Siglent SDG6000X series 200-500 MHz AWG's

[TurboTom]

Hi.
SDG6022X allows different digital modulations on both channels at the same time from internal source?
E.g. FSK 1k bp/s with internal data on CH1,
and FSK 10k bp/s with other internal data on CH2.

Or:
E.g. FSK 1k bp/s with internal data on CH1,
and PSK 1k bp/s with other internal data on CH2.

and other variations.


Thanks you.

Seems like any combination of modulations is possible, only I/Q modulation is available as a "single option" (obviously since both channels are required for this). Here's your inquiry of FSK and PSK. For FSK, you select the "shifted" frequency and the modulation rate (internal source) which may be up to 1MHz which appears to be common for all "simple" modulation schemes. Analog modulations like AM, FM, PM, DSB also allow arbitraries as modulation sources, thus it should be possible to even generate an FBAS-modulated UHF TV signal (didn't try). Might be boring to watch, though... 

[oz2cpu]

SDG6000 can play back wave files, so far so good
did anyone have a file, for a PAL composite test picture, colorbars, sharpness bars and the whole thing ?
interlaced and all ?
will it be too large/detailed for its wave memory ?

[fasset]

Seems like any combination of modulations is possible, only I/Q modulation is available as a "single option" (obviously since both channels are required for this). Here's your inquiry of FSK and PSK. For FSK, you select the "shifted" frequency and the modulation rate (internal source) which may be up to 1MHz which appears to be common for all "simple" modulation schemes. Analog modulations like AM, FM, PM, DSB also allow arbitraries as modulation sources, thus it should be possible to even generate an FBAS-modulated UHF TV signal (didn't try). Might be boring to watch, though... 

Ok, thanks.
I have read the whole thread.
You have mentioned a lot of problems and bugs here and also you have
personal experience with comparison to other AWG units.

Most of them are fixed (f.e. glitches).
Previously you said you have to verify waveform on oscilloscope every time when you changing setup.
It is still actual problem? Would you buy SGD6000x model again? It is reliable and are you using it for a now?

Not many AWG have also IQ option which can be useful in my future projects so Siglent is potentially good option.

[TurboTom]

Let's put it like this:

The performance (available frequency range, rise time, I/Q modulation -- though not with an arbitrary external data source), if you consider hacking and thus the price for the SDG6022X, is quite respectable. The previously troublesome incomplete initialization issue after some cold-starts or mode changes and thus the requirement to verify the integrity of the waveform has been addressed by Siglent in one of the recent firmware updates, so that's sorted.

So if you particularly need the high bandwidth or the I/Q functionality, go for it!

U/I of the instrument is mediocre at best, showing its age (heritage of the SDG1000X series), the encoder's performance is less than stellar and the display with the resistive touchscreen has a pitiful contrast. The touch functionality is not really worth talking about, it appears to be a reluctantly implemented afterthought, same as the frequency counter function.

If you're just in the market for "universal" AWG and the high bandwidth and I/Q options aren't requirements, I'ld get a different instrument anytime and skip Siglent products entirely! They have their strong points in o'scopes and RF gear but their "mature" AWG series isn't among them. Don't know about the new SDG7000A, though. But that's playing in a different league, also price-wise.

Would I buy it again? Clearly, NO!

[2N3055]

Let's put it like this:

The performance (available frequency range, rise time, I/Q modulation -- though not with an arbitrary external data source), if you consider hacking and thus the price for the SDG6022X, is quite respectable. The previously troublesome incomplete initialization issue after some cold-starts or mode changes and thus the requirement to verify the integrity of the waveform has been addressed by Siglent in one of the recent firmware updates, so that's sorted.

So if you particularly need the high bandwidth or the I/Q functionality, go for it!

U/I of the instrument is mediocre at best, showing its age (heritage of the SDG1000X series), the encoder's performance is less than stellar and the display with the resistive touchscreen has a pitiful contrast. The touch functionality is not really worth talking about, it appears to be a reluctantly implemented afterthought, same as the frequency counter function.

If you're just in the market for "universal" AWG and the high bandwidth and I/Q options aren't requirements, I'ld get a different instrument anytime and skip Siglent products entirely! They have their strong points in o'scopes and RF gear but their "mature" AWG series isn't among them. Don't know about the new SDG7000A, though. But that's playing in a different league, also price-wise.

Would I buy it again? Clearly, NO!

Well,

while I also complained to slow responding encoder, I would disagree with "clearly NO!".

