My only thought is:
If screws are good and nuts are nicely set up then:
X axis supported at both ends = good and solid. Immediate screw response.
Y axis only supported at one end = bad. Screw flexing causing delayed response.
After completing my 2 axis CNC Conversion on my Tawiwan BP clone knee mill, I've found what I believe is an unacceptable amount of "backlash" in my y axis, between 0.0018"-0.002". Both axes use new Hiwin ballscrews and my x axis is nearly perfect, only 0.0003"-0.0005". I've upgraded to a solid, and might I add beefy, steel ballnut yoke. I've also upgraded the bearing mounts to one's made of steel. I'm running SKF 7204 ac bearings on both axes, preloaded in back-to-back fashion. I can jog the y axis with tenths indicator in the spindle up against the fixed vice jaw, then set my DRO (glass scale) and indicator to zero. Next, I jog in the opposite direction exactly 1.0000" according to Acorn CNC. My glass scale reads 0.9980". Now I jog back towards the vise jaw 1.0000", according to Acorn, and my indicator and glass scale will zero out perfectly. I've repeated this test several times with the same result. I switched my way oil to ISO46 hydraulic oil hoping to cut down on stiction of the ways but that did nothing to help the backlash. I'm really stumped by this, any suggestions?
Similar Threads:
My only thought is:
If screws are good and nuts are nicely set up then:
X axis supported at both ends = good and solid. Immediate screw response.
Y axis only supported at one end = bad. Screw flexing causing delayed response.
I never thought about the y axis ballscrew being unsupported, I can see how that could influence the movement. I'm not ruling out that the ballnut is not preloaded correctly but that is my last ditch idea since it is brand new and Hiwin has a good reputation. How common is it that CNC converted knee mills have more backlash in Y than in X due to the inherent design of the axis arrangement? Anybody have any experience with this?
Mine is only a converted benchtop but I did a little test which led me to his thought initially.
I have 0.025-0.03mm backlash in both axis.
X is basic supported with a standard bearing, Y is unsupported. Booth drive ends are the same as yours with 2* AC BtoB.
Put an AC bearing in X supported end and got it butted up to the screw. X dropped down to 0.15mm.
Can't do anything more than I have already with Y.