Siglent SSA3000X and SSA3000X-Plus Spectrum Analyzers

[TheSteve]

Very nice, Thanks Dave. It will be quite interesting to see them in operation side by side.

[rf-loop]

Is this SSA3000X what you use for playing with SA equipped with new or old HW (afaik, main difference is TG)
If  last 6 numbers of serial number  is over  160100 (In some old information 160110) then TG is new improved version.
Is it normal selling version or distributors early phase demo version?

Serial 2160239, May 2016 build.
It is the new one with the revised hardware. From what I have heard it's the front end they had to revise.

Yep. I only ask so that readers know what are under test. (I have same rev HW)

Comment for image where can see noise floors and sweep speeds.

Sweep peed is quite complex compromise between things and one result is accuracy vs speed.

Here attached image.   Outside from image I can also note that just with this default trace and settings (144ms sweep time) it repeating speed is 3 sweep in second. (but this repeting time is really very much changing thing depending from many settings combination)

[EEVblog]

How'd you get the 4 difference traces with the different settings?
Buggered if I can do it...

[rf-loop]

I have seen many confusions when peoples wonder trace settings. There is many settings including peak hold max, min, averages etc. These can select individually for all 4 traces. And then there is other "mystery": Detector. Also it can of course set individually for all 4 traces.

Trace average and video average is very different. In simply: Video average works inside one sweep but trace average averages sequential sweeps. But, when run long time Trace average it also result nearly same as video average but... still video average is very very different is look its result and Detector mode Pos Peak and with it done trace averaging. Video average may give "nice looking" image but there is serious trap where user can drop if do not be careful. Video average is nice for noise but it may loose signal if signal is quite narrow on the screen. Mostly, use Detector mode Positive Peak! Until you really NEED something else. In some cases even Neg Peak may be  handy in right place and in right hands what know what are doing.

Note that I have selected 10Hz RBW but 30Hz WBW just for give more clear visibility to these things and get more wide p-p so that Neg Peaks hit this spectrum absolute bottom border around 940pV.

Perhaps these images may clarify even something.  (with these  images can also get imagine about  DANL)

[rf-loop]

How'd you get the 4 difference traces with the different settings?
Buggered if I can do it...

It is customary first to study and to work afterwards.

Perhaps important question for users. (and at this time User Manual is ...really "poor".  I hope some day chinese manufacturer(s) do real "Users Reference Manual" what include more deep explanations ancluding also some detailed look how equipment works in some details - exactly. aka "Theory of operation")
Every trace have individual settings in trace menu and individual detector setting in detector menu where you select what trace use what detector. And when you want "freeze" one trace just in trace menu select "View". It need work sequentially (do one trace ready and freeze it, do next and freeze..etc. With this method can collect different things to same display).

[EEVblog]

Every trace have individual settings in trace menu and individual detector setting in detector menu where you select what trace use what detector. And when you want "freeze" one trace just in trace menu select "View". It need work sequentially (do one trace ready and freeze it, do next and freeze..etc. With this method can collect different things to same display).

Yes, I'm trying to do that (get a different RBW for each trace) but it seems buggy or something and doesn't want to leave the trace after I switch. It's weird.

[rf-loop]

Every trace have individual settings in trace menu and individual detector setting in detector menu where you select what trace use what detector. And when you want "freeze" one trace just in trace menu select "View". It need work sequentially (do one trace ready and freeze it, do next and freeze..etc. With this method can collect different things to same display).

Yes, I'm trying to do that (get a different RBW for each trace) but it seems buggy or something and doesn't want to leave the trace after I switch. It's weird.

As example here in Reply #148



Works just simply and rock solid without any hassle.

But if your fingers works faster than your brain and more importantly faster than equipment brain.....  this is not way how we use equipments in lab. Every selection and every enter need do systematically and deliberately after cogitation. It is not gameboy where fastest finger win.

[EEVblog]

After much struggle with the stupid Trace system, I got this.
1MHz, 100KHz, and 10KHz RBW

[rf-loop]

After much struggle with the stupid Trace system, I got this.
1MHz, 100KHz, and 10KHz RBW

Do you still think Siglent have done copy from Rigol menu system? Or how long it takes to change opinion?
Every equipment is different. There is also learning curve. When I was use half year one Tek then I get (similar) HP to my hands I feel ...oh my god this is stupid and arrghs... but. After I feel familiar with it my opinion was changed.

But so or so. I hope Siglent and who ever manufacturer use more power to develop useability. This can not do if there is some programmers who have perhaps nearly total lack of knowledge about any real work in lab with T&M equipments and then he design UI as alongside the main job.
It need deep knowledge about human and it need lot of experience with real work using T&M equipments to design front panel and menu system so that it best serve equipment use in real work. Needing a broad interdisciplinary knowledge (including many science Sectors). Only then can design a good user interface. And it is not easy. It is also lot of work. This is why example in (old) HP they really work hard and invest lot of for useability and it have been also expensive. But also not without mistakes.

[nctnico]

@rf-loop: can you tell something about the user interface when comparing an Advantest spectrum analyser and the Siglent SSA3000X series? If you have ever worked with an Advantest spectrum analyser.

[PartialDischarge]

Only:
IF is stays hackable
IF the advanced features are actually usefull/working and not checkbox features / pray-ware
IF it is free of bugs

I count on SW bugs, but im more interested in the HW specs. The former can be fixed with time, but the latter can't.

True but how much time does Siglent need to get the firmware to an acceptable level if they aren't there yet? Their software team working on their oscilloscopes is horribly slow; you can count progress in years. OTOH a spectrum analyser is far less complicated from a software point of view so it shouldn't take too long to come up decent firmware but seeing is believing. As a general rule of thumb you have to make sure you can live with the device as it is out of the box with the most recent firmware and not wait for the firmware to mature because that may never happen within a reasonable timeframe.

