Siglent SDG5000 vs Rigol DG1000z - Siglent better buy?

[Wuerstchenhund]

Since my Siglent SDG1020 has too many limitations for my liking, I'm looking into getting another AWG. After checking what's available on the market it essentially comes down to either the Rigol DG1062z or the Siglent SDG5082 (the Rigol DG4000 is out because it essentially combines a high price with low sample memory).

The Rigol does have the large memory (8Mpts standard, 16Mpts option) and LAN as a benefit, as well as slightly better specs in sine mode. The Downsides are the price (the DG1062z is quite a bit more expensive than the SDG5082), and that there doesn't seem to be any AW software available from Rigol.

The Siglent SDG5000 on the other side is cheaper (for the price of a DG1062z I could even buy the SDG5122), the better paper specs in pulse/square mode and overall higher bandwidth in all modes, and that there is some software (although its crap) available. The downsides are the small (512kpts on Ch2) and really small (16kpts on Ch1) sample memories, the absence of LAN (really, Siglent?), and (judging by Dave's Teardown video) the questionable build quality (corrosion). Coming from the SDG1000 I have also concerns re the UI (which isn't very good on the SDG1000).

But the lower price and the higher bandwith is just too tempting, so I'm almost decided on the SDG5082/5122 (or even the 5162, we'll see). But before I buy I would like to get some comments from people that do have a SDG5000 and can give some comments/feedback.

What I'd like to know in particular:

- How good is the real-world performance, especially with very short pulses (i.e. <100ns) with low PRF? What about jitter (the SDG1000 had some issues there)?

- How is the build quality of newer exemplars (I hope the corrosion in Dave's video is a thing of the past)?

- How good is the UI, especially in comparison with the Rigol DG1000z Series?

- Any major hardware or firmware bugs?

- On the SDG1000, the rotary knob is also a push button which serves no purpose. Is that the case with the SDG5000 as well?

- How good is file management (which was cumbersome on the SDG1000)? I assume FAT(32) is still the only supported file system? How about memory sticks larger than 4GB? Long file names?


I'd also like to hear if anyone has experience with both AWGs.

[rodpp]

I don´t have these AWG, but I was searching for one sometime ago and chose the Rigol DG1032Z that will arrive in the next month. The memory length was decisive to me.

So my answer is only based on the datasheets, sorry!

The relevant differences in specifications to my personal use are:

- the SDG5082 has a better sample rate (500MSa/s vs 200MSa/s);
- SDG5082 can output 10Vpp until 40MHz vs 10MHz on the Rigol
- SDG display is 4.3" vs 3.5" on DG.
- DG1062Z has much better Harmonic Distortion specifications (DC-10MHz -65dBc vs -46dBc) and THD (0.075% vs 0,2%)
- the square wave duty cycle is 0.01% to 99.99% on DG vs 20% to 80% on SDG;
- the PWM modulation frequency range is 2 mHz to 1 MHz on DG vs 1mHz to 50KHz on SDG;
- the arbitrary waveform memory on the DG1062Z is 8Mpts (16Mpts optional) for each channel vs 16Kpts ch1 and 512Kps ch2 on the SDG5082;
- two features of the DG1062Z that I don´t know if are presents in the SDG5082 are waveform summing and channel track;
- LAN connection on Rigol.

[idpromnut]

I am curious: what applications require this large amount of AW memory?

[Wuerstchenhund]

I don´t have these AWG, but I was searching for one sometime ago and chose the Rigol DG1032Z that will arrive in the next month.

Where did you order that shipment takes a months?

The memory length was decisive to me.

That's something that's very tempting for me, too. As is the LAN port. And that it seem to offer more built-in waveforms.

So my answer is only based on the datasheets, sorry!

I've seen the datasheets. The much better specs of the Siglent are tempting (and from what I've found it seems the actual performance is good), but the small sample memory (512kpts isn't much, but even worse it's only for one channel).

I'd really like to hear from someone with a DG1000z how good it is, especially when producing very short pulses at low PRF. It seems the DG4000 does have some issues with the rise time, I hope the DG1000z is more like the Siglent.

[rodpp]

I am curious: what applications require this large amount of AW memory?

I´ll use the AW memory to store a reference signal for a control system. The reference signal can be generated in Matlab to perform specific tasks or for performance tests.

Another application is to reproduce some behavior captured from an equipment, like the output of a sensor when the system is running in a specific situation. Reproducing this signal in the laboratory one can develop and validate circuits that analyses this equipment.

[rodpp]

I don´t have these AWG, but I was searching for one sometime ago and chose the Rigol DG1032Z that will arrive in the next month.

Where did you order that shipment takes a months?

The memory length was decisive to me.

That's something that's very tempting for me, too. As is the LAN port. And that it seem to offer more built-in waveforms.

So my answer is only based on the datasheets, sorry!

I've seen the datasheets. The much better specs of the Siglent are tempting (and from what I've found it seems the actual performance is good), but the small sample memory (512kpts isn't much, but even worse it's only for one channel).

I'd really like to hear from someone with a DG1000z how good it is, especially when producing very short pulses at low PRF. It seems the DG4000 does have some issues with the rise time, I hope the DG1000z is more like the Siglent.

I´m from Brazil. The authorized rep is importing it from China. The initial schedule was 30 days, now they are asking 15 days more.

[Sebastian]

I have the DG1062z and I like it. I already had a DS1004z and a DP832 and because I didn't have any problems with them, I decided to go with Rigol again.
The UI and the UX in general are very good and I am pretty sure it has the claimed specs.
If you don't need the higher bandwidth right now, I would go with the DG1062z and get a separate RF signal-gen when you need it. (Except if you need arbitrary waveform generation at higher bandwidth of course)

[Wuerstchenhund]

I have the DG1062z

Hallelujah, someone does actually has one 

and I like it. I already had a DS1004z and a DP832 and because I didn't have any problems with them, I decided to go with Rigol again.
The UI and the UX in general are very good and I am pretty sure it has the claimed specs.

Since you have one, could you do me a favor and have a look at this Youtube video?


Could you check if the DG1062z has the same problem as the DG4062 in the video?

If you don't need the higher bandwidth right now, I would go with the DG1062z and get a separate RF signal-gen when you need it. (Except if you need arbitrary waveform generation at higher bandwidth of course)

The higher bandwidth of the Siglent would certainly be nice, but I don't really need it. I'm currently reading through the user manual, and the more I read the more I actually favor the Rigol.

What about the software? Rigol seems to offers some Windows tool (UltraStation) for DG4000 and DG5000, but I couldn't find anything for the DG1000z. Do you know if UltraStation works with the DG1062z as well?

[rodpp]

The bandwidth differences are not high in all waveforms, but are relevant in sine, ramp/tringular and noise:

Code: [Select]

Model SDG5082 DG1062Z
sine 80MHz 60MHz
square 30MHz 25MHz
pulse 20MHz 25MHz
ramp/triangular 2MHz 1MHz
noise 100MHz 60MHz
arbitrary 20MHz 20MHz

[tautech]

the absence of LAN

From my information the LAN connection is available as an option.
Sorry I do not have any costings for it.
Might be worth an email to Siglent to have this info for a full comparison.

[Sebastian]

I checked the issues from the video and also made a few screenshots for you.
The minimum rise time in pulse mode is 10ns.
The duty cycle goes from 0.001% to 99.999%.
The longest rise time with 31.25ms can be archived at 50% duty cycle. The maximum rise time gets progressively shorter if you go up or down in duty cycle.
All of this is valid from 25MHz down to 1Hz at 0dBm with 50Ohm output impedance  and also applies for the fall time.
No guarantee for correctness.

Hope I could help you.

[Wuerstchenhund]

I checked the issues from the video and also made a few screenshots for you.

Thanks a lot for the effort and the screenshots. That's really helpful. 

The minimum rise time in pulse mode is 10ns.
The duty cycle goes from 0.001% to 99.999%.
The longest rise time with 31.25ms can be archived at 50% duty cycle. The maximum rise time gets progressively shorter if you go up or down in duty cycle.
All of this is valid from 25MHz down to 1Hz at 0dBm with 50Ohm output impedance  and also applies for the fall time.
No guarantee for correctness.

