Hacking the Rigol DHO800/900 Scope

[AndyBig]

By the way, if anyone is interested, this is what the graph of the NanoVNA-F V2 output power looks like in milliwatts versus frequency.

[Fungus]

914+BW15T25 and 924 - yes, they should be absolutely identical, but it's interesting to see their BW compared to 825

It needs confirming, obviously, but there should be no difference between a DHO800 and a DHO900 with the same vendor.bin file.

[AndyBig]

It needs confirming, obviously, but there should be no difference between a DHO800 and a DHO900 with the same vendor.bin file.

No, of course I mean a software change of the oscilloscope model
There will be no difference between the 800 converted into 900 and the real 900, I am also absolutely sure of this.

[iMo]

Be careful with nanovna as a sig gen (especially the cheaper older models) - it outputs square wave (the amplitude is settable in 4 or 5 levels afaik) and the frequency is limited to some 300MHz (with an o'clocked generator chip), all freqs above are the harmonics only. This may not apply to the latest models and the older F V2..

[AndyBig]

Be careful with nanovna as a sig gen (especially the cheaper older models) - it outputs square wave (the amplitude is settable in 4 or 5 levels afaik) and the frequency is limited to some 300MHz (with an o'clocked generator chip), all freqs above are the harmonics only. This may not apply to the latest models and the older F V2..

Yes, that's true, but in theory, the RMS measurement doesn't care what shape the signal is.

[AndyBig]

I measured the bandwidth of the remaining models and options. Now the table and graph contain all the models and possible options that expand the bandwidth.
The surprise was that the 824 bandwidth is absolutely identical to the 924 bandwidth. That is, by declaring the 824 bandwidth at 200 MHz, Rigol actually gives an open bandwidth of 250 MHz (and in practice even a little more than 300 MHz)
As last time, I am attaching an archive with an Excel file, in which all the original measurement results are saved.

[shapirus]

We need a mode between DHO814 and DHO824/924...

[AndyBig]

We need a mode between DHO814 and DHO824/924...

8(1.5)4 )))

[fearas38]

Hello everyone, could you please help me with a question? I updated my DHO814 from version 00.01.02.00.02 to version 00.01.04.00.02. Is it possible to revert to 00.01.02.00.02, and how can I do that? Thanks.

[Fungus]

Hello everyone, could you please help me with a question? I updated my DHO814 from version 00.01.02.00.02 to version 00.01.04.00.02. Is it possible to revert to 00.01.02.00.02, and how can I do that? Thanks.

Yes. Just copy 1.02 .GEL file onto a USB stick and install it.

[stj]

so it doesnt have a forward-only policy like the 1000z does?
interesting.

[norbert.kiszka]

On my DHO924S kernels 5.x and 6.x had small problems with graphic drivers - it didn't work, no matter what I did - including testing 100 kernel forks from GitHub that was made to work on a RK3399. Once I did workaround (just commented out one 'if' that generates error) in a mipi-dsi driver (internal LCD is connected via mipi) and result of it was... HDMI started to work, instead of mipi 

So... without almost any knowledge about DRM Linux subsystem (bridge between kernel, graphic drivers and user space) or graphic drivers, I decided to do something really crazy - to make a port of mipi driver from Linux kernel 4.4 into 5.10. 4.4 and 5.10 internally are almost two different kernels - half things are different or deleted (from a regular user perspective it's almost the same - 5.x is much faster and safer). With many DRM api functions, I needed to read tons of documentation, articles and compare code from both kernels. Only mipi driver in 4.4 has 2195 lines of code.

With that crazy and almost impossible thing to do and after about two weeks of fight, I did it  Both mipi-dsi and HDMI works.

However, it crashes when CONFIG_ROCKCHIP_DRM_DEBUG is enabled (null pointer in a VOP driver) and it needs to cleanup huge mess in my code. After fixing those issues, I will put it on my GitHub.

Belive or not, practically I did a Rockchip job without documentation of anything in RK3399 - only pure code with almost no comments in it (code made by Rockchip).

Im not 100% sure, but there is no huge changes in Linux 6.x DRM subsystem (between 5.x and 6.x), so probably it should work on it too. Funny thing, theoretically now it's possible to do a split screen - but that code is probably only for a Android.

so it doesnt have a forward-only policy like the 1000z does?
interesting.

Rigol firmware updates are just zip files without any encryption. Once I did manual update of half of their firmware, from 1.02 to 1.03 - FPGA, kernel modules and scripts - everything without app. It was working fine.