Following up on this topic, I did report a bug 20 days ago on the sdg2042x ( actually tautech reported it to them  ), today a new firmware is available addressing that isssue and adding the capability of combining both output channels... So from my point of view they were pretty responsive.

[bson]

Following up on this topic, I did report a bug 20 days ago on the sdg2042x ( actually tautech reported it to them  ), today a new firmware is available addressing that isssue and adding the capability of combining both output channels... So from my point of view they were pretty responsive.

Agreed! They even fixed the external reference issue I had.   

[rf-loop]

How much does the noise floor change with the preamps enabled?

Here is two examples. Using RBW 1MHz and 10kHz.
Results are quite comparable with other RBW settings. (With full span most narrow RBW is limited to 1kHz)




Added Note: In this image RBW 10kHz and 1MHz use different "Sweep Mode". 10k sweep mode is FFT and 1M sweep mode is "Sweep".
Also sweep time is really very different and it is possible that it affect also bit specially for visible noise levels.

[nctnico]

Following up on this topic, I did report a bug 20 days ago on the sdg2042x ( actually tautech reported it to them  ), today a new firmware is available addressing that isssue and adding the capability of combining both output channels... So from my point of view they were pretty responsive.

Agreed! They even fixed the external reference issue I had.   

Signal generators where the first real instruments they made (SDG1000 series) and the early versions where useable but far from complete. I'm also under the impression that Siglent has different software teams on different instruments so the update frequency on one product line does not necessarily says something about the update frequency on the other product lines. Because they have been doing signal generators for such a long time it is to be expected they can fix bugs quickly. However the spectrum analyser is (AFAIK) the latest addition so the software team is likely new as well.

[rf-loop]

However the spectrum analyser is (AFAIK) the latest addition so the software team is likely new as well.

Siglent SSA1010 and SSA3030 are previous models so they are not just new in SA's.

[1design]

It is very interesting to see a difference greater than 20dB between the 1MHz and 10kHz RBW. This might be due to a difference in filer shape?

BR

[rf-loop]

It is very interesting to see a difference greater than 20dB between the 1MHz and 10kHz RBW. This might be due to a difference in filer shape?

BR

1M RBW sweep mode is  "Sweep"  and with 10kHz RBW  sweep mode  is "FFT".  (In Auto mode it select automatically)
FFT sweep mode is not visible in image because first I take 10k traces and after then 1M traces.

Also sweep mode "Sweep" have two selections: "Speed"(default) and "Accuracy". In image it was default and also sweep mode selection was in Auto so it change automatically from "Sweep" to "FFT"

[1design]

Makes sense, thank you for your quick reply.

BR

[EEVblog]

Phase noise for a 1GHz 0dBm carrier from my Rubidium + Marconi 2030 for a 1MHz offset on both the Rigol and Siglent.
Significantly different PLL responses.

[EEVblog]

Works just simply and rock solid without any hassle.

No, it's a hassle if you don't know the exact sequence to get it working to do that. It is non-obvious IMO.

[tautech]

Works just simply and rock solid without any hassle.

No, it's a hassle if you don't know the exact sequence to get it working to do that. It is non-obvious IMO.

Swatting up a bit on these myself and the Quick Start guide gives little instructive help while the User manual is much better. There's a inbuilt Help system that after selection any further keys pressed will give assistance on that topic to the user and to exit press Help again.
While you indicate that usage (for you) is not intuitive plastered all through the manuals and marketing blurb is: friendly user interface, Intuitive menu, easy-to-use and so on.
I'm not taking the piss Dave, just trying to explore possible deficiencies in the UI.

It would be useful for those watching at home to know if the Help is any help at all. 

http://www.siglentamerica.com/USA_website_2014/Documents/QuickStart/SSA3000X_Quick%20Guide.pdf
http://www.siglentamerica.com/USA_website_2014/Documents/UserManual/SSA3000X_User%20Manual_UM0703X_E02A.pdf

Latest datasheet: (different to that posted in reply#2)
http://www.siglentamerica.com/USA_website_2014/Documents/DataSheet/SSA3000X_DataSheet_DS0703X-E03A.pdf

[EEVblog]

It would be useful for those watching at home to know if the Help is any help at all. 

The Help menu for the Trace function is not a huge help, but has a short description if you select help for each option. It's not better than the manual, practically the same short text.
The Rigol works identically because the Siglent UI is a complete ripoff.

[Ivan7enych]

Phase noise for a 1GHz 0dBm carrier from my Rubidium + Marconi 2030 for a 1MHz offset on both the Rigol and Siglent.
Significantly different PLL responses.

Very strange picture on Siglent.
Why does Siglent show no warning that input power 0dBm is higher than Ref -40dBm, Rigol shows UNCAL warning.
Why trace peak on Siglent is much much lower (-65dBm) than actual input power (0dBm) ?

[EEVblog]

Why does Siglent show no warning that input power 0dBm is higher than Ref -40dBm, Rigol shows UNCAL warning.

That's not what the UNCAL warning is for on the Rigol, it has to do with sweep time.
The Siglent will eventually show an "ADC overload" warning message and beep continually if the input is overloaded.

[hendorog]

Why does Siglent show no warning that input power 0dBm is higher than Ref -40dBm, Rigol shows UNCAL warning.

That's not what the UNCAL warning is for on the Rigol, it has to do with sweep time.
The Siglent will eventually show an "ADC overload" warning message and beep continually if the input is overloaded.

Can you explain what those images are supposed to show please, cos it's making no sense to me to set the ref to -40dBm and feed in an 0dBm signal.