Rigol DG1022Z or Siglent SDG2042X，it's hard to choice.

[uinika]

Is there any budy give some advice ? THX...

[Noy]

DG811 and hacked to DG992...

[bd139]

DG811 and hacked to DG992...

I'll second that.

[Mechatrommer]

Uni-T UTG962 $130 60MHz

[uinika]

THX for ur reminding，but can the DG812 easily hacked to DG992 ?

[bd139]

Yes easily but only on some FW revisions. I think you can downgrade. You need a magic USB stick an image of which is available on this forum. Then you plug the thing into your PC with USB cable and issue an SCPI command to it with Rigol's software to change the model to DG922. All the buttons and screen goes red after reboot and it's a DG992. I don't have a link to the thread handy but should be easy enough to find. Worked for me!

[Noy]

DG811 is enough you can save the few bucks up to DG812

Downgrade is possible regarding the forum thread here. Still waiting for arrival of mine.

[uinika]

Does the DG81x lose functions, precisions or another sth when I hacked and both of it have same Circuit ? 

[Noy]

You should read/look at this thread:

https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/

Conclusion above 70MHz DG811 is losing 2-3db against DG992. This is probably caused by missing calibration up to this frequency. Maybe we are in future able to calibrate it later up to the 100MHz.

But currently up to 70MHz everything is fine, above there are the 2-3db lost...

[KungFuJosh]

I have the Uni-T UTG932 and the Siglent SDG2042X. They're both good, obviously the Siglent is more powerful, with more options. It depends what you need/plan to use it for. If you have a Siglent scope like the SDS1104X-E, I would get the SDG2042X so you can connect them for Bode plotting. If you're into that kinda thing.

[uinika]

U said it 

I have a Siglent SDS1204X-C，I try to use it with Rigol DG1022Z but I found a problem when we try to exports & use .csv files.
Obviously, Siglent 1104X-C's .csv did not work with Rigol DG1022Z simply.... I'm crazy about it.

And another problem is that: A 25MHz square wave generated by Rigol DG1022Z occurs serious distortion in my Siglent 200Mhz bandwith oscilloscope. Just as below ... I am even distinguish sine and square wave...
Another friends recommended me to hack Rigol DG812，it is a option for me. But just as U say another option is Siglent SDG-2042X, because I ve already have a Siglent oscilloscope.

There are more questions here:

- What is the price(exclusive of tax) of your SDG-2042X in the United States ?
- What kind of a Square Wave generated by SDG-2042X in 200Mhz bandwith oscilloscope ?

[tautech]

You've said it 

I have a Siglent SDS1204X-C，I try to use it with Rigol DG1022Z but I found a problem when we try to exports & use .csv files.
Obviously, Siglent 1104X-C's .csv did not work with Rigol DG1022Z simply.... I'm crazy about it.

Did you miss this in the datasheet:

SDS1000X-E can control the USB AWG module or control an independent SIGLENT SDG instrument, scan a devices amplitude and phase frequency response, and display the data as a Bode Plot.

The same applies when uploading files to be regenerated into the AWG from the scope.
Both instruments need to be of the same brand.   

[tautech]

But just as U say another option is Siglent SDG-2042X, because I ve already have a Siglent oscilloscope.

There are more questions here:

- What is the price(exclusive of tax) of your SDG-2042X in the United States ?
- What kind of a Square Wave generated by SDG-2042X in 200Mhz bandwith oscilloscope ?

$499
https://siglentna.com/waveform-generators/sdg2000x-series-functionarbitrary-waveform-generators/

Proper termination is everything when examining square waves especially when rise times are fast however SDG2042X is just 9ns which shouldn't challenge any 200 MHz scope.
It's later released little brother SDG1032X OTOH is much faster ~4ns and will do square wave to full 30 MHz frequency and for a good bit less cost at just $319.
These are commonly paired with SDS11/1204X-E.

[rf-loop]

U said it 

I have a Siglent SDS1204X-C，I try to use it with Rigol DG1022Z but I found a problem when we try to exports & use .csv files.
Obviously, Siglent 1104X-C's .csv did not work with Rigol DG1022Z simply.... I'm crazy about it.

And another problem is that: A 25MHz square wave generated by Rigol DG1022Z occurs serious distortion in my Siglent 200Mhz bandwith oscilloscope. Just as below ... I am even distinguish sine and square wave...
Another friends recommended me to hack Rigol DG812，it is a option for me. But just as U say another option is Siglent SDG-2042X, because I ve already have a Siglent oscilloscope.

There are more questions here:

- What is the price(exclusive of tax) of your SDG-2042X in the United States ?
- What kind of a Square Wave generated by SDG-2042X in 200Mhz bandwith oscilloscope ?

