Video Review, Comparison and Teardown of Siglent SDG5162 & Rigol DG4162 Arbs

[Hugoneus]

In this episode Shahriar compares the performance and features of the Siglent SDG5162 and Rigol DG4162 Arbitrary Waveform Generators. Siglent has recently opened a new North America office which would certainly help them bring their instruments to the US more quickly and provide better local customer and sale support.

The Siglent SDG5000 teardown is presented. Aside from various GUI features of these two units, the THD, phase noise and maximum output power of each unit is measured. The ability of the Siglent generator to produce 12ns pulse width independent of the output frequency is also presented with some practical applications. Finally the Easy Wave application us is used to generate some custom waveforms.

Watch the video here: [1 Hour & 5 Minutes]
http://youtu.be/xqYQqXz7DZA

More videos at The Signal Path:
http://www.TheSignalPath.com/

[pickle9000]

Great video, as usual.

I don't have the SDG5162 but one of the lower end ones. My initial thoughts on the interface where yuck. Having said that the layout is good, logical and fast. Took about 5 minutes to get used to.

One thing about gear I say over and over. Every thing has faults and strengths, it's not better or worse it's different. I do like the fact that you show gear in this light.

Pet peeve, the Easy Wave application should include a pile of premade waveforms. Or perhaps Siglent should offer "Waveform Packs" for download.

[tautech]

Pet peeve, the Easy Wave application should include a pile of premade waveforms. Or perhaps Siglent should offer "Waveform Packs" for download.

Great idea, will pass it along. 

[rstoer]

Excellent review.
Shahriar is currently producing some of the best technical videos out there, be they reviews, repairs, or instructional.
In this one I particularly liked the demonstration of how the Siglent's ability to produce a narrow pulse with a long period could be beneficial in a real-world situation. Although over an hour long I would have liked a little more actually. I would have liked to hear about the sweep and modulation functionality, or at least if there were any significant differences between the Siglent and the Rigol in that area. Still, it was a most informative review. Keep 'em coming!

[Hugoneus]

Excellent review.
Shahriar is currently producing some of the best technical videos out there, be they reviews, repairs, or instructional.
In this one I particularly liked the demonstration of how the Siglent's ability to produce a narrow pulse with a long period could be beneficial in a real-world situation. Although over an hour long I would have liked a little more actually. I would have liked to hear about the sweep and modulation functionality, or at least if there were any significant differences between the Siglent and the Rigol in that area. Still, it was a most informative review. Keep 'em coming!

You are welcome. I will have some microwave reviews in the future as well. Stay tuned!

[DJ]

A very useful review. Perhaps this was the one where I saw a pattern on the pcb for an unloaded OCXO/TCXO?

Did note the rtv/hotmelt or whatever did seem to be covering some of the capacitor vents,  if I remember the video correctly.  Otherwise appeared to have decent construction.

As for waveform "packs", perhaps there is somewhere people can upload waveforms to some sort of a public library? Off the top of my head, I would imagine the waveforms for 12-lead EKG's would be useful, particularly simulating various problems. I'm sure there are many other useful waveforms out there. Something like grabcad, maybe grabwave or grabforms?


Length was good, although may be some use to breaking them up to 30 minute segments should a review run much longer. Trying to find an uninterrupted 60 minutes can be challenging sometimes.

Again, very clearly presented. I may well purchase one of these and the video was most helpful. Keep up the good work!

[pickle9000]

As for waveform "packs", perhaps there is somewhere people can upload waveforms to some sort of a public library? Off the top of my head, I would imagine the waveforms for 12-lead EKG's would be useful, particularly simulating various problems. I'm sure there are many other useful waveforms out there. Something like grabcad, maybe grabwave or grabforms?

Not to go too far off. Some waveforms for decoding (spi, rs-232) purposes, bells and so on. Good for those that have never set up a scope to decode spi. A waveform library is not a bad idea.

[tautech]

As for waveform "packs", perhaps there is somewhere people can upload waveforms to some sort of a public library? Off the top of my head, I would imagine the waveforms for 12-lead EKG's would be useful, particularly simulating various problems. I'm sure there are many other useful waveforms out there. Something like grabcad, maybe grabwave or grabforms?

