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

[Johnny B Good]

 I'm guessing it's more likely that this data, along with other setup data, is stored in sram ready to be copied at an instant's notice into the flash when the soft power button is pressed to initiate the shutdown process controlled by the OS. It may be being saved to sram every 60 seconds but this only gets written to the flash on each soft power-down cycle.

 Of course there's always the alternative strategy of triggering a short  (10 to 30ms?) write to flash cycle when the PSU PG signal is negated by the loss of mains supply utilising the few hundred milliseconds worth of reserve power remaining in the HT smoothing caps to complete such a short write cycle to guard against even this most extreme of shutdown scenarios.

[pagrei]

Does any one have a vna factory cal file from a genuine SVA1000X they can share?

I’d like to have a go at understanding the internal file format…

[pagrei]

Specifically, I am after the following files:

/usr/bin/siglent/usr/config/usr_vna_o_calib_log.bin
/usr/bin/siglent/usr/config/usr_vna_s_calib_log.bin
/usr/bin/siglent/usr/config/usr_vna_calib_log.bin
/usr/bin/siglent/usr/config/usr_vna_ref_log.bin
/usr/bin/siglent/usr/config/enh_rsp_calib_log.bin
/usr/bin/siglent/usr/config/usr_vna_e_calib_log.bin
/usr/bin/siglent/usr/config/usr_vna_e_ref_log.bin
/usr/bin/siglent/usr/config/usr_vna_o_calib.bin
/usr/bin/siglent/usr/config/usr_vna_s_calib.bin
/usr/bin/siglent/usr/config/usr_vna_calib.bin
/usr/bin/siglent/usr/config/usr_vna_ref.bin
/usr/bin/siglent/usr/config/enh_rsp_calib.bin
/usr/bin/siglent/usr/config/usr_vna_e_calib.bin
/usr/bin/siglent/usr/config/usr_vna_e_ref.bin

[CoolNamesAllTaken]

After upgrading my SVA1032X to firmware version 3.2.2.6.2R5.r19, I seem to no longer be able to save .CAL files, to the local file system nor to an external USB drive. Has anyone else experienced this issue? Any attempts to save .CAL or .CSA result in a "MSG: Save file failed!" warning. It seems that I can still save CSVs and other file types.

For what it's worth, the calibration does seem to stick between reboots after it's done, but it's weird to not be able to save it to a separate file.

[NY2KW]

If you can give me some basic instruction how to download those files, I can try.

Jerry

[pagrei]

Hi Jerry,

Thanks for the offer.  I’m waiting to see if another option “pans out”.  If not, I’ll post some instructions to extract the calibration data to a USB stick.

Pa

[pagrei]

Hi Tautech,

Did you get any feedback from Siglent regarding the write to NAND every 60 seconds in the latest firmware?

In particular, does it only occur if you don’t select boot to default?

[IM3]

Yes, I share the same concerns.

This is preventing me from proceeding with the update. I know that the hardware does have flash memory with a write limit of 100,000 cycles.

It would be great if Siglent acknowledged the concerns of technical users and provided an explanation that inspires confidence.

I have an SSA3075X-R, and this 60-second save feature has been implemented there as well.

[TERRA Operative]

Wow, that's theoretically only 28 hours of runtime needed to start having issues of worn memory.....

[IM3]

The calculation is a bit different: 100,000 flashes / (24 x 60 minutes) = 69 days. 

It would be quite strange for Siglent to flash data every minute, so I hope it was just poorly phrased in the firmware notes.

However, the only reason to save data every minute would be for power-down purposes, which must be done in non-volatile memory, such as flash.

I recently photographed a flash memory chip in an SDS5000X. This flash was located near the processor. You can assume that this type of flash is used in all Siglent devices.

[spacehen]

After upgrading my SVA1032X to firmware version 3.2.2.6.2R5.r19, I seem to no longer be able to save .CAL files, to the local file system nor to an external USB drive. Has anyone else experienced this issue? Any attempts to save .CAL or .CSA result in a "MSG: Save file failed!" warning. It seems that I can still save CSVs and other file types.

For what it's worth, the calibration does seem to stick between reboots after it's done, but it's weird to not be able to save it to a separate file.

