detection pin, so even though the scope has a "10x probe detection ring", it does not work with the probes... 
- Power button is a soft button and the scope uses almost 4W when turned "off"! Though, there is no apparent benefit (e.g. faster boot) from this substantial idle power use. For comparison EU regulations require consumer equipment to use less than 0.5W in off mode, so I guess oscilloscopes are exempted from this... A hardware button on the back would also have helped instead of having to pull the plug.
- Boot time of 45 seconds is just acceptable.
- The fan in the scope is not exactly silent. Although the scope has do dissipate ~55W, the airflow does not feel to be in proportion with the sound level from the fan. Also has a bit of "cricket" sound beyond the "whoosh"
- I would much prefer to have a "Force trig" button instead of the useless "Touch", "Default" or "Auto setup" buttons. The scope is pretty useless without touch and the other two could easily have been buried in the utility menu - especially since you have to confirm them by touch anyway... (Edit: actually they are already present in the Acquire menu)
- Bug: If touch is disabled (Press the "touch" button) and either "Auto setup" or "Default" is pressed, the scope locks up... (Software 1.3.5R5)
- Memory depth: At 10ms/div you get both maximum sample rate 2Gs/s (1 channel) and maximum memory depth 200Mpts. Above 10ms/div the sample rate is lowered because of the maximum memory depth. Below 10ms/div the memory depth is lowered. Ideally the user could choose between using full memory depth (enabling "zoom out" functionality) or the current "automatic" mode resulting in a higher waveform capture rate. Yes, you can get almost the same functionality by never going below 10ms/div and using the "zoom in" functionality, but that is a bit of a work-around and clutters the screen.
- The AWG seems to be pretty good hardware wise, however, the software implementation is just a bit too rudimentary. I miss functionality like: modulation, double pulses and manually trigged pulses. I am not convinced it is worth the €200+ price. It should be easy to add in a firmware update, though.

I have not really used the scope yet, but all in all, it seems like a very nice scope, which is comparable to the regular "A-brands" but at a great price/performance ratio. 

[Martin72]

Potential for improvement:

Hehe...

The probes

Well...No comment. 

Power button is a soft button and the scope uses almost 4W when turned "off"! Though, there is no apparent benefit (e.g. faster boot) from this substantial idle power use. For comparison EU regulations require consumer equipment to use less than 0.5W in off mode, so I guess oscilloscopes are exempted from this... A hardware button on the back would also have helped instead of having to pull the plug.

For me there are only two states: Scope is plugged in and running or off and plugged off, so I don´t worry a thing about it.

Boot time of 45 seconds is just acceptable.

The more features a scope got, the longer it will boot - 45 seconds is in nower days a good value.

Memory depth:

Also a thing discussed a hundred times - They will work on it, but today I understand their way to do it so far, so let´s hope they won´t make it worser when allowing the full memdepth at every time.

The AWG seems to be pretty good hardware wise, however, the software implementation is just a bit too rudimentary.

Agree to this.

I have not really used the scope yet, but all in all, it seems like a very nice scope, which is comparable to the regular "A-brands" but at a great price/performance ratio.

This scope is a killer for it´s price, even when the entry model would cost 2000 bucks.
Although the siglent sales mangement confirms, that the rigol 5000 is a serious opponent to it, but I couldn´t agree to this, because I got a rigol5000 before.

[Elasia]

Mine has been on 24/7 since i got it... still ticking

[Vestom]

The more features a scope got, the longer it will boot - 45 seconds is in nower days a good value.

Yeah, somehow we will accept that a €100000 scope will take 4 minutes to boot and sound like a jet engine However a daily driver should just be fast and quiet. The Siglent is OK, but could have been slightly better.

This scope is a killer for it´s price, even when the entry model would cost 2000 bucks.
Although the siglent sales mangement confirms, that the rigol 5000 is a serious opponent to it, but I couldn´t agree to this, because I got a rigol5000 before.

I agree the scope is very good value compared to the low end models from Tek/KS etc. Obviously they are hesitant to put too many features into their low-end scopes and Siglent can easily compete with them by listening to the users and providing lots of features with a low-mid range front end.

IMHO the Rigol 5000 looks like one rung down the ladder and not really something I would consider - especially if the firmware updates are a bit "hit and run".

Personally, I am a very satisfied customer with the SDS2104X+ and consider my list of oddities to be minor. My main gripe is probably the simplistic AWG. However, if Siglent continues to provide improvements with firmware updates, they will do something the big brands don't and we will really have a winner 

[thinkfat]

The MSO5000 series puts some serious punch in hardware department. 8 GSa/s, holy ...!

Just, from what I read the MSO frontend is a bit noisy and the firmware looks a bit cobbled together - this will be Siglents path to compete, but they need to get their stuff together. The MSO5000 has a _ton_ of triggers, this is where I feel Siglent is lacking. More triggers, more analysis functions to make the scope more than a mere waveform display and a serious test instrument.

The AWG I don't mind, I just wish it had shorter rise times for pulse and rectangle functions. I have an external AWG anyway which I prefer.

[Martin72]

Just, from what I read the MSO frontend is a bit noisy

It is and as they told me, the frontend of the 7000 aren´t that much better, I´ve turned back to siglent.
Now I got "only" 2GSa/s, but could live with it...

[Tighermine83]

Hi Guys,

I've juste bought this model Siglent SDS2104X Plus 4Ch 100MHz. Is it upgraded to more bandwidth? if yes can any one help me to do it.

Thanks,

[Martin72]

Hi,

Just click my signature...

[pascal_sweden]

Where can I find the details on the noise levels for Siglent SDS2000X versus Rigol MSO5000?

I guess 500 microVolt vertical division is only available to users of the venerable Tektronix 2225.

