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

[Bicurico]

I'm pretty sure the brick was not a fault of the user, but rather a poor implementation of the web update on behalf of Siglent.
@Kibabalu: I would like to know what exactly happend and what you did? Did you just select "online FW update"? Did you have a USB disk inserted? Did you remove it after first reboot?

[tautech]

I'm pretty sure the brick was not a fault of the user, but rather a poor implementation of the web update on behalf of Siglent.

Possibly and maybe I should've checked that when updating my SVA1032X to V3.2.2.3.3R1 last evening just before Kibabalu found himself in a pickle. Thanks again for helping him out chaps.

Anyways by USB stick it was uneventful. 

[hendorog]

So got my SDR Kits Rosenberger "Calkit" i know it's a generic one but original ones are to expensive for "Hobby use" :-(
And i think they are decent enough for my useage (hopefully)..

Now my problem is, how can i get the Calkit Parametes inside my SVA1032x? I saw user calkit buttons but the Parameters were unclear for me.
Or should i telnet in and refine / manipulate one of the saved "original" Calkits with my parameters?

Here are the calkit parameters:
https://www.sdr-kits.net/documents/Rosenberger_Female_Cal_Standards_rev5.pdf

Can somebody elp me to verify which parameter inside the pdf is the correct parameter for the SVA user calkit (C1,C2,C3...)?

The SDR Kits are only supplied with partial calibration parameters. The only params you can use on the SVA are the delay figures in ps.

Essentially based on the params they do supply you are missing out on the limited characterisation of the load they have done.

SDR kits do a 4 wire resistance measurement of each individual load at DC and they also supply a generic parallel capacitance value in the data sheet you listed. The SVA cannot natively use either of those.
The SVA assumes that User supplied loads are perfect, and for the pre-loaded cal kits listed in the device menu it uses a generic measured data file for the load.

(Generic meaning that the parameters/data file is the same for all cal kits of the same type. The assumption they make is that all cal kits are the same within some margin of error.)

[Noy]

Hm and now?
What can i do? How can i use the provided cal parameters?
Or simply unuseable?

[hendorog]

Hm and now?
What can i do? How can i use the provided cal parameters?
Or simply unuseable?

Well you _can_ enter the delay figures supplied for the short and the open into a User cal kit.
But you can't use the load DC resistance or the CII value as there is nowhere to enter them. Bear in mind that they might not improve things very much anyway.

I have one of their kits and from memory the load was pretty good even though it was marked as 48.6 ohms on the DC resistance measurement. Certainly it measured better than some 18GHz loads I got from eBay - which had better DC resistance measurements.

If you want the best accuracy then you can to get someone with a proper cal kit to measure your cal kit.
Then you can use those to do error correction of your measurements on a PC. This also has the advantage of allowing the use of lots averaging, which is not supported for SVA calibration sweeps. It is a bit of a cumbersome process though.

[Kibabalu]

I'm pretty sure the brick was not a fault of the user, but rather a poor implementation of the web update on behalf of Siglent.
@Kibabalu: I would like to know what exactly happend and what you did? Did you just select "online FW update"? Did you have a USB disk inserted? Did you remove it after first reboot?

The procedure was:

1. download the FW on my machine (iMac)
2. selecting the online FW update offered from the web Front end inside the browser (Safari)
3. choosing the downloaded FW file (iMac)
4. reboot
5. brick!

The manual is stating very clearly, that a FW update has to be made by using a usb stick and initiated by the menu function of the unit, otherwise you loose warranty.

Thus, no question, it was my fault. I haven’t RTFM ;-)

But on the other side, siglent should not offer such a „trap“ inside their web front end.

In general, the web front end seems very sloppy when using browsers under macOS (tested with Safari and FireFox). It's sloppily developed, means they do not test properly, e. g. testing all functions against other operating systems than Windows too.

[tautech]

Try a Chrome browser.

[Kibabalu]

Try a Chrome browser.

Chrome seems to have the same minor glitches like FireFox, e. g. switching into full screen mode etc., but nothing to scary about.

But regarding the web front end the new FW is slightly worse than the old one. On the other side, the bouncing effect inside safari seems to be disappeared.

[tv84]

1. download the FW on my machine (iMac)
2. selecting the online FW update offered from the web Front end inside the browser (Safari)
3. choosing the downloaded FW file (iMac)

Maybe it's me but I definitely didn't understand this part.

Nonetheless, maybe the online web update still points to a lower FW version which is not compatible with the new FW 3.x and that's the cause of the brick...

If the procedure of online web update is getting a file into the USB disk to flash later, then it's pratically the same as downloading it first into the USB stick and doing a local update. UNLESS the file is different or corrupted.

[Kibabalu]

1. download the FW on my machine (iMac)
2. selecting the online FW update offered from the web Front end inside the browser (Safari)
3. choosing the downloaded FW file (iMac)

Maybe it's me but I definitely didn't understand this part.

Nonetheless, maybe the online web update still points to a lower FW version which is not compatible with the new FW 3.x and that's the cause of the brick...

If the procedure of online web update is getting a file into the USB disk to flash later, then it's pratically the same as downloading it first into the USB stick and doing a local update. UNLESS the file is different or corrupted.

After clicking on "FirmwareUpdate" a file browser dialog opens up and you can search and choose a FW file. This file should then be copied somewhere, probably. Maybe this copy process didn't work properly.

Thus, it's not really an online firmware upgrade. It's more a web based front end controlled FW update process.

[tv84]

Thus, it's not really an online firmware upgrade. It's more a web based front end controlled FW update process.

 

OK, then you did nothing wrong! BTW, download again the .ADS via that method BUT don't flash it and compare with the latest .ADS that we get from the web (in the PC). Or tell me the CRC32.

[Kibabalu]

Thus, it's not really an online firmware upgrade. It's more a web based front end controlled FW update process.

 

OK, then you did nothing wrong! BTW, download again the .ADS via that method BUT don't flash it and compare with the latest .ADS that we get from the web (in the PC). Or tell me the CRC32.

It's the same. I choosed the .ADS file I downloaded earlier from the Siglent support page.

The question is: What happens then? The file has to be copied and expanded from my machine somewhere into the device, most likely to /usr/bin/siglent/usr/usr/upgrade. Controlled by the web server inside the SVA. Maybe this went wrong?

[tautech]

Had too many  to check and as it's early 2021 now will rollback FW and try the web FW update tomorrow to see if I can brick mine also. 

Watch this space. 

[Bicurico]

The "online" update is a missleading name. What I understood is that you open the webpage of the SSA/SVA and select FW upgrade. This opens a file dialog pointing to your computer where you can select the ADS file. It is then transferred to the SSA/SVA and installed.

I think the problem is that this only works if a USB disk is attached to the SSA/SVA to store and extract the ADS file. Otherwise, it will be stored in the same folder, which is not mounted to the USB device! Apart from the lack of free memory space, because the device is not syncing this data, after a reboot it will be lost. Which is why the bricked device was complaining about the missing U-Disk0 device.

This is a serious bug! Why would you do an "online" update, if you have to insert a USB disk? Why not insert the USB disk with the ADF already copied onto it?

[tautech]

I'll document my experience with it tomorrow. 

[Kibabalu]

Had too many  to check and as it's early 2021 now will rollback FW and try the web FW update tomorrow to see if I can brick mine also. 

Watch this space. 

Uh, already new year in NZ?

Happy new year!

[tv84]

Those kiwis always like to go ahead of us!!! 

Happy New Year!

[hendorog]

Those kiwis always like to go ahead of us!!! 

Happy New Year!

We get the hangover early too damn it...    => =>

[tautech]

The "online" update is a missleading name. What I understood is that you open the webpage of the SSA/SVA and select FW upgrade. This opens a file dialog pointing to your computer where you can select the ADS file. It is then transferred to the SSA/SVA and installed.

Correct, it works just like that.

I think the problem is that this only works if a USB disk is attached to the SSA/SVA to store and extract the ADS file. Otherwise, it will be stored in the same folder, which is not mounted to the USB device! Apart from the lack of free memory space, because the device is not syncing this data, after a reboot it will be lost. Which is why the bricked device was complaining about the missing U-Disk0 device.

This is a serious bug! Why would you do an "online" update, if you have to insert a USB disk? Why not insert the USB disk with the ADF already copied onto it?

You don't however you can browse to a USB stick and select the ADS update file from there if you wish.

In the present GUI the firmware update labels/buttons and called Firmware Update whether they are in the instrument UI or the virtual key in the remote web sever so to call it an 'online' update as you are is misleading as the firmware update is totally done locally and mostly* works exactly as expected. 
No bug found.

* Mostly as in after the update the SVA reboots and if connected to the web server a refresh of the browser page is needed for the SVA to reconnect.

[Kibabalu]

your testing scenario was upgrading from Firmware V3.2.2.3.2 to V3.2.2.3.3R1 using the web gui and everything went's fine?

Then maybe in my case I didn't wait long enough before I tried another reboot and interrupted the copying process or I just had "bad luck".

[tautech]

your testing scenario was upgrading from Firmware V3.2.2.3.2 to V3.2.2.3.3R1 using the web gui and everything went fine ?

Yes.

Then maybe in my case I didn't wait long enough before I tried another reboot and interrupted the copying process or I just had "bad luck".

This ^
From the update instructions:

When the device upgrading finished, it will restart by itself.

[Kibabalu]

Yes, I was aware of this. But it took that long, at least in my feeling, that I suspected some issue.

In future I will try to school myself in patience 😉

[tautech]

Yes, I was aware of this. But it took that long, at least in my feeling, that I suspected some issue.

In future I will try to school myself in patience 😉

Yes the auto reboot seems to take longer to happen if using the webserver to update, however it does work even though it seems like a freeze.
Maybe some display 'please wait for reboot' message need be added for the impatient. 

[tautech]

Firmware update for all SVA1000X models....please ensure you select the correct version for your instrument.

Version： V3.2.2.4.0
https://int.siglent.com/download/firmwares/?ProId=30
~19.3 MB

Release notes
Notice: This version can not be rolled back
This firmware must be upgraded from V2.1.1.1.12a or later. If your SVA1000X has an earlier version, please update to V2.1.1.1.12a first and then proceed to update to the latest version.
 
Fix TG switch error at small RBW
Fix Help and Limit file lost when Factory reset
Fix Auto open delay unit error in VNA mode
Fix CNR calc error when no peak found
Fix OBW result overflow
Fix Harmonic calc error at 100Hz RBW
Fix Reflection Open+Short calibration action
Fix spectrum amplitude error in DMA mode
Optimize EMI .csv file content to all signal list
Optimize EMI peak search
Optimize Auto cal process
Add Peak search to Spectrum in DMA mode

[tubularnut]

Hmmmm 

Since updating to V3.2.2.4.0, I can't use the old telnet ADS files to get back in and re-enable telnet in the startup script? This worked fine on the last update to R1.

Have they removed / moved the telnet binary?