At last, Siglent's SDS5054X touchscreen DSO...

[0xdeadbeef]

We can only refer to the ** in P8 of the datasheet where it's specified for all models.

But the 350MHz model has just one asterisk which says the 200MHz limit kicks in <= 1mV/div. So ironically, the 350MHz model seems to have a higher bandwidth in the 2mV/div range than the two higher BW models. Actually this is another indication that the 350MHz model has a different frontend.
Side note: the data sheet differentiates between "setting range" starting at 500µV/div and "specified range" starting at 1mV/div. This sounds as if the 500µV/div range might not fulfill other specifications as well.

So yes, if auto BW limiting is used automatically at higher sensitives there must indeed be BW limit indicators for the operator.

Best solution would be of course to remove the problem that makes this limitation necessary. I.e. one way to look at this is that the scope's electronics create noise in the frequency range > 200MHz which is of a magnitude that it would be clearly seen in these low voltage ranges. What really puzzles me is that the WS3000(Z) doesn't seem to have this limitation (e.g. 1GHz bandwidth specified down to 1mV/div). Actually even the SDS3000X datasheet doesn't seem to mention anything like this but then again, I could only find Chinese datasheets which leave some room for interpretation without appropriate language skills.

[tautech]

We can only refer to the ** in P8 of the datasheet where it's specified for all models.

But the 350MHz model has just one asterisk which says the 200MHz limit kicks in <= 1mV/div. So ironically, the 350MHz model seems to have a higher bandwidth in the 2mV/div range than the two higher BW models. Actually this is another indication that the 350MHz model has a different frontend.

Possibly IDN.
However what I do know is the datasheet is revision D however I'm unaware of the revision timelines and how they might apply to latest production models.

Side note: the data sheet differentiates between "setting range" starting at 500µV/div and "specified range" starting at 1mV/div. This sounds as if the 500µV/div range might not fulfill other specifications as well.

Specified range is the Fine V/div adjustment and something I've not looked at in the 500uV/div range.
Seems like an unusual use case to need fine adjustment in the 500uV/div range.

So yes, if auto BW limiting is used automatically at higher sensitives there must indeed be BW limit indicators for the operator.

Best solution would be of course to remove the problem that makes this limitation necessary. I.e. one way to look at this is that the scope's electronics create noise in the frequency range > 200MHz which is of a magnitude that it would be clearly seen in these low voltage ranges. What really puzzles me is that the WS3000(Z) doesn't seem to have this limitation (e.g. 1GHz bandwidth specified down to 1mV/div). Actually even the SDS3000X datasheet doesn't seem to mention anything like this but then again, I could only find Chinese datasheets which leave some room for interpretation without appropriate language skills.

And yet as I have mentioned, the datasheet specifies automatic BW limiting for low V/div ranges yet there is no visual scope notification that this is happening unlike any older DSO I have ever seen.
I see this datasheet limitation as correctly informing the prospective buyer and yet in operation of the DSO there is NO clue BW limiting is occurring.
If we take this nugget of display information omission and overlay it onto WS/SDS3000/X/Z information and operator warnings or not, there is no wonder there is confusion as to which model does what wrt auto BW limiting.

As a user all I care about is the DSO must inform me if it has to go into another mode when I change any settings.....I need to be in charge.....not the DSO !
Unless the proper information is given to the operator via the display how can he/she make correct usage decisions ?

Siglent, are you listening ?

[2N3055]

There are many scopes that have variable bandwidth specifications on various input sensitivity.
For instance Tektronix MDO3000 and MDO4000 series have several different specs in several ranges, and different from 50 Ohm and 1 MOhm..

I doubt it is bandwidth limiting (deliberate). It is more likely that front end amplifiers have finite gain bandwidth product, and when they start pushing certain gain bandwidth runs out....

I personally don't mind this, 500uV/div is picking up so much crap from environment that you usually enable some sort of bandwidth limiting anyways.

That being said, it would be a nice thing if scope would mention it somewhere on the screen, just because it is nice as a reminder.
Not that you are not required to know scope specs in detail to be able to use it properly.

