If by coordinate system offsets you mean X,Y,Z,A,B,C applied to G54-G59.3 and also to a G52 local, then a G68 should not be needed. The rotation offsets on the G54-G59.3 would handle part/program rotations and the G52 would handle rotations of pocketing subs and such stuff within the Gcode program. May even be better as on some controls mixing G52 and G68 stuff can produce unwanted results.
Example of how I am thinking it could be used. Say you had subs for milling letters and numbers and you wanted to mill text into the part in a circular patterns at different lotations on the part. You would clamp the part to the table and set the g54 X,Y,Z work offset to the programmed zero on the part and the A,B,C rotated to the datums of the part. Then in the program move the g52 offset around to the centers of the features on the part around which you wanted the text. At each feature call the subs for each letter incrementing the g52 C offset between each letter. Scaling can if needed be done with the existing macro stuff. #500=1.5(X scale) #501=2.0(Y scale) #502=0.5(Z scale) G0 X[0.0*#500] Y[0.0*#501] Z[0.0*#502] G1 X[1.0*#500] Y[0.0*#501] Z[0.0*#502] G1 X[1.0*#500] Y[1.0*#501] Z[0.0*#502] G1 X[1.0*#500] Y[1.0*#501] Z[1.0*#502] If the three scale values are the same you can also use G2,3 moves. If scaling were built in to EMC I would like to see it done so it was possible to use G2,3 with unequal scale values. Used to make ellipses on wire EDMs. At 08:03 AM 5/20/2008, you wrote: >There was quite a bit of work done at NIST with locating of and >discovery of the pose of a casting using probing. Much of it became a >part of the CMM system they wrote. The advantage of it was that they >could throw a part up there anywhere on the table and the CMM would find >the part, find the pose, and then measure the things it had been told to >measure. > >There was also thinking that information like that found by the CMM >could be translated into the six values assigned to coordinate offsets >g54-g59.3 and used to locate and rotate a milling program. I'm pretty >certain that they did not get that far. They did get a rudimentary six >axis interpreter at the start of the process and six axis motion system >at the other but IMO there was a lot of work yet to be accomplished >between. > >As far as moving or rotating a part program in Cartesian space, I see no >need to go beyond simple coordinate system offsets. I know that this >does not take account of scaling that part program but what else am I >missing here? > >Rayh > > >On Tue, 2008-05-20 at 07:18 -0500, Andre' Blanchard wrote: > > At 06:16 PM 5/19/2008, you wrote: > > >On Mon, May 19, 2008 at 05:41:25PM -0400, xtra209 wrote: > > > > > > > > Is there any work happening on G68? > > > > > >Can you find a link to a manual or similar source that describes > > >definitively how G68 works? I found some examples, and people > > >asking questions about it, but none of them seemed like the full > > >story. > > > > > >Chris > > > > There is this Fanuc manual. > > > <http://www.compumachine.com/Support/Downloads/Fadal/GE%20Fanuc/0010__GE_FANUC_User_Manual.pdf>http://www.compumachine.com/Support/Downloads/Fadal/GE%20Fanuc/0010__GE_FANUC_User_Manual.pdf > > It describes a 2 block 3D rotation which I think is interesting but have > > never seen a machine that supported it. > > > > A manual for an older Mit M300V control. > > It also describes other types of rotations such as pattern rotation > done in > > an M98 sub call and rotations done with the G10 code. And parameter > > rotations which is the one used to align and entire program to a part > > placed on the table, it is done outside the program as part of the > setup of > > the job. Unfortunately not a lot of info on that one but on the > machines I > > have used it on (wire EDMs) it rotates around the work offset zero point > > and is just an angle you enter on the screen as either degrees or a j,k > vector. > > > <http://www.meau.com/functions/dms/getfile.asp?ID=010000000000000001000000460900000>http://www.meau.com/functions/dms/getfile.asp?ID=010000000000000001000000460900000 > > > > I have not used the other types much, mostly just used the 2D G68 > rotations. > > > > More Mit books. > > > <http://www.meau.com/eprise/main/sites/public/DOWNLOADS/-search_results?DocType=010&ManualType=0055>http://www.meau.com/eprise/main/sites/public/DOWNLOADS/-search_results?DocType=010&ManualType=0055 > > > > __________ > > Andre' B. Clear Lake, Wi. > > > > > > > > ------------------------------------------------------------------------- > > 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 > > >------------------------------------------------------------------------- >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 __________ Andre' B. Clear Lake, Wi. ------------------------------------------------------------------------- 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