Saving a calibration file works fine on my unit with the latest firmware.

[antenna]

Is there any more information about the excessive writes with the latest firmware?  I had updated my 1032x and came across this thread before using it again, and now I am afraid to use it until I hear more about this issue.  Thanks!

[wma]

The SSA3032X-R is the cheapest of the SSA X-R series and costs €4830, but you can expand it to 7.5 GHz for free, including all options with the key generator that you can download here.
http://tpcg.io/_SV2SRE  (Many thanks to the guys who made this script)

In my opinion, the SSA3032X-R is the device that gives you the most for the least money.

And real time is nice to have, really.

Hi I just want to double check for scripts http://tpcg.io/_SV2SRE  , Do I have to modify 3075 to 7075 and 3050 to 7050 ?
as mentioned in https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg5572021/?topicseen#msg5572021


The other thing, how about model SSA3032X-R,

Scripts only have following
#Model = 'SVA1000X'
#Model = 'SSA3000X'
#Model = 'SSA3000X_Plus'
#Model = 'SSA3000R'

Seems scripts does not have particular section for SSA3000X_R

Which Model  that I should use if I want to upgrade it from SSA3032X-R to SSA3075X-R

Many Thanks

[IM3]

Yes, you have to change 3075 to 7075. Skip the 7050 because upgrade in one step wil do the job.

The SSA3000R is the one you need for the SSA3032X-R upgrade.

It’s as easy as it sounds. You will be happy with this upgrade for sure.

[wma]

Thank you

[james38]

Maybe I misunderstood you, but I don't see that saving every 60 seconds reduces/minimizes writing to the flash memory.
Quite the opposite.

I'll see what I can find out.....if not immediately, in a few weeks when I visit HQ.

However, I personally deal with many NAND failures worldwide and have a multitude of recovery packages to repair problems with NAND management which are getting less by the day.
I can in no way see Siglent changes to SVA/SSA SW impact on NAND wear as they are very conscious of these issues.

I like you highly value SVA and it's my fav RF instrument yet have never had or seen problems with NAND wear in any Siglent analyzers.

I would like to come back to the topic of saving for 60 seconds.
A few months have passed now, but there is still no official statement about what exactly that means.

@Tautech
Were you able to find out anything about this at HQ?

[EE4all]

Any update on the flash writing every 60 seconds? I am holding off on the firmware update until this is clarified.

[antenna]

I too am hoping for a response here.  Mine has been, for the most part, unplugged since I did this upgrade to make sure it does not damage my 1032x.  I assume one way around this is to not have the power on option set to "Last", and if the feature is wanted, it could be switched on just prior to turning off the unit.  PITA though, always having to go switch it on and off.

[antenna]

I just turned mine on to change it from last to user and noticed it would freeze in SA mode after one sweep.  The little red crossshair would get to the end and all functionality just stops.  I had to quickly change it to VNA mode before the sweep finished and then find the factory reset in the menu.  After the factory reset, a SA sweep between 1 and 30MHz resulted in a sawtooth noise floor.  Thankfully after powering down and restarting again, it seems to be good.  This is ridiculous though, I still have 53 hours on my trial features.  Barely used and already having issues.

[IM3]

The 60-second Flash mystery has yet to be solved.

So, I decided to contact Siglent via email and hope to receive an answer any time soon.

I emailed the European headquarters at:  [email protected]

There is also the Asian headquarters at:   [email protected]
And the American headquarters at       :   [email protected]



Subject: Concern Regarding Firmware Update V3.2.2.6.2R5 for SVA1000, SSA3000X Plus, SSA3000X-R Series Spectrum Analyzers

    Dear Siglent,

    I would like to share my concerns with you regarding the firmware update V3.2.2.6.2R5 for the spectrum analyzer series SVA1000, SSA3000X Plus, SSA3000X-R.

    In the firmware release notes, the following new functionality is mentioned:

    “Save the configuration every 60 seconds, and load it when Power On is turned Last.”
   
    On the EEVblog, users have expressed their concerns about the meaning of this statement.

