sensitivety of E2(pin9) and GND(pin2) on expansion as to E1

Yao

Is the EM sensitivety of a simple endstop switch connected to E2 (pin9) and GND(pin2) on the expansion higher compared to the same endstop switch connected to E1_Stop (E1 and GND)?
I am getting unexpeceted results in reading the switch connected to E2. That wire is 60cm longer,,, (260cm compared to 330 cm... that extra 60cm length is unshielded but twisted.

Running Duet wifi with 3.4.6

T3P3Tony

@Yao the headers on the Duet 2 board designed to have external things directly connected to them (e.g. E0 and E1 endstop, have protection and they also have a diode to help indicate if the endstop is triggered or not.

The 50 way connector is designed to have a Duex5 or EBOB or similar board connected to it, the expectation is any required protection would then be implemented on that expansion board. So E2_STOP etc connect directly to the processor.

The additional protection is probably minimising the impact of noise. A quick check would be to swap the switches between E1 and E2 and see what happens.

Edited to add a link to the schematic:
https://github.com/Duet3D/Duet-2-Hardware/blob/master/Duet2/Duet2v1.06/Duet2_Schematic_v1.06.pdf

Yao

@T3P3Tony Thanks! That explains it all. i now know how to mitigate this

dc42

@Yao the other reason why the expansion connector inputs are more sensitive to noise is that the only pullup resistor on them is the one in the MCU which has a value of about 100k. You could add a lower value pullup (e.g. 3.3K or 10K) between the input and +3.3V to reduce the sensitivity to noise.

Yao

@dc42 any way to do this without adding parts?
Is there another (set of) free pin(s) I can use with the switch wired as is? (E0 and E1 stop are used, as are all standard endstop pins.)

I did switch to the shielded cable to the E3 (pin9/ GND-pin2) but this dd not solve it. another machine layed out like this with much shorter wires does work.

The strange thing is that when I test wire it on E3 (pin9/ GND-pin2) with a10cm wire+NCswitch. I also get inconsistent response. echo states are Always 0 pressed or unpressed. However, when I switch, the board does trigger...

Yao

@dc42 pullup resitor between expansion pin 43 and E3_stop, switch between E3_stop and ground? (I already used the nearby 3.3v pin 3)
.
.
.
5min later.
Yes, This works with a 10KΩ ....