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

[eloso]

Thanks for the note. Yes, spent more hours than should be necessary on checking all cabling parts, including reaching for the good stuff that I try to keep on the shelf. Claned and checked everything as per the well knonwn Siglent Video on checking your connectors.

SA, Tracking generator are perfect. VNA is ok on S21. Just S11 which is making me think that there is some hardware issue with the directional coupler in the reverse direction. I would obviously  love for it to be a software issue, even better if its just an oversight but I've been using VNAs for a while now and am quite comfortable with them.  Given the amount of head scratching I'm doing on this problem  I am confident its not just an oversight.

I have a little NanoVNA for portability and certain types of measurements outside the lab and normally its not a patch on the Siglent - but at the moment its outperforming it dramatically. The Siglent doesn't settle down even with Averaging.

Cheers


Eloso

[gf]

*** SIGLENT FEATURE REQUEST ***

The new SVA1032X firmware upgrade 3.2.2.6.0R7 upgrade works, BUT...
  The field "IF bandwidth" for the VNA is still grayed at 10 kHz and this seriously limits the performance of the instrument.
So, please, if somebody here can contact Siglent, ask them to change this! Even the NanoVNA can configure lower BWs, improving the measurement noise and hence the dynamic range.

The DANL is specified in the data sheet with -156 dBm/Hz. At 10kHz ENBW, this should result in a noise floor of -116 dBm. Is that so bad?

OTOH, the DANL of the NanoVNA is certainly worse, so a narrower RBW is needed to compensate. However, this does not come for free, but it comes at the cost of a slow sweep.

Based on the hardware performance in SA mode, the instrument should be able to reach 100 or better 10 Hz BW.

I would draw a different conclusion: Hardware performance allows for a low noise floor even with a larger RBW, which in turn allows for a faster sweep.

[ Note that with a DSP-based filter/detector, narrow RBW is easy to realize and not a matter of hardware performance, but only a matter of patience
For instance, 10 Hz RBW would imply a measurement duration in the order of a few hundred ms -- for each frequency point.
There is no way to get around the tradeoff between RBW and measurement duration. So if you want a fast sweep, then narrow RBW is not what you want. ]

[RoV]

The DANL is specified in the data sheet with -156 dBm/Hz. At 10kHz ENBW, this should result in a noise floor of -116 dBm. Is that so bad?

DANL is specific to the SA mode. In VNA it is much worse: preamp is missing and the bridge introduces a loss in S11 mode.

OTOH, the DANL of the NanoVNA is certainly worse, so a narrower RBW is needed to compensate. However, this does not come for free, but it comes at the cost of a slow sweep.

Any serious VNA has the BW setting: I have worked with several models from hp (Agilent/Keysight) and Copper Mountain Tech. All of them have it and is useful. It is clear that Siglent had in program to introduce it, otherwise they wouldn't have put the relevant menu option that, sadly, is grayed.

[gf]

DANL is specific to the SA mode. In VNA it is much worse: preamp is missing and the bridge introduces a loss in S11 mode.

Oh, that's unfortunate then...

[tautech]

In VNA it is much worse: preamp is missing and the bridge introduces a loss in S11 mode.

Show us this with a load and Mag Loss trace.
 

0dB stimulus should return a ~0dB trace.

[gf]

In VNA it is much worse: preamp is missing and the bridge introduces a loss in S11 mode.

Show us this with a load and Mag Loss trace.
 

0dB stimulus should return a ~0dB trace.

The question is, what's the S21 noise floor when you measure S21 of a "zero-gain DUT" using a stimulus of (say) 0dBm?
(an example of a zero gain DUT is if you terminate port 1 and port 2 with 50R each)

[ And (any) bridge or directional coupler certainly does introduce some additional dynamic range loss for S11 readings, since only a fraction of the signal is coupled to the receiver. ]

[tautech]

In VNA it is much worse: preamp is missing and the bridge introduces a loss in S11 mode.

Show us this with a load and Mag Loss trace.
 

0dB stimulus should return a ~0dB trace.

The question is, what's the S21 noise floor when you measure S21 of a "zero-gain DUT" using a stimulus of (say) 0dBm?
(an example of a zero gain DUT is if you terminate port 1 and port 2 with 50R each)

[ And (any) bridge or directional coupler certainly does introduce some additional dynamic range loss for S11 readings, since only a fraction of the signal is coupled to the receiver. ]

Member RoV has no issue with S21 measurements, a through measurement and only makes comment on S11 measurement, a reflection measurement.
I asked for a loaded Mag Loss trace which will show if there is a problem or not.

Currently we do not have SVA1032X but 7.5 GHz SHA852A arrived today so after we get to know it some maybe I can offer some screenshots.

[RoV]

Member RoV has no issue with S21 measurements, a through measurement and only makes comment on S11 measurement, a reflection measurement.
I asked for a loaded Mag Loss trace which will show if there is a problem or not.

I mentioned S11 because it tends to be more noisy due to the additional loss introduced by the internal bridge. S21 doesn't use the bridge, because the analyzer is a T/R, not a full 2-port, so in S21 mode is equivalent to the SA in tracking mode (however, in VNA mode it misses the BW setting respect to SA mode).
If you measure a load you will find S11=0 with some noise. If you look at it in Log Mag mode, you'll see a noise level around -something (perhaps -50 dB?). Decreasing the BW, this would improve.
This is particularly critical if some device is inserted. For example, when measuring an external antenna impedance, in order to avoid the risk of damaging the VNA due to external signals, I prefer to insert a limiter in series to the port (full limit @5-10 dBm), followed by a 10 dB attenuator . Then I calibrate AFTER those devices. In this way the VNA port is very well protected, but >20 dB attenuation are added to the measurements (the chain limiter+attenuator is passed twice, first by the test signal and then by the reflected one), so traces tend to be very noisy (even during calibration) unless BW is kept small. This procedure worked very well with an old hp 8714ES, but doesn't with the SVA1032X.
 

[tautech]

Member RoV has no issue with S21 measurements, a through measurement and only makes comment on S11 measurement, a reflection measurement.
I asked for a loaded Mag Loss trace which will show if there is a problem or not.

I mentioned S11 because it tends to be more noisy due to the additional loss introduced by the internal bridge. S21 doesn't use the bridge, because the analyzer is a T/R, not a full 2-port, so in S21 mode is equivalent to the SA in tracking mode (however, in VNA mode it misses the BW setting respect to SA mode).
If you measure a load you will find S11=0 with some noise. If you look at it in Log Mag mode, you'll see a noise level around -something (perhaps -50 dB?). Decreasing the BW, this would improve.
This is particularly critical if some device is inserted. For example, when measuring an external antenna impedance, in order to avoid the risk of damaging the VNA due to external signals, I prefer to insert a limiter in series to the port (full limit @5-10 dBm), followed by a 10 dB attenuator . Then I calibrate AFTER those devices. In this way the VNA port is very well protected, but >20 dB attenuation are added to the measurements (the chain limiter+attenuator is passed twice, first by the test signal and then by the reflected one), so traces tend to be very noisy (even during calibration) unless BW is kept small. This procedure worked very well with an old hp 8714ES, but doesn't with the SVA1032X.

Understood however I have never used such practices for S11 antenna measurements of which I have done many.

Your HP might need additional protection and caution in use however SVA is not the same beast and a quick look at the datasheet states this for TG spec.
Average safe reverse power Total : 30 dBm (1 W)
Maximum safe reverse level Voltage: ±50 VDC

How any ordinary antenna we might sweep might exceed these levels is unclear. 

[RoV]

Average safe reverse power Total : 30 dBm (1 W)
Maximum safe reverse level Voltage: ±50 VDC

How any ordinary antenna we might sweep might exceed these levels is unclear. 

Antenna measurement was just an example.
I don't think the 30 dBm rating applies to port 1. It does, for a brief time and -I believe- with at least 20 dB internal attenuation inserted, for port 2 (SA port).

[tautech]

Average safe reverse power Total : 30 dBm (1 W)
Maximum safe reverse level Voltage: ±50 VDC

How any ordinary antenna we might sweep might exceed these levels is unclear. 

Antenna measurement was just an example.
I don't think the 30 dBm rating applies to port 1. It does, for a brief time and -I believe- with at least 20 dB internal attenuation inserted, for port 2 (SA port).

