On Tue, 2012-09-25 at 13:20 -0500, Igor Chudov wrote: > I have a routine that lets me mill a deep circle in several passes, > typically to cutout circles from plates. > > Something strange is going on and the resulting shape is not exactly round. > It can be best described as a circle with humps. > > The routine is below. It is basically a while loop, milling progressively > deeper until Z reaches the required depth. Every loop consists of two half > circles milled helically, with the end mill going down along a spiral. > > Any idea why the shape may not end up being circular? > > Thanks > > Routine attached. > > (Copyright - Igor Chudov, released under GNU Public License V3) > > O<circulargroove> sub (circular groove, good for milling through holes too) > #<xc> = #1 (X center) > #<yc> = #2 (Y center) > #<safez> = #3 (safe height) > #<depth> = #4 (depth of milling) > #<radius> = #5 (outside radius) > #<milld> = #6 (end mill diameter) > #<frate> = #7 (feed rate) > #<zstep> = #8 (z step, per circle, optional) > > O<if> if [ #<frate> NE 0 ] > F#<frate> > O<if> endif > > O<if> if [ #<zstep> EQ 0 ] > #<zstep> = [#<milld>/3] > O<if> endif > > #<z> = #<safez> > #<r> = [#<radius> - #<milld>/2] > > G0 Z#<safez> > G4 P0 > G0 X[#<xc> + #<r>] Y#<yc> > > O<loop> while [ 1 ] > > O<if> if [ #<z> - #<zstep> LT #<depth>] > #<zstep> = [#<z> - #<depth>] > O<if> endif > > G2 X[#<xc> - #<r>] Y#<yc> Z[#<z> - #<zstep>/2] R#<r> > G2 X[#<xc> + #<r>] Y#<yc> Z[#<z> - #<zstep>] R#<r> > > #<z> = [#<z> - #<zstep>] > > O<if> if [ #<z> LE #<depth> ] > O<loop> break > O<if> endif > > O<loop> endwhile > > G2 X[#<xc> - #<r>] Y#<yc> R#<r> > G2 X[#<xc> + #<r>] Y#<yc> R#<r> > > G0 Z#<safez> > > O<circulargroove> endsub > > M2 Hi Igor, Just off the top of my head I suspect that your program is fine; but the machine is well used and has measurable backlash. In a perfect machine you get real axis reversal ( X | Y ) every 90 degrees. If your ballscrews (lead screw) have backlash then you have to catch up with that before the axis actually changes direction. Since the slope is small around the inflection points not much happens until you take up the slack and then you see the actual reversal. Might be as much as 10 degrees off the ideal. Multiple passes may make it look better but will not necessarily improve the precision. Backlash comp does not fix this; only throwing $$$ at the machine. :-(
I hope it turns out to be something fixable without the injection of many $. Dave > ------------------------------------------------------------------------------ > Live Security Virtual Conference > Exclusive live event will cover all the ways today's security and > threat landscape has changed and how IT managers can respond. Discussions > will include endpoint security, mobile security and the latest in malware > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
