Siglent SDG1000X Waveform Generators

[BillyO]

Is the firmware up to date?

I've only ever used mine through Ethernet and have had no issues.

[tautech]

I have just received my SDG1032X and everything looks good so far, except: the device freezes when I try to control it from the PC through USB (with LAN it works fine).

Are you working from the latest Programming guide ?
Which version of NIVISA do you have installed ?

https://int.siglent.com/u_file/download/22_08_26/SDG_Programming%20Guide_PG02-E05A.pdf

[nedeb]

The firmware is up to date (1.01.01.33R3).

The latest guide from Siglent makes use of Python 2.7 and an outdated version of PyVISA (1.4) which cannot be installed anymore (at least with PIP) but I'm using the equivalent code with Python 3 and PyVISA 1.12.0.

Thanks for pointing to the VISA backend: I'm using pyvisa-py (0.5.3) and it seems to be the only backend with this problem: the *IDN? query is successful in the following cases:

• Ethernet with pyvisa-py on Linux
• USB with NI-VISA application on Windows (NI-VISA 2022 Q3)
• USB with python-usbtmc on Linux

I'd really like to get this working in PyVISA on Linux: it's by far the easiest and most maintained solution. But no luck so far:

• With pyvisa_py backend it crashes the device
• With librevisa backend it doesn't work (known issue: this library doesn't implement viParseRsrcEx and the library is unmaintained)
• With NI-VISA backend: I cannot install NI-VISA on my system as it's not compatible with the latest Linux kernels. Haven't tried the Keysight version but I expect similar problems since they don't support the latest Linux versions.

I'll bring this to the PyVISA developers, hopefully the pyvisa_py backend can be fixed (though I suppose the bug could also be in the device firmware and only triggered by pyvisa_py for some reason).

Here's a minimal example to reproduce the freeze if someone wants to try (you need to install the pyvisa, pyvisa_py and pyusb Python packages):

Code: [Select]

#!/usr/bin/python3

import pyvisa as visa

rm = visa.ResourceManager('@py')

# Replace the serial number with your device's
addr = 'USB0::0xF4EC::0x1103::SDG1XDCX6R2773::0::INSTR'

device = rm.open_resource(addr)
(This is basically the initialization code from https://int.siglent.com/resource-detail/27/ . See also https://www.siglenteu.com/application-note/programming-example-create-a-stair-step-waveform-using-python-and-pyvisa-using-lan/ for another example.)

Even if the bug is in pyvisa_py maybe Siglent should fix the firmware to avoid the freeze... If I look at the Linux kernel messages when I connect the device with USB, there's definitely something that looks wrong:

Code: [Select]

[17263.890859] usb 1-4: new high-speed USB device number 25 using xhci_hcd
[17265.203020] usb 1-4: device descriptor read/64, error -71
[17265.497789] usb 1-4: config 1 interface 0 altsetting 0 bulk endpoint 0x81 has invalid maxpacket 64
[17265.497804] usb 1-4: config 1 interface 0 altsetting 0 bulk endpoint 0x1 has invalid maxpacket 64
[17265.538695] usb 1-4: New USB device found, idVendor=f4ec, idProduct=1103, bcdDevice= 0.20
[17265.538702] usb 1-4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[17265.538705] usb 1-4: Product: SDG1032X
[17265.538707] usb 1-4: Manufacturer: Siglent
[17265.538708] usb 1-4: SerialNumber: SDG1XDCX6R2773
(See errors in lines 2,3,4)

[tautech]

I still have some doubts however thank you for a good write up of the issue you are getting.

If we look at the first few pages of the programming manual it specifies functionality with NI-VISA 5.4 or the Run-Time Engine, which is just the connectivity part of the NIVISA package and much smaller.
Just checked and it's ~75MB however I don't know if it has any LINUX compatibility issues.
https://ni-visa-runtime.software.informer.com/download/

As this is the 5.4 version specified in the programming manual I recommend you investigate it first rather than we dive off on some bug chase just yet.

[nedeb]

I could only find the Windows version on that page... I can't find an official way to get the Linux version of NI-VISA 5.4 but https://github.com/pyvisa/pyvisa-py/issues/108 has a link to http://ftp.ni.com/support/softlib/visa/NI-VISA/5.4.1/Linux/NI-VISA-5.4.1.iso . Unfortunately it seems this ISO only contains a 32-bit version the visa library.

However your answer prompted me to try a finer-grained install of NI-VISA, since it's only the kernel modules that fail to install. Indeed installing only ni-visa-passport-usb works fine! And I can use it with pyvisa to communicate with my SPD1168X.

But pyvisa with the NI-VISA 2022 Q3 backend still doesn't work with the SDG1032X (now I get the error "The resource is valid, but VISA cannot currently access it").

I also realized that there's also a problem when using python-usbtmc: the first time I execute *IDN? I get the expected result but the second time I get a buffer size error (while it works fine with the SPD1168X).

Any suggestion what to try next?

[tautech]

Any suggestion what to try next?

None other than wait for some guidance from HQ Tech support from the email I've sent them.

[tautech]

Any suggestion what to try next?

None other than wait for some guidance from HQ Tech support from the email I've sent them.

Actually LXI Tools could be the solution you need.
Thread:
https://www.eevblog.com/forum/testgear/open-source-lxi-tools-v2-0-released/
Siglent App notes with LXI guidance to work through:
https://siglentna.com/application-notes/waveform-generators/page/2/

[nedeb]

None other than wait for some guidance from HQ Tech support from the email I've sent them.

OK, thanks a lot!

Regarding LXI Tools: I tried and it works well indeed, but it's only for LAN connection unfortunately. LAN communication also works well with pyvisa in my limited testing, it's really USB that I can't get to work on Linux. (I want to use USB because of a lack of Ethernet ports in the office, and so I don't have to worry about securing the devices against mischievious students).

[tautech]

None other than wait for some guidance from HQ Tech support from the email I've sent them.

OK, thanks a lot!

Regarding LXI Tools: I tried and it works well indeed, but it's only for LAN connection unfortunately. LAN communication also works well with pyvisa in my limited testing, it's really USB that I can't get to work on Linux. (I want to use USB because of a lack of Ethernet ports in the office, and so I don't have to worry about securing the devices against mischievious students).

Yes I did wonder it mightn’t suit the USB need you mentioned earlier. 
Tech support have been in touch so I pointed them to another of your posts.

[nedeb]

It turns out I can get something working by removing some USB device resets and config calls in pyvisa, see https://github.com/pyvisa/pyvisa-py/issues/333 for details. So maybe we can fix it there (I guess it depends on whether the pyvisa developer can find a fix that doesn't stop other devices from working). Though I must say, the fact that the SDG freezes makes testing a bit cumbersome, and I worry of damaging the device through excessive power cycling.

[NickKUK]

Interesting thread. I have a SDS1104X-E upgraded to 200Mhz and looking at purchasing a SDG1000 but couldn't decide between the 30Mhz and the 60Mhz version. The target use would be some RF tube radio, testing for a ADC using a 24.756Mhz clock and some DSD clocking plus audio sine wave bode plots.

If I can hack the 30MHz to 60MHz without severe impact on the squarewave then I'll put my order in for the 30Mhz version.

[james38]

I have a serious phaenomen with my SDG1032X (now SDG1064X).
Latest FW: 1.01.01.33R3
HW: 02-01-00-24-00

When i set a frequency for example sine 30Mhz and check this with my SVA everything is fine.

But when i enable the 10 Mhz external clock reference than the signal is overlaid with disturbances.
It seems like some intermodulation products on it.

I have checked the reference signal separately but it is ok an clean.
The input level is about +6dbm. I have also generate a signal with an SSG3021x to check as an alternate but see the same result.

I have attached some pictures.
One with the 10 Mhz sine reference signal.
The other one are with and without external reference enabled.
Once for 30 Mhz and 60Mhz.

Does anyone have an explanation of what could be the cause?

Regards Chris

[Wrenches of Death]

Mine has the same problem. It was basically useless at higher frequencies when using an external clock. I dropped back to the previous B6 firmware version and the problem disappeared. Go back a few pages in this thread and you'll find a number of posts on it.

Hopefully, it will get fixed in the next firmware revision.

WoD

[james38]

Yes I have read the last posts.
If I understood it correctly, it was more about the frequency accuracy.

In my case I have no problem with it but rather with the intermodulation signals when i enable the external reference.

Even if I use 15 Mhz or 20 Mhz, I always have these interferences when the external clock is active.

I don't know where it come from.

PS: I will check it with 33R1B6

[Wrenches of Death]

I'd call the problem with mine a MAJOR stability problem. The output of the generator wildly fluctuates in frequency when using an external clock. It is stable with the internal clock. A frequency counter on the output to attempt to measure it is useless. I didn't bother to put a scope on it.

Reinstalling the previous firmware release cured mine. Your mileage may vary.

Hopefully this will get fixed in the next release.

WoD

[james38]

You are absolute right.
These are the results with FW: 1.01.01.33R1B6

The noise level is now greatly decreased.
I am only have harmonics in 5Mhz steps.

I agree with you.
Hopefully it will be fixed in the next release.

PS:
The harmonics has fluctuation in his value.
They are every 5Mhz you only seen a part of it in the picture.
It goes from base level upon to the values in the screen.
An it ends at 60Mhz.

[Wrenches of Death]

We have to thank "Ringmodulator" for this temp fix. He was the first to post that the problem disappeared with the firmware rollback.

I didn't think of trying the spectrum analyzer. I've only had it a few days and am still in the "crawling" phase. 

Hopefully, in a few weeks I'll advance to toddler!

WoD

[rf-loop]

I have checked the reference signal separately but it is ok an clean.
The input level is about +6dbm. I have also generate a signal with an SSG3021x to check as an alternate but see the same result.

Just sidenote.
About input level 6 dBm.
If it is real 6dBm (4mW) it means around 12.6 Vpp in this case because reference input impedance is 5000 Ω (only reference output impedance is 50 Ω).

To avoid confusion, my opinion is that it is not good to tell levels using power unit dBm if impedance is not clear typical 50 Ω for avoid mess.
Now if you set 50 Ω output impedance generator level to 6dBm but your load impedance is other than 50 Ω then need calculate voltage level. Because reference input level is specified as voltage levels.

If you have terminated (example using 50ohm feed thru) this SDG reference input, then with generator (50 Ω)  6 dBm level there is roughly 1.25 Vpp and this level  is below specified minimum 1.4Vpp.


If you have 50 Ω generator level set for 6dBm and this is connected to Hi-Z load (or in this case 5 kΩ) its level is around 2.5Vpp and naturally this is ok level to SDG reference input. Totally "fun" thing is that there is not specification (in SDG data sheet) for maximum level at all.

I use still FW version 33R1 because after it have not happened anything useful. I will keep this 33R1 version until Siglent do some useful FW rework for real improvements and bug fixes and without generating new errors.

[nedeb]

I'm having difficulties with my SDG1302X (running firmware 1.01.01.33R3) getting a sync signal for the AM modulating frequency.

I'm generating a 5MHz sine carrier, amplitude modulated (internally) with a 1kHz sine, and enable the "Sync" option in the Utility menu. The manual says "For AM, [...], the frequency of the sync signal is the modulating frequency". But instead, when I connect the aux out to the scope I get pulses at the carrier frequency.

It seems the manual is not quite up to date: it describes only the sync types CH1 and CH2, but the Sync menu offers CH1, CH2, MOD-CH1 and MOD-CH2. But selecting CH1 or MOD-CH1 makes no difference, I always get the carrier frequency.

Am I doing something wrong? Did someone manage to get a modulation frequency sync from the aux out?

[Mortymore]

There's any way to take a screen capture from SDG1000X?

I want to place in a report the wave images and parameters set in both channels of the SDG, and it would be rather nice to have them screen captured instead of some photos taken of the SDG screen.

The EasyWave and EasyWaveX software seems that can only import waveforms from a scope (Siglent), and though I had built myself the data waveform used for this work, I would prefer a screen capture with all the info gathered in one nice image.

Thanks

[Hexley]

There's any way to take a screen capture from SDG1000X?

Check out JHenderson's work at https://www.eevblog.com/forum/testgear/visa-scanscreen-capture-utility/msg3947497/#msg3947497, or ColoradoRob's work at https://www.eevblog.com/forum/testgear/the-siglent-sdg2042x-thread/msg2346945/?topicseen#msg2346945.

[Mortymore]

Thank you both, Hexley and jhenderson0107 for his work

Couldn't manage to get it work from TCP/IP but it worked from USB

[nedeb]

Update on the freeze issue when controlling through USB: this is fixed with the latest pyvisa-py version (0.6.0)!

[BillyO]

I dont think, that there is an issue with my setup, because with FW 1.01.01.33R1B6 this problem is not there and the unit works as expected, also above 30MHz.

There is a chance, that it is related to the hardware version.
My unit is 02-01-00-24-00 and identifies as SDG1062X. It was born as SDG1032X.

For what it's worth, mine has the same hardware revision. I just installed the older B6 firmware revision and it's stable as hell. I'd bet money that this is a firmware "issue" with the R3 revision.

thanks!
WoD


.

I have to agree.

Now that I need to use an external reference I got some very strange results.  Connecting it up to my SDS2104XP I saw continuous glitching.  Back dating the firmware from 33R3 to 33R1B6 fixed the issue.

Bottom shot is normal, top shot is what you get 1 out of every 5 captures.


Is it possible Siglent are unaware of this?

PS:  How do you post images so they appear full sized and in-line?

[james38]

I have to agree.

Now that I need to use an external reference I got some very strange results.  Connecting it up to my SDS2104XP I saw continuous glitching.  Back dating the firmware from 33R3 to 33R1B6 fixed the issue.

Bottom shot is normal, top shot is what you get 1 out of every 5 captures.


Is it possible Siglent are unaware of this?

PS:  How do you post images so they appear full sized and in-line?

Look at my posts above.
https://www.eevblog.com/forum/testgear/siglent-sdg1000x-waveform-generators/msg4434094/#msg4434094
https://www.eevblog.com/forum/testgear/siglent-sdg1000x-waveform-generators/msg4434403/#msg4434403

I can acknowledge the issues with 33R3 when an external reference is connected.
I am also still on  33R1B6 because I always use an exernal refernce on all my devices.

As rf-loop says in the following post you must aware that the impedance for external In ist 5000 Ohm NOT 50 Ohm as expected.
I was a bit suprised and checked the datasheed in detail. He was right.
The +6db Level that I mentioned are measured directly from the GPSDO.
I forgot to say that I had a 6 db attenuator on it.
But anyway when the input impedance is not 50Ohm the level is not the measured level.

I has then tested it again with lower levels but the results are the same.
At least I can rule that out for myself that the cause is not the Clock-input level.
However, the minimum voltage must be at least 1.4V Vpp like the datasheed said.

With FW 33R1B6 i have no problems.

I can't say if it will ever be fixed or not.
I hope they will do ... sometime. ;-)
But I think it is known to Siglent and was also reported there.