I flashed the latest jetpack 6.1 to my Jetson Orin Nano 8GB developer kit. It now shows I can have MAXN instead of limiting to 15W.
I then went to modify jetson-io.py to add my two csi cameras. However this does not open, it just flashes for an instance to the application (shows an empty bordered box) and then closes without error. A search of this forum shows lots of entries about this happening with a mis-configured SPI. But this is an absolutely clean install onto my NVMe.
I tried renaming it as backup but this then made jetson-io.py not even flash briefly.
Has the wrong dtb file been installed? What can I do to fix this?
I recall there used to be an option for developer verison 3 and version 5. Could it be that my unit is a version 3 and not compatible? How would I know what version it is? My part number in JTOP implies it is correct and version 5
Can I downgrade within the Jetson Ubuntu instance or do I have to go back to SDK Manager and do it from there? Does this mean I have to start again and re-install everything to the NVME?
deb https://repo.download.nvidia.com/jetson/common r36.4 main
deb https://repo.download.nvidia.com/jetson/t234 r36.4 main
deb https://repo.download.nvidia.com/jetson/ffmpeg r36.4 main
with
deb https://repo.download.nvidia.com/jetson/common r36.4.1 main
deb https://repo.download.nvidia.com/jetson/t234 r36.4.1 main
deb https://repo.download.nvidia.com/jetson/ffmpeg r36.4.1 main
(i I tried r36.4.0 but it said no such files at the repo site)
then running
sudo apt update
sudo apt install --only-upgrade
It returned with
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
The following packages were automatically installed and are no longer required:
gdal-data gir1.2-gst-plugins-bad-1.0 libaec0 libarmadillo10 libarpack2 libblosc1 libcfitsio9 libcharls2 libdc1394-dev libexif-dev libfreexl1 libfyba0 libgdal30 libgdcm-dev libgdcm3.0
libgeos-c1v5 libgeos3.10.2 libgeotiff5 libgl2ps1.4 libglew2.2 libgphoto2-dev libhdf4-0-alt libhdf5-103-1 libhdf5-hl-100 libheif1 libilmbase-dev libkmlbase1 libkmldom1 libkmlengine1
liblept5 libminizip1 libmysqlclient21 libnetcdf19 libodbc2 libodbcinst2 libogdi4.1 libopenexr-dev libpq5 libproj22 libraw1394-dev librttopo1 libsocket++1 libspatialite7 libsuperlu5
libsz2 libtesseract4 liburiparser1 libvtk9.1 libxerces-c3.2 mysql-common proj-data unixodbc-common
Use 'sudo apt autoremove' to remove them.
0 upgraded, 0 newly installed, 0 to remove and 89 not upgraded.
So did not downgrade. If I downgrade to 36.4.1 do I lose JetPack 6.1 and the new MAXN power function?
jp6.1 will work. maxn will not work on the 36.4.0.
The dev manual has instructions on the downloading the image and installing. sdkmanager will only install 36.4.2. And yes, all will be lost either method.
We had 1 board and it crapped out during testing so I just sent it back. Waiting for USA stock so the board can be replaced.
Nothing wrong with the 36.4.0 release. Just depends on what you are doing with the board and where you need to go.
The USB camera will plug in and openCV built with CUDA handles if very well. Only thing I noticed is their USB is not very fast. If you go from a big box with usb 3.x to the nano you will see latency. You only have overlays for the imx219 and some other one, both are lame compared to a good usb cam.
Hi @simon.bunn ,
I had the same issue with latest version of Jetpack. If you other post, you can see that this is some bug related to the dtb file of the new “super” mode. Nvida team seems that will fix in next release. In the meantime I found the following workaround:
create a new boot configuration in /boot/extlinux/extlinux.conf copying from the default primary
add in new configuration the following line
FDT /boot/dtb/kernel_tegra234-p3768-0000+p3767-0005-nv.dtb
where the file /boot/dtb/kernel_tegra234-p3768-0000+p3767-0005-nv.dtb is a copy of /boot/tegra234-p3768-0000+p3767-0005-nv.dtb
reboot the board and select the new boot option
when the system start you can succesfully run jetson-io.py
Basically with this fix you use the kit without the new super mode
I will live without my CSI cameras. They seem crap image quality at best. I will use a spare USB CAM I have lying around until a patch comes out. Thanks!
Well this is unexpected. I was running some deepstream 7.1 apps and accidentally ran one that launched my CSI camera! I then ran source1_csi_dec_infer_resnet_int8.txt and that also launched my second camera. So does this mean we don’t ever need to run jetson-io.py ever again?
Hi,
The default CSI camera enabled on Orin Nano developer kit is RPi camera v2(imx219). If you use this camera, you don’t need to use jetson-io to enable it.