It seems it's not as fast re pulse edges as the Siglent, but it should be good enough for my purposes.

One last thing though is the software. Do you by chance know if Rigol UltraStation works with the DG1000z?

[Wuerstchenhund]

From my information the LAN connection is available as an option.
Sorry I do not have any costings for it.
Might be worth an email to Siglent to have this info for a full comparison.

Yes, LAN is listed as option in Service and User Manuals. But when the SDG1000 came out Siglent did list LAN as an option for it as well, which never materialized. There also seems to be no code in the firmware which supports LAN.

I guess it's just an option for OEM customers. The SDG1000 and SDG5000 do have cutouts in the chassis for a LAN port. But I guess they never developed the firmware support for it as even LeCroy's variant doesn't offer LAN.

BTW, is the rotary knob on the SDG5000 also a push button that does nothing as on the SDG1000?

[tautech]

From my information the LAN connection is available as an option.
Sorry I do not have any costings for it.
Might be worth an email to Siglent to have this info for a full comparison.

BTW, is the rotary knob on the SDG5000 also a push button that does nothing as on the SDG1000?

Sorry, I do not have a 5000 series, so I do not know.
I suspect the rotary encoder is a push type commonly used in DSO's and it may have been cheaper/simpler to use those on hand rather than increase inventory..
It is probably wise to have this type installed for possible future functionality stemming from firmware upgrades.(programmable default waveform?)

Interestingly, I had a local inquiry today, I suspect originating from this thread.
The rise times for the 1000 and 5000 series are very similar.
Powered my SDG1010 up for a rise time spec check and it is as advertised @ 7 nS

[rf-loop]

[/quote]
BTW, is the rotary knob on the SDG5000 also a push button that does nothing as on the SDG1000?
[/quote]

SDG5000 series rotary knob push button is in use.
Example with it can fast jog between parameters what need adjust.

About jitter. SDG5000 do not have at all same corner glitch and jitter (in square wave if period is long) as it is in SDG1000 series spoecially in square wave mode. Square wave is produced totally different way.

Then, of course SDG5000 have less jitter than this Rigol - least in "theory".
This is because both are still "simple machines" and fixed clock. 200MSa/s leads to 5ns jitter and 500MSa/s leads to 2ns jitter. With same reason SDG1000 have 8ns jitter due to 125MSa/s system.

About square wave duty limits.  Why square wave need produce narrow pulses or wide pulses. If need narrow pulses use pulse mode - of course. I do not even understand why there is separate "square wave" and then "pulse" mode. In pulse mode there can set duty % or pulse time.
And here in pulse mode my opinion (based from datasheets about Rigol and real tests with Siglent) is that SDG5000 beats this Rigol as 10-1.

And with jitter do not drop to rms trap.

[Wuerstchenhund]

BTW, is the rotary knob on the SDG5000 also a push button that does nothing as on the SDG1000?

SDG5000 series rotary knob push button is in use.
Example with it can fast jog between parameters what need adjust.

So what function does it have? Something like a Enter button i.e. to confirm a selected parameter? Or does it just jump to the next setting?

Re. Jitter: I'm not worried about the SDG5000's RF signal performance. But what worries me is the lack of LAN (a big minus) and the small memory (especially on Ch1, another really big minus), and also the lower number of integrated Arbitrary Waveforms (another minus, even though not that big as for the two other points).

The Rigol is much more expensive (when counting in that it also comes with lower bandwith) but at least from the manual and spec sheet it looks like a much more versatile device.

It's really not an easy decision.

[Sebastian]

Do you by chance know if Rigol UltraStation works with the DG1000z?

It does work but I haven't used it a lot yet and I think it is not very intuitive. You could download it yourself and play around with it.

[Wuerstchenhund]

It does work but I haven't used it a lot yet and I think it is not very intuitive.

Well, it can't be worse than the Siglent program for their AWGs (EasyWave?).

You could download it yourself and play around with it.

I tried once but apparently it requires a Rigol SG connected as a dongle.

But at least it seems to work with the DG1000z.