Average safe reverse power Total : 30 dBm (1 W)
Maximum safe reverse level Voltage: ±50 VDC

How any ordinary antenna we might sweep might exceed these levels is unclear. 

Antenna measurement was just an example.
I don't think the 30 dBm rating applies to port 1.

RTFM/datasheet.
Port 1 is the TG port however in VNA mode the reflection path is protected by the internal bridge just like how Port 2 (RF In) is protected when using an external bridge that provides additional attenuation when in SA mode.

Isolation for Port 1 is from the internal bridge of which we only have the TG reverse spec to work from which the internal bridge is identified shortly after Dave did the teardown early in this thread.
You could look that up and study its datasheet for its reverse voltage and attenuation spec of which Siglent has worked within so to provide the TG spec common to all these SVA/SSA designs.

Common sense would have Siglent null internal bridge attenuation from measurements to some degree so not to have uncalibrated measurements far off the display.

It does, for a brief time and -I believe- with at least 20 dB internal attenuation inserted, for port 2 (SA port).

Port 2 (RF In) has more defined specs as expected and in SA and adjustable attenuation: 0 ~ 50 dB
Maximum input DC voltage +/- 50 VDC
Maximum average power 30 dBm, 3 minutes, fc ≥10 MHz, att > 20 dBm, preamp off
Maximum damage level 33 dBm, fc ≥ 10 MHz, att > 20 dBm, preamp off

Have you ever tripped the RF In warning buzzer ?

[DL4JY]

When having longer antennas, eg. 40+ meter you get to certain times quite impressives levels. Especially when ROV is starting to transmit nearby - or somebody else e.g. in a car.

To that comes - and that is more important for magnitudes - a static charging of the Antenna and coax cable to several hundreds of Volts, which generates enormous HF-Spikes when the max Voltage of the insulation at a certain point in the structure is reached and an arc will occur.
This is not hypothetical - it's a matter of fact. I had a ground plane Antenna, which was not having an inductor to ground on the roof - for 2m only - I measured nearly a 1000v during a rain shower with an electron tube multimeter (ancient) .
I have seen cases where the input and power transistors of transceivers were killed by the peaks produced by static arcs.

[tautech]

When having longer antennas, eg. 40+ meter you get to certain times quite impressives levels. Especially when ROV is starting to transmit nearby - or somebody else e.g. in a car.

To that comes - and that is more important for magnitudes - a static charging of the Antenna and coax cable to several hundreds of Volts, which generates enormous HF-Spikes when the max Voltage of the insulation at a certain point in the structure is reached and an arc will occur.
This is not hypothetical - it's a matter of fact. I had a ground plane Antenna, which was not having an inductor to ground on the roof - for 2m only - I measured nearly a 1000v during a rain shower with an electron tube multimeter (ancient) .
I have seen cases where the input and power transistors of transceivers were killed by the peaks produced by static arcs.

From the User manual preface:
Electrostatic Prevention
Operate the instrument in an electrostatic discharge protective area environment to avoid damages
induced by static discharge. Always ground both the internal and external conductors of the cable to
release static before connecting.

[DL4JY]

From the User manual preface:
Electrostatic Prevention
Operate the instrument in an electrostatic discharge protective area environment to avoid damages
induced by static discharge. Always ground both the internal and external conductors of the cable to
release static before connecting.

but one time grounding does not help if the capacitors (coax line) are re-charged. I would never use port 1 or 2 of the VNA without external protection with an Antenna mounted outside.

[tautech]

From the User manual preface:
Electrostatic Prevention
Operate the instrument in an electrostatic discharge protective area environment to avoid damages
induced by static discharge. Always ground both the internal and external conductors of the cable to
release static before connecting.

but one time grounding does not help if the capacitors (coax line) are re-charged. I would never use port 1 or 2 of the VNA without external protection with an Antenna mounted outside.

Antenna static problems are nothing new and no different to what my pop faced with his end fed 30m SW antenna 75 years ago well before I was even a twinkle in his eye.
When a child his end feed anchor point and lightning arrestor were still fixed to grandmas house where when young he obtained QSL cards from all over the world and had them all pasted into a book.