Rigol DG1032Z 25MHz Square 1Vpp. Scope input 50ohm terminated.
( this Rigol test is from this msg:  )





Siglent SDG1032X, 25MHz square 1Vpp, scope SDS1202X-E input with 50ohm termination. (impedance matching is not perfect)

SDG2042X is overall bit better but Square  shape (risetime) performace is much more poor. This is because Siglent special Square wave improvement is based to more new technology. This was not ready when more old SDG2000X series was launched.

Example for BodePlot with Siglent SDS1104X-C  SDG1032X is very exellent. Of course also SDG2042X what is more expensive. And what perhaps can mod up to 120MHz when SDG1032X possible mod is limited to 60MHz.
Rigol DG800 / 900 can not work with Siglent scope for example BodePlot or exchange .CSV data.
SDG1032X have also lot of usable features including also waveforms combine between channels and well done modulations and also advanced level settings (example freedom to swap between level units on the fly). Also it have today "TrueArb" feature as also SDG2000X model but naturally with reduced frequency limits. In SDG1032X max TrueArb function clock is 75MHz.
-------------------------
may I ask if this SDS1104X-C is also English language available or is it only chinese.

[tautech]

may I ask if this SDS1104X-C is also English language available or is it only chinese.

Recently when I asked about some 'C' models they were domestic market only with some lesser specs than we see in the west. Eg, Sequence mode is just 200K wfps.

When I visited the factory in 2014 a SDS3000 in the reception lobby had only Chinese for language selection.

Deep analysis of the datasheet may reveal other differences.
https://www.siglent.com/products-overview/sds1000x-c/

[uinika]

Siglent SDG1204XC only for chinese and as mentioned above Sequence Mode is really 200K wfps.

I found there's only SDG1032X and no SDG1022X on Amazon, why ? And Can I hacked SDG1022X to 60Mhz ?

[rf-loop]

Siglent SDG1204XC only for chinese and as mentioned above Sequence Mode is really 200K wfps.

I found there's only SDG1032X and no SDG1022X on Amazon, why ? And Can I hacked SDG1022X to 60Mhz ?

Yes I also know C model is reduced speed "web shop" model as also this SDG1022X is "only for web shops". It looks very similar... difficult to find any signs it perhaps have different HW.  Same with SDS1000X-C series oscilloscopes. SDS1000X-U models have least some difference in HW... least unpopulated Si Bus for LA because it do not have MSO option at all. .

These special models for China domestic web shops have not been so much in western markets (also warranty is AFAIK only valid in mainland China) so I think here is not real knowledge with true experience how these things go with these special models FirmWare versions, least I have not seen.  Somehow I feel they do not differ so much but.  and because I can not read and find all chenese electronic hobbyists discussion forums I have not seen any info.

These C, U and SDG1022X are not for Export. They are for China mainland markets as far as I know so perhaps this is enough heavy reason you can not find it from western Amazon.

[KungFuJosh]

Rigol DG1032Z 25MHz Square 1Vpp. Scope input 50ohm terminated.
( this Rigol test is from this msg:  )





Siglent SDG1032X, 25MHz square 1Vpp, scope SDS1202X-E input with 50ohm termination. (impedance matching is not perfect)

SDG2042X is overall bit better but Square  shape (risetime) performace is much more poor. This is because Siglent special Square wave improvement is based to more new technology. This was not ready when more old SDG2000X series was launched.

What about an apples-to-apples comparison with similar rise/fall times?

[rf-loop]

What about an apples-to-apples comparison with similar rise/fall times?

You mean similar risetime generators or using similar risetime scopes. About modified Rigol scope risetime you need ask from Rigol fanboys... they can tell you amazing risetimes. For measure around 10ns risetime it must be well ok.

And Siglent SDS1202X-E risetime is enough for this Siglent generator risetime when we talk about this accuracy what is enough in this case. Of course roughly 1.7ns risetime do not give very accurate measured values for around 4ns risetime signal  but this is well enough here for this purpose.  And if Rigol 1000Z have even 5ns own risetime it is also well enough for this purpose here for this case.

Under Rigol image is link to Rigol test image original message. There user tell he have used hacked 1054Z scope and comparable connection to scope.

It do not change result any meaningful amount even if you take 1GHz Rohde&Schwarz for compare these signals.

Siglent generator SDG1032X and SDG1062X here can do up to 30MHz  or 60MHz square. Its specified risetime is for both models

Rise/fall times
4.2 ns 10% ~ 90%, 1 Vpp, 50 Ω load
3.8 ns 10% ~ 90%, 2.5 Vpp, 50 Ω load
Overshoot 3 % 100 kHz, 1 Vpp, 50 Ω load

And this Rigol 1022Z

Rise/Fall Time
Typical (1 Vpp) <10ns
Overshoot Typical (100 kHz, 1 Vpp) ≤5%

So what was your point. Or just for fun.

[KungFuJosh]

So what was your point. Or just for fun.

