Siglent SSA3000X and SSA3000X-Plus Spectrum Analyzers

[videobruce]

Upper image is from PC, running free PC software named EasySpectrum.
This EasySpectrum PC software come free with SA  or  it can also free download from Siglent.

Before install it, first need install  National Instruments VISA (Ni-Visa 15.0.1)
Also it is free. It can get from National Instruments.

For anyone not familiar with this 'add on' from NI, it is H-U-G-E!!! And I mean HUGE! I tried it once and removed it as quick as possible and that was no simple task since the installer did NOT completely remove everything, not even close.

It installs a ton of processes that runs continuously even if you aren't running the software which is unacceptable to me.
You might as well call it a M$ product, just as bloated, if not worse. I have no idea why on earth anything has to be this enormous for what should be a somewhat simple interface.

[tautech]

Upper image is from PC, running free PC software named EasySpectrum.
This EasySpectrum PC software come free with SA  or  it can also free download from Siglent.

Before install it, first need install  National Instruments VISA (Ni-Visa 15.0.1)
Also it is free. It can get from National Instruments.

For anyone not familiar with this 'add on' from NI, it is H-U-G-E!!! And I mean HUGE! I tried it once and removed it as quick as possible and that was no simple task since the installer did NOT completely remove everything, not even close.

It installs a ton of processes that runs continuously even if you aren't running the software which is unacceptable to me.
You might as well call it a M$ product, just as bloated, if not worse. I have no idea why on earth anything has to be this enormous for what should be a somewhat simple interface.

Yes it is large, it's a connection utility for many brands and types of instruments, however I've never had a problem with it or noticed it being hungry on resources. 

[videobruce]

Look at all the processes running when you aren't using it and the size of the program itself. IIRC, doesn't that need Net Framework also?

[Siglent America]

Hello,

We have posted a new application note on RBW/VBW and phase noise and how they effect the frequency resolution of the measurement in spectrum analyzers. I thought some people might be interested in reading it.

http://www.siglentamerica.com/USA_website_2014/FAQ/SSA/SpecAn_Bandwidth.pdf

[videobruce]

Has anyone come up with a 'wish list' or a 'fix this list' for these?

To start;

1. A Mechanical on-off switch to completely remove power w/o using a power strip or pulling the plug,
2. A lower speed fan to cut down on the noise, and/or one that has a thermistor to operate when it needs to, not continuously,
3. Option to remove (hide) the on screen display in the upper left of the grid which is: A. too large, B. usually not necessary,
4. Resign the standby mode to meet the Energy Star limit of under one watt, not FIVE watts which may of been ok 15-20 years ago, but shouldn't be now,
5. Expand the 'grid' by narrowing the side bars and merging the top to the bottom. Too much dead (wasted) space around the perimeter. To much of the larger screen (over the competition)  is wasted

The first three also apply to Rigols SA's. Number four does not. 

[rf-loop]

Has anyone come up with a 'wish list' or a 'fix this list' for these?

To start;

1. A Mechanical on-off switch to completely remove power w/o using a power strip or pulling the plug,
2. A lower speed fan to cut down on the noise, and/or one that has a thermistor to operate when it needs to, not continuously,
3. Option to remove (hide) the on screen display in the upper left of the grid which is: A. too large, B. usually not necessary,
4. Resign the standby mode to meet the Energy Star limit of under one watt, not FIVE watts which may of been ok 15-20 years ago, but shouldn't be now,
5. Expand the 'grid' by narrowing the side bars and merging the top to the bottom. Too much dead (wasted) space around the perimeter. To much of the larger screen (over the competition)  is wasted

The first three also apply to Rigols SA's. Number four does not. 

1. Important. But not only SSA. All X series equipment same. I do not think so much energy save but also safety.
3. Display menu, Screen text  --> OFF.   It is there.  Do you mean some other thing?
5. Number of display points is 751.  It need many things if rise this. It is not alone display question. Perhaps many internal things are based to this amount of data points.
If only "stretch" this 751 for example 800 or some other amount, how to easy map data for display without any adverse effects.

