E-cam30a_hexcutx2 cameras undiscoverable

i have the e-cam30a_hexcutx2 6 camera system for my jetson agx xavier. After i run the install_binaries.sh script everything finishes and the system reboots. after the reboot the cameras cannot be discovered using
$ dmesg | grep -i “Detected ar0330 sensor”
or with
$ ls /dev/video*

Jetpack version 4.6.0(rev 3)
driver version “e-CAM30A_HEXCUTX2_JETSON_L4T32.6.1_31-AUG-2021_R06”

i have attached the binary installation log to this post

binary_installation_log.txt.txt (25.4 KB)

Please consult with E-CON.

i have been, but their customer support has been very slow. so i put this here in case anyone else has run into this/ has any ideas.

Hi @ubunt2,

Sorry for the inconvenience! Could you share your dmesg log?

The problem has since been resolved with the help of e-con’s support.

according to support, the problem was caused by having an extra SSD installed on the board, which somehow prevented the install_binaries.sh script from working properly.

bellow are the steps that got it working for me:

==============
Pre-Requisites

  1. Jetpack 4.6 installed in host system ($HOME/nvidia/nvidia_sdk/JetPack_4.6_Linux_JETSON_AGX_XAVIER_TARGETS/) and user have already flashed Jetson kit atleast once with that.
  2. Type C Cable.
  3. Jetson Xavier Kit already flashed with and running Jetpack 4.6

===============================
Procedure to flash Kernel Image

  1. Open a terminal in the Host PC.

  2. Move to the path where the release package (e-CAM30A_HEXCUTX2_JETSON_L4T32.6.1_31-AUG-2021_R06.tar.gz) is present.
    $ cd <PATH_TO_RELEASE_PKG>

  3. Extract the release package in the Host PC.
    $ tar -xaf e-CAM30A_HEXCUTX2_JETSON_L4T32.6.1_31-AUG-2021_R06.tar.gz

  4. Move inside the release package.
    $ cd e-CAM30A_HEXCUTX2_JETSON_L4T32.6.1_31-AUG-2021_R06/

  5. Extract the binary package(NileCAM21_CUXVR_L4T32.6.1_JP4.6.1_JETSON-XAVIER_R02.tar.gz) in the Host PC.
    $ tar -xaf e-CAM30A_HEXCUTX2_JETSON_L4T32.6.1_JETSON-AGX_XAVIER_TX2_R06.tar.gz

  6. Move the binaries directory.
    $ cd e-CAM30A_HEXCUTX2_JETSON_L4T32.6.1_JETSON-AGX_XAVIER_TX2_R06/TX2_XAVIER/Kernel/

  7. copy the kernel image in the release package with “Image” present into the “$HOME/nvidia/nvidia_sdk/JetPack_4.6_Linux_JETSON_AGX_XAVIER_TARGETS/Linux_for_Tegra/kernel” directory in the Host PC.
    $ sudo cp Image $HOME/nvidia/nvidia_sdk/JetPack_4.6_Linux_JETSON_AGX_XAVIER_TARGETS/Linux_for_Tegra/kernel -f

  8. Move to the “Linux_for_Tegra” corresponding to Jetpack 4.6 in the Host PC.
    cd $HOME/nvidia/nvidia_sdk/JetPack_4.6_Linux_JETSON_AGX_XAVIER_TARGETS/Linux_for_Tegra

  9. Connect the USB Type-C cable to the host PC and the USB TYPE-C port of Jetson AGX Xavier kit.

  10. Set the board to recovery mode, use either one of the two methods below:
    1)
    In the jetson, run the command $ sudo reboot --force forced-recovery
    2)
    a) Press and hold the Recovery button of Jetson AGX Xavier kit.
    b) Press the Power button of Jetson AGX Xavier kit.
    c) Release both Recovery and Power buttons.

  11. Run the following command to verify whether the board is in recovery mode.
    $ lsusb

look for the device with nvidia as the vendor
  for the jetson AGX xavier, the number will be 7019 if the board is detected and in recovery mode
  if the board is on, but not in recovery mode, the number will be 7020
  1. Flash the kernel “Image” partition alone in the Jetson AGX kit using the following command.
    $ sudo ./flash.sh -r -k kernel jetson-xavier mmcblk0p1

    Note:
    The kit will be automatically rebooted. Check whether camera is streaming properly with updated “Image”.

after the jetson powers up and you are at the desktop. check the cameras with the following commands

$ dmesg | grep -i “Detected ar0330”
expected output:
Detected ar0330 sensor
Detected ar0330 sensor
Detected ar0330 sensor
Detected ar0330 sensor
Detected ar0330 sensor
Detected ar0330 sensor

also try

$ ls /dev/video*
expected output:
/dev/video0
/dev/video1
/dev/video2
/dev/video3
/dev/video4
/dev/video5

to see a camera and adjust its settings use
$ ecam_tk1_guvcview

when working with openCV or jetson.inferece reference the camera through it’s video for linux 2 (v4l2) node
replacing 0 with the id of the camera you want
/dev/video0

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.