The super wonky/sketchy Windows tool worked… got the boot-ROM updated and PD now works for me.
I guess it would be a great idea to either implement a working in OS upgrade path, update the U-boot before shipping or at least fix the toolchain needed to upgrade it.
I have been able to update with rkdeveloptool… Use this to set the correct permissions
echo 'SUBSYSTEM=="usb", ATTR{idVendor}=="2207", MODE="0666",GROUP="plugdev"' | sudo tee /etc/udev/rules.d/51-android.rules
Then, make sure you are in the same folder as the image you want to flash. Second, run all commands pretty fast after booting, or all of them in the same line, including the reboot one.
I have managed to both update spi and flash the emmc with this tricks. Now my laptop PD supply works!
what platform you use with Windows that works? e.g. CPU? Chipset? Run the tools with admin privilege?
I used Windows 11, a VM, created a snapshot, installed driver (needs admin) used the tool (admin too, super wonky) and rolled the snapshot back, the VM runs on the same Linux box that failed to do the job with the linux CLI tool.
I could not get it working even with fixing the device permissions @xtract I guess there is something else going on there.
I also manage to flash now using Ubuntu, and the solution is VERY STUPID: I used the very cheap USB-A 2.0 to USB-C cable, and don’t use anything in between. Then I just follow the official guide and everything seems fine. I used USB-A 3.0 to USB-C cable before (swapped from my USB HDD) and that won’t work. Having USB-C extension cable also won’t work.
Ohh well, I can confirm, works fine with a USB 2.0 USB AtoC cable, not working with C to C and also not with a 3.0 A to C… like what the fuck?
it is really stupid… I just found that out by opening two ssh sessions and monitor on what’s happening on the host side using dmesg -w, and see a lot of usb related messages that should be weird until I used that USB-A 2.0 to USB-C cable, then the usb message become “normal”.
@jack may I get a v1.3 dev board for finding this out? hahaha 
With rock-5b-ubuntu-focal-server-arm64-20220930-1451-gpt.img all my problematic USB PD chargers work now and the SPI flash is accessible from inside the running OS since this little device-tree snippet is active:
root@rock-5b:/home/rock# ll /dev/mtd*
crw------- 1 root root 90, 0 Oct 16 19:05 /dev/mtd0
crw------- 1 root root 90, 1 Oct 16 19:05 /dev/mtd0ro
brw-rw---- 1 root disk 31, 0 Oct 16 19:05 /dev/mtdblock0
See also https://github.com/radxa/build/commit/5e03c182130740ecf7904b9730613b27548ff490
They are advertising this at 12v, 15v, 20v. Is the problem being discussed affecting the 5B 1.42.2022.08.29 ?
Also unclear but is it a software / firmware problem, or hardware? I’ve got a 5V, 9V, 12V, 15V, 18V, 20V usb-c 65W adapter coming tomorrow; hoping to stick an NVMe in with wifi.
I’ve also stuck passive coolers on the RAM and will put something 25mm x 25mm x 24-50mm on the die with adhesive or a movable pad.
I guess, judging by the bootloader code, the Rock5B only accepts up to 12V.
Advertising was not the point here… At least not so far.
Here is the link I’m referencing. It might be a translation issue
https://wiki.radxa.com/Rock5/hardware/5b#:~:text=USB%20PD%20power,USB%20C%20plug
So long as it negotiates and doesn’t burn itself out I don’t mind. Just seems a strange communication glitch. Excited to see how bloated I can go with 16GB RAM 
please be so kind to explain: is this version updating the SPI from the OS or is the OS doing the neg? just got the package today (btw. if someones in germany an desperately waiting for one: i have a spare 8gb version, that i can lend out for the duration of my parental leave/“no tinkering time”) and want a fast start tomorrow. else i’ll just update the bootloader
well that might take ages to merge… but yeah, good to know.
Have You tried with nvme? On my side with m.2 it hangs and are not able to run. I’ll give a try latest image and more chargers but so far I was not able to run with nvme (few different).
Has been merged 20 hours ago but not arrived in Radxa’s package repository. I updated their Ubuntu Focal image right now and after a reboot still no USB PD negotiations exceeding 12000 mV.
Nope, still only crappy SSDs lying around here.
Mine boots fine without anything in m.2 -
I think it booted once, but drive was not found in system.
As far as I understand it’s safe to supply something like 12V 4A to usb-c (without usb-pd)?
If I understood it correctly then the way Radxa is doing this USB PD stuff currently is still flawed. They enabled lower DRAM clocks to
‘have enough power booting to kernel PD voltage negotiation’
Yes.
