[elogicca]

Ok if i put a diage gauge on my mill test 0" -> 1" -> 0" i can get the backlash right down, but my question is this if i do a 0" -> -1" -> 0" it doesn't match ???

Is this normal?

[HuFlungDung]

What kind of screw and thrust bearing are you testing? Is this a precision ground ballscrew and precision thrust bearing block?

There are certainly instances where screw wear needs to be mapped for different positions along its length. A cnc controller then automatically adjusts the backlash for additional error due to non-periodic wear of the mechanical components, and these adjustments would be location specific on the machine.

[elogicca]

using acm, non ground ... so i gues the best bet is to start at one end with a dial and go 1/2" by 1/2" and map the values?

[elogicca]

Ok if i take out backlash move left by 0.1" my dial indicates 0.982" i can repeat this quite close, so is this screw mapping or steps per inch??

10tpi screw microstepping 1/16's 32000 steps , i get different results on the x and y

[HuFlungDung]

If it seems to be relatively uniform everywhere, then .018 would be the backlash setting. That is a very large amount for a 10 tpi screw. I would make every attempt to eliminate looseness in all thrust shoulders, and maybe make a springloaded nut, in an attempt to keep the screw bearing always on the same side of the nut. A new, tighter nut would also be good, or a double nut, with a fixed shimmed distance between them. You should be able to reduce a uniform backlash down to a thousandth or two.