Author Topic: New Rigol 16-bit function generators DG800/900 series  (Read 192411 times)

JimboJack and 1 Guest are viewing this topic.

Offline commongrounder

  • Frequent Contributor
  • **
  • Posts: 372
  • Country: us
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #125 on: May 16, 2019, 04:44:35 pm »
My DG952 works fine with a cheap USB 3.0 Gigabit LAN dongle.
One reason I purchased a 900 series instead of the 800 series is the difference in the arb bit depth.  Both series have 16-bit main function generation, but the 800 series only has 14-bit arbitrary output.  Loaded in 16-bit arb files just get the lowest two bits truncated.  I also was attracted to the additional arb memory of the 900 series which allows for lengthy sequences of low distortion audio tests.
I’ve had my unit for several months, and have found it easy to learn.  I almost never use the touch screen, since the main knob plus arrow keys can access all of the screen icons.
And, I don’t give a crap about what the case looks like.  It’s dead silent, and performs very well for my applications. :-+
 
The following users thanked this post: thm_w, Sighound36, suicid

Offline maxwell3e10Topic starter

  • Frequent Contributor
  • **
  • Posts: 869
  • Country: us
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #126 on: May 16, 2019, 06:08:15 pm »
Can you post some performance data, like a noise spectrum or time jitter measurements?
 
The following users thanked this post: Sighound36

Offline Kodovo

  • Newbie
  • Posts: 6
  • Country: fi
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #127 on: May 16, 2019, 07:13:24 pm »
One reason I purchased a 900 series instead of the 800 series is the difference in the arb bit depth.  Both series have 16-bit main function generation, but the 800 series only has 14-bit arbitrary output.  Loaded in 16-bit arb files just get the lowest two bits truncated.

Is that really true?  I asked clarification for the User's Guide page 2-74 from Batronix last November, and Jan-Steffen Schwiemann answered:

  "we are sorry for the late reply.
    But we have finally gotten an answer from Rigol.

    For the DG800 series it is definitely 16bit resolution.
    Looks like they took the description directly from one of their other generators and forgot to correct this part."

Has anybody checked if those last bits really get truncated or not?
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 9057
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #128 on: May 17, 2019, 02:49:11 am »
Rigol's data sheet for the 900 series also has errors in the maximum carrier frequency in burst mode. Unless there's something wrong with the generators, the max frequency in burst mode should match the max frequency of the generator. However, the data sheet shows them as 5 MHz less than the corresponding 800 series. It looks like someone changed the least significant digit from 5 to 0, but forgot to change the other digits.
TEA is the way. | TEA Time channel
 

Offline frozenfrogz

  • Frequent Contributor
  • **
  • Posts: 936
  • Country: de
  • Having fun with Arduino and Raspberry Pi
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #129 on: May 17, 2019, 11:36:17 am »
So, if you convert your 8xx to 9xx you should have the 16-bits.  But, somebody will have to really test it.

I am willing to test! :)

DG812 is what I have here. What other per-requisites are there?
He’s like a trained ape. Without the training.
 

Offline Kodovo

  • Newbie
  • Posts: 6
  • Country: fi
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #130 on: May 17, 2019, 12:10:26 pm »
We now have a discrepancy between what Rigol says (according to Batronix), and what is actually printed in the Manual.  So, I'd like to see whether the official firmware has 14 or 16 bits for arb. wave forms.  If you could produce an arb. form with, say, values 0x4000, 0x4001, 0x4002, 0x4003, and 0x4004, and check with a scope whether there are five different voltage levels or just two.  I think that should close the case.

Screenshot would be nice!
« Last Edit: May 17, 2019, 12:28:42 pm by Kodovo »
 

Offline timber23

  • Contributor
  • Posts: 47
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #131 on: May 17, 2019, 01:41:14 pm »
We now have a discrepancy between what Rigol says (according to Batronix), and what is actually printed in the Manual.  So, I'd like to see whether the official firmware has 14 or 16 bits for arb. wave forms.  If you could produce an arb. form with, say, values 0x4000, 0x4001, 0x4002, 0x4003, and 0x4004, and check with a scope whether there are five different voltage levels or just two.  I think that should close the case.

