Siglent SDG6000X series 200-500 MHz AWG's

[TurboTom]

if anyone wants a waveform file that generates the standard Philips PAL TV picture
perfect and correct to the last detail, I got it :-)
one of my friends made it, he also shared a free tool so you can make your own TV pictures of what ever you like.

And the SDG6000 even allows RF modulation...

[2N3055]

Mike, you can also add custom noise waveforms. We did this when Martin wanted Pink noise. I don't remember any other details though.

We are seeing some strange behavior wrt the Noise output and wish to better understand what's going on. For example the Noise BW setting changes characteristics completely from the lower frequency where the -3dBV BW seems like ~500 times the Noise BW setting to where the BW setting and the waveform -3dBV agree somewhat at the higher frequency end. In between below 100KHz the Noise BW setting changes characteristics completely, even becoming non-monotonic!!

Understanding how the Noise File implemented and how it relates to the final output would be beneficial for our understanding. This "could be" a quite useful waveform for many applications but we are reluctant to use such because of our limited understanding of the Noise characteristics and implementation.

In addition the SDG2000X has a lower BW Noise limited imposed as 20MHz which doesn't make sense to us!!

So we seem to be in a fog wrt to the Noise implementations on the 6000X as well as the 2000X, hopefully someone will provide some guidance thru this Noise implementation fog we've encountered

Best

Mike,

First, for noise we setup Stdev and Mean.
Stdev is RMS of the noise with that BW. NPD will change with BW to keep constant Stdev.
Mean is average DC offset.

But you know all this.

I don't remember using noise with sub Mhz BW on SDG6000X before.

So I went and tried.

It works as expected, but only if you enter more than 100 kHz BW.
If you enter anything less, it is still 100kHz BW.

It is a bug.
Will report..

[mawyatt]

Thanks, now try 1Hz, 10Hz, 100Hz, then you'll see the 500X 3dB corner we mentioned!!

Best

[EvgenyG]

if anyone wants a waveform file that generates the standard Philips PAL TV picture
perfect and correct to the last detail, I got it :-)
one of my friends made it, he also shared a free tool so you can make your own TV pictures of what ever you like.

Please share and the tool as well

[oz2cpu]

here you go EvgenyG
please follow this link

https://blog.dzl.dk/2024/10/20/generating-loopable-pal-test-signal/

the file it generate simply upload and playback, the file name ends with some numbers
that is on purpose, since this file format dont contain correct playback speed,
we keep this info in the file name so you can much easier remember to put this number into the bit pr sec playback setting
now the picture will be correct.

[TurboTom]

Since the inverting function of the SDG6000X only affects the physical outputs and not the internal modulation link between the channels, you need to invert the composite video by modifying the parameters of the tool. An "invert" function would have been easily added after taking all the effort to write up such a great tool, but the work-around to modify the parameters as shown in the screenshot will just do the same trick.

Btw, this tool should work with all the AWGs that provide sufficient sample size playback (just short of 3MSa) in loopback mode at 35.46895 MHz sample clock. The sample generated by the tool just contains four half frames. The file generated by the tool can be renamed to "*.bin" or "*.raf" and directly imported to any of these AWGs as binary. As it seems, so far this nice gadget is only available for the PAL colour modulation scheme, so those who need NTSC or SECAM appear to be out of luck...

Attached you will find a photo where I used a Rigol DG800, albeit not working as a modulator -- this is beyond this instrument's capabilities.

Cheers and all the best,
Thomas


PS.: As you see I always try to find the the right Color CRT TV to match the size of the AWG's display...  I'm kind of a sucker for rare vintage miniature CRT gear. 

PPS: The apparently crappy picture quality of the TV test picture is owed to the spacial interference of the CRT's shadow mask with the resolution of the digital camera's sensor chip -- I already took the photo at a weird angle and rotated it back post-process, otherwise the moiré patterns would have been completely unbearable... 

[2N3055]

here you go EvgenyG
please follow this link

https://blog.dzl.dk/2024/10/20/generating-loopable-pal-test-signal/

the file it generate simply upload and playback, the file name ends with some numbers
that is on purpose, since this file format dont contain correct playback speed,
we keep this info in the file name so you can much easier remember to put this number into the bit pr sec playback setting
now the picture will be correct.

Yes,

when you load the file, switch AWG to TrueArb mode, and set Srate at 35468950 (35.468950 Msa/s).

Your friend is clever chap.. Nice work.
 
