Gantry homing move in same direction when siwtch is hit on one side

Did you verify that the Y2 joint home switch is working and connected up in hal?

Open Show Hal Configuration and put the home switch pin in a watch window.

Post the ini file as well.

JT

Are you sure that Joint_1 home switch physically corresponds with Joint_1 motor, and the same for Joint_2?

Attach the ini file.

FrankyBoy wrote:

ng that someone proposed is that the switch double trigger when it hit's the magnet (it's a magnet triggered limit switch).
Is it possible that it does not backup due to that?

It would surprise me, but it's possible. 

It rather depends on your home search and latch directions. You could try a (software) debounce on the home switches. 

I tend to debounce them two different ways, a short debounce for homing and a longer one for hard limits. (because a nuisance trip of the hard limits can spoil your whole day, and a 10mS delay in detection is unlikely to be a problem unless the axis is moving _really_ fast. 

FrankyBoy wrote:

How would I do a <software debounce>

There is a HAL component for the job. 

linuxcnc.org/docs/stable/html/man/man9/debounce.9.html

Though that manpage isn't likely to be much help without some context:
linuxcnc.org/docs/stable/html/hal/intro.html
 

Forget debounce for now.
Is the X and Z homing OK?
Is Y1 or Y2 homing OK?
You have to be sure what side is Y1 and make sure the sensor pertaining to it is wired to the correct input referenced in hal..
I went through your files twice, to many things are not at default and/or consistent.
The simplest way to explain gantry/tandem homing is :
-make X axis work properly
-copy the values to Y1 and Y2 from hal and ini files
-put both joints under the AXIS Y section, name them as joint 1 and joint 2 (assuming X is 0)
-set the home_sequence to -1
- if it ends up as Y1 not working, try removing 1, so just Y, leave Y2 as is <<<< i am sure this is how PncConf sets them up. not sure if it matters though.