Siglent SVA1015X and SVA1032X 1.5, 3.2GHz Spectrum & Vector Network Analyzers

[antenna]

Anyone know the size of the file it writes?  If it is like they say and writes to a new address each time, I imagine by the time it cycles through whats left of that 2Gb of storage and repeats that 100k times, we will all be dead anyhow. 

[Johnny B Good]

 Actually, that highlighted in bold statement was precisely what I was trying to get across to IM3. I thought rewording it might help convey the meaning of Siglent's message.

[dmytre]

After several months of operation, the reflection coefficient at the startup of the device looks as shown in the attached photo. The device in VNA mode can still be calibrated, but the dynamic range has decreased to S11 levels of -15...-20 dB. Has anyone encountered a similar issue?

[IM3]

Try checking if you have the correct levels in spectrum analyzer mode. It's possible that the input is blown.

[IM3]

I think you misunderstood and missed the highlight portion below:

This file system has a balanced management of Flash, which means it is NOT writing to the same physical address.
Our application determines whether the configuration has been modified, and if not modified, it will not be saved.

I’m genuinely trying to understand this.

It seems to me that the configuration changes almost immediately after powering on. The settings are probably adjusted right away—that's just how it goes.

I’m willing to believe that the configuration is stored in RAM, although Siglent doesn’t say that at all in their response. And how do you quickly save it to Flash during a power-down?

After that, I don’t understand what problem Siglent thought they were solving with this option. I could already choose to use the last configuration at startup.

And now Siglent is even saying that in the next update, this new feature can be turned off again.

( sake of the discussion, nothing more.)

Actually, that highlighted in bold statement was precisely what I was trying to get across to IM3. I thought rewording it might help convey the meaning of Siglent's message.

 A reverse variant of all the alternative descriptions used by the disgruntled customer in the Monty Python "Dead Parrot" sketch as in "This flash won't be joining the choir invisible any time soon".

Tone can be hard to discern in written text, and it’s possible you intended it as humor rather than criticism. But it might also carry an undertone of "explaining things to someone who doesn’t get it."
And that doesn’t feel right, especially when Tautech follows up with a “Thanks.”

That shouldn’t be the tone we have here.

[tautech]

I think you misunderstood and missed the highlight portion below:

This file system has a balanced management of Flash, which means it is NOT writing to the same physical address.
Our application determines whether the configuration has been modified, and if not modified, it will not be saved.

I’m genuinely trying to understand this.

It seems to me that the configuration changes almost immediately after powering on. The settings are probably adjusted right away—that's just how it goes.

Good.

Let's discuss this further as I understand this.

Always we have had the option to boot from Last settings.
As I see it Siglent has modified how the Last settings are saved to memory in that:

Our application determines whether the configuration has been modified, and if not modified, it will not be saved.

This implies saving was previously done differently and now does a check every 60s when settings are set to Last.

This would be useful for those that suffer power failures for a reboot to provide the same or settings close to those last used.


I have seen NO memory issues with SSA/SVA models since their release and expect Siglent to have made these changes to reduce flash use/wear not increase it.

[IM3]

Wow, do you ever sleep? 

Thanks for your addition. It really makes a difference in understanding.

I now also believe that this is not a thoughtless memory destructive option. But it was a concern for many and had been lingering for a while.

In the absence of clarification, I wrote a note to Siglent and had already received an interim response, but now they've answered through you. And that's fine.

[Johnny B Good]

Actually, that highlighted in bold statement was precisely what I was trying to get across to IM3. I thought rewording it might help convey the meaning of Siglent's message.

 A reverse variant of all the alternative descriptions used by the disgruntled customer in the Monty Python "Dead Parrot" sketch as in "This flash won't be joining the choir invisible any time soon".

Tone can be hard to discern in written text, and it’s possible you intended it as humor rather than criticism. But it might also carry an undertone of "explaining things to someone who doesn’t get it."
And that doesn’t feel right, especially when Tautech follows up with a “Thanks.”

That shouldn’t be the tone we have here.

 My apologies for that "humorous" line. I've snipped it out. I'm afraid I'd gotten the impression that Tautech had been responding to my post despite the quoted text showing he was addressing your apparent confusion. His post had been the final one on page 41 with Antenna's being the first on page 42 of this topic thread and mine the second.

 It wasn't until well after I'd made that post that I turned back to the previous page to re-examine Tautech's post that I realised my mistake.

 As for saving the latest changed version of a configuration data block to the flash, such a short operation can be completed within a matter of a few milliseconds using the energy remaining in the psu's smoothing capacitors (typically a few hundred milliseconds worth).

 This is an already solved problem which I'm sure Siglent must have taken care of, one way or another, otherwise they wouldn't have said what they'd said on the matter.

[tautech]

Wow, do you ever sleep? 

Thanks for your addition. It really makes a difference in understanding.

I now also believe that this is not a thoughtless memory destructive option. But it was a concern for many and had been lingering for a while.

In the absence of clarification, I wrote a note to Siglent and had already received an interim response, but now they've answered through you. And that's fine.

Late arising today.....

If we look back at the release notes for V3.2.2.6.2R5 they are a large set where many features have been tweaked or fixed which gave me confidence the complete FW version had been worked on for some time and some changes had been in the plans for a good while but not given to us until they were confident all was well.

Working with Siglent on their private forum they have several branches all working on improvements and from different parts of China but work remarkably well as a team.
Throw in some western input and the complete machine works really well.

Look what your query to the EU GM achieved > Q to SW team > back to EU GM > to me to make a statement here on EEVblog.

Anyways, I hope everybody now has a handle on how Last saves work and that it's NOT a step backward but a improvement in how these analyzers work.

[dmytre]

Try checking if you have the correct levels in spectrum analyzer mode. It's possible that the input is blown.

The SA mode seems to be OK. And tracking generator is also OK. The output power corresponds to the set value.

This positive S11 could be calibrated but the dynamic range degrades.

I have attached the measurement in VNA mode of 8 dB and 20 dB attenuator. S11 ~ -40 dB could not be measured and -16 dB is more noisy than it was before this HW error.

Has anybody seen similar issue?

[IM3]

Try checking if you have the correct levels in spectrum analyzer mode. It's possible that the input is blown.

The SA mode seems to be OK. And tracking generator is also OK. The output power corresponds to the set value.

This positive S11 could be calibrated but the dynamic range degrades.

I have attached the measurement in VNA mode of 8 dB and 20 dB attenuator. S11 ~ -40 dB could not be measured and -16 dB is more noisy than it was before this HW error.

Has anybody seen similar issue?

I repeated your measurement as you described it and got roughly the same result.

I don’t think there’s anything defective, but you may be making a measurement error. The span of 3 GHz is quite large. Also, the measurement with a 20 dB attenuator can give strange results because the reflection should then be 40 dB. If not everything is set up correctly, you won’t achieve that.

Do this:

    Make sure your stimulus level is -10 dBm. It might have been set much lower in your initial measurement and lowers dynamic range.
    Reduce the span for now because it’s quite large.
    Set a start frequency at for example 250 MHz.
    Set a stop frequency at for example 500 MHz.
    Perform a 1-port calibration with a good LOS (Load-Open-Short).

After that, don’t use your attenuators first for a control measurement but use the Load from your LOS. You should then be able to achieve at least 50 dB Return Loss.

If you achieve this, I don't think there’s anything wrong with your VNA.

[vsilves]

I was trying to get screenshots from converted Siglent SVA1032X remotely via SCPI like I do for my Siglent SDS800X-HD and failing.  For the Siglent SDS800X-HD I use "PRIN? PNG" as documented in the programming manual.

Searching online found this link from Siglent documenting another command to get screenshots from Siglent SDG2042X.  Command "SCDP" not documented in latest programming manual for SDG2042X.  Not really interested on screenshots from my signal generator but that opened my curiosity.

https://siglentna.com/operating-tip/sdg-remote-scpi-command-to-request-a-screen-image/

Asked ChatGPT if he was familiar with the SCPI command SCDP for SDG2042X (yes he was) with follow up question if there was a similar command for SVA1032X.  Answer yes with example of a "new" HCOPy command not on latest programming guide for Siglent SA or SVAs either.  Using google found this Siglent App Note documenting HCOPy.  Full command ":HCOPy:SDUMp:DATA?" (only for SSA) or ":INNer:ALL:DP:DATA?" (for SSA+ and SVA).

https://siglentna.com/application-note/analyzer-screen-image-python-lan/

I guess I will give the command a try and it will be a blessing to be able to quickly get screenshots from the command line, if it works. Tired of sticking the USB stick in the pie hole. 

But, what the heck is going on with new SCPI commands not documented in the respective programming guides?  I mean, the app note was published in early 2020 and we are in 2025 already.  Or why make so many variations of the command?  Shaking my head, but grateful if it works!

[tautech]

I was trying to get screenshots from converted Siglent SVA1032X remotely via SCPI like I do for my Siglent SDS800X-HD and failing.  For the Siglent SDS800X-HD I use "PRIN? PNG" as documented in the programming manual.

Searching online found this link from Siglent documenting another command to get screenshots from Siglent SDG2042X.  Command "SCDP" not documented in latest programming manual for SDG2042X.  Not really interested on screenshots from my signal generator but that opened my curiosity.

https://siglentna.com/operating-tip/sdg-remote-scpi-command-to-request-a-screen-image/

Asked ChatGPT if he was familiar with the SCPI command SCDP for SDG2042X (yes he was) with follow up question if there was a similar command for SVA1032X.  Answer yes with example of a "new" HCOPy command not on latest programming guide for Siglent SA or SVAs either.  Using google found this Siglent App Note documenting HCOPy.  Full command ":HCOPy:SDUMp:DATA?" (only for SSA) or ":INNer:ALL:DP:DATA?" (for SSA+ and SVA).

https://siglentna.com/application-note/analyzer-screen-image-python-lan/

I guess I will give the command a try and it will be a blessing to be able to quickly get screenshots from the command line, if it works. Tired of sticking the USB stick in the pie hole. 

But, what the heck is going on with new SCPI commands not documented in the respective programming guides?  I mean, the app note was published in early 2020 and we are in 2025 already.  Or why make so many variations of the command?  Shaking my head, but grateful if it works!

What's wrong with the webserver where you have full control of the instrument and the Save Screenshot feature that pushes images into your browsers Download folder ?

[vsilves]

What's wrong with the webserver where you have full control of the instrument and the Save Screenshot feature that pushes images into your browsers Download folder ?

Nothing wrong with web interface and do use from time to time but not my cup of tea in most occasions. I find I work faster from command line.  Not so much messing with the mouse and it can be easily automated, like I could literally whistle in to the computer while my hands are busy and the computer would take a screenshot if I wanted to.

In any case, I made it work.  Key was using :HCOPy:SDUMp:DATA? command (on SSA3032X Plus) which does return a BMP binary blob. 

The comment in Siglent application note's code states for SVA and SSA+ should use :INNer:ALL:DP:DATA? but that does NOT work, returns a binary blob that is not a BMP, PNG, or anything that I could figure out.

> Code on SIGLENT App Note code:
> qStr = SocketQuery(s, b':HCOPy:SDUMp:DATA?') #Request SSA screen image
> #qStr = SocketQuery(s, b':INNer:ALL:DP:DATA?') #Request SVA/SSA Plus screen image

Happy camper now.  I can quickly get screenshots from command line. Python and vxi11 is the fastest/shortest way (no need for any VISA installation).

[vsilves]

Took a lot of screenshots today and after every screenshot, even though the web interface is not connected or open, the instruments enters its "Remote" sate/mode upon receiving any SCPI command.  That means that I have to remember to press the "Esc" key in the control panel of the instrument before being allowed to go anywhere else in the physical UI.  I forgot so many times that it was not even funny...

Just found out that I can send this SCPI command, right after getting the screenshot, and that reverts back to local mode.  Tested and works.  No more annoying forgetting to press the "Esc" key!

     :SYSTem:LOCal

[MRF237]

Hi all!

I've a question to anyone who upgraded to the latest official firmware, V3.2.2.6.2R5.

I own a SSA3032X-R, which came with V3.2.2.6.0R10.
The VNA functionality was usable out-of-the-box. Without any additional user calibration, it showed correct S11/S21 on the N-connectors.
I.e., when leaving port 1 open, I saw 0dB reflection, and with a good 50 ohm terminator I saw < -40dB reflection.

However, after upgrading to V3.2.2.6.2R5, this wasn't the case anymore. S11 and S21 show meaningless values without user SOLT calibration.
After SOLT calibration, everything works as expected. But this takes time, and must be repeated after changing the settings. Often, it's fine for me when it's calibrated at the N-connectors.

I wonder if this is due to something I did, or if it's an issue in 3.2.2.6.2R5? For me it was a reason to revert to the previous version. After downgrading, the 'factory' calibration was still missing  , but using the 'factory default' function in the System menu restored it  .

Could someone with a SVA1000X or SSA3000X-R confirm this behavior?

[DL4JY]

Did you enable the calibration in the Measure menu ? I have the same issue with a 3021x plus after powering on - I need to enable the calibration in the submenu (e.g. enhanced through)

cheerio

[RoV]

My upgraded 3021X+ restores my last calibration on poweron. I normally leave a couple of cables on to avoid wear on the instrument connectors and I calibrate on the cable ends. After executing the calibration, I enabled it and saved it on a local file in the analyzer, this may be one difference.
Also, in the System menu, under Power ON Preset, I enabled "Power on":"user" and "Preset type":"user" and saved a user config. But I'm not sure this affects the VNA calibration.

[MRF237]

Did you enable the calibration in the Measure menu ? I have the same issue with a 3021x plus after powering on - I need to enable the calibration in the submenu (e.g. enhanced through)

Yeah, I'm quite sure I tried that with 6.2R5. AFAIK this enables the 'user calibration', which requires manual SOLT calibration first. After doing that, 'Cur' or 'C?' appears on screen to indicate that a user calibration is active.
However, I see quite correct results (although not perfect) when leaving it disabled. And, regardless the frequency range or number of points. See the screenshot, showing S11 for a good dummyload directly on the N-connector.
When I manually do a SOL calibrate, the result is even slightly worse! This is because I only have a SMA female cal kit, and the adapters causes some mismatch.

to avoid wear on the instrument connectors

Have you actually seen this happening  ? But it's wise, I think. I'm always afraid of ruining the connector threads with a cheap adapter or something.

and I calibrate on the cable ends.

Clear! But, such user calibration is only valid for a single frequency range / number of points combination, not?
I'm still curious, what do you see when you do _not_ apply the calibration? Are the results usable?
I can imagine that with a converted SSA3021X+, this 'factory' calibration is missing at all ? It might be difficult to find someone on this forum who has bought an genuine SVA1032X :DD.

It sounds to me like Siglent has changed the file location of these factory cal files in 6.2R5, or something like that . Or, it only fails on SSA3000X-R devices? I assume it's about files na_1p and na_2p, which are in firmdata0/cali.

[DL4JY]

Roberto,
you are right - after applying the preset parameters for user - it stores the last setup and the VNA is calibrated without any additional interaction.

Many thanks for this hint, I hadn't checked this menu in depth.

[MRF237]

Did you enable the calibration in the Measure menu ? I have the same issue with a 3021x plus after powering on - I need to enable the calibration in the submenu (e.g. enhanced through)

Ok, I tried it one more time. I installed 6.2R5 again, but could by no means not get my correct default calibration working.
Without a previous manual calibration, the 'Apply Calibration' is grayed out and cannot be selected. Reset-to-defaults or reset to factory did not help. Distance to Fault is also not working without user calibration! With manual calibration, it works, but then (only) measures from the calibration point, where in the previous version by default the distance is measured from the N-connector.

So, I reverted, again, to 3.2.2.6.0R10,.. (which, for the SSA3000X-R, also has a way faster trace update rate, but also a few annoying bugs which are solved in 6.2R5).

[tautech]

Dunno what's happened to 6.2R5 FW and it was online until a few days back. Monday here and in a few hours Siglent HQ will be online and maybe upload a new version. 
I'll ping them when I get a chance .......

With SVA1032X in the past I have saved a Cal file to local memory to recall at any time.
Not tried this for a while as back then quality Cal kits were borrowed. 

To do these Cal file saves you need change the file save type to Cal and later load the same file.
Just be sure to name them so to know which you are loading....

[tautech]

Dunno what's happened to 6.2R5 FW and it was online until a few days back. Monday here and in a few hours Siglent HQ will be online and maybe upload a new version. 
I'll ping them when I get a chance .......

A new SVA1000X FW version is coming.

[tautech]

I can feel another RF antenna project coming on similar to my Antenna Project Log posted some years ago.

First we need know what we're dealing with so out comes one of my fav instruments, SVA1032X and a RF loop from just a small loop of coax terminated like a near field probe, that is the center conductor soldered to the coax shield forming a loop, in this case only some 20-30mm dia.
Add your fav coax fitting to the other end and you're ready to go with a non-contact RF sniffer, a RF loop that will save your analyzer front end unless signal amplitude is extreme. 

So what's tautech up to ?
Just a simple 433 MHz keyfob remote range extender by means of a good antenna where I'll do another homebuilt J-Pole.
Requirement is for a sliding gate controller that already has a short range 433 MHz receiver and keyring remotes.

This time we require 75-200m connectivity for best convenience, a good bit less than the 500m required on the last antenna project. 
Attached is the 1st sniff of what I'm dealing with, 50mm from the remote.

[spacehen]

Ok if they are developing new fw why has the old fw been deleted then?

I've never encountered that.