EDIT:  When file is created with the online tool,  there , on the top, you specify sample rate. Same sample rate needs to be entered in AWG for it to work.
Default is 35468950 , but as oz2cpu say, you can use other frequencies. Default creates very nice and clean signal.

[oz2cpu]

to Turbotom, about file size, and playback speed
we tried to use much less clock setting and it also works
i think 5 times color burst frequence was good and stable,
this will give you a very compact file,
and now it can be playback on much cheaper generators,

[TurboTom]

You are right, the size of the sample files is substantial. But since most of the test pictures are highly redundant, they can be pretty efficiently compressed. My file that I used for the modulated test picture shown above shrinks from 5.675MB to 150kB (attached if someone would like to experiment...). Reducing the sample frequency/count from eight to five times the colour burst frequency helps a little but doesn't make too much of a difference. Cutting the data width by half (16 bit to 8 bit) may work as well, but I think in binary mode, the AWGs expect 16 bits of data per sample. Anyway, most contemporary AWGs will cope pretty well with the original file size so for my own part, I'll leave it at that. 

Whatsoever, this tool is something I had been thinking (and dreaming) about for a long time but couldn't justify to spend the time on such a "toy project" -- moreover the involved programming work also isn't something that I'm particulary good at  . So I was very happy when you refered to your and your friend's project in the other thread some time ago and I enjoyed playing around with the tool and the available AWGs a lot... Thanks so much for sharing! 

Cheers,
Thomas

[oz2cpu]

you are welcome, we had so much fun digging deep into pal systems
and make the sw as a hobby project
please also download the page with all files behind it,
so you have a local backup, just in case the internet brake one day,
as we know not all pages stay up and online forever

[Performa01]

Thanks, now try 1Hz, 10Hz, 100Hz, then you'll see the 500X 3dB corner we mentioned!!

As 2N3055 already stated, there is something wrong in the SDG6000X. This has been reported.

Just to demonstrate that it is not a problem for ALL Siglent AWGs, I've tried it with an SDG7000A. Just like the SDG6000X, noise bandwidth can be set as low as 1 mHz (millihertz, just to be clear, as some folks like to write mHz when they actually mean MHz...)

I have not the patience and possibly also not the means to test a single digit mHz bandwidth, but attached is a screenshot for 500 mHz noise bandwidth and 50 mV AC-RMS level.

First measured with the new SDS5000X HD and 20 MHz physical 1st order bandwidth limit.
Second measured with the SDS7404A and 125 Hz bandwidth limit by a FIR filter (for multi GHz DSOs, physical low frequency LPF are not feasible anymore).

We can see, that the 125 Hz FIR LP gives us a noticable lower noise floor on the 4 GHz machine than the physical 20 MHz 1st order LP on the 1 GHz SDS5000X HD. Consider the 20 dB/div for the FFT...

[NickJL]

Hi all - we recently bought some Siglent SDG6022Xs, which are working very well for our needs apart from one issue. I am using it to apply arbitrary waveforms. I would like to upload a long list of waveforms to the internal storage and then switch between them using remote control. I can upload the waveforms (using pyvisa and SCPI) and then I can select them from the front panel of the instrument, but I can't then select them by remote control via visa. I can however select built-in waves. So    'C1:ARWV NAME,Duty46'    and      'C1:ARWV INDEX,167'    both work. but     'C1:ARWV NAME,mywave'    doesn't work, even though 'mywave' can be selected and applied on the front panel. I have the output on a scope, and it is doing exactly what it tells me it's doing - it just ignores attempts to select user generated waveforms.     'C1:ARWV NAME,"/Local/mywave.bin"'       also doesn't work. I feel like I'm not giving it the right format for the wave name, but according to the programming manual I'm doing the right thing. I also cannot use waveforms that are on a USB stick programmatically, although again this works from the front panel.

Firmware version: "6.01.01.37R6"
Hardware version: "03-00-00-81-39"

Any advice on how to solve this problem would be very welcome.

[tautech]

Welcome to the forum Nick.

Have you studied Cpt 11 (P85) in the User manual ?
https://int.siglent.com/u_file/download/25_03_11/SDG6000X_UserManual_EN02A.pdf

[NickJL]

Thanks very much for the welcome! :-)

I have read that - and it all makes perfect sense in that selection of waveforms from the front panel works fine. I can either upload a number of waveforms remotely from VISA or put them on a USB stick and then select them from the front panel. What I can't do is select one of those waveforms remotely. If I send it "*ESR?" after trying to switch arbitrary waveform, I get back '0' - ie none of the Status Event Register bits are set. So it doesn't hate the commands I send, but it doesn't actually take any notice of them either.