[tautech]

Where can I find the details on the noise levels for Siglent SDS2000X versus Rigol MSO5000?

I guess 500 microVolt vertical division is only available to users of the venerable Tektronix 2225.

Old model, this thread is about SDS2000X Plus !

Anyways, SDS2000X Plus noise measurements are on P3 of this thread in the thumbnails of this post:
https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2840268/#msg2840268

[pascal_sweden]

I meant the Siglent SDS2000X Plus. Sorry for the mistake.

Is the Rigol much noisier?

How do their vertical sensitivity compare?

[mawyatt]

Just received the Siglent SDS2102X Plus from Saelig thanks to Rob at Tautech and Amanda at Saelig

First impression is "Wow"

This is really a nice instrument, the larger touch screen is excellent. UI is nice and I've been able to navigate without using the manual, things are intuitive for the most part so far.

Need much more testing and evaluation, but first impression is very good

Best,

[Tom45]

Where can I find the details on the noise levels for Siglent SDS2000X versus Rigol MSO5000?

I guess 500 microVolt vertical division is only available to users of the venerable Tektronix 2225.

Or you can go with the even more venerable Tek 7A22 in a 7000 series mainframe and get down to 10 μV/Div. Just one of the reasons I have a Tek 7904 on my bench along side my SDS2204X Plus.

[Vestom]

Where can I find the details on the noise levels for Siglent SDS2000X versus Rigol MSO5000?

I guess 500 microVolt vertical division is only available to users of the venerable Tektronix 2225.

You can find the details about Siglent here: https://www.siglenteu.com/wp-content/uploads/2020/04/SDS2000X-Plus_Datasheet_DS0102XP_E01A.pdf

Noise is specified as <80 uV RMS @ 500MHz BW and you can get down to 500 uV/div when running with X1 probe or direct connection. I find it to be quite OK.
You have to google the Rigol yourself 

[pascal_sweden]

Can you refer to the source where you read about the Rigol MSO front-end being noisy?

The MSO5000 series puts some serious punch in hardware department. 8 GSa/s, holy ...!

Just, from what I read the MSO frontend is a bit noisy and the firmware looks a bit cobbled together - this will be Siglents path to compete, but they need to get their stuff together. The MSO5000 has a _ton_ of triggers, this is where I feel Siglent is lacking. More triggers, more analysis functions to make the scope more than a mere waveform display and a serious test instrument.

The AWG I don't mind, I just wish it had shorter rise times for pulse and rectangle functions. I have an external AWG anyway which I prefer.

[Howardlong]

Where can I find the details on the noise levels for Siglent SDS2000X versus Rigol MSO5000?

I guess 500 microVolt vertical division is only available to users of the venerable Tektronix 2225.

I have this little spreadsheet I just made up this afternoon. The MSO5000 isn't great when you look at the Vrms.

I picked the common lowest and highest channel attenuator ranges of all the scopes I tested (2mV/div and 5V/div), as well as the most common attenuator settings I use with a 10x probe (in 1x terms, 100, 200 & 500mV/div).

At the 500uV setting I'm looking at now (not in the spreadsheet) I get

Code: [Select]

MSO5000   Vpp:  1.19mV
SDS2000X+ Vpp:  0.158mV

MSO5000   Vrms: 0.485mV
SDS2000X+ Vrms: 0.255mV


[Vestom]

Can you refer to the source where you read about the Rigol MSO front-end being noisy?

I don't know the spec or have the scope, but in this video around the 27:00 you can see some low level performance.
To quote Dave: "You don't want to get this scope for any low level measurement"...

[Howardlong]

Can you refer to the source where you read about the Rigol MSO front-end being noisy?

The MSO5000 series puts some serious punch in hardware department. 8 GSa/s, holy ...!

Just, from what I read the MSO frontend is a bit noisy and the firmware looks a bit cobbled together - this will be Siglents path to compete, but they need to get their stuff together. The MSO5000 has a _ton_ of triggers, this is where I feel Siglent is lacking. More triggers, more analysis functions to make the scope more than a mere waveform display and a serious test instrument.

The AWG I don't mind, I just wish it had shorter rise times for pulse and rectangle functions. I have an external AWG anyway which I prefer.

I received an MSO5000 this week, the first thing that jumped out at me was the noisy front end. Subsequent measurements have shown that to be the case, with not even 5 bits of effective resolution.

Edit: effective resolution in bits below

[Vestom]

From Daves video it looks like the MSO5000 has a resolution of about 150uV/bit.

I just tested my SDS2104X+ and found it to have a resolution of about 33uV/bit (8 bit mode). Strangely it seems we get 8 levels more by going to 10 bit mode so that is ~4uV/bit. Is 10 bit mode really 11 bit mode??

[Howardlong]

From Daves video it looks like the MSO5000 has a resolution of about 150uV/bit.

I just tested my SDS2104X+ and found it to have a resolution of about 33uV/bit (8 bit mode). Strangely it seems we get 8 levels more by going to 10 bit mode so that is ~4uV/bit. Is 10 bit mode really 11 bit mode??

We'd need to know quite a bit more. What attenuator setting, what bandwidth setting, how is the channel terminated, etc etc.

[Martin72]

I just tested my SDS2104X+ and found it to have a resolution of about 33uV/bit (8 bit mode).

How do you test it ?

[Vestom]

I just tested my SDS2104X+ and found it to have a resolution of about 33uV/bit (8 bit mode).

How do you test it ?

I prefer dot-mode... exactly to know when I am measuring at the limits of my scope (been fooled more than once by vector mode)

Edit: Apparently it is only at 500uV/div we get 8 extra levels in 10 bit mode. At 1mV/div the resolution is still 33uV/bit in 8 bit mode but we only get 4 extra levels in 10 bit mode. Funny...