Author Topic: New Rigol 16-bit function generators DG800/900 series  (Read 192457 times)

0 Members and 1 Guest are viewing this topic.

Offline battlecoder

  • Regular Contributor
  • *
  • Posts: 156
  • Country: cl
    • A BIT of Mystery
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #625 on: October 17, 2020, 09:48:00 pm »
I've been meaning to purchase a 822 for a while now, and this thread is not helping me abandon the idea.  ;D
I like that they can be hacked to unlock more features, but I guess I wouldn't make it behave as a model that exceeds its calibration point if I don't have how to make it work correctly in the full range (So I guess I'd transform it into a DG952 at most).

Anyway, I'll be following this thread now because I have a problem and can't stop buying devices I don't really need.  :-DD

(Also, thanks to all the amazing people in this threat who has contributed to unearth the secrets of these devices, and unlock their capabilities)
 

Offline SMB784

  • Frequent Contributor
  • **
  • Posts: 421
  • Country: us
    • Tequity Surplus
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #626 on: October 17, 2020, 10:22:35 pm »
I've been meaning to purchase a 822 for a while now, and this thread is not helping me abandon the idea.  ;D
I like that they can be hacked to unlock more features, but I guess I wouldn't make it behave as a model that exceeds its calibration point if I don't have how to make it work correctly in the full range (So I guess I'd transform it into a DG952 at most).

Anyway, I'll be following this thread now because I have a problem and can't stop buying devices I don't really need.  :-DD

(Also, thanks to all the amazing people in this threat who has contributed to unearth the secrets of these devices, and unlock their capabilities)

Don't spend the extra on the 822, just get the 811.  The second channel is enabled through software described above.  No reason to pay extra when the 811 does what you need!
 
The following users thanked this post: battlecoder

Offline battlecoder

  • Regular Contributor
  • *
  • Posts: 156
  • Country: cl
    • A BIT of Mystery
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #627 on: October 17, 2020, 10:37:43 pm »
Don't spend the extra on the 822, just get the 811.  The second channel is enabled through software described above.  No reason to pay extra when the 811 does what you need!
I would normally agree on that, but if -for whatever reason- I'm unable to perform the "upgrade" I would not like to be stuck with only 1 channel. I will ask the store if they have stock of the 811 if (when) I do the purchase. If the price difference is considerable then I'll definitely go for the 811, but considering how wildly different pricing is here, that may not be the case.
 

Offline TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #628 on: October 17, 2020, 11:36:43 pm »
@rea:

Thanks a lot for pointing me to the right direction. The instruction how to convert the supplied private key file to the format PuTTY is expecting, was what I had been missing... Now the login works a treat and I already transfered the complete file system to my PC using "psftp".

I went through most of the files and as it seems, there is no interactive calibration function available in these AWGs (DG800/900/2000). It appears that Rigol only made provisions for calibrating these units via SCPI commands. If we would be able to retrieve a set of cal files from a genuine DG992, we might be able to assemble a hybrid cal file by patching the resident cal files with the additional information (always assuming that the hardware won't differ too much -- tolerance-wise that is). I may be completely in error with this assumption, though. But it may anyway help us to understand how the cal files are structured. For comparison, I attached the cal files of my AWG that started life as a DG811.

@battlecoder: Don't worry and just get the DG811. The hardware of the latter and the DG900 is identical except for the casing colour (...). This can be officially verified since considerable time ago, a chinese language service manual for these instruments surfaced, stating exactly the same P/N for the main PCB for these two instruments. Moreover, AFAIK there even is/was an update option by Rigol to add the second channel to single-channel versions of the AWG.

Edit: both attached cal files had the extension *.dat, had to rename them to meet the forum attachment requirements
« Last Edit: October 18, 2020, 09:00:01 pm by TurboTom »
 

Offline SMB784

  • Frequent Contributor
  • **
  • Posts: 421
  • Country: us
    • Tequity Surplus
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #629 on: October 18, 2020, 12:57:26 am »
@rea:

Thanks a lot for pointing me to the right direction. The instruction how to convert the supplied private key file to the format PuTTY is expecting, was what I had been missing... Now the login works a treat and I already transfered the complete file system to my PC using "psftp".