On other thing is if left side can also use better for some other useful purposes. Example more informations.  One possible is develop markers information better when use marker table, but also when screen is not split to table and SA trace. Also visual ergonomy may develop bit more.

[videobruce]

1. Important. But not only SSA. All X series equipment same. I do not think so much energy save but also safety. All T&M equipment should have this!
3. Display menu, Screen text  --> OFF.   It is there.  Do you mean some other thing? Not the entire display text, just what is inside the grid
5. Number of display points is 751.  It need many things if rise this. It is not alone display question. Perhaps many internal things are based to this amount of data points. No, the area of the grid.

See the attachment for clarification.

What is with all these trailing zeros, especially with RBW, VBW, Sweep & Span? 

[videobruce]

For comparison, here is Rigols screen with the same problems. At least Siglent's font size in the right pane is smaller.
I added the comment about the duplicate text that I didn't have room for in the previous posts attachment. The problem with the oversized, duplicates text within the grid is, many times it's in the way of the waveform I'm trying to see. Though it can be moved, that's annoying and has to be moved back later.

[PartialDischarge]

What is with all these trailing zeros, especially with RBW, VBW, Sweep & Span? 

those trailing zeros are needed because they may not always be zero, see my post with a photo of a rohde crtu
https://www.eevblog.com/forum/testgear/the-siglent-sdg2042x-thread/msg948669/#msg948669

[videobruce]

Then, show when needed. 
Your example was only the center frequency. When would those extra decimal places be needed for RBW, VBW, Sweep & Span? 
How often do you see a figure with a decimal point with training zeros unless it was in reference to another figure with additional trailing digits?

[nctnico]

What is with all these trailing zeros, especially with RBW, VBW, Sweep & Span? 

those trailing zeros are needed because they may not always be zero,

But if they are zero then just cut them. It just clutters the screen and makes the numbers difficult to read. Also there are very few situations where the numbers have to show so many digits anyway.

[PartialDischarge]

What is with all these trailing zeros, especially with RBW, VBW, Sweep & Span? 

those trailing zeros are needed because they may not always be zero,

But if they are zero then just cut them. It just clutters the screen and makes the numbers difficult to read. Also there are very few situations where the numbers have to show so many digits anyway.

ok, fair enough. Those zeroes dont bother me, but tell all other manufacturers first, rohde, anritsu, agilent.... Dont make it look like Siglent is screwing it up.

[rf-loop]

What is with all these trailing zeros, especially with RBW, VBW, Sweep & Span? 

those trailing zeros are needed because they may not always be zero,

But if they are zero then just cut them. It just clutters the screen and makes the numbers difficult to read. Also there are very few situations where the numbers have to show so many digits anyway.

I like more these fixed lenght. When I adjust (not only with this SSA) example frequency using small increments, it is really terrible if length is all time jumping when adjust. Fixed positions are much more for human eye. I have also trucck load of instruments what really keep wholöe length without turnin these zeroes on and off. Counters, signal generators atc. When there is 1.000 009999 GHz on the display I really do not like if +1Hz and it turns last 0...00 off and then if +1 Hz agen on these last 0...01.   I throw it out from window if they change this. I do not want look this kind of "numbers jumping".  This have also very carefully designed and thinked example in Hewlett-Packard when they tink watching ergonomy in some instruments imho.

---

Then about horizontal width. Signal have 751 data points (1:1 with TFT resolution) , no less and no more and stretching this for more wide on the display do not give any advantages. Perhaps not disadvantages but also it may happen.

Adding real data points and then get more width, of course nice but also without more processing power it may slow.

Keysight N93222C  have 461 data points. (and tiny 6.5" 640x480 display) (also 1:1 TFT resolution)

Keysight N9040B UXA  have more, it have adjustable amount of data points, up to 40001 (this is quite powerful  and expensive tool)

Rigol DS800 serie have 601.  (and also 1:1 TFT resolution)

[videobruce]

Why do the data points have to change? I'm talking about enlarging the active area as one would enlarging a photograph.

[Pinkus]

3. Display menu, Screen text  --> OFF.   It is there.  Do you mean some other thing?

Yeah, but then you are not able to see your data entry when typing in e.g. "FREQ" and then "433.1750" "Mhz". You have to type the whole number blindly.
I think the display off feature is nice, but should be enhanced: when pressing "FREQ/SPAN/RBW" or any number key, the display for this should be switched on again. By this you are able to see your entry. Then after a few seconds (after the last pressed key) it should be switched off again (only when the corresponding display option is "off" of course).

[videobruce]

Your choice is the idea. I'm asking for a option to turn it off. My other point it is too large.
I don't remember ever looking at that when I enter a frequency, so I wouldn't miss it (assuming it works the same way as Rigols).

[videobruce]

Your right, I had no idea, is it in the manual, I never saw it? 

[rf-loop]

Why do the data points have to change? I'm talking about enlarging the active area as one would enlarging a photograph.

Yes I undestand well what you want but..  there fixed real compued data points. You want magnify it. It means you add some fake "data" inside real tata points queye. Is it entertainment image machine or measurement equipment? Between what data points you want add these fake's so that trace is bit more long on the screen. After then, what we get more? What is advantage of generated visual arts.

[tautech]

I think the "real estate" balance is about right.
OK, I'm no expert but if you keep bashing around in the UI you'll find that the "little" additional space that could be gained will only clutter the GUI and possible confuse newbies to spectrum analyzers more.
For those that are proficient with such tools, yes some more enlargement/compaction might be advantageous but who is the target market ?
If the UI is used to a full extent like some of rf-loops screenshots the GUI is then getting quite busy.
A quick look back a few pages to P13 finds some screenshots to illustrate both sides of the discussion:





Of course Siglent will see/read these comments and no doubt have some discussion about your suggestions Bruce, whatever the outcome it's good to examine these things so thanks for bringing them up. 

[videobruce]

Coming from a video background, when I look at a T&M display, I look at how large the most important part of it is; the actual test result area, the 'grid'. On this, more so than Rigol (since it is a larger screen) I see unused area around the perimeter that tells me, 'why not expand the grid'. I didn't take into account the 'data points' equation which I would figure is the same as pixels in a photo. Enlarge the photo in editing software, something has to 'fill in' to make up the difference.

My bad, learned from my error. 

[hendorog]

I think the "real estate" balance is about right.
OK, I'm no expert but if you keep bashing around in the UI you'll find that the "little" additional space that could be gained will only clutter the GUI and possible confuse newbies to spectrum analyzers more.
For those that are proficient with such tools, yes some more enlargement/compaction might be advantageous but who is the target market ?
If the UI is used to a full extent like some of rf-loops screenshots the GUI is then getting quite busy.
A quick look back a few pages to P13 finds some screenshots to illustrate both sides of the discussion:





Of course Siglent will see/read these comments and no doubt have some discussion about your suggestions Bruce, whatever the outcome it's good to examine these things so thanks for bringing them up. 

There is an opportunity to change the data under the graphic - in between 'Centre' and 'Span' or 'Start' and 'Stop' in your images.
Personally, I would add 'Start', 'Centre' and 'Stop'.

I can see why they have done it that way, and the (blunt) message here is that innovation and improvement should trump copying.

[nctnico]

On most spectrum analysers (the ones I have used) the frequency range is either start/stop or span/centre. It depends on how you input the frequency range. IMHO there is nothing wrong with the way it works. Sometimes I want to look at a range (start/stop) and sometimes I want to look at the spectrum around a certain frequency (span/centre).