[nctnico]

I personally don't mind this, 500uV/div is picking up so much crap from environment that you usually enable some sort of bandwidth limiting anyways.

It can be a problem. Many HF probes have an attenuation which can be 20, 50 or even 100x so having a limit on the bandwidth on low V/div settings can limit the useful signal range when using HF probes.

[0xdeadbeef]

I personally don't mind this, 500uV/div is picking up so much crap from environment that you usually enable some sort of bandwidth limiting anyways.

It's not only the 500µV range, but everything <= 2.45mV/div for the better models or <= 1mV/div for the 350MHz model. But you're right, the Tek MDO3000 scopes have similar, partially even worse limitations (couldn't find something like this for the MDO4000C but maybe I overlooked it). Actually the Tek MDO3000 also has limitations which differ from the two selectable bandwidth limits, so I guess it also can't/won't display an accurate BW limitation in the lower ranges!?

Anyway, even with an active differential 20:1 probe, I rarely go below 20mV/div, so the bandwidth limitation in the lower ranges isn't really a showstopper for me. So, yes, it might not be important for lots of people, but it's something to keep in mind when comparing scopes.
Side note: both selectable bandwidth limits (20MHz and 200MHz) on the SDS5000X are specified with +/-40% accuracy. This seems to be quite a lot but I couldn't find such a specification for any other scope.

[2N3055]

Excerpt from specs MDO3000/4000...

Let me say this: Keysight 3000T is full bandwidth all the time..
Except it doesn't have 500µV/div. Or 1mV, or 2mV/div range. It's lowest real range is 4mV/div. I wish it had 500µV/div even if it meant lower bandwidth..

[maginnovision]

The 3000 is quite an old scope by now though. Over 10 years old.

[Mr Nutts]

The 3000 is quite an old scope by now though. Over 10 years old.

Neither MDO3000 nor Keysight 3000T are that old. The MDO is I think from 2013 and the 3000T from 2015

[Performa01]

I doubt it is bandwidth limiting (deliberate). It is more likely that front end amplifiers have finite gain bandwidth product, and when they start pushing certain gain bandwidth runs out....

I think you are not too far off.
It actually is not a noise issue.
Analysis of the frequency response reveals that the PGA (Programmable Gain Amplifier) doesn’t behave very well with high gain factors at frequencies above some 300MHz. See the following FR plots.

First the frequency response up to 2GHz at 5mV/div (grid is 200MHz/div and 5dB/div):


SDS5104X_FFT_FR_2GHz_5mV_01

The frequency response at a lower gain like 5mV/div is rather nice already, in fact within +0.7/-1dB up to 880MHz and +0.7/-3dB up to 1.17GHz.

Now let’s have a look at the frequency response up to 2GHz at 500µV/div (grid is 200MHz/div and 5dB/div):


SDS5104X_FR_500µV_full

Here we can already see that it doesn’t exactly look like a first order lowpass filter. Instead it shows some nasty peaking that might sum up to more than +2dB, even above 1GHz. With this, we might also exceed the large signal bandwidth of that PGA. Reason enough to have a bandwidth limit at the highest sensitivities. Yes, it could have been set a bit higher – as we can see for the 350MHz model, where we don’t need an explicit BW limit above 1mV/div.

Quite obviously, Siglent did not want to bother with individual filters and just enforced the existing 200MHz bandwidth limit instead. I fully agree that this should be indicated in the UI though and I have already requested this a while ago. So hopefully we’ll get that with the next FW-Update.

Side note: both selectable bandwidth limits (20MHz and 200MHz) on the SDS5000X are specified with +/-40% accuracy. This seems to be quite a lot but I couldn't find such a specification for any other scope.

The bandwidth limiter is an integral feature of the PGA. As you all know, on-chip resistors may have good thermal tracking, but also very high tolerances. This is the reason why the bandwidth limit cannot be any more accurate than that.

[maginnovision]

The 3000 is quite an old scope by now though. Over 10 years old.

