After homing X and Y, can I auto move to a non 0,0 location?

Well reading the (correct, even) documentation hasn't helped.

Clearly, my expectations are wrong.  I have an axis that goes from 0 to 12.  I want to home at the "12" end.  It's a "bed moves" mill, so the motions are all backwards from the POV of the config, which is clearly expecting a "spindle moves" mill.  

Here's what I have.  The SEARCH and LATCH phases work fine, both searching negative.  But when the axis engages HOME_FINAL_VEL, it also heads in the negative direction to find HOME.  To be clear: HOME_SEARCH_VEL, HOME_LATCH_VEL, and HOME_FINAL_VEL are all moving the axis in the same, negative, direction.

With the INI as shown, it moves .5" before stopping.  (I've replaced "HOME=12" with "HOME=.5" for testing, simply so the axis doesn't move so far in the wrong direction.  I'm testing in the middle of the bed, simulating by activating the limit switch by hand.)

What do I need to do?


 

Just to be sure about this...where are the limit switches physically located in relation to the travels?

With the table all the way back and to the right?

You have to remember you are dealing with cartesian coordinates.
You can home to any corner you wish. Homing to the lower left of the worktable means all x,y coordinates are in the positive quadrant.
Homing to another corner will mean at least one axis will count in the negative direction. Think of the graphs you drew at school

I did have a machine where the home switch was at max Y but I added a home offset to make the home position at min Y
Thinking about it after using it, I realized I could have homed it at the home switch position so the Y axis would count in the negative direction beecause once you zero G54 offsets at lower left on your workpiece, everything would count in a positive direction anyway.

@spumco may be onto something? My limit switches are shared: both switches on one circuit. Maybe that's the problem?

@rodw -- My impression and hope was that if I homed to the upper right, I could somehow let the mill know that that point was 12,12 , and that therefore both axes would, yes, count in the negative direction, but not actually ever be negative. They would still go from 0 to 12, but it wouldn't be the 0,0 corner that homing was finding.
Does that make sense?

Switches shared on one input isn't the issue.

I'm trying to determine where the switches are in relation to the table movement so we can make sure that everything moves the correct direction during homing, and if need be move back to a position you want after homing is complete.

See how you go with the position.

Regards shared himing switches, there is a pin joint.n.is_homing ot similar (check man motion).
If you and2 that with your physical limit switch and use the output to your homing input, the non homing joint will ignore the switch input.