I went through most of the files and as it seems, there is no interactive calibration function available in these AWGs (DG800/900/2000). It appears that Rigol only made provisions for calibrating these units via SCPI commands. If we would be able to retrieve a set of cal files from a genuine DG992, we might be able to assemble a hybrid cal file by patching the resident cal files with the additional information (always assuming that the hardware won't differ too much). I may be completely in error with this assumption, though. But it may anyway help us to understand how the cal files are structured. For comparison, I attached the cal files of my AWG that started life as a DG811.

@battlecoder: Don't worry and just get the DG811. The hardware of the latter and the DG900 is identical except for the casing colour (...). This can be officially verified since considerable time ago, a chinese language service manual for these instruments surfaced, stating exactly the same P/N for the main PCB for these two instruments. Moreover, AFAIK there even is/was an update option by Rigol to add the second channel to single-channel versions of the AWG.

Edit: both attached cal files had the extension *.dat, had to rename them to meet the forum attachment requirements

Apparently the cal directory (under /root/cal) is mounted at some point.  See the bash script file I attached and look for the /root/cal section

Also, can you do a search of the file system for a file called ServCalibration.cpp?
« Last Edit: October 18, 2020, 01:08:22 am by SMB784 »
 

Online tv84

  • Super Contributor
  • ***
  • Posts: 3213
  • Country: pt
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #630 on: October 18, 2020, 09:03:05 am »
If we would be able to retrieve a set of cal files from a genuine DG992, we might be able to assemble a hybrid cal file

That's an excellent, and probably the best, idea.

Also, can you do a search of the file system for a file called ServCalibration.cpp?

That is the source file in development (C++). You only get its compiled output.
 

Online tv84

  • Super Contributor
  • ***
  • Posts: 3213
  • Country: pt
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #631 on: October 18, 2020, 10:22:38 am »
Focusing on CalibrationData.dat, and trying to reduce our analysis area:

The file has 2 exact halves (maybe for backup). So we can split it in 2 and disregard the 2nd half (#2).

Just looking at 1st half (#1):

If we take into account that the first 2-bytes are CRC/checksum, we can further split the file in 2, assuming that there are 2 channels calib, at the 0x5824 mark.

If we compare these remaining 2 files (1a and 1b), we see lots of similarities. Maybe we can even swap both blocks and that would swap the calib of both channels...

Attached is a visual mapping (raw) of half #1 and the splittings that I talked about.

If anyone wants to try assembling something I may have a look at the CRC/checksum to see how that's done.
 
The following users thanked this post: TurboTom

Offline TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #632 on: October 18, 2020, 03:54:09 pm »
The calibration files appear to consist of 32 bit signed integers. I removed a few bytes at the start to align the four-byte boundaries properly and converted them to decimal, keeping 16-byte groups so it's comparable with the typical hex editor arrangement. There are surprisingly many "round" numbers in the list, whatever that may mean. But have a look yourself...
 

Offline SMB784

  • Frequent Contributor
  • **
  • Posts: 421
  • Country: us
    • Tequity Surplus
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #633 on: October 18, 2020, 04:22:13 pm »
Focusing on CalibrationData.dat, and trying to reduce our analysis area:

The file has 2 exact halves (maybe for backup). So we can split it in 2 and disregard the 2nd half (#2).

Just looking at 1st half (#1):

If we take into account that the first 2-bytes are CRC/checksum, we can further split the file in 2, assuming that there are 2 channels calib, at the 0x5824 mark.

If we compare these remaining 2 files (1a and 1b), we see lots of similarities. Maybe we can even swap both blocks and that would swap the calib of both channels...

Attached is a visual mapping (raw) of half #1 and the splittings that I talked about.

If anyone wants to try assembling something I may have a look at the CRC/checksum to see how that's done.

If indeed one of the halves is a backup, I wonder if that backup is common to all models of with the same model number.  I have attached my CalibrationData.dat file to compare.  I tried doing a diff between the hexdumps of TurboTom's file and my own, and I saw some significant differences, as well as some areas that were common between both.  Not sure what to make of that just yet.