Neither MDO3000 nor Keysight 3000T are that old. The MDO is I think from 2013 and the 3000T from 2015

The 3000t underlying hardware is the same as the A.

[tautech]

The latest production firmware is now available on the Siglent factory website.

Version 0.8.2R1
40.1 MB
http://old.siglentamerica.com/USA_website_2014/Firmware&Software/firmware/SDS5000X_0.8.2R1_EN.zip

Release notes
1. Supported search across history frames
2. Optimized zone trigger and mask test accuracy in zoom mode;
3. Supported editing a trigger zone after creating it
4. Supported MIL-STD-1553B trigger
5. Improved the input frequency upper limit of holdoff by event from 20 MHz to 120 MHz
6. Solved the defect that the scope possibly does not trigger on the first edge of a burst train with carrier frequency
    above 120 MHz
7. Added Reboot and Shutdown function under “Utility” top bar menu, so the instrument can be remotely rebooted
    and shutdown by web
8. Supported saving the decode list as a CSV file
9. Optimized response time of mask test when disabling/enabling it or changing the type
10. Optimized webserver response when dragging a trace in vertical direction; Added alternative VNC port for
      webserver
11. Added automatic clear of measurement statistics when changing horizontal/vertical/trigger settings
12. Supported editing selected measurement parameter
13. Fixed several bugs
     a) Cursors: Unexpected jump when changing horizontal settings
     b) Webserver: Incorrect mouse position with IE in full screen mode

[2N3055]

Release notes
1. Supported search across history frames

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

[tautech]

Release notes
1. Supported search across history frames

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

Dunno, I don't currently have a unit to play with. 

[maginnovision]

Release notes
1. Supported search across history frames

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

Dunno, I don't currently have a unit to play with. 

Come on pusher man, just giving us a taste of patch notes. We need more!

[tautech]

Release notes
1. Supported search across history frames

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

Dunno, I don't currently have a unit to play with. 

Come on pusher man, just giving us a taste of patch notes. We need more!

I won't claim anything I can't prove for myself so you'll have to wait for JPortici or Performa01 to answer 2N3055's query. 

[JPortici]

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

AFAIR, not yet

[2N3055]

I won't claim anything I can't prove for myself so you'll have to wait for JPortici or Performa01 to answer 2N3055's query. 

It's okay, I just remembered that you don't have one. I'm in no hurry, it is just fantastic that Siglent can add it no problem and R&S can't.. 

If it can search historic buffers for both analog events and protocols, that makes it king of the hill in that regard.
Keysight 3000T has many protocols and measurement and search, but small memory, no histograms and color persistence modes. Most of the stuff is options.
R&S 3000 has memory, nice spectrum mode and decodes, but no search on segments, no search on common serial protocols and in general still seems kinda "not finished yet". And all of it is very expensive with even segmented memory being an option.
Lecroy Wavesurfer 3000 is just very basic. Nothing to be really excited about. Expensive for what it is. Everything is an option.
Rigol 7000 is still doing the "soul searching". Rigol will get there, but Siglent did a smart thing: they delayed product sales until they ironed out at least most important hardware platform things. Applications and measurements can be added later. Releasing scope that has basics working fine increases credibility with customer. Rigol was in a hurry to get out first, and they still have problems with autocal and such.

Big plus for Siglent here.
 
So yeah, as I said before 350Mhz 4ch SDS5000X is shaping up to be a class leader. It is cheapest of the scopes with similar capabilities, or has best capabilities for the scopes in that price range, however you choose to look at it. AFAIK this is first firmware update for it, already adding serious capabilities.. If they keep up the good work, with time it will become world class instrument.

[2N3055]

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

AFAIR, not yet

Thanks for that, much appreciated!!

Well, it is a shame, but let's hope they are working on it.
Even on Keysight 3000T with max 1000 segment, i use search all the time, both for analog parameters and sifting through the packets. Even 1000 segments in 2 Mpoint memory is too much to go through it manually. Now imagine that on a scope with 100s of Mpoint and 50000 segments.

