The Siglent SDG2042X Thread

[jpep]

It would be nice if others running the latest firmware version (V2.01.01.37R3 released past month) could check for the frequency counter bug and post their hardware version and results, so we can report it to siglent for them to fix it.

To check for the bug, just run the frequency counter:

Code: [Select]

Utility -> Counter -> State "ON"
The device should start to feel laggy/unresponsive as the time passes and once "num" reaches between 700-1200 counts the device freezes.

To speed up the process you can enable fast counting under setup:

Code: [Select]

Utility -> Counter -> Setup -> Type "Fast"

[djadeski]

Just checked on my enhanced 2042 with hardware 02-01-00-40-00 and I experienced the same thing - the counter starts out as not being very responsive to key presses and by 1200 or so samples it continues counting but is completely unresponsive to any keyboard input and requires a power cycle to recover.

[bicycleguy]

02-02-00-40-00     V2.01.01.37R3 enhanced
Definitely some kind of bug but:
Laggy at first but after a 5 to 6 presses of say the < or > buttons it seems to get more responsive.
Pressing Waveforms then Utility and Counter gets back to startup lag or better.
With counter running in background the generator functions seem to work with normal response.

" alt="" class="bbc_img" />

After posting this it froze at 11,472 counts

[bicycleguy]

In it's frozen state SCPI can still do a screen dump.  Note the counts in the bottom row don't match as normally.

" alt="" class="bbc_img" />

[thinkfat]

Someone with telnet access to the instrument might want to check memory stats - things getting laggy over time and then eventually freezing sounds like a memory leak to me

[jpep]

Thanks for the testing guys. Support told me they have also managed to reproduce the bug, and now they are forwarding it to the HQ. Hopefully we will get a firmware update with a fix.

Someone with telnet access to the instrument might want to check memory stats - things getting laggy over time and then eventually freezing sounds like a memory leak to me

I thought the same at first, then I took a quick look and iirc the ui app was stuck on a futex syscall. So now my guess is that it is a synchronization/locking problem in their app or their kernel code.

[Neepa]

Ok after now a year of owning the SDG2042X I tried the firmware hack and immediately encountered a problem.

DHCP doesn't work. Initially I reverted back to the oldest available formware version 17R5 and there DHCP turned on but the IP, Subnet and Gateway all turned to zeroes upon turning it on.

The Default IP adress being a weird 10.11.9.230.

At this stage at least DHCP was ostensibly still turning on....

However upon reverting to the old firmware 35R3B2 DHCP doesn't even turn on anymore. Tried the newest one 37R3 and no luck here also. The cursor jumps into the DHCP field but pressing it doesn't do anything at all. No IP change nothing.

Could this be a problem on the PC side?
Hardware is 02-02-00-40-00.

So what do?

[bson]

Can you check the DHCP server to see if it actually got a lease, or an error of some sort was logged?  Can you check if the DHCP client is running on the instrument?  It may have crashed due to a memory hardware problem, or it's out of memory or there's a bug or something.

It can also be something more complex, like it gets a lease for an address it already has an ARP entry for, meaning it thinks it's already in use by something else, so it's rejected.  It then asks the server for another one, but some DHCP servers will never return a new lease and will just keep repeating the one they already have.

You may have more than one DHCP server active, both trying to hand out leases on a default 192.168.0.0 net.  (Some super basic routers can't even be configured to use anything else.)  So the address might already be in use.  I've used the 10-net since time immemorial, with subnets for static addresses, DHCP, and VPN; if I ever see a 192.168 address I know immediately it's a configuration problem (something isn't set up correctly).

[Martin72]

Hi,

I´ve bought a 2042X (now 2122X..) since last month.

Firmware untouched so far, 2.01.01.37R2 , Hardware the same as you(02-02-00-40-00).

On my way to hack this thing, I didn´t have any problems with DHCP.
BUT:
It allows any changes only when LAN is active AND a network was found.
Tried it a few minutes ago without having a connection everything fails, DHCP couldn´t be activated.

Hope that helps.
Martin

[Neepa]

Yup that helps.

You having similar symptomps means it's not entirely the fault of the instrument but something in conjunction with my PC and the ethernet connection.

HAHA!
Connected the generator to my router and there DHCP works now. The Python hack should be doable over this indirect network connection right?

Something to try out tomorrow!

[Martin72]

Hi,

Connected the generator to my router and there DHCP works now.

My meters are always connected via the router.

The Python hack should be doable over this indirect network connection right?

The script doesn't care.
It is only to generate, not to transfer anything.

[Performa01]

Yup that helps.

You having similar symptomps means it's not entirely the fault of the instrument but something in conjunction with my PC and the ethernet connection.

So it's not "entirely the fault of the instrument" - how generous! Could it be that it's also a teeny-weeny bit the fault of a clueless user?

HAHA!
Connected the generator to my router and there DHCP works now.

That's what routers are made for. Guess where the DHCP server sits?

[Neepa]

I dunno honestly.

It's just that all the instructions for performing the hack simply state "connect via Ethernet"; "turn DHCP on" followed by basically "it's go time".

So reading this my basic thought was connect PC directly to generator and it should work.

[Performa01]

Connecting a device directly to the PC is a very special case and cannot be considered as a network - it's only a point to point connection after all. For a network, we need at least a hub (went out of fashion a long time ago) or a router. Only the latter contains a DHCP server, which of course is the prerequisite for DHCP to work.

A direct PC connection can be made to work with manual IP settings.

[tv84]

Only the latter contains a DHCP server, which of course is the prerequisite for DHCP to work.

Don't want to nitpick but you can configure a DHCP server in the PC.

[2N3055]

I dunno honestly.

It's just that all the instructions for performing the hack simply state "connect via Ethernet"; "turn DHCP on" followed by basically "it's go time".

So reading this my basic thought was connect PC directly to generator and it should work.

Unlike Performa, I will nitpick..

Siglent is hardly responsible for the fact that instructions how to hack instruments are not friendly to users that don't know much about networking....

And also Siglent in it's manual explains it all:

Note:
⚫ If the generator is connected to the PC directly, set the IP addresses,
subnet masks and gateways for both of the PC and generator. The subnet
masks and gateways of PC and generator must be the same and the IP
addresses of them must be within the same network segment.
⚫ If the generator is connected to the LAN of your PC, please contact with
your network administrator to get an available IP address. For details,
refer to the TCP/IP protocol.

4) DHCP Configuration Mode
In DHCP mode, the DHCP server in the current network assigns LAN
parameters, e.g. IP address, for the generator. Press DHCP to select “On” or
“Off” to turn DHCP mode on or off. The default is “Off”.

[Performa01]

Only the latter contains a DHCP server, which of course is the prerequisite for DHCP to work.

Don't want to nitpick but you can configure a DHCP server in the PC.

Right. I was talking about the default conditions though. It's hard to believe that in this particular case anything other than the default would be of any significance

[JOFlaherty]

About the frequency counter bug in the SDG2042X - it seems to be something associated with displaying the data, or the repeated calculations needed to do that.

If you start the frequency counter, and then press the waveform key, going back to that screen, the FC continues to run. I tested it last night, and when I went back to the counter screen (UTIL,COUNTER), it showed N = 115,450. 
Then leaving it on the FC screen, the instrument had frozen up by N = 116,409.

So, it's possible to collect counter statistics for a long time, as long as you're staying on another screen.

[Vit G]

Hello! I have a 2042x after trying to downgrade the firmware (16R2) stopped loading. On the screen only a Logo presents. how to bring it back to life now? Help please!
I saw on Dave's video how to restore 2122x using sd card. Where can I get the files for it for the 2042x model? perhaps this would help me.

UPD
My 2042X is 05-00-00-46-00 hw version. Dont have SD slot on board. it seems i need USB recover stick ?

UPD2
After about 10 minutes, a message appeared on the screen:  "No external clock source!".
 I connected an external 10MHz from my HP8920A and repeated the power-up procedure again. It did not help. The message reappeared after 10 min.

 The HL1 LED near CPU does not light up.

UPD3
 Thanks  for Siglent support. I make restore USB stick but this not helped .  I see some numerous errors in the bootlog. I am attaching it here.
 Need opinion from more experienced people  :'(

[mvlc]

As far as I remember, at least firmware version 23R7 need to be installed to be able to use system recovery from U-disk.

[tautech]

Hello! I have a 2042x after trying to downgrade the firmware (16R2) stopped loading. On the screen only a Logo presents. how to bring it back to life now? Help please!
I saw on Dave's video how to restore 2122x using sd card. Where can I get the files for it for the 2042x model? perhaps this would help me.

UPD
My 2042X is 05-00-00-46-00 hw version. Dont have SD slot on board. it seems i need USB recover stick ?

UPD2
After about 10 minutes, a message appeared on the screen:  "No external clock source!".
 I connected an external 10MHz from my HP8920A and repeated the power-up procedure again. It did not help. The message reappeared after 10 min.

 The HL1 LED near CPU does not light up.

UPD3
 Thanks  for Siglent support. I make restore USB stick but this not helped .  I see some numerous errors in the bootlog. I am attaching it here.
Need opinion from more experienced people  :'(

You do as your actions broke this working AWG.

Member mvic is quite correct in that attempting to install 16R2 took the OS back to before 23R7 when USB recovery was introduced.
You report your later unit has no SD card slot which results from a time when the firmware was so stable it was removed as any boot freeze could be recovered with the USB tool.

Any further progress relies on you installing a SD card holder and when done contact me again for the SD recovery package.
BTW, also get an 8GB SD card for it in readiness.

BTW, I am not Siglent support but I give my time freely to help owners even if they have done darn silly things with their instruments.
FYI Firmware revision history attached.

[Vit G]

Yes, I know it's my own fault. =( I did not think that the hardware had changed so much and did not even look inside the device to find that the SD slot was missing.
  I would install it, but is it technically possible? on this board there is not even a wiring for a slot and, of course, I do not have circuits from these devices.
  On the board, I see one connector that is possibly used for booting, but I don’t know. Is it possible to connect an SD card to it in SPI mode?

 UPD.
 No. It's most likely configuration of power management. tps65400

[tautech]

From the SD card recovery instructions, please see below pic with the location of the SD card holder.
It is likely the footprint is still there to be populated with a new SD card holder.

[thinkfat]

Nobody finds it weird how it's even possible to install a firmware that is too old to support the hardware? Preventing this must be trivial. 

[Vit G]

At the end of the firmware installation there was a message about a failed to upload firmware, so I did not expect that something was loaded there. Is it possible that there is some kind of backup before overwriting?