Greetings; 1: I am trying to get past the infamous, permanently programmed into LCNC error of:
"length of cutter compensation entry move is less than the tools radius" The tool diameter in the tool_table is 0.030 just for test. and the tool table entry contains no TLO, nor am I using an G41 D0, plain old G41. The code has a G40 in its preamble. And the next x or y move after invoking a G41 is several inches away. There is also an intervening z move of a few inches, so it has literally feet of motion to bring in the compensation. The first move after its invocation is I believe, a z move that could be as little as 50 thou depending on where the machine is sitting. OTOH it could be more than 7" in Z, and 18 in x, or 3" in y while its still up in the air above the workpiece. Despite quite a few attempts over the years, I have never actually made it work. Until today, but gouging problems despite my coding in plenty of room for a tool direction reversal. What am I missing? 2: In the manual, it speaks of left and right compensation, and I am having a hard time wrapping my head around the corelation of that language to "climb" or "dig" cutting. As all this wood carving is being done in climb direction to reduce any dig splintering, it seems like I'd have to issue a few of the opposite flaver to assure it doesn't plow up the end of the board on the retrace back to the left edge. That, or pick up the tool for move clearance. In which case having blending on could damage the workpiece. That too is do-able but Y is faster without the tool comp move. And it appears from its behaviour that a G41 is a "climb" cut, while G42 I assume is a "dig" cut. So I think have that sorted. But the gouging exits are a problem yet because it will not draw the backplot to show one where the gouge is. I don't have it when doing the end of a "side" board, but it sticks up its ugly red x hand if I turn side off to do the ends of a end board in this box. If side is on, I do see 2 tiny red x's at the left edge of the backplot, both would occur well off the workpiece at it approaches the starting location of a pass. Beyond that, I am in the dark. So my final question is: Is there a detective method that will show me where it is? Since LinuxCNC doesn't (yet) follow subroutines, blaming everything wrong on the call to the subroutine, do I have to copy/paste a copy of my 2 corner routines so the first iteration of each is inline before it will tell me where the problem is? Since it was the left corner that was the majority of the trouble, I did copy/paste it as inline, twice, and was able to fix enough of it until I hit the right corner code, which I'll have to do the same to, but its getting sleepy out, and way past beer thirty so I am still my one a day behind. If tool compensation can really be made to work, it would simplicate at least 100 LOC out of this program. Already has, but I'm now adding more stuff to make sure there is turn around room at the ends of the cut sweeps. But tomorrow is another day. I am pleased with todays progress. Comments? Thanks. Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