    The only reason to save data every minute would be for power-down purposes, which must be done in non-volatile memory, such as flash memory. The concern is that flash memory has a limited lifespan and will degrade after approximately 100,000 write cycles.

    This would mean:
    100,000 write cycles / (24 x 60 minutes) = 69 days of operational time.

    It would be quite unusual for Siglent to write data to flash memory every minute, so I hope this was simply a case of poorly worded release notes that are being misunderstood.

    Currently, many users are refraining from installing this update until more details about this issue are clarified.

    I look forward to your response, which I will gladly share on the EEVblog forum.

   
    Kind regards,

[tautech]

The 60-second Flash mystery has yet to be solved.

Statement on this matter from HQ via EU HQ:

We use UBIFS.
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.

In our next FW version, we will add a user option that allows users to decide whether to enable the periodic check and save function.

[Johnny B Good]

The 60-second Flash mystery has yet to be solved.

Statement on this matter from HQ via EU HQ:

We use UBIFS.
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.

In our next FW version, we will add a user option that allows users to decide whether to enable the periodic check and save function.

 Thanks for winkling that nugget about the 60 seconds auto-save feature out of Siglent - it only took five months  

 That's not too far off my own hypothesis I'd posted here:

https://www.eevblog.com/forum/testgear/siglent-sva1015x-1-5ghz-spectrum-vector-network-analyzer-(coming)/msg5585383/#msg5585383

 The only bit I'd left out being about if no changes had been made to the sram buffered data, the flash would remain unmolested on the next shutdown whether a user instigated shut off via the on/off button or an unexpected power loss.

[IM3]

Even though UBIFS is designed to work efficiently on raw flash memory, it still contributes to flash wear. UBIFS relies on the UBI layer for critical functions like wear leveling and bad block management, which help extend the lifespan of the flash. However, frequent write operations, such as making a backup every minute, can still exhaust the flash memory over time.

For example:

Writing every minute results in 1,440 writes per day, or over 525,000 writes annually. Even with good wear leveling, this can wear out blocks, especially if your flash has a typical endurance of 100,000 Program-Erase (PE) cycles.

UBIFS operates on large logical erase blocks, so even small data writes can trigger full block erases, increasing wear.

While UBIFS includes compression and buffering to reduce writes, it can't eliminate the fundamental limitations of flash memory.

In short, while UBIFS is efficient and reliable, heavy write patterns will still eventually wear out flash memory.


I skip this firmware update and wait for the next one.

[Johnny B Good]

 From Siglent's reply to Tautech's enquiry, it seems the autosave every 60 seconds is only applied in ram and it is only written to the flash if the config data had been changed since the previous session on powering down or the result of the ac power being lost.

 Even if the config data has been changed, such flash updates will only be applied on power down, not 1440 times a day. I doubt anyone will be power cycling their SA every 15 minutes or so, let alone every minute. A more likely usage pattern would be power up for an hour or three perhaps once to three or four times a day.

 I'd very surprised if anyone ever power cycles such kit more than an average of three times a day. At that rate, assuming a 100K block write/erase cycle endurance, it would take something like a century to wear it out, even without wear leveling.

[tautech]

Even though UBIFS is designed to work efficiently on raw flash memory, it still contributes to flash wear. UBIFS relies on the UBI layer for critical functions like wear leveling and bad block management, which help extend the lifespan of the flash. However, frequent write operations, such as making a backup every minute, can still exhaust the flash memory over time.

For example:

Writing every minute results in 1,440 writes per day, or over 525,000 writes annually. Even with good wear leveling, this can wear out blocks, especially if your flash has a typical endurance of 100,000 Program-Erase (PE) cycles.

UBIFS operates on large logical erase blocks, so even small data writes can trigger full block erases, increasing wear.

While UBIFS includes compression and buffering to reduce writes, it can't eliminate the fundamental limitations of flash memory.

In short, while UBIFS is efficient and reliable, heavy write patterns will still eventually wear out flash memory.


I skip this firmware update and wait for the next one.

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.

If we go back to the FW release notes, this discussion only applies to Save Last settings every 60s when enabled and when no changes are made to settings, nothing is written to flash.

If there is no desire to reboot to Last used settings this feature can remain turned OFF so the analyzer boots to default settings.