Screen is touch screen, and resistive type. Very good ? No. But I don't use it much. Buttons are quite quick.
So I just disregard touch screen. It is not a primary U/I anyways.
Honestly, any touch screen that is so small won't be good U/I for touch, especially if you don't keep on the desk right next to you. Screen contrast is not best but personally never had problems reading.

And yes bandwidth and I/Q modulation are very important features. Very much so.

So to me it is more that it could have been so much better if it was made today, differently.. But that would be different product then. So a moot point at this time..

None of the stuff you mentioned would be "clearly NO!" to me.

[oz2cpu]

Would I buy it again?

YES, i love it..
True there are a few minor bugs, that is a bit frustrating they dont get solved
most of it should be very simple to solve for the sw people involved.
i just learned to work a round the bugs..
again : PLEASE siglent... fix it.. release a new sw for us, we are ready for it

[TurboTom]

@fasset - and here you've got it, three individuals, three different attitudes  .

Point is, if you need the performance, there's no other choice in its price range, and despite its shortcomings otherwise, you will get the job done with it.

After I received two R&S SM300 signal generators (albeit repair jobs) for quite little money, my SDG6000X gets rarely taken off the shelf; the I/Q function may be nice to have but as yet I didn't have to use it for paid jobs. As a "general purpose" AWG I was quite surprised how well instruments form other manufacturers perform that have the touch screen feature integrated in their design right from the start.

In the end, you're the one who knows best what features you need and what amount of "U/I comfort" you expect. So your milage may vary...   

[Plasmateur]

I was able to bandwidth upgrade my SDG6022X to SDG6052X using this script

Code: [Select]

import hashlib
# Replace this with your SN
SN = 'SDG00000000000'
Model = 'SDG6000X'
otheropt = ('500M' )
hashkey ='5zao9lyua01pp7hjzm3orcq90mds63z6zi5kv7vmv3ih981vlwn06txnjdtas3u2wa8msx61i12ueh14t7kqwsfskg032nhyuy1d9vv2wm925rd18kih9xhkyilobbgy'
def gen(x):
    h = hashlib.md5((
        hashkey + (Model+'\n').ljust(32, '\x00') +
        opt.ljust(5, '\x00') +
        2*((SN + '\n').ljust(32, '\x00')) +
        '\x00'*16).encode('ascii')
        ).digest()
    key = ''
    for b in h:
        if (b <= 0x2F or b > 0x39) and (b <= 0x60 or b > 0x7A):
            m = b % 0x24
            b = m + (0x57 if m > 9 else 0x30)
        if b == 0x30: b = 0x32
        if b == 0x31: b = 0x33
        if b == 0x6c: b = 0x6d
        if b == 0x6f: b = 0x70
        key += chr(b)
    return key.upper()
for opt in otheropt:
    print('{:5} {}'.format(opt, gen(SN)))
However I don't know the SCPI command for the IQ upgrade. Has anyone figured this out?

[oz2cpu]

IQ unlock is done via frontpanel manual input

[DL2XY]

SDG6000 can play back wave files, so far so good
did anyone have a file, for a PAL composite test picture, colorbars, sharpness bars and the whole thing ?
interlaced and all ?
will it be too large/detailed for its wave memory ?

No problem, the channel bandwidth for PAL B is 7MHz. 20MS/s Samplerate is sufficent, that will give you a 10s videoclip with the available memory.

Walter

[tautech]

New firmware for SDG6000X models.

Version 6.01.01.36R1
27 MB
https://siglentna.com/download/27353/

Release notes
1. Control the IQ output amplitude to avoid distortion
2. Optimize phase noise when using external reference clock
3. Fix bug in FM modulation on square waveform

[TurboTom]

  after one year, I had hoped for more than that! Looks a lot like alibi action...

[Martin72]

I had hoped for more than that!

Are there still bugs "alive" actually ?

[TurboTom]

There are still many "usability issues" that could be eliminated with small modifications that would make the instrument much more pleasant to deal with.

1. Rotary encoder - stay on the digit! IMO, it's an absolutely horrible approach to jump the decimal when reaching the "end" of the adjustable range while incrementally adjusting a parameter. Example: I want to adjust a pulse width between 100µs and 2ms in 100µs steps by means of the rotary encoder. So I enter 1ms via the keypad and set the adjustable digit to the 100µs decimal. Now everything's fine as long as I don't accidentally "try to" reduce the pulse width below 100µs, which results in the adjustable digit to jump to the 10µs decimal and to stay there, even if I increase the pulse width again above 100µs. I'ld expect to have at least the choice (by a U/I configuration switch) to simply have the the decrement function stop there and notify me by means of a different "key tone" and/or a short popup that I tried to exceed the adjustable range. This approach would be desirable for all parameters that can be adjusted incrementally. The funny thing about this "feature" is that the decimal jumping stops when the cursor reaches the decimal delimiter! It's simply inconsitent.

2. Adjustments to one parameter affecting the settings of other parameters: In pulse mode, select 50Hz frequency and 10ms pulse width. Now increase the frequency (incrementally or via keypad) to 100Hz or above -- the pulse width will be automatically decreased to a value to permit generating the requested frequency with the preset parameters. It doesn't work in reverse -- increasing the pulse width isn't accepted, the frequency is kept as a "hard limit" for the pulse width. I simply expect a good lab instrument to give me the choice if mutually dependent parameters will affect each other automatically or if adjustment attempts beyond borders that will affect these parameters will simply be blocked. Once again, this could be enabled/disabled as a switch in the system configuration. There may be other places where this issue is present, I picked the example just for ease of understanding.

3. Selecting higher digits for adjustment with the encoder: If a low-digit value has been entered for a parameter, say 1Hz, it's simply impossible to chose a digit for adjustment higher than 1Hz. Why doesn't the user interface permit to select for example the 100Hz digit by simply adding zeros to the left of the 1Hz digit and permitting to increment this digit? Couldn't be simpler than that!

4. Built-in arbitrary selection: Waveform -> Page2 -> Arb -> Arb Type -> Built-in, then I can finally chose the group of built-in arbitraries and the waveform itself. These are six/seven menus and keypresses to select the waveform. For the first selection, okay, be it like that, but if I find the selected waveform unsuitable for the intended job and I'ld like to try a few others, I've got to go through this procedure over and over again. This instrument has got a touchscreen! Touching the waveform window takes me back to the lowest level of waveform selection, identical to pressing the "waveform" hardkey. In Built-in Arb mode, it would be much more intuitive to take the user back just to the built-in arb selection page. This would make a big difference in usability comfort. Make decent use of the touchscreen! Simply replicating functions that are already present as hardkeys is nonsense! Implement (more) shortcuts as it has been done with the output impedance function.

I must have posted the same or a similar write-up some time ago already but what the heck, maybe it's got to be repeated to make Siglent "hear" it. Maybe I'm only too critical, and most users are happy with the instrument the way it is, and I'm aware that different usage cases favor different features. My point is mostly related to "eyes-off" operation, i.e. the user makes adjustments while observing the behaviour of a DUT on a different instrument, so a parameter can be easily "twisted" out of range.

5. In conjunction with this, additional upper/lower "limits" menus for parameters like pulse width, frequency, modulation depth... as it's already (globally) present for the output amplitude (upper), would be a nice feature (maybe with a global enable/override for the additional limits).

I'm sure many of these points are just common sense and won't be too difficult to implement. And they would make the SDG6000X much more enjoyable to use and attractive to the customer. One will get the job done without having these issues fixed, but IMO it leaves the impression of an unfinished product.

Edit:
6. In harmonic mode, the instrument still doesn't permit to enter the amplitude of the harmonics in Vrms even though it's possible for the fundamental.  Why, Siglent, why? 

[hpw]

Is this FW already gone??

while not present on any download side... also link do not work!!

waiting for the FW for SDG2K since a year O:((((((

[tautech]

 
It's been pulled for some reason.
Will hunt the reason why.....

[hpw]

It's been pulled for some reason.
Will hunt the reason why.....

Any the answer is.... while at

May ask also when the SDG2K gets this square wave modulation fixed...

[oewean]

New firmware for the SDG6000X released today:
SDG6000X Firmware - V6.01.01.36R3 (Release Date 09.01.22 )

1. Control the IQ output amplitude to avoid distortion
2. Optimize phase noise when using external reference clock
3. Fixed bug in FM modulation on square waveform
4. Fixed bug that the output amplitude of square wave, pulse
wave and PRBS is smaller than the set value
5. Fixed bug that VPP was still displayed on the UI when control
the output amplitude in dBm by the SCPI command
6. Fixed the problem of abnormal pulse output when pressing
output in burst mode
7. Resolve potential key and touch non response issues
8. Fixed the bug that 128QAM cannot be demodulated

[hpw]

New firmware for the SDG6000X released today:
SDG6000X Firmware - V6.01.01.36R3 (Release Date 09.01.22 )

1. Control the IQ output amplitude to avoid distortion
2. Optimize phase noise when using external reference clock
3. Fixed bug in FM modulation on square waveform
4. Fixed bug that the output amplitude of square wave, pulse
wave and PRBS is smaller than the set value
5. Fixed bug that VPP was still displayed on the UI when control
the output amplitude in dBm by the SCPI command
6. Fixed the problem of abnormal pulse output when pressing
output in burst mode
7. Resolve potential key and touch non response issues
8. Fixed the bug that 128QAM cannot be demodulated

Mira, mira, 3. Fixed bug in FM modulation on square waveform

As I noted on SDG2K, as given FW from Siglent support SDG2000X_P37R2 to test did not worked properly!

To check  Issue using 10MHz square & FM, 500Hz FM, 50mHz deviation:

500Hz signal fades out & in 30 second ... 1 Minute interval. May start first with 200mH deviations...

So how does this works on SDG6K & new FW?

[strahd_von_zarovich]

How can we use this script? I connect the device using command window and telnet, and send these messages, but I can not get any progress. Am I doing something wrong? Thanks in advance.

I was able to bandwidth upgrade my SDG6022X to SDG6052X using this script

Code: [Select]

import hashlib
# Replace this with your SN
SN = 'SDG00000000000'
Model = 'SDG6000X'
otheropt = ('500M' )
hashkey ='5zao9lyua01pp7hjzm3orcq90mds63z6zi5kv7vmv3ih981vlwn06txnjdtas3u2wa8msx61i12ueh14t7kqwsfskg032nhyuy1d9vv2wm925rd18kih9xhkyilobbgy'
def gen(x):
    h = hashlib.md5((
        hashkey + (Model+'\n').ljust(32, '\x00') +
        opt.ljust(5, '\x00') +
        2*((SN + '\n').ljust(32, '\x00')) +
        '\x00'*16).encode('ascii')
        ).digest()
    key = ''
    for b in h:
        if (b <= 0x2F or b > 0x39) and (b <= 0x60 or b > 0x7A):
            m = b % 0x24
            b = m + (0x57 if m > 9 else 0x30)
        if b == 0x30: b = 0x32
        if b == 0x31: b = 0x33
        if b == 0x6c: b = 0x6d
        if b == 0x6f: b = 0x70
        key += chr(b)
    return key.upper()
for opt in otheropt:
    print('{:5} {}'.format(opt, gen(SN)))
However I don't know the SCPI command for the IQ upgrade. Has anyone figured this out?

[berke]

Just got an SDG6022X.  Overall I'm pleased with it but I noticed a slow ~3.68 ns drift between the sync output and the channel output.  The firmware is V6.01.01.36R3.  See below, I connected the Aux to the scope's trig, and I'm generating a pulse from channel 1.  The timing of the channel 1 output slowly drifts by about ±1.8 ns.  Same thing happens if I connect Aux to the scope's second channel so it's not a triggering issue with the scope.

[XerXes777]

What is the KEY Word to get the IQ Key?
I use this Script:
I try these words:
IQ, I/Q, IQWAVE, AWG, SDG6000X-IQ, SDG6000XIQ, WAVE
I got the 500M Key its works.

Code: [Select]

# Keygen program for Siglent oscilloscopes

import hashlib

# You get this by running "SCOPEID?" at the SCIP prompt and removing the dashes
SCOPEID = '0000000000000000'
# Replace this with your SN
SN = 'SDS00000000000'
# This is one of the four options below
Model = 'SDG6000X'
# 'SDS1000X-E', 'SDS2000X-E', 'SDS2000X+', 'SDS5000X', 'ZODIAC-'

bwopt = ('25M', '40M', '50M', '60M', '70M', '100M', '150M', '200M',
         '250M', '300M', '350M', '500M', '750M', '1000M', 'MAX')
otheropt = ('AWG', 'WIFI', 'MSO', 'FLX',
            'CFD', 'I2S', '1553', 'FG', '16LA')

hashkey = '5zao9lyua01pp7hjzm3orcq90mds63z6zi5kv7vmv3ih981vlwn06txnjdtas3u2wa8msx61i12ueh14t7kqwsfskg032nhyuy1d9vv2wm925rd18kih9xhkyilobbgy'

def gen(x):
    h = hashlib.md5((
        hashkey +
        (Model+'\n').ljust(32, '\x00') +
        opt.ljust(5, '\x00') +
        2*(((SCOPEID if opt in bwopt else SN) + '\n').ljust(32, '\x00')) +
        '\x00'*16).encode('ascii')
    ).digest()
    key = ''
    for b in h:
        if (b <= 0x2F or b > 0x39) and (b <= 0x60 or b > 0x7A):
            m = b % 0x24
            b = m + (0x57 if m > 9 else 0x30)
        if b == 0x30:
            b = 0x32
        if b == 0x31:
            b = 0x33
        if b == 0x6c:
            b = 0x6d
        if b == 0x6f:
            b = 0x70
        key += chr(b)
    return key.upper()

for opt in bwopt:
    print('{:5} {}'.format(opt, gen(SCOPEID)))
for opt in otheropt:
    print('{:5} {}'.format(opt, gen(SN)))

