The more I read, the more I’m lost
I’m still trying to understand the current situation of Jetson Orin Nano dev board GPIO & latest jetpack (6.2.1) at the time of writing.
My objective is simple:
To drive an optocoupler in a 817 Module to press a button (connect a 5V high to gnd during 1s, to trigger a button press. The typical power button behaviour), using PIN 33 (GPIO13) and PIN34 as ground.
*Using the standard sdcard image.
The problem?
That there’s like a ton of info but as it is completely scattered.
I’m reading some people talking about:
- Re-creating the boot disk with a new DTB/BCT (nuts)
- SYSFS was deprecated (only god knows why)
- Some other talking about JP5 where it looks like gpio worked without any issue
- Some other that had a faulty carrier board
- Some commands talking about using “busybox” to read and write mem values… somewhat blindy (get the pinmux address, then run busybox devmem and write god knows what in hex)
- Plus there’s a nv github repository called “jetson-gpio” whose description is “A Python library that enables the use of Jetson’s GPIOs”
Starting from a fresh JP6.2.1 install (sdcard image):
- gpiod is installed.
- but it looks like it doesn’t work.
gpiofind detects “PH.00” as “gpiochip0 43”
gpioinfo | grep -i “PH.00”
line 43: “PH.00” unused output active-high
But when you execute:
gpioset --mode=time --sec=1 $(gpiofind “PH.00”)=1
gpioset --mode=wait $(gpiofind “PH.00”)=1
Nothing happens.
And you cannot monitor pin using gpiomon while touching it with gpioset :S
So, considering that GPIO is dangerous enough that you shouldn’t be playing around with it carelessly, now the question is…
How? (help 🥲)
Thanks in advance.
// DevKit is a Jetson Orin Nano 8GB, latest “july” batch.