Products > Test Equipment

The Siglent SDG2042X Thread

<< < (370/372) > >>

Martin72:

--- Quote ---Geez, fellas.
--- End quote ---
;D

I'm trying to imagine what you're doing and how it's set up.
Maybe like I sketched it out?

KungFuJosh:
No, I was being dumb and trying to input the AM503B directly from the SDG2122X output.

Instead, I pretended it was an actual DC supply, and used a 100Ω resistor in series with the AM503B (via voltage adapter - no probe), and a DMM. Now it's all working as expected. I knew I was doing something dumb. 😉

Now I can measure ~1.00071mA on my SDM3065X, and the AM503B is outputting ~49.948mV to the SDM3055X-E (50Ω terminated). That's exactly where I need to be to calibrate.

Thanks,
Josh

DavidPlatt:
I've been playing with my SDG2082X recently.  My goal has been to write a Python script which can generate a waveform for an FM-stereo composite test signal, download the waveform into the SG, generate the signal, and then use it to frequency-modulate a carrier (either 10.7 MHz for the IF, or directly up to the RF FM-carrier frequency I want).

I've got it working, but ran into some snags which seem to suggest that there are still some "can't do that via SCPI" limitations in the firmware, as of the version I'm using (37R3), and apparently some bugs.

What's missing:  I haven't found a command to set the modulation for C1 (or C2) to "ARB, with a user-specified waveform".  It's possible to modulate with a built-in ARB waveform (the MDWV command takes an INDEX parameter) but there's apparently no way to give a modulation waveform name via SCPI. 

What's buggy:  if I use the front panel to select FM via internal ARB, and then go into the ARB waveform selection and select "user waveform" and choose a waveform file I've downloaded, the SG loads this waveform and accepts it, and starts doing the frequency modulation as expected.  However, if I then send

  C1:MDWV?

and look at the parameters the SG says it's using, they're wrong.  Specifically, it will still say "INDEX,nn", giving the waveform index number of the most recent built-in waveform I've used, and there's no indication at all that I've switched over to a user-generated waveform (let alone which one it is).

So, it looks as if the "allow SCPI to specify the ARB waveform to use for modulation" feature implemented back in 2019 (in 35R3B1) was rather incomplete... neither the setter or the getter function deals with named user waveforms. 

Am I missing something?  Has this been fixed in more recent firmware?

The good news is that I was able to work around this by using the second channel.  Since 35R3 allows one channel to use the other as a modulation source, I've been able to configure C2 with the necessary ARB waveform, and then tell C1 to modulate using C2.  I've tested the resulting signal on my spectrum analyzer, and it looks just right, with a deep Bessel null at the expected deviation/frequency ratio. 

Although the SDG2082X won't do sine-wave output up in the commercial FM band (unless hacked, of course :-) ) I can frequency-modulate a square wave at 1/5 of the FM broadcast frequency I want, cut the deviation by 5:1, and my FM stereo receiver picks up the fifth-harmonic signal perfectly and the "stereo" light comes on and the tones come out of the correct speakers.

Just for the sake of elegance I'd love to be able to do this without tying up the second channel on the generator, and without having to do a bunch of the setup manually from the GUI.  I've been a bit reluctant to try upgrading firmware to 38R4 to see if the problem is fixed, since some people have run into the "upgrade fails and bricks the unit because there's a relict firmware file clogging the filesystem" bug.  I like this generator too much to want to risk bricking it.

tautech:

--- Quote from: DavidPlatt on November 20, 2024, 05:14:44 am ---I've been playing with my SDG2082X recently.  My goal has been to write a Python script which can generate a waveform for an FM-stereo composite test signal, download the waveform into the SG, generate the signal, and then use it to frequency-modulate a carrier (either 10.7 MHz for the IF, or directly up to the RF FM-carrier frequency I want).

I've got it working, but ran into some snags which seem to suggest that there are still some "can't do that via SCPI" limitations in the firmware, as of the version I'm using (37R3), and apparently some bugs.

What's missing:  I haven't found a command to set the modulation for C1 (or C2) to "ARB, with a user-specified waveform".  It's possible to modulate with a built-in ARB waveform (the MDWV command takes an INDEX parameter) but there's apparently no way to give a modulation waveform name via SCPI. 