Regards,

[Performa01]

Release notes
1. Supported search across history frames

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

My statement from reply #216 still holds true

https://www.eevblog.com/forum/testgear/at-last-siglent_s-sds5054x-touchscreen/msg2422620/#msg2422620

One that SDS50000X should be compared to is RTM3000K, but on that one promo is only for a fully loaded 1GHz one for 8600€.
Which is a good deal actually if you need 1GHz. Despite fact that on RTM3000K they don't even have search on basic protocols and segments so large memory they are promoting so much is pretty much useless...

I should point out that the SDS5000X cannot do a search for serial triggers right now. But I have requested this quite some time ago and it's on Siglent's to-do list.

Search across segments on the other hand – yet another request from me – has already been implemented in the latest firmware. Search is also reasonably fast, even with full 250MPts memory (<400ms per search) as I think I was able to demonstrate in an earlier post.

EDIT: I should add that the V0.8.2R1 firmware is already about two months old. But it's only now that it has been made available for public download.

[2N3055]

Release notes
1. Supported search across history frames

That is a great news. Now I really want to test it to see how it's implemented. Does it search on protocols too?

My statement from reply #216 still holds true

https://www.eevblog.com/forum/testgear/at-last-siglent_s-sds5054x-touchscreen/msg2422620/#msg2422620

One that SDS50000X should be compared to is RTM3000K, but on that one promo is only for a fully loaded 1GHz one for 8600€.
Which is a good deal actually if you need 1GHz. Despite fact that on RTM3000K they don't even have search on basic protocols and segments so large memory they are promoting so much is pretty much useless...

I should point out that the SDS5000X cannot do a search for serial triggers right now. But I have requested this quite some time ago and it's on Siglent's to-do list.

Search across segments on the other hand – yet another request from me – has already been implemented in the latest firmware. Search is also reasonably fast, even with full 250MPts memory (<400ms per search) as I think I was able to demonstrate in an earlier post.

EDIT: I should add that the V0.8.2R1 firmware is already about two months old. But it's only now that it has been made available for public download.

Thank you for clearing that out. I knew about your previous post, I just got confused a bit, thinking this was a step forward, since some time passed.
But as you said they are working on it, and when they do I would like to try it out to see how it compares .. At this moment, Siglent has earned certain level of respect in being realistic and keeping their promises (not in small part thanks to you !), and I'm looking forward to see what comes next. I'm also interested in FRA to higher frequencies, using SDG6000X, etc etc..

Interesting times.

Have a good one,

Siniša

[jemangedeslolos]

Hello  !
I have to make a choice. I need two scope, one at home and one in my own lab. I bought a Rigol MSO5074 few weeks ago for home and I have a Siglent SDS2204X that I want to upgrade in my lab.

The SDS5034X seems to be a very good scope but the price difference is huge compared to an hacked MSO5000.
To be honest, A hacked MSO5000 will be fine for 99 %  of my work but I already have one.
Do you have practical example of what the SDS5034X can do and the MSO5000 can't  ?

It is not so easy to compare them on paper....but it seems that their performance are similar  ?

Thank you for helping me choosing the right one  : )

[maginnovision]

If you're paying for it, hacking it, and it works for pretty much everything I don't think the Rigol 5000 is going to be your 2nd choice considering the price difference. If you just want something new and different get the Siglent.

[tautech]

Hello  !
I have to make a choice. I need two scope, one at home and one in my own lab. I bought a Rigol MSO5074 few weeks ago for home and I have a Siglent SDS2204X that I want to upgrade in my lab.

The SDS5034X seems to be a very good scope but the price difference is huge compared to an hacked MSO5000.
To be honest, A hacked MSO5000 will be fine for 99 %  of my work but I already have one.
Do you have practical example of what the SDS5034X can do and the MSO5000 can't  ?

It is not so easy to compare them on paper....but it seems that their performance are similar  ?

Thank you for helping me choosing the right one  : )

Use the SPL2016 from your SDS2204X.
Webserver ? ..... haven’t had a play with the 5kX one yet but if it’s like the webserver in the X-E models it’s a useful remote access tool. Performa01 can advise.
Autosense probes.
50 ohm inputs ?

[Martin72]

Autosense probes.

Still I can´t get the advantage of autosensig a probe - it tooks only seconds to adjust the coupling manually…

Where this:

50 ohm inputs ?

Is a really advantage for those who need it ( working with higher frequencies), a external termination couldn´t reach the performance of an internal one.
But is this worth a difference about say 1500€....
It´s a personal decision.
Although I´m looking for the SDS5000 for our needs, instead rigols 5000.
When it´s mostly debugged, the MSO 5000 (which I own in private)could be the first choice for hobbyists, it got outstanding performance for the price.
The SDS 5000 series are in a another class, in my opinion.

[Performa01]

Hello  !
I have to make a choice. I need two scope, one at home and one in my own lab. I bought a Rigol MSO5074 few weeks ago for home and I have a Siglent SDS2204X that I want to upgrade in my lab.

The SDS5034X seems to be a very good scope but the price difference is huge compared to an hacked MSO5000.
To be honest, A hacked MSO5000 will be fine for 99 %  of my work but I already have one.
Do you have practical example of what the SDS5034X can do and the MSO5000 can't  ?

It is not so easy to compare them on paper....but it seems that their performance are similar  ?

Thank you for helping me choosing the right one  : )

Regarding your particular question:

I cannot know what the MSO5000 can or cannot do, or how well certain features are implemented. It would be up to you to present a use case where the MSO5000 is not convincing (or fails altogether), then I could demonstrate how the SDS5000X handles it. Since the SDS5000X is not nearly feature complete yet, it might fail as well at this point.

Ultimately, I cannot comment on the MSO5000, but I can safely state that the SDS5034X would be a major upgrade to the SDS2204X.


Here is the more general answer:

The feature comparison of newly introduced instruments doesn’t tell the whole story. New features can be added later on, which might change the picture completely.

If you understand the instrument platform and its resources, then you should be able to know in advance whether you can expect a certain feature to be added at some point. And then it’s a matter of trust whether a manufacturer is willing to maintain their product and add new features after the introduction.

Since I happen to know the final feature set of the SDS5000X (because I have helped to define it), I certainly know what to expect – and that’s quite a lot. The current software platform is shared amongst all newer Siglent scopes (SDS1000X-E and SDS2000X-E with traditional UI, SDS5000X with touch UI) and it will also be used for future releases, including higher class instruments above the SDS5000X – to the effect that it expands the ongoing fruitful cooperation with LeCroy…

Knowing a little what’s going on behind the scenes, I have full confidence in Siglent’s DSO/MSO platform development. I am glad to confirm that they actually listen to their beta testers/consultants and they also maintain an excellent relationship with their Western partners (such as LeCroy), rather than pissing them off

Apart from the features, there are also topics not found in a spec sheet, like fan noise, quality of the (touch)screen, UI experience and responsiveness. In this regard, the SDS5000X absolutely doesn’t need to be afraid of the competition in my opinion. Yes, it cannot touch a Keysight DSOX3000T in terms of snappiness and speed, but offers plenty of memory and deep measurements in return, while still being plenty fast and responsive.

Because it has been mentioned before: the integrated webserver allows full instrument control as well as remote screenshots, data downloads and even firmware updates. This is a real convenience:


SDS5104X_Web 01

Still I can´t get the advantage of autosensig a probe - it tooks only seconds to adjust the coupling manually…
 

Yes, changing the probe factor is quick and convenient, especially with the touch UI, and in many test scenarios the same (x10) probes are used almost all the time.

On the other hand, at least a toggle between any probe and direct coax connection can be required quite frequently. If it’s just on a single channel, then it’s not a big deal, but it can become an annoyance when several channel have to be changed at once.

Apart from that, automatic probe sensing reduces the probability of user error. I’ve seen several instances (also on this forum), where incorrect probe factor settings spoiled automatic measurements and this was not immediately obvious for the operator.

To cut a long story short, probe autosensing is just a standard feature in all but entry-level scopes nowadays. In midrange scopes and beyond we expect active probe interfaces on top of that.