Here’s what I’ve found. Attached is the schematic for the limit switch connectors from BigTreeTech manual for the SKR Pro.
There’s a 4.7K pullup resistor and a 1K current limiting resistor to the LED. I measured 1.7V across the LED which is fairly typical for a red LED. Looking at the voltage divider of the 4.7k/1.0K resistor, the voltage for pin 3 (signal pin) calculates to 1.98 V which is pretty much what we have measured with the DVM.
So… the question is what changed? Why does this work for some and not others? Looking at the CPUs input/output characteristics, the minimum voltage for a signal to be considered high, Vih, ranges between 1.7 (theoretical) and 2.3 (production unit testing) volts. hmmm, right where we’re sitting. No wonder we’re seeing odd behavior.
Seems like this behavior may have been seen before: https://forum.v1e.com/t/skr-pro-1-2-strange-problems/32821 although there are some differences. Unfortunately there is no response resolving the problem.
I’m thinking that the on-board pullups are there to keep the pin from floating open but an external pullup (about 1.8k to 1.5K) is needed for real world operation.
hope this helps,
s