[TurboTom]

Just got an SDG6022X.  Overall I'm pleased with it but I noticed a slow ~3.68 ns drift between the sync output and the channel output.  The firmware is V6.01.01.36R3.  See below, I connected the Aux to the scope's trig, and I'm generating a pulse from channel 1.  The timing of the channel 1 output slowly drifts by about ±1.8 ns.  Same thing happens if I connect Aux to the scope's second channel so it's not a triggering issue with the scope.
(Attachment Link)

I guess that's the same behaviour I observed here -- see the description regarding the second photo of the post. It's all about rounding errors due to the finite accuacy of the internal math engine, combined wiht the waveform synthesis principle of these generators (called "True Arb" in this case, which basically refers to a dynamic sample adjustment at increments finer than the sync output can be adjusted). It may also relate to the oversampling principle utilized in siglent's generators, i.e. the ad-friendly 2.4GSa/s cut down to a real world's 600MSa/s with a few real benefits when generating sine signals, which basically get handled by the DAC internally. The total drift window of ~3.3ns suggests that the real internal sampling speed at those frequencies where a sync signal can be generated (up to 10MHz!), the sample frequency is further reduced by 2 (not that it would make any difference). I also more then once mentioned (and won't get tired of it... ) that Siglent's decision ot provide a single Aux connection on this series of AWGs, common for both channels, may not have been the brightest one...

Anyway, if you got this AWG for its I/Q and high frequency range capabilities, you probably made a decent choice. Enjoy your new instrument!

[tautech]

What is the KEY Word to get the IQ Key?
I use this Script:
I try these words:
IQ, I/Q, IQWAVE, AWG, SDG6000X-IQ, SDG6000XIQ, WAVE
I got the 500M Key its works.

It might be the SKU code here:
https://siglentna.com/product/sdg-6000x-iq-signal-generator-function/

[berke]

I guess that's the same behaviour I observed here -- see the description regarding the second photo of the post.

Yes that's exactly it.  Your A to B delay vs. time plot nails it.  I guess I should have read the whole thread before posting !

It's all about rounding errors due to the finite accuacy of the internal math engine, combined wiht the waveform synthesis principle of these generators (called "True Arb" in this case, which basically refers to a dynamic sample adjustment at increments finer than the sync output can be adjusted). It may also relate to the oversampling principle utilized in siglent's generators, i.e. the ad-friendly 2.4GSa/s cut down to a real world's 600MSa/s with a few real benefits when generating sine signals, which basically get handled by the DAC internally. The total drift window of ~3.3ns suggests that the real internal sampling speed at those frequencies where a sync signal can be generated (up to 10MHz!), the sample frequency is further reduced by 2 (not that it would make any difference).

Do you have a block diagram for this thing?

The drift is so regular I'm wondering if it can be fixed in the FPGA or at least "dithered away".  Maybe there are some frequencies where there is no drift?

Anyway, if you got this AWG for its I/Q and high frequency range capabilities, you probably made a decent choice. Enjoy your new instrument!

I bought it mainly to replace my old 10 MHz GW Instek analog function generator without breaking the bank too much (that thing can't stay on frequency for a second.)  So it's night and day.  Now should I have bought something else for the same price or less?  I considered some entry-level Keysight/Tek generators but right now I do need to look at ≥50 MHz-ish digital signals with some reasonably fast edges (say 3 ns), for which the PRBS output is handly.  The AUX sync output drift sucks but I can tap the channel or use the second channel as a sync output.

I also more then once mentioned (and won't get tired of it... ) that Siglent's decision ot provide a single Aux connection on this series of AWGs, common for both channels, may not have been the brightest one...

I agree a dedicated sync out connector would be better.  There should be a EU law mandating front-panel dedicated 50 ohm sync out connectors in all AWGs sold in the EU.  BTW the shape of the sync pulse is quite horrible and it refuses to output a sync signal above 10 MHz.