Forget the generators, I'm talking about for the scope screenshots. Yes, obviously the generator with the better rise time is nicer, but the comparison is odd since all things aren't even close to equal. Maybe that is just for fun.

[rf-loop]

So what was your point. Or just for fun.

Forget the generators, I'm talking about for the scope screenshots. Yes, obviously the generator with the better rise time is nicer, but the comparison is odd since all things aren't even close to equal. Maybe that is just for fun.

In this particular case these tests are well enough comparable. If not, I do not even put these together with my knowledge and experience with this kind of things. Of course if look both with exactly same instrument may give small (nonsense) difference  and also peoples without enough experience and knowledge can then think they can directly compare. When tests are made using different instruments it need understand what are differences and how they affect and importantly how much.  If you do not understand its not my job. I don't bother and don't want to twist it with iron wire.
 

[mawyatt]

Think the DSO and DUT rise times are independent and should follow as:

Measurement RT^2 = DSO_RT^2 + DUT_RT^2, or Measurement RT = Square Root (DSO_RT^2 + DUT_RT^2)

1st order model for DSO RT is ~ 0.35/BW where BW is DSO -3dB Bandwidth. Since this is Root-Sum-Squared measurement (common in many effects) the usual ~5X Tektronix recommended better scope rise time only causes ~2% measurement effect.

So one could get a pretty good estimate of the actual DUT rise time with different DSOs with different BWs following the above guide lines.

Best,

Edit added for those interested:

The 1st order DSO scope model assumes a single pole response, higher order responses are expected in higher BW DSOs which involve a more complex frequency response than a single pole.

If you assume a single pole response then the DSO BW follows:

 1/(1+(w/wc)^2) = 1/2, where wc is the DSO "corner" half power or -3dB frequency in radians/s, or 2*pi*BW where BW is in Hz.

The single pole DSO time step response will follow:

Step Response = 1-exp^(t/tau), where tau is the DSO time constant which equals 1/wc.

So the 10% step response is 0.1 = 1-exp^(t1/tau) and the 90% step response is 0.9 = 1-exp^(t2/tau), or

t1 = -tau(ln(0.9)) and t2 = -tau(ln(0.1)

10% to 90% rise time = t2-t1, or tau( ln(0.9) -ln(0.1)), or tau ln(9),

then 10-90% rise time = ln(9)/wc, or ln(9)/2*pi*BW, or 0.3497/BW

[omgoleus]

Hello, I know this is a bit of an old thread, but I'm wondering about the rise time specs posted here... I'm looking at the data sheet from Siglent for the SDG1000 series and it says for pulse, risetime is <7ns. I am curious where the 4.2ns and 3.8ns figures quoted by rf-loop in the message above came from?
Here's the datasheet where it is currently hosted at the Siglent web site, the rise/fall time specs for Pulse waveform are on page 5:
https://siglentna.com/wp-content/uploads/dlm_uploads/2017/10/SDG1000_DataSheet_DS02010-E08A.pdf
Thanks!

[Performa01]

Hello, I know this is a bit of an old thread, but I'm wondering about the rise time specs posted here... I'm looking at the data sheet from Siglent for the SDG1000 series and it says for pulse, risetime is <7ns. I am curious where the 4.2ns and 3.8ns figures quoted by rf-loop in the message above came from?
Here's the datasheet where it is currently hosted at the Siglent web site, the rise/fall time specs for Pulse waveform are on page 5:
https://siglentna.com/wp-content/uploads/dlm_uploads/2017/10/SDG1000_DataSheet_DS02010-E08A.pdf
Thanks!

You are looking at the old SDG1000, whereas rf-loop mentioned the SDG1000X.

https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2017/10/SDG1000X_DataSheet_DS0201X_E01G.pdf

And the transition times were quoted for the quare wave, not pulse.

[gf]

SDG2042X is overall bit better but Square  shape (risetime) performace is much more poor.

I think its square wave rise/fall time is even close to optimal with regard to the following objectives:

• jitter-free reproduction via Nyquist reconstruction, even at arbitrary frequencies which are not an integral fraction of fs/2
• don't exceed the specified 3% overshoot
• it has to run at the given sample rate

For a jitter-free Nyquist reconstruction it is necessary to band-limit an ideal square wave to frequencies < fs/2, when the prototype waveform is resampled by the DDS and then sent to the DAC. Discarding all harmonics beyond fs/2 with a brickwall filter were likely optimal in this regard, but leads to severe overshoot/ringing. In order to fulfill the < 3% overshoot requrement as well, a filter with a wider transition band is necessary, reducing the rise time even more. So if the mentioned objectives are honored, there is a theoretical limit for the achievable risetime, and I think the generator comes close to this limit. If (1) or (2), or both, do not matter, then then it's easy to brag about a faster rise time.