Hacking the Rigol DHO800/900 Scope

[Thripster]

I thought he stated at the start that the battery pack was 20,000mAh?

[Nikki Smith]

How practical would it be to hack the scope to display HDMI input?

I don't have a lot of workspace, and it would be nice to be able to have the scope display double up with a schematic or datasheet from my PC, or from my microscope. I guess it would need a 2nd HDMI port and some extra circuitry on a little PCB to switch between the HDMI input and the scope?

Thanks

[norbert.kiszka]

How practical would it be to hack the scope to display HDMI input?

I don't have a lot of workspace, and it would be nice to be able to have the scope display double up with a schematic or datasheet from my PC, or from my microscope. I guess it would need a 2nd HDMI port and some extra circuitry on a little PCB to switch between the HDMI input and the scope?

Thanks

You need to recompile Linux kernel, since display drivers are compiled as built in. Linux kernel is on GPV v2 license, so theoretically Rigol should give source code to You.

[Fungus]

How practical would it be to hack the scope to display HDMI input?

About 0.000001% practical.

I don't have a lot of workspace, and it would be nice to be able to have the scope display double up with a schematic or datasheet from my PC, or from my microscope. I guess it would need a 2nd HDMI port and some extra circuitry on a little PCB to switch between the HDMI input and the scope?

It's Android so there's lots of ways to do that with no hardware hacking needed:

It has a built in web browser to view web pages.

Installing a PDF viewer is possible.

You can probably install an app to use it as a second monitor for your PC - probably the most useful option but I haven't tried that one yet.

Start by attaching a keyboard and/or one of the "alternate UIs" that are mentioned in this thread so you can easily get to the Android home screen. After that you need an app launcher and your 'scope becomes and ordinary Android device.

You can also install/launch Android apps from your PC using ADB.

[norbert.kiszka]

You can probably install an app to use it as a second monitor for your PC - probably the most useful option but I haven't tried that one yet.

Probably You are user of a magic glass ball.

With a precompiler definitions (Linux kernel config file), screen mirroring is hardcoded. Both output screens are using exact same memory region as a frame buffer. Using it as a separate screen is only possible after recompiling the kernel.

[Fungus]

With a precompiler definitions (Linux kernel config file), screen mirroring is hardcoded. Both output screens are using exact same memory region as a frame buffer. Using it as a separate screen is only possible after recompiling the kernel.

An app can receive screen data over WiFi and display it independently of the kernel.

[Hix]

Does anyone have some decent quality photos of the PCB for the DHO804? I'm trying to ascertain if the board is different from the 814.

[Fungus]

Does anyone have some decent quality photos of the PCB for the DHO804? I'm trying to ascertain if the board is different from the 814.

It isn't.

[Duglum]

I apologize if this has been answered before, i've probably read half this thread by now and couldn't find a definite answer.

So, i have a DHO804 on the way. I know about adb with rigol_vendor_bin to change the model and unlock all options, that shouldn't be a problem. I also know about the alternative UIs.

But: Should i go with a DHO824 or a 924 as Model?
Some people wrote that going with the 924 decreases accuracy and makes the scope worse, others said that running the self calibration fixes all potential problems. I'm not sure what is the correct information.
What advantages would the 924 software even have? (Besides the theoretical 200Mhz vs. 250Mhz)


Have people that went the 924 route successfully connected a LA probe to the pin header that exists internally? (Not that i need a LA at the moment..)

[norbert.kiszka]

Personally I have DHO924S and I don't have problem with hacking from 800 to 900, but some people said that 800 with vendor_bin from 900, is not very good idea, because it's not working correctly. So my advise is to hack it up to the 824 with 200 MHz bandwidth and little less decoders.

Eventually You can modify libscope-auklet.so library file to gain everything from DHO924 without issues from vendor_bin. But this is far from simple job.

[Fungus]

What advantages would the 924 software even have? (Besides the theoretical 200Mhz vs. 250Mhz)

Extras: It has CAN and LIN decoders.

Downsides: More UI clutter on screen and higher bandwidth is more disproportional to the sample rate when you're using more than two channels.

[Fungus]

So my advise is to hack it up to the 824 with 200 MHz bandwidth and little less decoders.

The 824 has the 50M memory option - important.

[pict]

I have a newly acquired DHO804 with 1.03.00.06 firmware installed and wanted to make a clone of the SD card prior to firmware upgrade in case of corruption. I am using a MacBook and hoped to image the uSD card using Disk Utility or gdd in terminal. However the Rigol uSD card is not recognized and does not mount on the MacBook. Any idea what tools would I need to use in order to achieve this with macOS please (or alternative OS/software if not doable)? I understand I can pull the contents with adb, but I am concerned that in the event of formatting corruption of the card I will be stuck if I go this route. Would feel happier with a known bootable clone before I start to mess with it. Possible?

Also wondering how I can unpack the firmware .GEL file with a Mac (or again, alternative OS/software if not doable)?
 

[Fungus]

Also wondering how I can unpack the firmware .GEL file with a Mac (or again, alternative OS/software if not doable)?

You don't have to unpack it.

(It's a Unix .tar file if you just want to look inside)

[pict]

Also wondering how I can unpack the firmware .GEL file with a Mac (or again, alternative OS/software if not doable)?

You don't have to unpack it.

(It's a Unix .tar file if you just want to look inside)

Thanks! That was easy just changing the extension.

[norbert.kiszka]

However the Rigol uSD card is not recognized and does not mount on the MacBook

There is no system that will "recognize" it, because it doesn't have partition table. However, there are partitions inside.

[Fungus]

I have a newly acquired DHO804 with 1.03.00.06 firmware installed and wanted to make a clone of the SD card prior to firmware upgrade in case of corruption.

Not really necessary.

Firmware is downgradable and there's images out there if things go wrong.

The ONLY thing that makes your Rigol unique are two files in /rigol/data (vendor.bin and RKey.data).

Make a copy of those two files and you're good.

[pict]

I have a newly acquired DHO804 with 1.03.00.06 firmware installed and wanted to make a clone of the SD card prior to firmware upgrade in case of corruption.

Not really necessary.

Firmware is downgradable and there's images out there if things go wrong.

The ONLY thing that makes your Rigol unique are two files in /rigol/data (vendor.bin and RKey.data).

Make a copy of those two files and you're good.

Sorry new to this so not really getting it. If the SD card fails how do I rebuild one with a firmware release? My understanding is that the scope will not boot without the SD card, so I am under the impression that in this event I could not use the scope menu to install new firmware on a blank card as it would not start up. Assuming I had a backup of my original vendor.bin and RKey.data, a firmware release, and a blank 32Gb uSD card, how could I get it going again please?

[norbert.kiszka]

Now I have 62.5 Mpts memory depth in my DHO924S. However only in auto depth.

Originally it was hardcoded to do maximum 100k in auto.

Any questions?

[Proxy64]

Hi all,
Can we run rigol_vendor_bin directly on the scope, as for generate_all_options (I don't have any Windows system) ?

I tried but got this:
./rigol_vendor_bin: not executable: 32-bit ELF file 

[norbert.kiszka]

Linux 64 bit without 32 bit libraries?

https://github.com/norbertkiszka/rigol_vendor_bin

[shapirus]

Any questions?

Of course.

1) how?
2) any issues/side effects?

[norbert.kiszka]

Any questions?

Of course.

1) how?
2) any issues/side effects?

Ad. 1. I did many changes - most of it by low level changes in libscope-auklet.so. Im planing to do a separate and public repository for it, after I test how many changes are needed exactly. Especially because I have also experimental changes - most likely harmless, but I don't want to risk making somebody day bad because of it.

Ad. 2. When I increased limit to 200M and when I increased time base, at some point it was 80M and frequency measurement was bad (like ~10 MHz instead of ~17 MHz). At around 100M or 125M (I don't remember precisely and text with sample rate was hiding part of text of memory depth), waveform was flickering, but single shot (with frequency measurement) was ok - as far I tested. Going further with time base waveform disappeared (likely FPGA temporary stopped his work for some reason)  After lowering it to the 62.5M it looks ok - as far I tested.

80M or 100M maybe will work after some further changes. Some of compiled code is from other scopes (You can put vendor.bin from DHO1000 and DHO4000 but it's unusable practically) and that 80M was likely it (not intended by original developers, they just used 'ifs' instead of precompiler directives).

[norbert.kiszka]

Ok, so I made it from scratch and looks like only one change was necessary to do it. Exactly one very small and easy change in libscope-auklet.so.

Additionally I tested 78.125M (which additionally gives "strange" values of sample rate) but measurements was also bad like in 80M. So I reverted it back to 62.5M.

Im not 100% sure, but I think measurements are done by FPGA - which kinda explains why this happens with bad measurements results.

Bon appétit: https://github.com/norbertkiszka/Sparrow_DHO800_DHO900

[norbert.kiszka]

I did more experiments and turns out 125M limit is working ok. I pushed it into my repository.

Likely my other experimental changes (not published) caused problems with more than 62.5M.