I thought about that except that the data file is setup via the same method as the Y axis.
However the X axis screw is much worse than the Y axis. I went over the file with the guy who did the measuring, and he agreed that the data seems like it is in the proper order. Its weird that the Y axis is working really well, while the X axis is not working well at all. I have a small X-Y machine in my office that I may setup with a dial indicator and do some testing just so I can better understand the screw comp. I did some looking and Nook, a big ball screw maker in the US says that their rolled ball screws, which are not normally used for machine tools are accurate to +/- .004 per foot. In the first foot of this X axis screw, the screw is out 0.030 of an inch. Nook's ground ball screws, standard accuracy are good to +/- 0.001 per foot. So I wonder if I have reached the limit of LinuxCNCs ability to compensate for a junk screw? Dave On 6/28/2016 11:56 AM, andy pugh wrote: > On 28 June 2016 at 15:35, Dave Cole <linuxcncro...@gmail.com> wrote: > >> LinuxCNC is trying to use the x axis data since ball bar testing >> results change significantly if the X axis screw comp is enabled or >> disabled. However the screw comp is not effectively correcting the >> screw. > Is it possible that the compensation file is upside down? > ------------------------------------------------------------------------------ Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San Francisco, CA to explore cutting-edge tech and listen to tech luminaries present their vision of the future. This family event has something for everyone, including kids. Get more information and register today. http://sdm.link/attshape _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