[rf-loop]

This is just one example image for thinking display development for improve useability.
(I have lot of images with different setups for thinking how some things can fix and improve. It is, imho, not bad now but there is many details what can improve)

Perhaps signal area can move more right and then get more room to left and after then then bit new arrange and there is room for some more information.

But this work is not at all ready. When ready I will send My opinions to Siglent development team.

One thing is clear. Bottom: I like if there is always Start f, Center f,  Span f, and Stop f  visible.
Also I like if there is more information visible about markers including some marker functions results and some things with these is now, If I can say, bit poor.  One wink to Siglent, if you copycat, do not copy poor things, copy only good things. Best is design own, better than others.

Also when user change start, center, stop and span f. Logic is not best possible. User is master, and if user want Start f priority it need keep start f untouched if user change Span. It is frustrating if I want keep constant start and then change span, now I need always after then agen change start and  in some cases I need do many "iterations" before I get what I want. Just why.
It can be much more clever or is it better say, more user friendly.
These kind of things are not really bad in this model but why it is semi good when using same money it can be good or even excellent. 

Example this (f change) priority can select with some clever way. Then I can (example) underline what parameter is set for "keep" or can it say highest priority. If I underline Span. After then if I change start it also change stop and if I highlight start then if I set Stop or Span it keep this Start. If I underline Span  and I change Center then both, Start and Stop follow with this constant Span. And so on..

Without marker table open there is room for display all 4 marker freq and level. Now it show only one marker.
When I use marker functions and select NdB example if measure filters so that one marker show -3dB BW and next marker -60dB width.
I want both (and all markers) can show this BW freq (and perhaps more info) least in marker table.

But, this kind of story is "endless".  These my example opinions are not at all finished, just only imagine someting...for example.

All things need think carefully and think pros and cons. So I will not start debate here how and what need develop better. This kind of work need good and experienced working team. But, principle is of course, every idea, even maddest idea,  is good in first place... ideas selection, evaluation, processing, developing, grinding and polishing, it is an art.

It is not like this and that is nice to have or like santa-claus wish list. Of course I want all... or perhaps not... I want what I need and make things more usable.. and also in different environments.

[morris6]

It was mentioned earlier in this thread:

Unfortunately, the SSA somehow works erroneously on other traces if TG normalization is used so the following scans are taken without normalization. It appears the SSA occupies one of the traces for the normalization data but I didn't look at this in detail for the moment, I just disabled this function.

I investigated the behaviour somewhat, to find out what's going on.
Keeping it simple I connected TG output to SA input, Selected TG on. In the Normalize menu I entered Norm Ref pos 50%, Ref trace View on and then Normalize on.



Trace A (yellow) is nice and straight loopthrough 0dB respons and the green trace is the "nomalization reference trace", the correction substraced from the actual SA input, if I'm right.. ?
Now in Sweep menu I selected Single to observe what is happening without the results running "of screen".
Next in Trace menu I selected trace A and activated View, to keep it during further measurements, with B or C..
This triggers another sweep and..



It seems another normalize correction was done to trace A. This is not wanted here, we want to keep trace A while we select trace B for further measurements.
OK, select trace B to ClearWrite..
As expected this triggers another sweep and..



While trace B comes up as loopthrough 0dB, again a normalize correction is done to trace A in View mode..?
Now in the Sweep menu I did another 5 sweeps



The frozen A trace, in View mode, gets further "corrected". If you had added some attenuation in the TG output before normalization the frozen trace gets an extra "correction" wih the amount of attenuation.
IMHO the TG function and normalization only works correctly for one trace since a trace in View mode, to save it during further measurements, gets "corrected" with every sweep.

It would be nice if the trace function could be used while normalization is on in TG mode.

[videobruce]

I have used the TG function in any of the SA's I have owned all the time, but I have never put multiple traces up on the screen.
I wonder if Rigol has the same issue since these seem to 'mirror' one another in many ways?