Hi Chris, others. Really tool length is just part of the problem. A tool number is associated with both a tool length and a tool diameter which may or may not be nominal. So far I just grit my teeth and work thru the process. I have considered (briefly) a procedure to normalize the measured lengths against the reference tool and write that to the tool table.
Dave On Apr 29, 2007, at 8:03 AM, Chris Radek wrote: > On Sun, Apr 29, 2007 at 03:34:34PM +0200, Sven Mueller wrote: >> Chris Radek wrote: >> [probe moves] >> >>>> It would be nice to be able to set tool length offset this way - >>>> but >>>> currently there is no way to do that, since lengths need to come >>>> from the tool table. >> >> Hmm, I'm still mostly a newbie regarding EMC. But does the above mean >> that EMC can't automatically correct for tool length differences >> unless >> you measure these in advance and store them in the tool table? > > Well yes and no. In the part of my message you clipped, I mentioned > that you could set a G10 offset using the result of a probe. But if > you're talking about G43 tool length offset, that's right. > > As you can see here, > http://www.linuxcnc.org/handbook/RS274NGC_3/RS274NGC_33a.html#1013768 > G43 takes a tool number, not a length - so yes it is intended that > you measure your tools (which are in holders) and put that > information in the tool table. The G43 length is measured relative > to a "reference" tool, and you use the reference tool when you set > the coordinate systems, touch off to the work, etc. > > Yesterday sat down to change the interpreter so you can specify a > length to G43. This is a simple change to make but I quickly > found a couple problems/questions: > > Currently, G43 H- X- Y- Z- changes the tool offset and then causes > motion to XYZ with the previously set motion type (G0,1,2,3). I don't > see this behavior described in the ngc spec, but I'm afraid to use > XYZIJK for something else if people use the current behavior. So > that's problem 1 - I don't know how to allow specification of these > lengths at the gcode level without breaking existing programs. (If > you guys agree that an existing program that uses this behavior is > NOT ngc, I'd feel less bad about breaking it.) > > Even if G43 can take a tool length directly, I couldn't come up with a > straightforward way for the user to use that in gcode. Imagining the > mill case first because it's simpler: maybe you'd store in a gcode > variable the Z coordinate you get when you probe your reference > tool. Then when you probe a new tool you'd take the difference and > pass that to G43. But if someone doesn't have tool holders at all, > I don't know what the reference tool would be, and with no > reference tool I don't know what you'd use to touch-off the work. > > Tool length offset is just that - a length - and to measure a length > you need to find the difference between two points. If you probe a > tool tip, that's only one. This is problem 2 - I don't quite see a > good way for someone to use G43 in this way, after we solve problem > 1. > > Chris > > > ---------------------------------------------------------------------- > --- > This SF.net email is sponsored by DB2 Express > Download DB2 Express C - the FREE version of DB2 express and take > control of your XML. No limits. Just data. Click to get it now. > http://sourceforge.net/powerbar/db2/ > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users