Screenshot would be nice!

I did the test. Here is the result. This is the display of my DG811:


This is the measurement with my MSO5000:


See test2_1416.RAF (inside attached ZIP) for ARB file. I repeated the 0x4000 ... several times for better visualization.

Seems to be 16bit. No truncation.

Best regards
timber
 
The following users thanked this post: thm_w, bitseeker, PushUp, maxwell3e10, frozenfrogz, Mr. Scram, Kodovo, nikitasius

Offline Kodovo

  • Newbie
  • Posts: 6
  • Country: fi
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #132 on: May 17, 2019, 06:25:03 pm »
Great, timber23, thank you!    How stupid from Rigol: the mistake in the User's Guide has certainly affected on the sales.  At least I have postponed the purchase until now.  But not anymore, thank you :)   Just waiting for the details of tv84 before deciding for the model.
« Last Edit: May 17, 2019, 06:35:23 pm by Kodovo »
 

Offline timber23

  • Contributor
  • Posts: 47
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #133 on: May 17, 2019, 06:48:17 pm »
What does the attenuator go down to when you set it to 50 ohms impedance?

Does anybody know how to measure this?

Strangely the measurement doesn't change if I change from HighZ to 50 ohm impedance.



The red curve is BNC direct connected to scope. Doesn't matter if I select HighZ or 50 ohm impedance. The yellow trace is BNC connected to BNC-T-adapter with 50 ohm termination. As you can see, the 20Vpp is divided by 2, maybe because of a voltage divider. For my MSO5000, I cannot change input impedance. The button is grayed out and it says 1 M ohm.

Any suggesstions why there is no difference between HighZ and 50 ohm on DG800? What am I doing wrong?

Best regards
timber23
 
The following users thanked this post: PushUp

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23018
  • Country: gb
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #134 on: May 17, 2019, 07:34:36 pm »
I was just interested in what the minimum input value was.

I suspect this is a UI issue. Impedance has to match at each end to be accurate.
 

Online thm_w

  • Super Contributor
  • ***
  • Posts: 6276
  • Country: ca
  • Non-expert
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #135 on: May 17, 2019, 08:52:01 pm »
Strangely the measurement doesn't change if I change from HighZ to 50 ohm impedance.



The red curve is BNC direct connected to scope. Doesn't matter if I select HighZ or 50 ohm impedance. The yellow trace is BNC connected to BNC-T-adapter with 50 ohm termination. As you can see, the 20Vpp is divided by 2, maybe because of a voltage divider. For my MSO5000, I cannot change input impedance. The button is grayed out and it says 1 M ohm.

Any suggesstions why there is no difference between HighZ and 50 ohm on DG800? What am I doing wrong?

Yes MSO5000 has only 1M input, no switchable 50 ohm.

The spec of the DG800 is 10Vpp into 50 ohm, so that is the most output you will see into a matched load. In your case if you connect 1M, this doubles, as you have a 50 ohm source and no 50 ohm loading it down.
Usually, when you switch from 50 ohm to Hi-z, the output will not change. The UI will simply be updated to have a different displayed voltage (in this case, it should show 10V, if you switch to Hi-Z then 20V).
But maybe as bd139 is hinting at, could be a UI bug.

You can try a lower voltage and see if the result is the same (1Vpp say), the UI should update when switching modes.

https://www.keysight.com/main/editorial.jspx%3Fckey%3D1948055%26id%3D1948055%26nid%3D-11143.0.00%26lc%3Djpn%26cc%3DJP?&cc=CA&lc=eng
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 
The following users thanked this post: PushUp

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 9057
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #136 on: May 17, 2019, 09:48:45 pm »
Correct, the DG's output impedance is always 50Ω. The setting in the DG simply changes the amplitude value that's displayed, assuming that you're connecting it to a device with an input impedance that matches the setting you chose.

On an oscilloscope with switchable input impedance, the setting does change the input impedance.

It's a bit confusing, but that's the story.
TEA is the way. | TEA Time channel
 

Offline timber23

  • Contributor
  • Posts: 47
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #137 on: May 18, 2019, 09:30:40 am »
You are right. I did some tests on 1 Vpp. And as you said, if I change to 50 ohm than the output amplitude is automatically set to 500 mVpp.

Here are some screenshots. DG811 on HighZ:


Resulting measurement on MSO5000 without 50 ohm termination:


Now, DG811 on 50 ohms:


Resulting measurement on MSO5000 without 50 ohm termination


Resulting measurement on MSO5000 with 50 ohm termination


It was a litte bit difficult to obtain the screenshots from DG811 without using an USB disk. Here is the code which I wrote to do it with Python and PyVISA:
Code: [Select]
import visa
import datetime

def main():
    rm = visa.ResourceManager()
    res = rm.list_resources()
    my_instrument = rm.open_resource(res[0])
    my_instrument.write(':HCOP:SDUM:DATA:FORM PNG')
    print ('Screenshot format: '+my_instrument.query('HCOPy:SDUMp:DATA:FORMat?'))
   
    my_instrument.write(':HCOPy:SDUMp:DATA?')
   
    bmpdata = my_instrument.read_raw(391734+11)
    bmpdata = bmpdata[11:]
    my_instrument.close()
    fileName = datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
    fileName = 'Screenshot_'+fileName+'.PNG'
    with open(fileName,"wb") as f:
        f.write(bmpdata)
   
if __name__=='__main__':
    main()

Best regards
timber23
 
The following users thanked this post: PushUp

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23018
  • Country: gb
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #138 on: May 18, 2019, 09:40:53 am »
Yeah that's all correct  :-+
 

Offline maxwell3e10Topic starter

  • Frequent Contributor
  • **
  • Posts: 869
  • Country: us
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #139 on: May 18, 2019, 01:53:16 pm »
Thanks for testing. Even with 50 Ohm termination the square wave has some ringing. It seems fairly bad compared with other generators.
 

Online TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #140 on: May 18, 2019, 02:45:33 pm »
Well, I couldn't resist the possible opportunity to get a quite well performing, full featured generator at the price of the entry level model. Not that I'ld really need another AWG, but I'm probably a bad case of TEA and once in a while I've just got to get another piece...

Actually, this time I really followed the EEVBlog rules and took the unit apart before I turned it on. So here's a sequence of posts with just photos of the interior of a DG811. Actually, I found nothing to complain about, the build quality is first class and I really like how the metal chassis is integrated in the instrument. Disassembly is easy and all the boards seem to be well laid out. The quality of the switch mode transformer impressed me, it's actually wound with triple-insulated magnetic wire despite a decent layer insulation. The PSU appears to be intended to be used in several instruments just by the fact that a mains frequency coupler is included (hence it might be used in an oscilloscope, providing a source for mains trigger). Output voltages are 5V, +-7V and +-17V. The solder side of the PCB is actually conformally coated with the mounting pads spared out.

I'm quite surprised Rigol used a plug-in CPU board in this instrument. Maybe they intend to design several instruments around this platform. It's funny that there's an unpopulated LAN footprint on the PCB that supposedly requires a connector with integrated isolation transformer. Who will be the first one to install it in order to figure out if it works?

An now for the teardown photos -- Enjoy!
« Last Edit: May 20, 2019, 06:41:16 pm by TurboTom »
 
The following users thanked this post: thm_w, bitseeker, PushUp, nikitasius

Online TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #141 on: May 18, 2019, 02:46:52 pm »
More Photos...

