Power & Source of Big Ideas

You were right, Pavel, in rk3399-sd-friendlycore-bionic-4.4-arm64-20181219.img camera magically starts to work! Thanks for this info. I wonder which kernel version is it in this new release. As for the kernel - everything is simple - 4.4.143, which means no changes. There must be changes into the k...
I have this camera, but it was not working. Now a new camera is coming to me, I think it should earn. As soon as it arrives, plug it in and write about the results. For information - the manufacturer has released new versions of operating systems "... 20181219.img" Try it, maybe there are...
isp1 driver (ov13850.c) fails in the same way during camera register write attempt through "i2c_master_send".
Well, power on/off only means assignment of some PM and pin states. We don't really get any "power on" confirmation from the camera on this step. In the context of further camera register read failure, this leaves an option that an assigned PM or pins state is invalid. What also looks susp...
Camera can technically switch on/off. But what ultimately fails is a call to "i2c_transfer" from the following context: static int ov13850_check_camera_id(struct ov_camera_module *cam_mod) { u32 pidh, pidl; int ret = 0; ov_camera_module_pr_debug(cam_mod, "\n"); ret |= ov_camera_m...
More diagnostics with "cat /sys/kernel/debug/gpio" effectively reveals that none of camera gpios described in DTS node have been set. This probably suggests that error occurs earlier than they could be assigned.
I'm pretty sure though, that the driver loads at 0x10 register, which could be seen in "i2cdetect 1" command output. And if I change 0x10 to 0x30, i2cdetect output changes accordingly.

Maybe we also need to send proper signals to gpio ports, in order to power on the camera?
Does this actually work for you? I've tried this already, same "ERR: no active sensor configuration" as before. I believe FriendlyARM does not respect the wiring of the Rockchip spec. If you lookup for used lane numbers in the M4's own MIPI spec, they do not seem to properly correspond to ...
Hi, does anybody have a working Linux kernel DTS config for ov13850 camera?

Thanks!
Thanks for the info! OK, alas not so simple.

I'm interested particularly in OV5647 and IMX219, because their drivers are well-studied by community (e.g. 6to9 on github). I'd love to provide funding/equipment to support the discussed RPI "adapter" research work, if you are willing to do it.
30-pin CAM1320 has a "cell-phone" pluggable camera port onboard. Does it mean that in theory we can replace the OV13850 sensor with e.g. IMX219 sensor, also with a "cell-phone" plug?
Thanks. So my mistake was to try getting 15-pin cable into 30-pin port of completely different spec. Makes no sense indeed.

@ayaromenok, is it really only one camera module available for M4??
Dear FriendlyARM, Unfortunately, bad hardware experience with my new NanoPi M4 board. I'm using the NanoPi together with RaspPi camera modules - OV5647 and IMX213. Surprisingly, their standard 15pin cables do not even fit into MIPI-CSI ports of the board. I have to cut 0.5mm off the side to get the ...