With regards to your other thread on backlash here
And continuing my suggestion / thoughts
viewtopic.php?f=4&t=1361__________________________________________________
If you cannot reduce the backlash to acceptable levels
Why not choose a backlash value half way between your two axis backlash values that way your backlash will be halved .
For simplicity say you have 2mm on X axis and 1mn on X' (a slaved) set the backlash at 1.5 so the error is only + and - 0.5 between the axis
If you set it on 1 or 2 the error between the axis will be + / - 1
Your application is plasma.... How bad is your backlash really.... Is it that apparent in the cut?
Would it be that apparent with a midpoint backlash value?
Sometimes it's worth asking if people are chasing perfection in an imperfect world. I don't know your application or machine (is it commercial, is it a hobby, is the machine hobby built or commercial built etc)
I run a belt setup (with steel wire core) on a hobby machine, but I accept that running 0.3g on acceleration may induce backlash by way of belt tolerance and belt teeth compression over the pulley when the axis velocity changes, dispite my belts being tight, tensioned and spring loaded.
But where I came from was hand cutting with plasma or grinder with a whole load of hand measuring and marking and cleanup... It's also not Lazer but plasma.
I don't know your background but newcomers to CNC tend to chase perfection with backlash settings (plus plasma divots at end of cuts), but like I've posted above it is possible to half the error by choosing a midpoint between axis, but it's best to try to minimise it first before applying backlash compensation