Just another 'threadless ballscrew' 3D-printed... What I really wanted to figure out is this: How much slip does this ballscrew accumulates when travelling over the smooth rods, moving around back and forth a couple times...? Is it suitable for an open-loop design, or do you need position sensors for feedback..? Backlash is very low, which is nice. The essential geometry of the tilted bearings was done in OpenScad, nice and customizable. The roller bearings push flat on the shaft surface. The script is a remix from: http://www.thingiverse.com/thing:124706/#files Thanks for that one! The part from OpenScad was then further edited in tinkercad to add fastening screws, double sides, and more details: https://tinkercad.com/things/73GGA5MH9z7 3D-printed and CNC-machined at FabLab Zurich: http://www.zurich.fablab.ch See the parts moving: https://vimeo.com/76689145 This is still work in progress... But as for now: I got the final position after the test run in the video pretty accurate ~0.1mm (sorry, you can't check the scale), but if a disturbing force is applied during the movement, the ballscrew slips and things gets out of track very quickly. At the moment, I do not see a chance for open-loop control to function reliable. Has anyone else gone this way with success? I am very interested if anyone else could share test results of their builds! Future work: