I had my R4 powered off for a while. I’m about to convert it to a PoE model, but am still waiting on the BPI-5400B to come in, and so I am still powering it with a standard barrel jack. Today, I installed OpenWrt 24.10.4 for the PoE model onto an SD card and upgraded the bootloader on NAND. After a reboot, UART is dead:
GND pin is not showing connected to ground, tested with a multimeter between the pin and usb shield/switch shield/SFP cages.
TX is measuring 1.3V to 1.6V, nowhere near 3.3V.
Any ideas how to diagnose? FWIW, it does still boot successfully from eMMC.
What if you reverted back to non PoE / old working firmware and boot loader? Just to make sure first it’s a “soft error”, not permanent damage to the UART.
Personal thoughts: Perhaps we can try re-creating the official image via SSH to determine if it’s a problem with OpenWrt 24.10.4. If none of the other images work after several attempts, it might indicate a fault in the TTL circuit.
Strangest thing is happening. The board has been powered on all night and UART GND seems normal now. TX is still sitting at ~1.6V though, and I get no output, not even from early stages.
The version of OpenWrt I left on the eMMC has the standard BPi R4 (non-PoE) bootloader on it.
I noticed the same issue. The board is booting but i can not access the UART console. It’s just blank. I measure 3.0v on Rx and 1.3v on Tx.
It’s almost a year now that i had to connect to the UART console, because i manged to lock myself out by setting a interface wrongly after i plugged in a sfp module.
I use a Uart-Usb adapter and I have it plugged in all the time for convenience. So i can be sure that the connection was correct. I actually checked if the adapter works with different uart interface -it worked. So it seams to me like the uart controller on the board is broken?
I saw some other people had voltage regulators shortened. Maybe this is somehow related?