I just finished my 9x20 CNC conversion and after ditching my Acme Lead Screws and Dumpster CNC antibacklash nuts, I installed Roton Ballscrews and 2 ballnuts per axis for anti backlash.

So I'm getting .002" backlash on each axis. I think it has to do with the way the two nuts are held apart with the spring to create the anitbacklash, because when I setup a dial indicator and push and pull on the axis it doesn't move (aside from normal flex for such a small lathe) but it needs the backlash comp at .002. It's weird when I set the step jog to .0001" it moves right away but in the wrong direction for .0005 then starts to go in the correct direction.

Any suggestions, sounds like there are some force pusing on the axis, like I said probably the tension spring in the antibacklash. If anyone has any experience with that let me know. Thanks.

Maybe its the bearingblock the ballscrew is mounted in.

regards,

roy

I'm using 2 angular contact bearings, but it could be. I'm not sure how to tell where the backlash it coming from.

Could it be skewing of the slide?
Try taking the messurements close to the screw location, then again at a further distance away from it.
If the meassurements are much better closer to the screw you will have to tighten the slides clearances.
Gibscrews or wedge?
Good Luck

I've messed with those and they're pretty tight and no slope anymore. I added a ball bearing to the slide to ride along the bed to keep it from skewing. I'm not super concerned as when I do a test it goes back to zero and the distances are correct.