I tried tightening up my 3 axis mill's Z axis backlash, because I was getting around .002" and I figured it could do better than that.

I tightened the pulley/nut combo on the ballscrew.

The backlash went down to about 0.0013"..but something odd is going on...

I put my 0.0001" tenths indicator in the spindle and lowered it to the bed.

Using the tenths 0.0001" jog, I tried to examine how the backlash behaves.

What I noticed is, if I jog until the needle is regularly moving in one direction, then switch directions to go the other way (and measure backlash), the needle KEEPS GOING in the wrong direction for maybe 0.001"! Which is ten jog steps.

I thought maybe it was the head tilting on the gibs as the force of the ballscrew changed direction, but its the wrong direction for that, it should be the other way shouldnt it?

Is there some known reason why this would be happening?

When I first start off reversing direction, I can see the DTI responding even in the first couple tenths..so it makes me think that I could get the backlash much lower since apparently there is coupling to the stepper motor that soon.