The calibration files appear to consist of 32 bit signed integers. I removed a few bytes at the start to align the four-byte boundaries properly and converted them to decimal, keeping 16-byte groups so it's comparable with the typical hex editor arrangement. There are surprisingly many "round" numbers in the list, whatever that may mean. But have a look yourself...

Its possible that these 32 bit signed integers actually represent decimal numbers, as it isn't possible to tell where the decimal is intended to go in a binary/hexadecimal number.  I wonder if that is what we are looking at.
« Last Edit: October 18, 2020, 04:33:41 pm by SMB784 »
 

Offline battlecoder

  • Regular Contributor
  • *
  • Posts: 156
  • Country: cl
    • A BIT of Mystery
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #634 on: October 19, 2020, 09:17:14 pm »
@battlecoder: Don't worry and just get the DG811. The hardware of the latter and the DG900 is identical except for the casing colour (...). This can be officially verified since considerable time ago, a chinese language service manual for these instruments surfaced, stating exactly the same P/N for the main PCB for these two instruments. Moreover, AFAIK there even is/was an update option by Rigol to add the second channel to single-channel versions of the AWG.
Unfortunately the *ONE* company importing that line of devices here does not have the DG811. The lower they go is the DG822, from what they told me today.
 

Offline u271D

  • Contributor
  • Posts: 41
  • Country: ua
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #635 on: October 31, 2020, 07:51:07 pm »
bump  Looks like some progress was forthcoming on the calibration issue. Then the thread died.
Great work though!!

\Thanks/   
 

Offline Coder69

  • Newbie
  • Posts: 9
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #636 on: November 01, 2020, 09:13:16 am »
If there is a calibration file for a DG992 available it would be possible to compare ist with files for DG800.
One shot could be to use the DG992 file in a DG800 and check out the impact.
Thanks for your work, it is very interesting to read this thread  ;D
 

Offline SMB784

  • Frequent Contributor
  • **
  • Posts: 421
  • Country: us
    • Tequity Surplus
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #637 on: November 01, 2020, 07:05:26 pm »
bump  Looks like some progress was forthcoming on the calibration issue. Then the thread died.
Great work though!!

\Thanks/

It hasn't died, I have just been traveling for the past few days. But I'm back now and will be trying done SCPI commands for calibration purposes soon.

Offline Coder69

  • Newbie
  • Posts: 9
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #638 on: November 05, 2020, 06:22:08 pm »
As I have just successfully upgraded my DG812 to DG992 I would like to post yet another summary of the required steps.

0. Note your DG's serial number if you care.

1. If you have software version higher than 00.01.08.xx.xx you need to downgrade to version 1.08 using specially modified GEL file to allow downgrade (as described by evgbog here: https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/msg3017342/#msg3017342).
My contribution is that I've prepared modified GEL file, available here:
https://drive.google.com/file/d/16nC-uJ12MXilJElv4Eukvt1IW4RzGdb7/view?usp=sharing
Put the file on USB stick, plug it into your DG and accept "update" when popup shows. WARNING! This "upgrade" will reset your SN to DG80000000001 but convert your unit into DG992 immediately. This is the effect of the downgrade itself - the firmware is official Rigol firmware. The modification only changes the boot script to allow downgrade.

As a side note, I don't see a point in using hex editor in evgbog method of preparing the GEL. Those GELs are tar files as mentioned by tv84. I prepared the file by simply untaring it, decrypting fw4uboot.sh, modifying it, encrypting it back (you need to maintain file size) and finally TARing everything back into GEL file using 7zip.

2. Prepare a "magic" USB stick. Format USB stick as FAT32 (stick MUST BE BIGGER than 1GB). Open it in HxD editor as LOGICAL DISK (not physical disk - I did that mistake) Tools->Open Disk (uncheck "open as readonly"). Jump to sector 2000000 (sector field is at the top right corner). Enter value manually or use "Paste write" in edit menu (Ctrl+B). Magic value: 0B0A3B2E5F4CECBE

3. Insert the magic stick into DG (make sure it is recognized and USB icon is shown) and connect the instrument with the computer (using USB A-B cable and downloading Rigol's bloatware UltraSigma is the easiest way). Open SCPI command panel, check communication by issuing *IDN? command (use Send & Read). Send command :PROJ:SN DGxxxxxxxxxx with your correct SN number noted in step 0 (use Send command, DO NOT USE Send & Read). Verify if it worked by sending another *IDN? command - it should report back with correct serial number. Now send command :PROJ:MODE DG992. You can disconnect the instrument from PC and remove USB stick.