[havefun]

I have used both SDG1050 and SDG5162(also a SDG5000) though I haven't get one for myself. I think both of it is OK for my purpose.
1. I teardown the SDG1050 to check if the errosion is that serious. I find out that the errosion only happens on the horizontal section of the shell and it is caused by the die shop(and I wonder most of the errosion is caused by the shipment on seas). So it doesn't affect the AWG's funtion.
2. As for the memory depth, I think a 512Kpts one is long enough to show the details of the waveforms. The siglent SDG5162 also supports USB storage to save waveforms. It supports USB-TMC and GPIB to transform data.
3.The UI of SDG5000 is much better than the SDG1000. , though I still don't know what is the rotary button can be pushed down is used for.
4.Why do you say the Easywave is crap? I haven't used another one. Can you describe it ?
5.I havn't used the DG1000z I don't know if it is good. But the SDG5162 is OK for its price!

[havefun]

[/quote]

So what function does it have? Something like a Enter button i.e. to confirm a selected parameter? Or does it just jump to the next setting?

[/quote]
I have checked my SDG5162, the rotary button can both confirm a selected parameter and then jump to the next setting. Just like the enter key on the keyboard.

[Wuerstchenhund]

I have used both SDG1050 and SDG5162(also a SDG5000) though I haven't get one for myself. I think both of it is OK for my purpose.
1. I teardown the SDG1050 to check if the errosion is that serious. I find out that the errosion only happens on the horizontal section of the shell and it is caused by the die shop(and I wonder most of the errosion is caused by the shipment on seas). So it doesn't affect the AWG's funtion.

No matter how low the price is, corrosion shouldn't happen. If it corrodes while on transport then Siglent needs to improve its packaging to make sure this doesn't happen.

2. As for the memory depth, I think a 512Kpts one is long enough to show the details of the waveforms.

At 500MS/s the 512kpts give you 1ms of signal storage, and this is only on Ch 2. On Ch 1 it's only 16kpts so only 32us.

The Rigol's 8Mpts provide for 40ms which allows for much more complex waveforms.

The siglent SDG5162 also supports USB storage to save waveforms. It supports USB-TMC and GPIB to transform data.

True, but that doesn't help as the AWG can't play directly from USB. USB-TMC is pretty much standard nowadays, as should be a LAN interface.

3.The UI of SDG5000 is much better than the SDG1000. , though I still don't know what is the rotary button can be pushed down is used for.

Good to hear.

4.Why do you say the Easywave is crap? I haven't used another one. Can you describe it ?

In short: awful UI, written by someone who has no clue about how to do a proper user interface. Clunky control, limited manipulation methods, and in a few cases the resulting csv files didn't really result in a signal that even resembled what it should look like (and this wasn't because of any AWG limitation).

[Wuerstchenhund]

I have checked my SDG5162, the rotary button can both confirm a selected parameter and then jump to the next setting. Just like the enter key on the keyboard.

I see. At least it has a purpose, unlike on the SDG1000.

Thanks for checking!

[rf-loop]

..and in a few cases the resulting csv files didn't really result in a signal that even resembled what it should look like (and this wasn't because of any AWG limitation).

Please can you explain this more detailed. What exatly happends?

[Wuerstchenhund]

..and in a few cases the resulting csv files didn't really result in a signal that even resembled what it should look like (and this wasn't because of any AWG limitation).

Please can you explain this more detailed. What exatly happends?

For example, in one instance I put a simple sine signal (few hundred kHz) together in EasyWave, but the signal that came out of the AWG was garbage. I then opened the file in EasyWave, and it showed garbage, too.

EasyWave somewhat reminds me on old image editing program from the Windows 3.11 days (Paintbrush?). It's really primitive and needs a complete overhaul.

[havefun]

4.Why do you say the Easywave is crap? I haven't used another one. Can you describe it ?

In short: awful UI, written by someone who has no clue about how to do a proper user interface. Clunky control, limited manipulation methods, and in a few cases the resulting csv files didn't really result in a signal that even resembled what it should look like (and this wasn't because of any AWG limitation).
[/quote]


I downloaded the software made by Rigol(Ultra Sigma?) yesterday? but the software is always loading file. I cannot compare them. (At least SIGLENT has a steady one?)