Werner

You are not using Armbian but a fork. Ask where you got the image from. We don't support 3rd party OS.

We don't deal with Android here. Ask vendor or at some place like xda developer forums.

edit /boot/armbianEnv.txt and set verbosity to 7 to get an idea about what happens inside.

This should work too. Debug boot issues: https://debug.armbian.de Related perhaps: https://github.com/armbian/build/pull/8994

Can you provide logs?

Just boot from microsd and then use armbian-install to move image to spi/nvme directly attached?

We don't support 3rd party forks. Ask at the place where you got this OS from.

Because they sell you the hardware only. Their software support is very poorly made since for the price they cannot afford proper software support and - even more important - maintenance. They leave this burden to - mostly unpaid - random developers across the world like us. At least I'd go for a device with Standard Support which the Ultra and Max hasn't: https://www.armbian.com/download/?device_support=Standard support For the raspberry you would need additional hardware (an AI kit I guess). No clue though about how they both perform in comparsion. For RK3588 you have two choices for now: The rockchip bsp implementation and https://github.com/rockchip-linux/rknn-toolkit2. Or using 6.18 kernel using the reverse engineered Rocket driver and latest bleeding edge mesa to get npu access. I personally did not play with either of these options and have no intention to do so for the moment.

Not all possible kernel-userspace combinations are built and/or exposed. If you cannot find what you are looking for, DIY: https://github.com/armbian/build/

Don't use etcher. https://docs.armbian.com/User-Guide_Getting-Started/#flash-to-sd-card

debug boot issues: https://debug.armbian.de

current isn't a good option for rk3588/s devices. At this point of time barely any support for this SoC was mainlined. And it seems like later revisions introduced a regression causing a crash. Since 6.18 may become next LTS kernel it will replace current current kernel at some point. So better go for an edge based image instead.

orangepi5 and 6 have nothing in common. double-check what you are actually talking about. Also there is no 0.9.9 rknpu stuff. 5.10 bsp has been abandoned long ago.

not recommended for rockchip socs since it cannot handle the rather high baud rate of 1.5Mbaud. Use FT232, CH340 or CP2104 instead.

You should check uboot logs if the dtbo is loaded correctly. You can get those using a debug serial console. If you don't know what that is or how to connect: https://debug.armbian.de

https://github.com/armbian/build/blob/main/config/boards/retroidpocket-rp5.csc https://github.com/armbian/build/blob/main/config/boards/retroidpocket-rpmini.csc You can build images on your own. https://docs.armbian.com/User-Guide_Board-Support-Rules/#community-maintained

Armbian does not have a config for this board nor is it supported in some way.

The logs you can aquire this way: https://debug.armbian.de

@TonyMac32

That's an interesting question. The chip itself seems to communiate via spi or i2c. Since USB cannot do that they put a raspberry microcontroller (RP2040) on it to handle this. The communication then uses a simple serial connection with the ch341 chip. I assume the klipper thingies can handle this by themselves automatically while when connecting this to a different device, you have to come up with some way to communicate with this board.

Which of these boards are you refering to? https://github.com/armbian/build/tree/main/config/boards

The rk3588 soc has a fixed boot order which is spi, emmc, microsd, other I think. So if there is a boot loader found on spi it will ignore other options and try to boot from that. While Armbian uboot binaries on spi are instructed to check if there is a microsd card inserted and boot from that, vendors can (ab)use this circumstance to make it harder for other operation systems being used because tools like rkdevtool are necessary to get rid of it. If the spi is pre-flashed with crap, I guess that's why the step is necessary.

If you have patience, CEC may work at some point with mainline kernel. There is not fast solution.

didn't even realize you were on beta repo. Well when you use bleeding edge, you have to be prepared for breakage since it is expected ;).

I'm not sure bit IIRC there should be a 1 second delay which would allow interaction but again, not sure...