4. Edit: This step is only needed if your purchase version was higher than v1.08 and you did step 1 Download tv84's "Finish_Upgrade_2.04.zip" from this post: https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/msg3119264/#msg3119264
Put this GEL file on USB stick, plug into DG and confirm update when popup shows.

5. Download official 2.04 firmware from Rigol: https://int.rigol.com/Public/Uploads/uploadfile/files/ftp/DG/软固件/SardineII(ARM)Update-v00.02.04.00.00.rar
No logging in required :-DD
Put GEL file on USB stick (delete the previous one), upgrade DG and enjoy ;D
Serial number should be kept and model should be DG992.

All credits go to tv84, bulba99, evgbog and others who helped to create this hack!

I have a DG811 with firmware v00.02.02.00.00 on the bench and I am not sure how to proceed. I tried to follow the instruction above, but unfortunately step 1 doesn't work. he USB stick is recogniized and  a message box pops up and I confirm the update procedure. Then I can see a "Wait" message and can hear the clicking if the internal relais. After then the message box pops up and asks again for the update.
I read almost the complete thread and also read that the downgrade is locked in 2.02/2.04. That means that the instruction above only valid for V1.09, doesn't it ?
I have also read that there is a procedere for 2.02. Can anybody help ? Thanks in advance :-)
 

Offline jscm2000

  • Contributor
  • Posts: 25
  • Country: cn
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #639 on: November 07, 2020, 09:44:48 am »
As I have just successfully upgraded my DG812 to DG992 I would like to post yet another summary of the required steps.

0. Note your DG's serial number if you care.

1. If you have software version higher than 00.01.08.xx.xx you need to downgrade to version 1.08 using specially modified GEL file to allow downgrade (as described by evgbog here: https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/msg3017342/#msg3017342).
My contribution is that I've prepared modified GEL file, available here:
https://drive.google.com/file/d/16nC-uJ12MXilJElv4Eukvt1IW4RzGdb7/view?usp=sharing
Put the file on USB stick, plug it into your DG and accept "update" when popup shows. WARNING! This "upgrade" will reset your SN to DG80000000001 but convert your unit into DG992 immediately. This is the effect of the downgrade itself - the firmware is official Rigol firmware. The modification only changes the boot script to allow downgrade.

As a side note, I don't see a point in using hex editor in evgbog method of preparing the GEL. Those GELs are tar files as mentioned by tv84. I prepared the file by simply untaring it, decrypting fw4uboot.sh, modifying it, encrypting it back (you need to maintain file size) and finally TARing everything back into GEL file using 7zip.

2. Prepare a "magic" USB stick. Format USB stick as FAT32 (stick MUST BE BIGGER than 1GB). Open it in HxD editor as LOGICAL DISK (not physical disk - I did that mistake) Tools->Open Disk (uncheck "open as readonly"). Jump to sector 2000000 (sector field is at the top right corner). Enter value manually or use "Paste write" in edit menu (Ctrl+B). Magic value: 0B0A3B2E5F4CECBE

3. Insert the magic stick into DG (make sure it is recognized and USB icon is shown) and connect the instrument with the computer (using USB A-B cable and downloading Rigol's bloatware UltraSigma is the easiest way). Open SCPI command panel, check communication by issuing *IDN? command (use Send & Read). Send command :PROJ:SN DGxxxxxxxxxx with your correct SN number noted in step 0 (use Send command, DO NOT USE Send & Read). Verify if it worked by sending another *IDN? command - it should report back with correct serial number. Now send command :PROJ:MODE DG992. You can disconnect the instrument from PC and remove USB stick.

4. Edit: This step is only needed if your purchase version was higher than v1.08 and you did step 1 Download tv84's "Finish_Upgrade_2.04.zip" from this post: https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/msg3119264/#msg3119264
Put this GEL file on USB stick, plug into DG and confirm update when popup shows.

5. Download official 2.04 firmware from Rigol: https://int.rigol.com/Public/Uploads/uploadfile/files/ftp/DG/软固件/SardineII(ARM)Update-v00.02.04.00.00.rar
No logging in required :-DD
Put GEL file on USB stick (delete the previous one), upgrade DG and enjoy ;D
Serial number should be kept and model should be DG992.

All credits go to tv84, bulba99, evgbog and others who helped to create this hack!

I have a DG811 with firmware v00.02.02.00.00 on the bench and I am not sure how to proceed. I tried to follow the instruction above, but unfortunately step 1 doesn't work. he USB stick is recogniized and  a message box pops up and I confirm the update procedure. Then I can see a "Wait" message and can hear the clicking if the internal relais. After then the message box pops up and asks again for the update.
I read almost the complete thread and also read that the downgrade is locked in 2.02/2.04. That means that the instruction above only valid for V1.09, doesn't it ?
I have also read that there is a procedere for 2.02. Can anybody help ? Thanks in advance :-)

After booting up, immediately press the help button to downgrade.
 
The following users thanked this post: Coder69

Offline Coder69

  • Newbie
  • Posts: 9
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #640 on: November 07, 2020, 07:07:31 pm »
Thanks, jscm2000. It was possible to downgrade to firmware version 1.08.01
The device is now a DG992 with serial number 0. I tried out the next step to prepare the magic USB-stick, which didn't work. A new serial number is not accepted by DG. I tried three sticks (2GB, 8GB and 16GB). The 2GB stick wasn't recognised, but after connecting the 8GB and the 16GB a box opened with the message "Wait". I tried out to use HXD (Windows 10) and tried out the command line tv84 posted with the same negative result. At the moment I am done (worked 3h on this  |O)
I read that some guys had similar problems. Maybe one of you can give a broad hint  :D
 

Offline SMB784

  • Frequent Contributor
  • **
  • Posts: 421
  • Country: us
    • Tequity Surplus
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #641 on: November 07, 2020, 09:56:16 pm »
After booting up, immediately press the help button to downgrade.

Unrelated to your post, but are you willing to share your calibration procedure with us?  It would be most helpful  ;)
 
The following users thanked this post: CeD

Offline Noy

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #642 on: November 07, 2020, 11:47:45 pm »
Anybody thought about replacing the opamp with THS3491 one? Will it improve the device even more?


And another question, im considering of buying a DG811 but i already have a FY6800 modded (THS3491, TCXO, FAN).
Is it a worth upgrade? FY6800 has VCO Function, 20Vpp below 10MHz. Ramp up to 10 MHz.

But DG811 has probably better Software + higher ARB Function, higher sinus...

Maybe it is with the DG811 a better combination with my MSO5000? Regarding arb Function recording/replay? Or should i stay with FY6800 + internal MSO5000? But recently a Juntek DP-2698 amplifier for the MSO5000 but still quite limited...

Is it worth to buy a GPSDO to use with DG811 as timebase?
« Last Edit: November 08, 2020, 01:47:51 am by Noy »
 

Offline Coder69

  • Newbie
  • Posts: 9
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #643 on: November 08, 2020, 11:13:10 am »
It is done. I could finish the hack. Let me sum up the single steps, which worked for me:
My device was delivered with firmware 2.02

1.
Install "Ultra Sigma"-Software from Rigol homepage

2.
Note your serial number (press the "Utility" key and touch "System Info")

3.
Prepare an USB-Stick (FAT32) with the downgrade file. You can find the link here:
https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/msg3157600/#msg3157600
Copy only the GEL-file, not the complete ZIP-file !
Switch off your DG and insert the USB-stick.
Switch on the DG and press the "Help" key continously. The bootloader should now print a message if you want to program the firmware on the stick.
(In my case only this installation was possible. Plug in the stick during operation didn't work. The update was recognised but not processed)