Maybe I should also say that when I upload a waveform with VISA using "<channel>:WVDT ..." It uploads and stores the data then immediately sets the output to that waveform. So I can set the output remotely that way - but of course it is not ideal to re-upload a long waveform every time I want to use it.

The waveform names I am attempting to use agree with those reported by "STL? USER", which are the ones I set when uploading with "WVDT".
i
I did not try specifying the full path for the USB files (ie "U-disk0/wave".) That's something to try tomorrow.

[tautech]

Thanks very much for the welcome! :-)

I have read that - and it all makes perfect sense in that selection of waveforms from the front panel works fine. I can either upload a number of waveforms remotely from VISA or put them on a USB stick and then select them from the front panel. What I can't do is select one of those waveforms remotely. If I send it "*ESR?" after trying to switch arbitrary waveform, I get back '0' - ie none of the Status Event Register bits are set. So it doesn't hate the commands I send, but it doesn't actually take any notice of them either.

Maybe I should also say that when I upload a waveform with VISA using "<channel>:WVDT ..." It uploads and stores the data then immediately sets the output to that waveform. So I can set the output remotely that way - but of course it is not ideal to re-upload a long waveform every time I want to use it.

The waveform names I am attempting to use agree with those reported by "STL? USER", which are the ones I set when uploading with "WVDT".
i
I did not try specifying the full path for the USB files (ie "U-disk0/wave".) That's something to try tomorrow.

Feel free to mail or call tomorrow Nick....your last initial gives strong clues to whom you are. 

There are further tips in the Programming manual in Chp 3.9 (P62)
https://int.siglent.com/u_file/download/24_06_07/SDG_Programming%20Guide_PG02-E05C.pdf

[KungFuJosh]

Siglent makes some great AWGs. I'm not sad about this acquisition.

Thanks,
Josh

[KungFuJosh]

More tests. More nice flatness too.

Oh well, time to sell the SDG2122X.

[KungFuJosh]

There appears to be a "Dead Zone" when sweeping around 275~300MHz. Same effect on both channels, no output when swept across 275-300MHz. Seems to be dead across 250 to 300MHz.

I would guess that it's a known bug, and eventually somebody will remember that and say so. 🤔🤔🤔

Ha! That's funny. Because I forgot about the known bug, and I have it too. For fun, I did a lazy manual sweep, and the signal is as expected outside of the stupid sweep bug.

Thanks,
Josh

[Ulrich.G]

New firmware 6.01.01.38R3  is available, adressing this sweep issue.

[KungFuJosh]

Woohoo! We can sweep again.

The PNG/BMP thing is interesting. Anybody find where to actually set that?

I held in the knob and it saved a PNG, which is great. But the on screen popup says PNG or BMP file saved. 🤣

[Performa01]

The PNG/BMP thing is interesting. Anybody find where to actually set that?

Utility -> System -> Page 3/3 -> Capt Type

[Martin72]

New firmware 6.01.01.38R3  is available

1. Support IQ waveform generated by customIQ of
SigIQPro;
2. Screenshot files support PNG and BMP formats;
3. MOD, SWEEP, BURST add RUN/Stop control function;
4. Support display of current output frequency during
frequency sweep;
5. Fixed the bug that the waveform is discontinuous when
the length of the sequence wave is not an integer multiple
of 16;
6. Enhance the sequence function, and the number of loops
of segment no longer occupies storage space multiple
times;
7. Delete the true arb mode of arb wave, and only keep the
DDS mode. The true arb mode is implemented by the
sequence (sequence with only one segment);
8. Solve the problem of abnormal output when the frequency
between 250MHZ and 300MHz during sweep;
9. Modify the default output polynomial of PRBS and support
programmable polynomial;
10. Sync output supports 50% duty cycle output mode;
11. IQ waveform information diaplay add data source
information;
12. Solve the problem of duty cycle is changed during
frequency change when square or pulse output;
13. Added a new DC output mode to solve the problem that
when changing the voltage, the output will jump to 0
before outputting. This method will reduce the output
accuracy;
14. Support amplitude sweep function.

Seven years after its release, it still receives this level of support—that's what I call exemplary model maintenance.

[famalex]

Less than a month ago, I bought such device and I don't regret this purchase. Nice instrument.

[Ulrich.G]

Just 94 days until christmas, one more thing for my wishlist - lowercase letters for the file browser keyboard...

[GerardF]

Looking into ARB's.
Can anybody confirm that the feature scripts still work on the SDG6000 ?