What's buggy:  if I use the front panel to select FM via internal ARB, and then go into the ARB waveform selection and select "user waveform" and choose a waveform file I've downloaded, the SG loads this waveform and accepts it, and starts doing the frequency modulation as expected.  However, if I then send

  C1:MDWV?

and look at the parameters the SG says it's using, they're wrong.  Specifically, it will still say "INDEX,nn", giving the waveform index number of the most recent built-in waveform I've used, and there's no indication at all that I've switched over to a user-generated waveform (let alone which one it is).

So, it looks as if the "allow SCPI to specify the ARB waveform to use for modulation" feature implemented back in 2019 (in 35R3B1) was rather incomplete... neither the setter or the getter function deals with named user waveforms. 

Am I missing something?  Has this been fixed in more recent firmware?

The good news is that I was able to work around this by using the second channel.  Since 35R3 allows one channel to use the other as a modulation source, I've been able to configure C2 with the necessary ARB waveform, and then tell C1 to modulate using C2.  I've tested the resulting signal on my spectrum analyzer, and it looks just right, with a deep Bessel null at the expected deviation/frequency ratio. 

Although the SDG2082X won't do sine-wave output up in the commercial FM band (unless hacked, of course :-) ) I can frequency-modulate a square wave at 1/5 of the FM broadcast frequency I want, cut the deviation by 5:1, and my FM stereo receiver picks up the fifth-harmonic signal perfectly and the "stereo" light comes on and the tones come out of the correct speakers.

Just for the sake of elegance I'd love to be able to do this without tying up the second channel on the generator, and without having to do a bunch of the setup manually from the GUI.  I've been a bit reluctant to try upgrading firmware to 38R4 to see if the problem is fixed, since some people have run into the "upgrade fails and bricks the unit because there's a relict firmware file clogging the filesystem" bug.  I like this generator too much to want to risk bricking it.

--- End quote ---
Welcome to the forum.

You are 3 firmware releases behind current.
https://int.siglent.com/download/firmwares/?ProId=5

Check out the release and version notes to check if what you see has been solved.
Never bricked any unit with a FW update and done many 100's.....

DavidPlatt:
Well, firmware 38R4 doesn't seem to fill the bill, and it appears to have at least one new problem.

When I first start it up and enable FM modulation on Channel 1, its status reads out as

C1:MDWV STATE,ON,FM,MDSP,SINE,SRC,CH2,FRQ,100HZ,DEVI,75000HZ,CARR,WVTP,SINE,FRQ,10700000HZ,AMP,0.5V,AMPVRMS,0.17675Vrms,AMPDBM,-2.04251dBm,OFST,0V,PHSE,0

There's no NAME or INDEX shown.

If I use the GUI to manually enable ARB modulation, user waveform, and select the "Composite" waveform file I downloaded, it then shows

C1:MDWV STATE,ON,FM,MDSP,ARB,INDEX,0,SRC,INT,FRQ,100HZ,DEVI,75000HZ,CARR,WVTP,SINE,FRQ,10700000HZ,AMP,0.5V,AMPVRMS,0.17675Vrms,AMPDBM,-2.04251dBm,OFST,0V,PHSE,0

The "MDSP,ARB,INDEX,0" does show that it has switched over to using the arbitrary-waveform system as the source of modulation, and (perhaps) that it's not using one of the built-in waveforms.  There's still no indication as to which waveform was loaded, nor can I figure out any command option to force the loading of the waveform via SCPI.

The new bug:  GUI crash.

To reproduce:

-  Power up the SG with default settings
-  Unzip the Composite.bin from the attached ZIP, put it on a USB stick and copy to C: on the generator
-  Set frequency to 10.7 MHz, turn on channel 1 output.  Observe proper signal.
-  Select MOD.
-  Select FM.
-  Set deviation to 75 kHz.  Observe signal, confirm that the signal is being modulated by a built-in waveform.
-  Select SHAPE.  Go to page 2, select ARB.  Select STORED WAVEFORMS.
-  Select the Composite.bin you've previously copied to C:, press knob to select.
-  Confirm that the carrier is being modulated properly, and that the display says that SHAPE is ARB.
-  Select SHAPE.  Go to page 2, select ARB.
-  Nothing happens.

The GUI is now nonresponsive.  It'll confirm-click when you press ARB, and then stops responding.

At this point, attempting to connect to the interactive SCPI port (5024) fails with a "connection refused".  Ditto for port 5025.

The FPGA still seems to be running (modulation continues) but the firmware app seems to have undergone a controlled flight into terrain.

I have rolled back to 37R3 and confirmed that the crash doesn't occur.


Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod