Strange results when zeroing z axis

[millbilly]

Hi, I have searched for this in the forums but found nothing. I have noticed that when I manually "zero" the z axis on my Mach 3/Taig machine (using the software button on the main Mach screen), sometimes the x and Y axis jumps to some other value. What happens is that I carefully reference my x and y axis for the project I am working on, then later I move the tool outside of the area I have already milled so I can change tips and re-reference the z axis to the top of the material block I am working on. When I do this and touch the z axis to zero it, the x and y axis SOMETIMES changes. Not all the time, just often enough that I know I didn't imagine it! I also noticed similar behaviour when I loaded a new G-code file after carefully referencing the x, y, and z axes. My references change.

Any help appreciated...

[ger21]

How much do they change by?

[HuFlungDung]

Are you seeing the X or Y motors actually jump when you rehome Z? Or does the axis display reset for all three axis?

Have you tried returning all three axis to home? Jog off somewhere in all three axis and then return to home a few times, to see how repeatable the machine is.

If your program uses G92, this could cause a problem with the home position changing whenever the axis are positioned in a random location. This would most likely be a really gross error, though.

[millbilly]

To answer a couple of questions:

the motor doesnt' move. The references for x and y simply change and by a lot. Z axis does go to zero.

By coincidence, I just checked the homing performance. Absolutley perfect for y and z, a couple of thousnaths occasional variance for x. Dont khnow why. I had to fiddle with the debounce to get x to work correctly. However, i noticed the reference problem long befor i even put the homing switches on, so it isn't that.

[ger21]

Do you have one of the offsets turned on? G55,G56....??

[millbilly]

Not that I know of, that is I didn't turn them on. I'm not sure but I think the numbers for x and y that appeared are the machine coordinates for the reference poiont I am trying to zero out. I need to replicate the error to verify that.