FTDI chip leaks 30mA @ 3.3v into EC RX pin when machine-under-debug is off
This is a quandary.
According to FTDI I/O pin states, the FT230X pulls the I/O pins to VCCIO
with a 75kΩ resistor.
Since the I/O reference belongs to the target machine, a proper design would tie VCCIO
to the target machine's power supply. After all, a chip's I/O backpower protection and overcurrent protection is often stated with regards to its own voltage. When the MEC1521 is at 0V (off), it doesn't matter what percentage it can theoretically sink as backpower.
Unfortunately, the Framework Laptop does not provide VBUS
in CCD mode, so there's nothing for us to tie VCCIO
to.
Judging by Kieran's recommendation to use this method, I might suspect that they've added backpower protection?