• DepthAI-v2

  • Using Luxonis preped RPI OS images (V10)

legow00

legow00 Bus 001 Device 002: ID 03e7:2485 Intel Movidius MyriadX

Good, the OAK-SOM is discoverable physically. Try python3 -c "import depthai; print(depthai.Device.getAllConnectedDevices())"

Thanks
Jaka

      @jakaskerl

      python3 -c "import depthai; print(depthai.Device.getAllConnectedDevices())"returns the following: []

        legow00
        Hmm. What do you get if you try to open a device using dai.Device()? Like running a simple rgb_preview example? I'm interested in the dmesg output when you run the script. Perhaps a if you run it using XLINK_LEVEL=debug and post the results. This is very strange, can't remember seeing it before.

        Thanks,
        Jaka

          @jakaskerl

          I tried to run the RGB Preview (https://docs.luxonis.com/software/depthai/examples/rgb_preview/) and the Python script retured:

          Traceback (most recent call last):

          File "/home/pi/Desktop/rgb_preview.py", line 24, in <module>

          with dai.Device(pipeline) as device:

          ^^^^^^^^^^^^^^^^^^^^

          RuntimeError: No available devices

          The output of dmesg can be found here: https://pastebin.com/6kYZ7QXp

          Setting XLINK_LEVEL=debug flag results in the same output as above

          I just checked my DepthAI Python version and I am at version 2.22.0.0, should I update to the most recent package in pip? Version 2.30.0.0.

            legow00
            Dmesg shows no indication that the HW detects the USB device.

            Let me check what's broken upon update & I'll get back to you.

            Thanks,
            Jaka

              legow00 legow00
              Tried on a CM4 POE with x64 image (V10 I think).

                 12  sudo apt update && sudo apt -y upgrade
   13  lsusb
   14  cat /etc/os-release
   15  sudo apt autoremove
   16  sudo reboot
   17  lsusb
   18  sudo apt autoremove

              "lsusb" still shows the device is connected.

              Can you try running systemctl status cm4-rvc2-usb-prepare? Enable if disabled.

              Thanks,
              Jaka

                  @jakaskerl

                  I checked systemctl status cm4-rvc2-usb-prepare but this systemd service is enabled and active.

                  I also tried the same commands as you mentioned:

                  12 sudo apt update && sudo apt -y upgrade
                  13 lsusb
                  14 cat /etc/os-release
                  15 sudo apt autoremove
                  16 sudo reboot
                  17 lsusb
                  18 sudo apt autoremove

                  Nothing changed lsusb still returns:

                  Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

                    legow00
                    Hmm, does it work if you try to reflash a fresh POE v10 image? Can you try re-seating the SOM:

                    also check:

                    pi@luxonis:~ $ ls /lib/firmware/renesas/
K2026.mem

pi@luxonis:~ $ ls /usr/sbin | grep upd72020x
rvc2-upd72020x-check-and-init
upd72020x-load

                    Thanks,
                    Jaka

                      @jakaskerl

                      I executed the following commands on a fresh V10 image install:

                      12 sudo apt update && sudo apt -y upgrade
                      13 lsusb
                      14 cat /etc/os-release
                      15 sudo apt autoremove
                      16 sudo reboot
                      17 lsusb
                      18 sudo apt autoremove

                      Still same result as before, lsusb still returns:

                      Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

                      What would be the best practice to remove the SOM while minimizing the risk of damaging the connectors?

                      @jakaskerl

                      I re-seated the SOM and started again with a fresh install of th V10 image.

                      I did multiple sudo apt update && sudo apt -y upgrade, sudo apt autoremove and reboot.

                      The Intel Movidius MyriadX USB device is still detected. RGB example scripts also works!

                      Thanks for your help @jakaskerl ! It was probably a loose or bad connections somewhere in the SOM connectors.

                      @jakaskerl

                      Oops seems like I celebrated to early… after some reboots a pop up in the top right part of the screen showed which notified me of few updates. After these updates the Movidius device was gone again… grrrrr. Is there a way to look at a log file of the last updates executed on this system?

                        legow00
                        /var/log/apt/history.log and /var/log/apt/term.log
                        should have your apt history

                        eg:

                        sudo awk '/Start-Date/{p=1} p' /var/log/apt/history.log | tail -n 50

                        Thanks,
                        Jaka

                          @jakaskerl

                          One (or both?) of these packages broke the USB communication with the Movidius device:

                          Start-Date: 2025-05-08 14:05:22

                          Commandline: packagekit role='update-packages'

                          Requested-By: pi (1000)

                          Upgrade: rpi-eeprom:arm64 (27.6-1, 27.7-1), piwiz:arm64 (0.90, 0.91)

                          End-Date: 2025-05-08 14:05:27

                          What would be my best course of action?

                            legow00
                            Try with
                            sudo ln -sf renesas/K2026.mem /lib/firmware/renesas_usb_fw.mem, then
                            echo -n 0000:01:00.0 | sudo tee /sys/bus/pci/drivers/xhci-pci-renesas/bind. I was able to rebind mine and then it started working again.

                            Thanks,
                            Jaka

                              @jakaskerl

                              Thanks! I was also able to rebind it! Could this be updated in the next release of a new image? So that other people in the future don't have this same problem.

                              Thanks for your ongoing support, I will try this out for a week before updating our other cameras just to be sure.

                              I hope this was the root cause of my problems. 🙂

                                legow00
                                I still have to check why that happens - then I'll add it to the image.

                                Thanks for confirming the fix.

                                Regards,
                                Jaka