IDEX U-Axis Homing Issues

Atom

Hello I am having issues with the U-Axis homing on my IDEX setup. This just started today though I haven't touched the firmware in awhile and wasn't having issues before. Instead of moving to the end of the axis and hitting the enstop, backing off and hitting it again slower, only the back off move is happening and the U-Axis print head just steps along the axis away from the endstop every time the U-Axis homing is called. Here is my homeu.g file:

G91 ; relative positioning
G1 H1 U320 F3000 ; move quickly to U axis endstop and stop there (first pass)
G1 H2 U-5 F6000 ; go back a few mm
G1 H1 U320 F360 ; move slowly to U axis endstop once more (second pass)
G90 ; absolute positioning

It's exactly the same as my homex file just moving the opposite direction though I'm not having this problem with the X-Axis so I'm really not sure why this is happening.

Phaedrux

@atom said in IDEX U-Axis Homing Issues:

only the back off move is happening

That makes it sound like an endstop is already in a triggered state.

@atom said in IDEX U-Axis Homing Issues:

G1 H2 U-5 F6000 ; go back a few mm

Remove the H2 and see how it behaves. The H2 is forcing a movement. This shouldn't be needed if the axis is already homed.

Use M119 to check the endstop state when triggered and untriggered.

o_lampe

@phaedrux said in IDEX U-Axis Homing Issues:

Use M119 to check the endstop state when triggered and untriggered.

It would be a nice safety feature to automatically check the endstop-state before homing. Could be done in homeall.g
The check would echo a warning.

Not sure, what the real object model expression would be, so here's a little pseudo code. Only works for normally closed switches (NC)

; homeall.g
if (endstop[X] || endstop[Y] || endstop[U] || endstop[V] || endstop[Z])     ; checks endstop status = true?
    echo (" One or more endstops are already triggered ")
    break     ; stop homeall.g
m98 p"homeX.g"
m98....

Atom

Use M119 to check the endstop state when triggered and untriggered.

Turns out one of the leads on my endstop came unplugged, should have known to check that.

The automatic checking of the endstop-states is an interesting idea. I'll see if I can implement it.