BTW, now we know that Rigol's internal designation of the DG800/900 series is "Sardine"  :-DD
If that name somehow relates to the design of the instrument's enclosure? Gills everywhere...

P.S.The laser-etched output amp is most likely a THS3091. Or could it be a THS3491??
Edit: It's now verified it's a THS3095, so basically a THS3091 with shutdown.
« Last Edit: August 08, 2021, 07:52:31 pm by TurboTom »
 
The following users thanked this post: thm_w, bitseeker, PushUp, nikitasius

Online TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #142 on: May 18, 2019, 02:51:25 pm »
Still more Photos...

BTW, the DAC is an AD9747 -- I wonder why "some" companies don't understand that it's no use lasering the markings off a chip... except maybe being unable to find out that the company skimped with the speed grading...
« Last Edit: May 18, 2019, 08:49:09 pm by TurboTom »
 
The following users thanked this post: thm_w, bitseeker, PushUp, maxwell3e10, nikitasius

Online TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #143 on: May 18, 2019, 02:52:32 pm »
Yet still more Photos...
 
The following users thanked this post: bitseeker, PushUp, 2N3055, timber23, nikitasius

Offline tv84

  • Super Contributor
  • ***
  • Posts: 3212
  • Country: pt
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #144 on: May 18, 2019, 03:10:46 pm »
BTW, now we know that Rigol's internal designation of the DG800/900 series is "Sardine"  :-DD
If that name somehow relates to the design of the instrument's enclosure? Gills everywhere...

The executable is also "Sardine"!  :)

BTW, nice job!  :clap:
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23018
  • Country: gb
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #145 on: May 18, 2019, 03:25:07 pm »
Indeed. Very nice job. Thanks for this  :-+
 

Offline tv84

  • Super Contributor
  • ***
  • Posts: 3212
  • Country: pt
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #146 on: May 18, 2019, 08:45:02 pm »
HOW TO CHANGE RIGOL DG800/DG900 MODELS

1. Get a blank formatted USB disk (capacity >= 1GB)

2. Write signature in the disk with the following command (in linux):

        sudo dd if=DG800_sardinha.bin of=/dev/sda1 bs=512 seek=2000000

3. With special USB disk inserted, send the SCPI command (for example) to change model to DG992:

        :PROJ:MODE DG992

Possible model types (as seen in the first 4 bytes of sd.xx configuration file):
00 - DG811
01 - DG821
02 - DG831
03 - DG812
04 - DG822
05 - DG832
06 - DG952
07 - DG972
08 - DG992

ATTENTION: Use a blank disk and be sure it's mounted as /dev/sda1. Don't blame me if you do things wrong!!

(I included some pics kindly sent by a voluntary member)

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23018
  • Country: gb
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #147 on: May 18, 2019, 08:51:29 pm »
LOL. Wondering if I should grab a DG811 now before they all disappear in a puff of smoke :-DD
 

Offline timber23

  • Contributor
  • Posts: 47
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #148 on: May 18, 2019, 09:14:42 pm »
Did someone managed to get a USB to ethernet adapter working? I have a super cheap adapter from action.com here, but DG811 aka DG992 is not getting an IP-address via DHCP.

Any suggestions? I had a look on amazon's offers, but I didn't find any information about the used chipsets.
 

Offline frozenfrogz

  • Frequent Contributor
  • **
  • Posts: 936
  • Country: de
  • Having fun with Arduino and Raspberry Pi
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #149 on: May 18, 2019, 09:15:16 pm »
HOW TO CHANGE RIGOL DG800/DG900 MODELS

1. Get a blank formatted USB disk (capacity >= 1GB)

2. Write signature in the disk with the following command (in linux):

        sudo dd if=DG800_sardinha.bin of=/dev/sda1 bs=512 seek=2000000

Thanks for posting the info! :)
Where do I get the binary file? Is it already in this thread and I missed it?
He’s like a trained ape. Without the training.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf