Siglent SDG6000X series 200-500 MHz AWG's

[TurboTom]

I wouldn't say the ringing (alias "undulation" if its "walking through") isn't the result of the scope's sampling engine. It's also observable with scopes of different, much higher sampling rates or even analog scopes. I re-read the previous message and realized that I might have misinterpreted it. I'm just not used to the term "Sinc Interpolation" in combination with AWGs. It's rather some kind of signal reconstruction by filtering, so to speak "reverse Sinc interpolation". The ringing is clearly related to the AWG's reconstruction filter that's optimized for  slopes of 2ns rise time, see the first screenshot (ref trace symmetry 0%, ch1 trace 1.2% symmetry to get in the 2ns rise time ballpark). This peculiarity had been discussed here in several contributions. You can blame Siglent to permit slopes that are faster than what the hardware has been designed for, but IMO that's a problem one can live with, you've just got to keep it in mind. Rarely the DUT is fast enough for this ringing to cause any trouble.

The reported level inaccuracy would make me much more nervous. I didn't find anything like this on my AWG, see the second screenshot. I blame the very tiny amplitude droop towards higher frequencies to the vertical amp characteristics of the scope I used ("pimped" Rigol DS2000A). With an HP RF power meter, I found the AWG's level accuracy to be better than 0.1dB over the whole frequency range.

[oz2cpu]

thanks once again DL2XY
YES correct, I double checked, the ringning is possible to almost remove, by change the symetry values to 1.1%
now the walking phenomenon is also almost gone

[DL2XY]

I wouldn't say the ringing (alias "undulation" if its "walking through") isn't the result of the scope's sampling engine. It's also observable with scopes of different, much higher sampling rates or even analog scopes. I re-read the previous message and realized that I might have misinterpreted it. I'm just not used to the term "Sinc Interpolation" in combination with AWGs. It's rather some kind of signal reconstruction by filtering, so to speak "reverse Sinc interpolation". The ringing is clearly related to the AWG's reconstruction filter that's optimized for  slopes of 2ns rise time, see the first screenshot (ref trace symmetry 0%, ch1 trace 1.2% symmetry to get in the 2ns rise time ballpark). This peculiarity had been discussed here in several contributions.

The unit calculates intermediate values from a 500MSmp/s synthesis engine to the 2.5Gsmp/s DAC rate via sync algorithm. Why won't you call it "Sync-Interpolation" ?
The reconstruction filter is hardware (low pass filter) behind the DAC, and his rolloff is about 1Ghz. It is very unlikely to cause ringing.

You can blame Siglent to permit slopes that are faster than what the hardware has been designed for, but IMO that's a problem one can live with, you've just got to keep it in mind. Rarely the DUT is fast enough for this ringing to cause any trouble.

I very like siglent to let the user go to the (nyquist) limit, you don't have to use it, but you can if necessary!

The reported level inaccuracy would make me much more nervous. I didn't find anything like this on my AWG, see the second screenshot. I blame the very tiny amplitude droop towards higher frequencies to the vertical amp characteristics of the scope I used ("pimped" Rigol DS2000A). With an HP RF power meter, I found the AWG's level accuracy to be better than 0.1dB over the whole frequency range.

I think oz2cpu may have not realized that the AWG automaticaly reduced the output level due to high frequency.

[tautech]

I think oz2cpu may have not realized that the AWG automaticaly reduced the output level due to high frequency.

Yes, clearly outlined in the opening post of this thread. 

Again here for those that have missed it in the datasheet:

10 Vpp 40 MHz ~ 120 MHz ( included ), HiZ load
5 Vpp 120 MHz ~ 160 MHz ( included ), HiZ load
3 Vpp 160 MHz ~ 350 MHz ( included ), HiZ load
1.28 Vpp above 350MHz, HiZ load

[TurboTom]

I think oz2cpu may have not realized that the AWG automaticaly reduced the output level due to high frequency.

Yes, clearly outlined in the opening post of this thread. 

Again here for those that have missed it in the datasheet:

10 Vpp 40 MHz ~ 120 MHz ( included ), HiZ load
5 Vpp 120 MHz ~ 160 MHz ( included ), HiZ load
3 Vpp 160 MHz ~ 350 MHz ( included ), HiZ load
1.28 Vpp above 350MHz, HiZ load

oz2cpu clearly reports that he used 1Vpp for his test, which the SDG6000X should be easily capable of supplying into a 50 Ohms load in the 200MHz range (350MHz actually).

[oz2cpu]

he he, YES this time i did read a bit more before "complaining"
I think the bw limitation is kicking in a little bit too erly for the 200MHz version,
this explain why the orher bigger versions works so much better.

I have seen others test the bw, much more often done using much less output voltage,
maybe that is a trick to get the response much more flat in %
but a more realistic senario should be to choose a voltage just a litte bit under the specified at max frequency,
that is why i picked 1v for all my tests, i hope you agree this is most usefull,

I really like the way you set the output voltages on this unit,
either low and hi points,
or a hi and offset, the normal way on all other brands i had,
so the two modes to set this gets a big thumbs up from here.

[TurboTom]

The hardware is identical on all the SDG6000X models and you can "improve" your AWG into the top-notch version as well, just start reading this thread from round about the third page. Mine also started life as a 6022X... 

When I did initial level tests some two and a half years ago, I didn't find the droop on my -then original- instrument that you're reporting now. Thus, I feel a little worried about that. You may want to check for resonance via the ground / PE lead. Also, make sure that the shield/screen contacts of your BNCs are properly spring-loaded against the sleeve of the sockets.

[oz2cpu]

thanks alot tom, i did try a few different BNC cables, and also the one that came with the unit (apear to be of good quality)
i could almost not detect any difference from cable to cable,
DL2XY´s suggestion to fine trim the slope, solved the ringing, and the tech explain was spot on.

about upgrading : i did search alot but did not find a solution that i could understand and perform
it is in tiny bits and a bit cryptic written, so only the expert who writes and understand it himself :-)
so i gave up for now.  EDIT it is the HOW to enter scpi commands, and how to write the whole command i need correctly,
what PC software do i need ?

[tautech]

For any Siglent AWG's the free EasyWaveX SW is a powerful tool.
https://int.siglent.com/download/softwares/?CateIdss=2

Just be sure to study the install requirements especially before attempting to connect via USB to a Windows based PC.

[oz2cpu]

thanks alot, i am done with the NI-VISA seleced yes to it all

I had the idea it was EasySDL i should and use for SCPI access ?

[tautech]

thanks alot, i am done with the NI-VISA seleced yes to it all

I had the idea it was EasySDL i should and use for SCPI access ?

EasySDL is the SW package for SDL1000X electronic loads.

Look up the SDG Programming guide for SCPI access:
https://int.siglent.com/u_file/document/SDG_Programming%20Guide_PG02-E04A.pdf

[oz2cpu]

PROBLEM SOLVED, YOU SIMPLY USE PUTTY to telnet via ETH,
find the IP number and input the correct port, in this case it was 5024
the unit will say hello with its type number so you know the connection is ok..
now it deliver 500MHz and 1nS risefall,

[oz2cpu]

idea for a feature:
in sinewave mode, 50 ohm load mode, offset 0, to be able to select dBm instead of voltage.

[DL2XY]

too late,
it's already implemented!

Just choose Amplitude, type in the number and press the hotkey "dBm".

[rf-loop]

Also there can toggle between units using "Amplitude" long press. So if user enter 0dBm and it display 0dBm then long push Amplitude and next it display 632.6mVpp and next 223.6mVrms. (note: it display it without offset even if there is offset ./  )

But about level flatness. SDG6022X is specified for ± 0.3dBm

Here sweeps using 7.50, 6.00, 3.00 and 0.00 dBm output level setting. Sweep range 2MHz - 200MHz.
Level absolute accuracy is not quaranteed here but it give quite good imagine about flatness. Also there is (quite good quality) 70cm BNC - N cable what attenuation is not corrected. With some experience it can assume flatness is least inside ± 0.15 dBm if use calibrated level meter and measure directly from SDG output port (even power total include all spurs and harmonics).

Flatness is not in "State of Art" class, not even "High End" class but it very good.

@oz2cpu some previous message is used just level as example 1v   
Example SDG6kX can handle 200MHz 1.5Vpp or 0.530Vrms @50ohm

It is good to use right values for tell level and in some cases including impedance if talk example SDG displayed level. Vpp, Vp, Vrms or dBm etc just for avoid any misinterpretation.
And when talk SDG displayed levels also load impedance setup is important if not fully self-evident, it is not same if it is set for 50, 61 or 77 or 93, or 128 ohm or what ever user have set (setup range is free between 50ohm - 100kohm or Hi-Z. 


SDG6022X just as factory new out from box. Output 50ohm, set for 50ohm load, 50ohm chinese high quality coax 70cm, and 50ohm input and all these have some ± tolerances.

[oz2cpu]

SUPER DL2XY ha ha YES dbm is there.. stupid me :-)

Thanks rf-loop your curves reveal that level flatness is not a compression level thing
and it is infact not a big deal, as long as you are a bit under max out level
please add to your curve / measurements
if the used unit for the experiment is stock 6022 ? or upgraded ?
since i believe it do makes a lot of difference, if you scroll back a lot of pages, you find curves of stock versus upgraded,
and the upgraded perform much better, more flat, even in the 0-200Mhz range

[rf-loop]

SUPER DL2XY ha ha YES dbm is there.. stupid me :-)

Thanks rf-loop your curves reveal that level flatness is not a compression level thing
and it is infact not a big deal, as long as you are a bit under max out level
please add to your curve / measurements
if the used unit for the experiment is stock 6022 ? or upgraded ?
since i believe it do makes a lot of difference, if you scroll back a lot of pages, you find curves of stock versus upgraded,
and the upgraded perform much better, more flat, even in the 0-200Mhz range

If you read this message just before image there is "SDG6022X just as factory new out from box. Output 50ohm, set for 50ohm load, 50ohm chinese high quality coax 70cm, and 50ohm input and all these have some ± tolerances."
Also in image clearly read SDG6022  if it is modified of course it is also told in image as have been practice mostly in my messages if this information is anyhow important/meaningful in context.

and sidenote: I know every single message in this thread.

[oz2cpu]

oh sorry.. that one important line : "SDG6022X just as factory new out from box" i did not see, not your fault :-)
great to see yours perform this good, or your test method / equipment is better.

[Performa01]

With regard to amplitude accuracy and flatness, I want to remind you of this old reply #162, showing the response of my (original) SDG6052X and comparing it to an older high performance signal generator.

https://www.eevblog.com/forum/testgear/siglent-sdg6000-series-awg_s/msg2621457/#msg2621457

At the sweet spot of -30 dBm, I got a frequency response of 100 kHz - 500 MHz  +0.06 / -0.04 dB.
That's not only excellent, I'd call it even spectacular.

[rf-loop]

With regard to amplitude accuracy and flatness, I want to remind you of this old reply #162, showing the response of my (original) SDG6052X and comparing it to an older high performance signal generator.

https://www.eevblog.com/forum/testgear/siglent-sdg6000-series-awg_s/msg2621457/#msg2621457

At the sweet spot of -30 dBm, I got a frequency response of 100 kHz - 500 MHz  +0.06 / -0.04 dB.
That's not only excellent, I'd call it even spectacular.

Yes in this individual unit it can say, it is spectacular.
Btw, usually with RF work do not need this accuracy. If we look example older HP some "State Of Art" class RF generators specifications... often decimal point can shift right. Some time ago I have in my homeland "lab" some this kind of RF gen where also was high accuracy flatness/accuracy option. And it was not as tight as you measured SDG6kX.  But example one very important thing was that it did not leak RF so it was also suitable to RF isolated "silent room".

Btw, how did you eliminate spurs and harmonics from this level measurement... (for avoid fun mess etc... of course in practice totally nonsense with these levels of harmonics and spurs, but your accuracy was amazing) /

[rf-loop]

SDG6022X just as factory new out from box. Output 50ohm, set for 50ohm load, 50ohm chinese high quality coax 70cm, and 50ohm input and all these have some ± tolerances.

For compare. Same individual SDG mod temporarily just for this one single test and returned then back to original. Just for check change (for better compare. just same 2 - 200MHz sweep, and only two level due to lack of time. (These sweeps are s l o w !)

...if you scroll back a lot of pages, you find curves of stock versus upgraded,
and the upgraded perform much better, more flat, even in the 0-200Mhz range

I  can not see noteworthy difference in this tiny test.

[oz2cpu]

1
if you need to stack your units close, the handles are in the way
the manual say pulll hard, but really ?? how hard ?
mine is stuck, or am i weak, or am i affraight to damage something ?

2
if you want to design your own equipment, some special stuff that just dont exist,
and you like to use a case that is as identical as possible, and also with the same front back protective plastic and so on,
is that possible to get somewhere ?
i am like very close to get the cheapest unit, and rip out the electronics just to get the case,
but it is a bit against my good feeling to do it that way, if i could only get an empty case ?
or maybe a defective unit ?

[rf-loop]

1
if you need to stack your units close, the handles are in the way
the manual say pulll hard, but really ?? how hard ?
mine is stuck, or am i weak, or am i affraight to damage something ?

They come out very easy if all is ok. Just turn 90º up. Pull axially both sides (or one side first then after it is out then other) as in image and same time when pull (not too hard for avoid this plastic teeth damage) when you pull also bit change angle ± some degree because angle tolerance is small.

[oz2cpu]

ok i try again tonight, i gave up last night, it was feeling wrong, like i needed way too much force

[TurboTom]

As rf-loop explained, it's important to have the handle in the right position to remove it. Put the generator flat on a table in it's "ready to use" position, rotate the handle straight up and pull at the joints gently, maybe slightly swinging the handle fore and aft, until they unlatch. It's really easy.

I'm not aware of a source for separate casings for Siglent gear. You may want to check their spare parts list and send an inquiry. But since the enclosures of all of Siglent's AWGs are identical, you may get the "bottom of the barrel" unit (SDG1032X). Maybe you can get a broken one second hand for less to salvage the casing??