Not to go too far off. Some waveforms for decoding (spi, rs-232) purposes, bells and so on. Good for those that have never set up a scope to decode spi. A waveform library is not a bad idea.

I proposed this to Siglent a few weeks back. they thought it was a good idea and forwarded it to R&D.
I suggested a sample library in Easywave and/or a "Waveform Pack" available for download.
No doubt it may be protected in some way for use in only Siglent products.
I wll ask how this is progressing and hopefully we don't have to wait too long.

[rf-loop]

As for waveform "packs", perhaps there is somewhere people can upload waveforms to some sort of a public library? Off the top of my head, I would imagine the waveforms for 12-lead EKG's would be useful, particularly simulating various problems. I'm sure there are many other useful waveforms out there. Something like grabcad, maybe grabwave or grabforms?

Not to go too far off. Some waveforms for decoding (spi, rs-232) purposes, bells and so on. Good for those that have never set up a scope to decode spi. A waveform library is not a bad idea.

I proposed this to Siglent a few weeks back. they thought it was a good idea and forwarded it to R&D.
I suggested a sample library in Easywave and/or a "Waveform Pack" available for download.
No doubt it may be protected in some way for use in only Siglent products.
I wll ask how this is progressing and hopefully we don't have to wait too long.

Also these are quite easy to make.
Siglent SDG's read nearly "what ever" text file what include data stream as plain text  for vertical values. These are CSV files, "comma separated values".

(note: It looks like numerical data in header do not matter but still needed something). Also there is different way to write data and header. Example from old EasyWave it is different (this exampole is with older EasyWave), from oscilloscope, example SDS2000 it is different etc... and SDG can read and compile these all and also it can automatically adjust lenght to fill 16k if waveform data is less.)

<this is header part>
data lenth,16384
frequency,1000.000000000
amp,20.000000000
offset,0.000000000
phase,0.000000000

<here start data pairs>
xpos,value
0,0.000000000
1,0.007669903
2,0.015339802
3,0.023009691
.
<clip><clip><clip>
.
16381,-0.023011835
16382,-0.015341945
16383,-0.007672047

Also if one can write math equation for waveform, EasyWave translate it for .CSV and transfer to SDG from EasyWave with USB connection or store to USB stick and read from USB stick with SDG's.

(--> draw menu --> Equation draw. There write or copypaste these examples and look what happend. With good math skills peoples can do of course much more complex things....)

For very simplified example, something like this:

(with these use default property setting 20V (-10 ~ +10)

10*(sin(2*(sin(2*(sin(2*x))))))    (this is very simplified)

(two cycles "square" using full scale (-10 - +10) made using sinewaves (harmonics))  (this is also exaple what happend if we look square wave with oscilloscope what can show only very limited amount of square wave harmonics...)

or just

10*(sin(10*x))   
10 cycles of sine.  (run this using 5MHz arb frequency and out is 50MHz sine (not same quality as equipment specifications for normal sine function)


Let's hope Siglent want later publish full and accurate documentation about SDG's  .CSV files structure details in different cases and how it handle these files in different cases... how it adjust (compress or stretch)  these to Arb waveform memory.

[papousek]

there seems discount for Rigol DG4000, but still higher than Siglent SDG2000X.
does any one have compared DG4000 and SDG2000X?

[slimbobaggins]

I just downloaded and installed Easywave, and I was trying to create an arbitrary waveform by defining points in a table, but it defines times in ms, and won't allow me to enter decimal places for microseconds (frequency is 60 Hz).

So for example, at 60 Hz, I want a 250 us pulse, off for 100 us, then a 4 us pulse.  Easywave's table definition didn't allow this, unless I'm doing something wrong.

I see the earlier post that says you can write an arb. function in csv, so I'll try that.  But if anyone has any suggestions for Easywave, please send them my way.  I also couldn't get my CPU to talk to it via a USB cable.  Telling Easywave to read the existing waveform results in the program crashing.  I don't have time to much with it, so my plan is to store a waveform on a USB drive and xfer it that way.

This is one shorting coming of the Siglent IMO... at least Rigol allows you to program from the front panel, even if it can be cumbersome.

[tautech]

I just downloaded and installed Easywave, and I was trying to create an arbitrary waveform by defining points in a table, but it defines times in ms, and won't allow me to enter decimal places for microseconds (frequency is 60 Hz).

So for example, at 60 Hz, I want a 250 us pulse, off for 100 us, then a 4 us pulse.  Easywave's table definition didn't allow this, unless I'm doing something wrong.

I see the earlier post that says you can write an arb. function in csv, so I'll try that.  But if anyone has any suggestions for Easywave, please send them my way.  I also couldn't get my CPU to talk to it via a USB cable.  Telling Easywave to read the existing waveform results in the program crashing.  I don't have time to much with it, so my plan is to store a waveform on a USB drive and xfer it that way.

This is one shorting coming of the Siglent IMO... at least Rigol allows you to program from the front panel, even if it can be cumbersome.

Have you installed the NI VISA package also for the LAN and USB connectivity drivers?

[slimbobaggins]

I just downloaded and installed Easywave, and I was trying to create an arbitrary waveform by defining points in a table, but it defines times in ms, and won't allow me to enter decimal places for microseconds (frequency is 60 Hz).

So for example, at 60 Hz, I want a 250 us pulse, off for 100 us, then a 4 us pulse.  Easywave's table definition didn't allow this, unless I'm doing something wrong.

I see the earlier post that says you can write an arb. function in csv, so I'll try that.  But if anyone has any suggestions for Easywave, please send them my way.  I also couldn't get my CPU to talk to it via a USB cable.  Telling Easywave to read the existing waveform results in the program crashing.  I don't have time to much with it, so my plan is to store a waveform on a USB drive and xfer it that way.

This is one shorting coming of the Siglent IMO... at least Rigol allows you to program from the front panel, even if it can be cumbersome.

Have you installed the NI VISA package also for the LAN and USB connectivity drivers?

I did but its possible its my machine as well.

[tautech]

I just downloaded and installed Easywave, and I was trying to create an arbitrary waveform by defining points in a table, but it defines times in ms, and won't allow me to enter decimal places for microseconds (frequency is 60 Hz).

So for example, at 60 Hz, I want a 250 us pulse, off for 100 us, then a 4 us pulse.  Easywave's table definition didn't allow this, unless I'm doing something wrong.

I see the earlier post that says you can write an arb. function in csv, so I'll try that.  But if anyone has any suggestions for Easywave, please send them my way.  I also couldn't get my CPU to talk to it via a USB cable.  Telling Easywave to read the existing waveform results in the program crashing.  I don't have time to much with it, so my plan is to store a waveform on a USB drive and xfer it that way.

This is one shorting coming of the Siglent IMO... at least Rigol allows you to program from the front panel, even if it can be cumbersome.

Have you installed the NI VISA package also for the LAN and USB connectivity drivers?

I did but its possible its my machine as well.

OK, you might need to drill down into Win Device Manager and point the AWG USB driver to the NI VISA driver that'll most likely be ausbtmc.inf.
Search your PC for it.
Maybe in C:\Program Files\IVI Foundation\VISA\IVI USB Staging


Little doc for DSO connection problems attached, hope it helps.

[slimbobaggins]

I just downloaded and installed Easywave, and I was trying to create an arbitrary waveform by defining points in a table, but it defines times in ms, and won't allow me to enter decimal places for microseconds (frequency is 60 Hz).

So for example, at 60 Hz, I want a 250 us pulse, off for 100 us, then a 4 us pulse.  Easywave's table definition didn't allow this, unless I'm doing something wrong.

I see the earlier post that says you can write an arb. function in csv, so I'll try that.  But if anyone has any suggestions for Easywave, please send them my way.  I also couldn't get my CPU to talk to it via a USB cable.  Telling Easywave to read the existing waveform results in the program crashing.  I don't have time to much with it, so my plan is to store a waveform on a USB drive and xfer it that way.

This is one shorting coming of the Siglent IMO... at least Rigol allows you to program from the front panel, even if it can be cumbersome.

Have you installed the NI VISA package also for the LAN and USB connectivity drivers?

I did but its possible its my machine as well.

OK, you might need to drill down into Win Device Manager and point the AWG USB driver to the NI VISA driver that'll most likely be ausbtmc.inf.
Search your PC for it.
Maybe in C:\Program Files\IVI Foundation\VISA\IVI USB Staging


Little doc for DSO connection problems attached, hope it helps.

Great, thanks for the help, will check this at work tomorrow

[rf-loop]

I just downloaded and installed Easywave, and I was trying to create an arbitrary waveform by defining points in a table, but it defines times in ms, and won't allow me to enter decimal places for microseconds (frequency is 60 Hz).

So for example, at 60 Hz, I want a 250 us pulse, off for 100 us, then a 4 us pulse.  Easywave's table definition didn't allow this, unless I'm doing something wrong.

I see the earlier post that says you can write an arb. function in csv, so I'll try that.  But if anyone has any suggestions for Easywave, please send them my way.

What Siglent SDG you have?
This can do many ways using EasyWave including also .csv editing. (.csv edit can also use for some small detail fix in EasyWave made .csv)
I ask model because there is differencies, specially between fixed memory lenght models and flexible memory lenght models.   Also, you do not need write wanted frequency (and level, you can use example 2Vp-p what give +1V and -1V or some other value) just right if you do waveform with EasyWave. For 60Hz you can use (example 60kHz). When you use SDG then you adjust frequency (and level)

But, as you told you want 250us high and then 100us low and then 4us high and rest time low.
Main principle is that Arb frequency mean whole used Arb memory repeating frequency.
If have SDG1000 there is 16384 data points and this value is fixed.

If have SDG5000 there is also fixed lenght 16384 data points for CH1 and 2  but Ch2 have also 524288 data points fixed lenght.

But, SDG2000X is different, it have flexible 8pts - 8Mpoints lenght memory. This make things very different.

60Hz period is 16666.6666... us. 
If there is 16384 data points there is not possible to do perfect 250-100-4us
With 60Hz frequency one step in memory is 1.017us.

For 250us you need roughly 246 data points high, then for 100us you need roughly 98 points low. Then for 4us.. well due to rise and fall time I think 3 data points high is some kind of compromise. After then all low to end of memory.

Btw, this is quite easy to draw with EasyWave. If You want different scale set freq to example 60kHz and you get us scale instead of ms. And then, when drawing use Zoom (Horizontal or Vertical) for small details or small time increments (in this case Zoom horizontal in to first around 400us then drawing is more easy and accurate (after learn how this poor draw feature works). Always it show your cursor accurate x,y values.

When use this waveform in SDG just set SDG Arb freq 60Hz.

Then (bug in P34 version). Least in my computer I need store .csv file using menu "Save As"
After then if close EasyWave it ask do you want save. If answer Yes, it "smudge" your saved .csv.  If use "Save As" after then close EasyWave, answer No if it ask save.

But, with SDG2000X flexible memory. Now you can set memory lenght.
For 1us steps you can set for 60Hz just 16667 (or 16666) points. But for more resolution/accuracy you can set 1666667 memory lenght and now for 60Hz you can think every step is 10ns.

Attached .csv for SDG2000X. Using only 16667 data points. Hand draw using EasyWave.
(remove .TXT)
This can also Edit just using even Notepad (or some other what do not add any extra)
When edit do not edit xpos. Only value column.

SDG800, 1000 and 5000 are very similar BUT inportant. They are fixed memory lenght!

[slimbobaggins]

Well I was about to try Tautech's suggestions, and fired the computer machine up today, and for whatever reason now I'm able to connect to the Siglent.  Go figure.  At least now I know if it doesn't work next time, I've got troubleshooting steps.

Rf-loop, I have a SDG5162.

Thanks for all the helpful info.  I opened your file and modified it in Excel, and was able to get what I need.

I abhor freehand sketching with a mouse (at least when I'm laying out a board it's got a snap grid!).  That's why I was attempting to use their coordinate draw function, especially since my waveform is so uncomplicated.  I just don't understand what input the coordinate draw function is looking for.  It gives you 50 data points, and regardless what data format I attempt to enter in the X column, it barks at me that I'm violating the allowed timespan.  Doesn't matter anyways now, since I've got your file to modify.   Thanks again!