4.
Prepare the magic USB-stick (> 1GB). This was the most difficult part. I tried it out with Windows 10 and HXD and it didn't work. Even writing the bytes with Linux didn't work. In the end I tried out to format the stick with Linux and write then the bytes. This worked !
You can find the binary file which contains the byte sequence to be programmed in the attachment (DG800_sardinha.zip). Unzip it to place which you can access to from your booted Linux. This could be another USB-Stick
Start a Linux installation, format the USB-stick and write the byte sequence.
I prepared an USB-Stick with Rufus
https://rufus.ie/en_IE.html
and used a small linux ISO:
http://blog.puppylinux.com/fossapup64-release
(I had to disable the security boot in BIOS to be able to boot from the stick)

Format the USB-stick (check out that you choose the correct device !!! In my case it was sda1)
The Linux command is:
sudo mkfs.vfat /dev/sda1

Mount the drive on which you have copied the file with the byte sequence (DG800_sardinha.bin). I used an additional USB-stick which was device sda2.
The linux command to write the bytes was in my case:
sudo dd if=/mnt/sda2/DG800_sardinha.bin of=/dev/sda1 bs=512 seek=2000000

5.
Switch on your DG (without the magic stick). After booting plug in the stick. A message box pops up with the text "Wait". The box is visible for ca. 6s. The time was much longer with my other sticks which failed.
Start "Ultra Sigma" and connect the USB-cable with your PC. The DG should now be recognised and now you can send the SCPI command to reprogram your serial number you have noted in step 2 (Only press the SEND-button):
:PROJ:SN DGxxxxxxxxxxx
You can check if this worked with the command:
*IDN?
The result should now be your original serial number.
I also sent the command to write the model (again only SEND-button)
:PROJ:MODE DG992
This command failed and after that there were no further commands accepted. The DG is after the downgrade to v1.08 already a DG992 so I am not sure if this command is really necessary

6.
Restart your DG again and check out if the serial number and model are correct

7.
Prepare an USB stick with GEL-File which is in the attached file "Finish_Upgrade_2.04.zip"
Plug in the stick. A "Wait"-message box pops up and after that the update-fie is recognised you need to confirm the update with "OK"

8.
Prepare an USB stick with update 2.04. Here is the link:
https://int.rigol.com/Public/Uploads/uploadfile/files/ftp/DG/%E8%BD%AF%E5%9B%BA%E4%BB%B6/SardineII(ARM)Update-v00.02.04.00.00.rar
Plug in the stick. A "Wait"-message box pops up and after that the update-file is recognised you need to confirm the update with "OK"

9.
Check out the firmware version in your DG whether everything is OK


Thanks to tv84, Vtech and all others for this hack !

Now we are all looking forward to calibrating the DG up to 100MHz  8)
« Last Edit: November 08, 2020, 05:23:26 pm by Coder69 »
 
The following users thanked this post: Gandalf_Sr, fjalvingh

Offline Noy

  • Frequent Contributor
  • **
  • Posts: 361
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #644 on: November 10, 2020, 08:29:10 am »
So i ordered one DG811. And i will sell my modded FY6800 hopefully with not to much loss...

I like the fact that it is with a AM335 with Linux. Huge benefit. I know secure boot is not doable on an AM335 (only with a special automotiv version) so from this side we are safe :-D
And the LCD signals from AM335 are the boot pins. So we can easily change the bootsource to serial loader or something else.
It looks like even the ethernet phy is mounted on the som. so only a rj45 with integrated magnetics is missing.. :-) Nice little device.

Will there be a benefit to change the Opamps to THS3491 if they are really THS3091 ones? Regarding the loss over 70MHz?
 

Offline CeD

  • Contributor
  • Posts: 19
  • Country: be
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #645 on: November 10, 2020, 11:13:33 am »
I may hacve missed something in this long thread, but FYI
- I just got a DG812 shipped with fw 2.04
- I tried downgrading it to 2.02, which went fine.

But the model is kept, and the serial is kept. If this worked for TK, it clearly doesn't seem to work all the time  :D
So unless i'm missing something, downgrading from 2.04 to 2.02 does always not change the features of a DG812.

« Last Edit: November 10, 2020, 03:08:28 pm by CeD »
 

Offline CeD

  • Contributor
  • Posts: 19
  • Country: be
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #646 on: November 10, 2020, 04:58:50 pm »
As I have just successfully upgraded my DG812 to DG992 I would like to post yet another summary of the required steps.


First of all thanks for sharing this method and the time spent on this,

I've successfully updated my DG812 to DG922 with this method, altough i'd like to leave a few comments

- My DG812 was shipped with 2.04
- It accepts to downgrade to 2.02, but that doesn't make it a DG922 :) (my message above)

So from now on, I'll be talking about vtech's procedure :
It was mandatory for me to first downgrade from 2.04 to 2.02 before anything else, as version 2.04 completely refused to downgrade to 1.08, may it be from the booted device or by pressing "help" during boot.

- Contrary to what  it says above, from 2.02, update 1.08 does not start from a booted device. If you try, it makes a popup, you click update, then a bit of "waiting" popup, then a click and then... Nothing. It refuses to downgrade. It was mandatory for me to downgrade from 2.02 to 1.08 by hitting "help" while booting.

The rest of the procedure works. ( sigma, commands, etc)

- Final comment is the "post update" script spits an "error script" if you try to load it from a booted device. Again, I had to "help-help-help" while booting to get the "post-update" script to do something. NOt sure what it does, but well, it worked and flashed smth.

Then, update to official 2.04 (can be done from an already booted device) and indeed, it keeps 922 feature set + your serial.

Here are my comments :)  thanks again to everyone involved.
« Last Edit: November 10, 2020, 05:03:30 pm by CeD »
 

Offline DaveR

  • Regular Contributor
  • *
  • Posts: 165
  • Country: gb
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #647 on: November 10, 2020, 06:29:56 pm »
Will there be a benefit to change the Opamps to THS3491 if they are really THS3091 ones? Regarding the loss over 70MHz?

Well, when your DG811 arrives, perhaps you can swap the opamps with the THS3491s from your FY6800 and let us know what happens?  Until someone actually does it, any answers you get here will just be speculation.  Perhaps the 3491s will make a difference at >70MHz but they may also alter the calibration below that frequency as well.  If the existing opamps are 3091s, your FY6800 will be quite happy to use them in exchange for the 3491s  ;)
 

Offline TurboTom

  • Super Contributor
  • ***
  • Posts: 1388
  • Country: de
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #648 on: November 10, 2020, 06:55:24 pm »
I wouldn't touch the hardware of the DG800/900 unless something's broken or it's absolutely necessary for "scientific" purposes  ;).

Removing these exposed pad SO8 amplifiers from a multi-layer PCB, assembled in a lead-free process, isn't something for the faint of heart and easily results in collateral damage if there isn't a professional rework station available and someone who knows how to use it.

Moreover, even if you swap in a THS3491, you may end up having to calibrate the low frequency range as well and not just the upper 30MHz... And who tells that Rigol doesn't already use the THS3491 -- the markings are lasered off, and while it's possible to drive the output with a single THS3091, it actually appears to be a little on the weak side and may not be able to provide the observed low distortion. The DG800/900 isn't an FY6800, the amount of sophistication is a little higher IMO.
 

Offline SMB784

  • Frequent Contributor
  • **
  • Posts: 421
  • Country: us
    • Tequity Surplus
Re: New Rigol 16-bit function generators DG800/900 series
« Reply #649 on: November 10, 2020, 07:38:07 pm »
I wouldn't touch the hardware of the DG800/900 unless something's broken or it's absolutely necessary for "scientific" purposes  ;).

Removing these exposed pad SO8 amplifiers from a multi-layer PCB, assembled in a lead-free process, isn't something for the faint of heart and easily results in collateral damage if there isn't a professional rework station available and someone who knows how to use it.

Moreover, even if you swap in a THS3491, you may end up having to calibrate the low frequency range as well and not just the upper 30MHz... And who tells that Rigol doesn't already use the THS3491 -- the markings are lasered off, and while it's possible to drive the output with a single THS3091, it actually appears to be a little on the weak side and may not be able to provide the observed low distortion. The DG800/900 isn't an FY6800, the amount of sophistication is a little higher IMO.

Seconded.  Its already hard enough to calibrate this instrument, and it is already pretty good quality with some good engineering knowhow put into it (unlike the FY6xxx).  The potential benefit is not worth the risk, in my opinion.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf