Kirk Wallace wrote: > On Sun, 2008-01-27 at 09:53 -0500, Stan Blosser wrote: >> Dredging up a slightly old thread, but rather than start a new one... >> >> I'm in the design/planning stages of CNCing a Sieg X3 mill and will be >> using ball screws. At this point, I am planning to use Nook XPR >> screws as they have a lead error of .001 in/ft. The main reason being >> is that I do not have an accurate method (that I know of) to map a >> less accurate screw for error compensation in EMC2. If I did, I would >> likely be looking at Roton screws at .004 in/ft error. Regardless of >> the screws, I'll be using double nuts to adjust out the backlash. >> >> Does anyone know of a "home shop" class method to map a screw? >> Perhaps using a dial indicator and moving along an inch at a time. >> That sounds a bit tedious and prone to mistakes, but it's the only >> thing I can come up with off the top of my head. >> >> Also, I'm assuming EMC2's screw compensation interpolates between data >> points in the compensation file. Can anyone confirm this? If so, is >> it just a simple linear interpolation? >> >> Thanks, >> Stan > > The only ideas I have come up with so far are, to use gauge blocks which > could get expensive in a hurry, and maybe laser interferometry like > here:
For measurements every inch, you can often find jig-borer micrometer sets on ebay or other surplus places, often mis-listed as inside mics like this one: http://cgi.ebay.com/PRATT-WHITNEY-10-PIECE-INSIDE-MICROMETER-SET_W0QQitemZ140200200097QQihZ004QQcategoryZ41937QQssPageNameZWDVWQQrdZ1QQcmdZViewItem (7 hours to go, $22.50) They were used with Pratt & Whitney jig borers - those machines have V-grooves next to the table in X and Y. You set the mic head to the fractional part of an inch you need, and place it plus some combination of the whole inch rods in the V-groove. Then bring the table to position. A protrusion on the table engages the stack of rods and mic-head, and pushes it against an indicator. Turn the crank till the indicator reads zero, and lock the table down. It was a good way to do work that was more accurate than the screw on the machine. The rods in the set on ebay will let you build any number of whole inches from 1 to 35. They aren't quite as accurate as gage blocks, but they are more accurate than just about any home shop machine or reasonably affordable ballscrew. I picked up a similar set a while back for about $40, and used it to map the error on my machine. In my case, they sit quite well in the t-slots of the machine table. I put a solid stop at one end (clamped to the table) and a 0.0001" indicator in the spindle (locked out the spindle motor so I wouldn't do anything stupid). I wrote a g-code program that would approach the rods, pause for me to write down the indicator reading, then go past zero and approach from the other side, pause again for me to take a reading, then move forward an inch plus some finger room and pause for me to swap rods to a stack that is 1" longer, then repeat. I used M0 and (MSG, message-to-operator) in the G-code to tell me what I was supposed to be doing at each pause. A little time consuming, but works nice. Regards, John Kasunich ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
