Looking fo help setting my backlash on my Microkinietics CNC Express

[PHSS]

Piece of crap mill, it was in the shop when I inherited it (high school metal shop) and never worked. Finally today I got the power hooked up to it (work order was in for over 2 years, I love my school districts idea of what is a priority, not me, but the school boards office needed new lights...) anyways When I got it running, low and behold I have over 30 thou backlash on the Y axis. Now I know that Microkinetics put their own ball screws and tapered roller bearings in the machine, and its not a normal backlash problem, the problem is this:

When you move the saddle towards the column (a round column mill drill for those that don't know, sucky I know) the outer roller bearing by the motor is tight to the housing but inside there is some room between the inner bearing and its race. When we reverse the direction the ball screw, bearing assembly moves outwards 30 thou before the bearing contacts and rides against its race. I have tried everything, made sure I wound the screw out so the inner bearing was tight to the race, then put on the pulley and lock nuts on, using a pipe wrench of all things to get the bearings preloaded as best I can. No difference the best I get is 30 thou, useless for cnc. I am going to talk to the manufacturer tomorrow but have found them to be useless in the past. Anyone know if anyone else uses a ball screw system like this and any ideas. All I can think of is the inner bearings rests on a turned shoulder pressed on and the outer bearing slides up to this shoulder but not far enough to contact the race. I will try shimming the races and see if that gets me anywhere. Anyone got any ideas?

[BobyKeen]

The best thing you can do right now without dismembering and trying to shim the bearing is to use the software's built in backlash compensation feature if you have a constant backlash, of course and your software has the feature. You might have to tighten a bit more the slides too and avoid climb milling because you will break the tool otherwise. If you decide to eliminate the backlash, you can try and use some liquid metal instead of shims because to shim that bearing you will need to turn or cut pretty precise spacers out of sheet metal... Another idea is to try and extract the pressed shoulder and file it or better turn it 30 thou or a bit more (I use metric sistem so for me is about 0.75mm if I'm right). If you can't extract it you have to take out the entire screw and turn that shoulder down... I see no other reliable solution, honestly.

[PHSS]

Found out some things today. The back bearing was not seated correctly for alignment reasons. I got it all aligned but now I have on both axis 7 thou backlash. The company says to tighten the nuts/pulleys up until the backlash is gone, no more, well I have 7 thou and they are so tight that the 470 oz/inch motors stall out and the gibs are not even tight. I don't know what to do but accept the backlash. The software has s backlash compensation but it doesn't seem to work from what I can tell, it says to enter the distance and select either positive or negative. I am not sure which to select. I know this is a POS idea of a mill but I don't know what to do as its what I am stuck with, there is no way my high school will replace it with somthing else. I guess the next step is to see if they will tell me how to adjust their ballscrews.

[BobyKeen]

As far as the software goes, the best idea is to test it; you can't break anything... The easiest way to check if it works is to use a dial gage and just move the axis 1/4 inch back and forth and you will get the backlashfrom the diference in reading. Try both positive and negative and see what is going on... I suppose they offer you both to set up the backlash after or before the 0 absolute of the axis, if you use the home in the midle of the table. Anyway, trial and error... You will still have backlash but the axis should come back to the same point when the feature is used corectly. With used ballscrews you might get a baclash anyway so you have to learn how to use the compensation feature anyway. 7 thou is a lot of backlash tho', check if the problem is with the nuts or the bearings. If is the bearing assembly, you can make shims from a piece of paper; two normal paper sheets are about 6-7 thou thick. If the nuts are the problem... may Force be with you... you might get a better chance using cheap threaded bars and PTFE nuts as I use on my router... By the way, search some forums for the software you use, maybe someone figured it out...