Siglent SSA3000X and SSA3000X-Plus Spectrum Analyzers

[tautech]

Good news, looks like the SSA3021X Plus has been transformed into the SVA1032X 

It appears the SMA cable is installed also

I'm not so sure.

Best as I can see mawyatt's settings and after a Preset to clear mine here's what a real SVA1032X looks like in the same Smith chart mode. FW = latest dot 8 version.



Maybe rig up a 50 ohm resistor on a cable and LF sweep it in Smith R+jX mode and see if it returns a result near the center of the display. 

I am waiting to order a cal kit (see if this conversion actually worked), and made up a simple test cable as you suggested. Used a 47 Ohm on a banana adapter to a BNC to the N to BNC cable I have (I'm waiting on other cables and adpaters that I've ordered). The trace didn't move to the center of the Smith Chart for a sweep of 100K to 10MHz as it should, so evidently I do not have the interior cable installed 

Looks as if I'll have get a cable, or make one up.

Best,

Data point for you, much the same setup as yours but with N-BNC adapter and BNC to croc clips cable with 47 ohm resistor.

[tomud]

Good news, looks like the SSA3021X Plus has been transformed into the SVA1032X 

It appears the SMA cable is installed also

Thanks to to Elasia, tv84 and Emo for all the help

Best,

I think the trace you are seeing is the result of the missing cable

This is a normal view without performing any calibration. The lack of a cable can be seen, for example, in this post: https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg3087608/#msg3087608

[mawyatt]

I dissembled the unit and these are images of the inside. The "missing cable" is present

For a test I used a simple leaded 1/8 Watt 47 ohm resistor on a Banana to BNC adapter (between the screw terminals), hanging on the end of a N to BNC 700mm cable. Two plots, one from 100K to 3.2GHz (Full span), the other from 100K to 100MHz.

Maybe a proper cal will make this look better?

Best,

[tautech]

I dissembled the unit and these are images of the inside. The "missing cable" is present

For a test I used a simple leaded 1/8 Watt 47 ohm resistor on a Banana to BNC adapter (between the screw terminals), hanging on the end of a N to BNC 700mm cable. Two plots, one from 100K to 3.2GHz (Full span), the other from 100K to 100MHz.

Maybe a proper cal will make this look better?

Best,

Yep, should do.
The full span shot suffers from straight line syndrome as it's only sweeping 201 points.
Trace interpolation could be improved however straight lines are a flag that the resolution needs be better for wide spans.
Change the setting to max 751 and it will look nice and smooth.

[kahuna0k]

Yep, should do.

Did you finish the experiments comparing the NanoVNA (1 and 2) cal kits and the official Siglent one? It seems that the latest firmware lets you export and import calibration data, maybe there is a way to get real calibration data into the "improved" SSA.

[tautech]

Yep, should do.

Did you finish the experiments comparing the NanoVNA (1 and 2) cal kits and the official Siglent one? It seems that the latest firmware lets you export and import calibration data, maybe there is a way to get real calibration data into the "improved" SSA.

Member hendorog has my Cal kit for comparisons against his HPAK ones on his HPAK legacy VNA so no, not had a session with it yet. He's also cross checking results against his SSA3075X-R which has SVA functionality.
Work to date has indicated all SVA user calibrations are only corrections to the inbuilt factory calibration file standards yet we have asked for implementation of support to import Touchstone files so any properly characterized calibration kit files can also be imported into SVA's.
Deeper discussion best for the SVA thread:
https://www.eevblog.com/forum/testgear/siglent-sva1015x-1-5ghz-spectrum-vector-network-analyzer-(coming)/

[tomud]

Maybe a proper cal will make this look better?

Best,

For high frequency you need a good calibration kit. Something like you test, you can use up to a few megahertz (100MHz is definitely too much for such a resistor plugged into the BNC). Once you get the calibration kit, you should be fine.

You should also use good test leads. With bad cables, we will observe phase shifts when bending them.

Converting SSA to SVA is fun. However, the money saved should be spent on good cables and a calibration kit - without it, it is difficult to use a VNA correctly.For SMA connectors, it is also worth buying a torque wrench.

[Billy33]

Just two new topics:

* there is quite a difference in power down behaviour between the SA/VNA SVA1032X+ and the DSO SDS2000X+:
DSO: pressing the on/off button leads to a shutting off process.
VNA: here there is no feedback and power seems to be hard switched off.
Any ideas?

* VNA data download: is there a way for aquiring not only the screenshot graphics, but also the data itself, e.g. the .s2p data like you can get easily from the nanoVNA series?
I want to aquire this data for feeding it into some calculation progs. Menu File/Save can save into CSV on the USB storage, but i would like to aquire .s2p directly if possible.
Also the Easy-Spectrum SW does not support the VNA yet?
Does somebody have some experience here?

[noreply]

Here is a suggestion for Siglent FW developers ...

The Keysighr MXA signal analyser has a 'phase noise' function - see attached screen shot

Would this be possible to do with the SSA3000x range?

There is enough ability in the SSA to get the raw data - so it just needs to be processed and displayed in the same domain as the Keysight MXA

Or am I missing something here - where the SSA simply does not have the HW to perform such a plot?

Its a very interesting feature to have - especially if its just a bit of code in the end without any HW enhancements.

Just Saying ...

[tautech]

FYI, the Phase measurement feature is available standard in SVA models.

[rf-loop]

Here is a suggestion for Siglent FW developers ...

The Keysighr MXA signal analyser has a 'phase noise' function - see attached screen shot

Would this be possible to do with the SSA3000x range?

There is enough ability in the SSA to get the raw data - so it just needs to be processed and displayed in the same domain as the Keysight MXA

Or am I missing something here - where the SSA simply does not have the HW to perform such a plot?

Its a very interesting feature to have - especially if its just a bit of code in the end without any HW enhancements.

Just Saying ...

SSA3000X own phase noise is not so nice for direct PN measurements
20 °C to 30 °C fc=1 GHz
<-95 dBc/Hz @10 kHz offset
<-96 dBc/Hz @100 kHz offset
<-115 dBc/Hz @1 MHz offset



MSA X performance is in tis own class and have lot of better own phase noise (but not true excellent)
Noise sidebands
(20 to 30 °C, CF = 1 GHz)
10  Hz  –80 dBc/Hz, nominal
100 Hz  –91 dBc/Hz
1   kHz –112 dBc/Hz, nominal
10  kHz  –113 dBc/Hz
100 kHz –116 dBc/Hz
1 MHz  –135 dBc/Hz
10 MHz –148 dBc/Hz, nominal

[rf-loop]

FYI, the Phase measurement feature is available standard in SVA models.

FYI, this  have nothing to do with phase noise measurements.  But other ways of course good to know for these who perhaps do not know.

[hendorog]

Bit of a phase noise comparison in this thread, for the SSA which I think has similar PN performance to the SVA.

https://www.eevblog.com/forum/testgear/phase-noise-head-to-head-siglent-ssa3032x-vs-hp-8566b-vs-signalhound-sa124b/msg1559449/#msg1559449
(note there is an approx 2.5dB error in my SSA chart, which rf-loop pointed out at the time)

Also, if you have a clean enough source to use as a reference, then the phase noise floor of the instrument can be lowered by maybe 10dB using 'noise floor enhancement'.

[Sighound36]

Regarding the the PN ability of the MXR it is add on for those looking a multi faceted instrument that can cover the majority of lab parameters. It produces a fair stab at this function. However its not real the real McCoy by  long shot.

Very low phase noise floor is required for genuinely accurate phase noise measurements and the MXR just does not have this specific trait to really dig deep in the noise to show meaningful results with a lot of the clocks we work with.

For others it will be helpful to a certain  degree though I suspect the Siglent SSA would be very close in performance.

[noreply]

Here is a suggestion for Siglent FW developers ...

The Keysighr MXA signal analyser has a 'phase noise' function - see attached screen shot

Would this be possible to do with the SSA3000x range?

There is enough ability in the SSA to get the raw data - so it just needs to be processed and displayed in the same domain as the Keysight MXA

Or am I missing something here - where the SSA simply does not have the HW to perform such a plot?

Its a very interesting feature to have - especially if its just a bit of code in the end without any HW enhancements.

Just Saying ...

SSA3000X own phase noise is not so nice for direct PN measurements
20 °C to 30 °C fc=1 GHz
<-95 dBc/Hz @10 kHz offset
<-96 dBc/Hz @100 kHz offset
<-115 dBc/Hz @1 MHz offset



MSA X performance is in tis own class and have lot of better own phase noise (but not true excellent)
Noise sidebands
(20 to 30 °C, CF = 1 GHz)
10  Hz  –80 dBc/Hz, nominal
100 Hz  –91 dBc/Hz
1   kHz –112 dBc/Hz, nominal
10  kHz  –113 dBc/Hz
100 kHz –116 dBc/Hz
1 MHz  –135 dBc/Hz
10 MHz –148 dBc/Hz, nominal

Yes, I see you point - but if Keysight thought it worthwhile to have a PN plot function and they ONLY have 20dBc over the SSA - this would still be a great feature for Siglent to include in the functionality IMHO.

If I do a PN measurement - for the sake of argument the DUT has worse PN than the SSA - then I would like to see a log 'plot' for the frequency range of interest - rather than make individual 'on freq' measurements.

I still think its a great feature to have.

If I were to 'buy' a new low cost SSA today - and one had this function and the other did not (all other things being the same) - thern without douubt the one with the PN plotting function would get my attention to 'buy'.


Just saying ...

[mawyatt]

Regarding PN measurements, as mentioned by rf-loop the SA must have a significantly lower internal PN that the subject. Achieving this is difficult and expensive and why specialized instruments capable of accurate PN measurements are available (and very expensive).

An interesting story on PN measurements back in late 80s we were developing a new tiny low PN microwave oscillator based upon a newly developed "Over-moded Bulk Acoustic Wave Resonator" created by Dr Ken Lakin. We didn't have the proper PN measurement capability in-house and I had just returned from an IEEE MTT conference where MA-COM described a PN measurement technique with some custom developed circuits that could be preformed with in-house equipment generally available. We engage our university with a grant to fund the development of this new type PN instrument technique, I later hired the brilliant grad student that did the work

The concept basically works like this. You need two identical subject oscillators which are coupled into a high quality passive mixer at the RF and LO ports, the signal level needs to be sufficient to turn on the mixer diodes. The IF port is low passed filtered with a passive LC filer which has low DC resistance (resistance contributes additional noise) and has a BW of the desired PN measurement BW. We initially used a 4th order capacitive input Butterworth LC at 100KHz. The output of the filter then is measured with a low frequency high dynamic range analyzer. You may need to include a low noise buffer amplifier before the analyzer, we used 40dB gain custom designed ultra-low noise amplifier with very low 1/f noise.

Since the two oscillators will injection lock thru the mixer port to port leakage, the mixer IF port output will be the sum of the two oscillators PN and thus 3dB higher since F1-F2 is zero and F1+F2 is filtered by the low pass filter. So each individual oscillator has a PN that is 3dB below the measurement, plus the mixer loss must be added (~7dB).

The setup required battery power and the two oscillators placed in a thermally isolated box to get reliable measurements which were about -140dBm (~22nv/rtHz) at 10KHz offset. Later we gained access to a proper HP PN measurement instrument and confirmed this measurement.

Best,

[mawyatt]

Good news, looks like the SSA3021X Plus has been transformed into the SVA1032X 

It appears the SMA cable is installed also

I'm not so sure.

Best as I can see mawyatt's settings and after a Preset to clear mine here's what a real SVA1032X looks like in the same Smith chart mode. FW = latest dot 8 version.



Maybe rig up a 50 ohm resistor on a cable and LF sweep it in Smith R+jX mode and see if it returns a result near the center of the display. 

I am waiting to order a cal kit (see if this conversion actually worked), and made up a simple test cable as you suggested. Used a 47 Ohm on a banana adapter to a BNC to the N to BNC cable I have (I'm waiting on other cables and adpaters that I've ordered). The trace didn't move to the center of the Smith Chart for a sweep of 100K to 10MHz as it should, so evidently I do not have the interior cable installed 

Looks as if I'll have get a cable, or make one up.

Best,

Data point for you, much the same setup as yours but with N-BNC adapter and BNC to croc clips cable with 47 ohm resistor.

Cal kit arrived and I did a cal. The results look better now with the 50 ohm resistor hanging off a ~700mm RG223 cable with a BNC to Terminal adapter

Now waiting on some quality cables and adapters

Best,

[indeterminatus]

Long time lurker, first time poster ... just wanted to report that the steps to cross-flash a SSA 3021X Plus to a SVA 1032X worked for me as well. Thank you all for the efforts involved.

I am now operating in "pro" mode, but I'd like to restore a proper serial number with the software options re-installed. Pleading for PM with insights on how to do that, since that seems to be private knowledge 

[olc]

Let me add my SSA+ to the list of successfully crossflashed machines 
SMA cable is installed (calibrated for testing using the NanoVNA plugs).
Thanks to the forum, especially @nike75, @tv84, @Elasia, @maximevince and much much more!
If possible, I'd like to restore the serial number as well (for fun).

--
Olivier

[tautech]

New firmware for SSA3000X Plus models

Version V3.2.2.3.2
14 MB
https://int.siglent.com/download/firmwares/?ProId=48

Release notes
Fix avg power bug in EMI Mode
Fix amplitude bug in AMA Mode
Fix peak search bug in SCPI command

[TurboTom]

I wonder if Siglent will still provide updates for their "non-Plus" models of the SSA3000X. Since they released their improved model, it seems the previous one has been dropped from their support list. There are still a few bugs to address, even if they may be minor ones, like reversing "upper" and "lower" 3rd order intercept points in the corresponding measurement function (well, I shouldn't complain since I didn't pay for that function...).

Anyway, not continuing to maintain these "more mature" products may prove to turn expensive in the long run... I for my own part have made my decision regarding Siglent gear, at least unless the way they deal with flawed firmware and well-meant improvement suggestions by their "early" customers changes considerably.

[mrprecision]

Hello,

I want to by a Siglent SSA3021X PLUS. Can be the device hacked to SSA3031X PLUS with the actual firmware?

Regards

[tv84]

 

Why put the same question in 3 threads? ? ? ?   

Specially after it has been answered in the others threads!!!

[mrprecision]

Can you please tell me how to to this?

[tautech]

SSA/SVA File management.

An Export <file> feature request is in a draft however before we flick this to Siglent I'm looking for more/better ideas.

Background.
When screenshots or such are saved without a USB stick inserted, files are auto saved locally and presently the only obvious way to export them is to Copy/Paste them onto a USB stick.
Saved files are accessed in the File menu and to perform tasks with them the Operate sub menu is used.

Request proposal.
To add a new Export <file/s> feature into the File>Operate sub menu where selected files are simply highlighted and quickly saved to a USB stick with a new Export key.
Further, when the SSA/SVA is accessed via the webserver Instrument Control page and export <file/s> functionality is required the addition of one new Export button to accompany the Firmware Update and Screen Shot buttons to allow for selected file export to the PC's browser Download folder.

In summary, 2 new types of user saved file export of only <user> selected files to either USB stick or PC Download folder and both dependent on the available connection path; USB or LAN.
Export feature is denied if USB or webserver is unavailable.

Please discuss.