Hi thagrol,
You make a fair point but I wanted to check everything on the Pi side of things first in case I missed anything.
I did have it set in cmdline.txt but original using 'console=ttyS4,115200' based on the entry generated when enabling the default serial console via raspi-config. As you suggested changing it to ttyAMA4 did indeed work.
I did some testing by using a bunch of wires to connect the Pi to UGV header rather than being directly mounted on top. I started off just with the power connections and found that this time the Pi did not unexpectedly boot up again. The second involved adding the serial connections and again it did not reboot. Therefore I was able to narrow my search to gpio2 and gpio3 as the likely cause given there are no other physical connections.
After a bit more digging, I found out that gpio3 can be used to wakeup the Pi, even from a supposed powered down state. I can only guess that the i2c activity from the esp32 controller board is causing this to be triggered. Fortunately I have found that updating the eeprom configuration to set 'WAKE_ON_GPIO=0' now means that the Pi will not reboot when directly mounted on the UGV.
You make a fair point but I wanted to check everything on the Pi side of things first in case I missed anything.
I did have it set in cmdline.txt but original using 'console=ttyS4,115200' based on the entry generated when enabling the default serial console via raspi-config. As you suggested changing it to ttyAMA4 did indeed work.
I did some testing by using a bunch of wires to connect the Pi to UGV header rather than being directly mounted on top. I started off just with the power connections and found that this time the Pi did not unexpectedly boot up again. The second involved adding the serial connections and again it did not reboot. Therefore I was able to narrow my search to gpio2 and gpio3 as the likely cause given there are no other physical connections.
After a bit more digging, I found out that gpio3 can be used to wakeup the Pi, even from a supposed powered down state. I can only guess that the i2c activity from the esp32 controller board is causing this to be triggered. Fortunately I have found that updating the eeprom configuration to set 'WAKE_ON_GPIO=0' now means that the Pi will not reboot when directly mounted on the UGV.
Statistics: Posted by Ryzer58 — Sat Nov 02, 2024 9:16 pm