BBC, Switzerland and Texas are just a few I can remember he had those SW directional transmission QSL cards for using primitive valve radios of his own builds and some of his own design.

I'll leave this 2 minute Google find here FYI:
https://ham.stackexchange.com/questions/2276/protect-rig-from-static-charge-on-antenna

[DL4JY]

Understood however I have never used such practices for S11 antenna measurements of which I have done many.

I wanted only to stress the elaboration of ROV for using safety precautions to protect his SVA.

[DL4JY]

and yes I prefer being one time faint-hearted than being dead forever 
Means use limiters and attenuators instead of replacing the input attenuator and other integrated circuits. Or the output of the S11 port.

BTW I bought a spare HMC1118 and have fixed it with Scotch tape on the back of my Siglent for all cases. Not knowing when this IC will be not available from a reliable source as Mouser anymore. I heard about guys, which replaced this IC twice within a week or two. 

But my hobby is playing with 65V LDMOS Fets running with almost 50 Amps.

[tautech]

New ECal units just released.

SEM5000A range.
4.5, 9, 13.5 and 26.5 GHz, 2 and 4 port N type, SMA and 3.5mm variants.
https://int.siglent.com/u_file/images/23_09_25/SEM5000A_datasheet_DS60050_E01A.pdf

SEM5000A series electronic calibrators include 5002A / 5012A / 5022A / 5032A / 5004A / 5014A / 5024A / 5034A, covering the frequency bands of 9kHz to 4.5GHz / 9kHz to 9GHz / 100kHz to 13.5GHz / 100kHz to 26.5GHz, with USB interface for communication and power supply, and the RF connectors come in three forms: SMA, 3.5mm, and Type-N. They are suitable for calibration and error correction of SHA850A series, SVA1000X series and SNA5000A series vector network analyzers of SIGLENT. The SEM5000A series are simple, fast, efficient, accurate, and widely applicable.

Prices from $2,345

[Martin72]

Hi...
I am still missing a spectrum analyzer for home, but I also want to be able to measure the overtagunsfunktion of filters, so I will probably need the combi device from siglent:
https://www.batronix.com/versand/spektrumanalysator/Siglent-SVA1015X-VNA.html
Are there any reservations about this device ?
I had skimmed the manual earlier, the VNA section.
If I have interpreted correctly, the start frequency is at 10Mhz ?
That would be high, if one wants to represent e.g. the transmission line of a 10Mhz filter.

[tautech]

Hi...
I am still missing a spectrum analyzer for home, but I also want to be able to measure the overtagunsfunktion of filters, so I will probably need the combi device from siglent:
https://www.batronix.com/versand/spektrumanalysator/Siglent-SVA1015X-VNA.html
Are there any reservations about this device ?
I had skimmed the manual earlier, the VNA section.
If I have interpreted correctly, the start frequency is at 10Mhz ?
That would be high, if one wants to represent e.g. the transmission line of a 10Mhz filter.

Compare specs carefully against SVA1032X. 

[Martin72]

And now ?

[tautech]

And now ?

SA mode also.

Will send PM.

[robkar02]

Hello,
I have just bought a sva 1015x and i am a newbie to this stuff. Im going to use it for hobby purpuse ( Ham radio) .

I have all the options on trial now, but would really make them Permanent without changing to much in the unit,

Is it possible for someone to guide me how to proceed?  I have tried to search the forum, but havent found what im looking for.

Best regards

Robert

[tcottle]

 Since you are already starting with a SVA 1015X all you have to do is generate the keys and enter them.  The process is pretty much goof proof

Here are the instructions (modified from https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg4692437/#msg4692437)

IV) Enable full options on SVA

1) Launch online python tool: https://www.online-python.com/PX3v0J8nbs
2) Uncomment  #Model = 'SVA1000X' line (remove #)
3) In your SVA navigate to system info and note HOSTID
4) Input HOSTID into python tool
5) Generate license codes and with SVA menu option System -> Load Option install and options: either one at at time (I did this) or choose ALL

[robkar02]

Thanks a lot for the information, will try this when coming home:)

After this, is it possible to update the FW in the future, and still keep the options?

Best regards

Robert