On Tuesday 08 March 2016 19:24:33 Gene Heskett wrote: > Greetings; > > I have gone thru this machine, starting with nuking linuxcnc.var, > homed it and zeroed it until the DRO display, even when asked to show > offsets, is reporting a whole screen full of 0.0000 for all axis's. > > Then I used a pad of pcb to measure, and put into the tool table, all > the Z axis offsets from there to contact with the the pcb surface > which was laying nominally 0.070" above the material surface. > > I load my code to carve the needed clearances for a Rockler lid stay > hinge, code which assumes a touch off at the material surface is > 0.0000, and likewise the XY are touched off at the center of the > workpiece, but when I do, from the MDI line, t17m6g43, which loads a > 1/4" round nosed mill, the dro says the tool tip is at the surface of > the material just fine when it is run there by hand. And I can run it > clear to that home/limit switch, 5" above the the backplots workpiece > location, just fine. > > Just one problem, the limit box in the backplot says the tool tip is > only about 150 thou below the z axis upper limit!, and my code refuses > to run because it runs up 4" to a tool change height, which there are > 3 of in the code.. > > This does not make sense, the machine, when at the material top > surface, is almost 5" below the upper limit/home switch. > > From the .ini file: > > MIN_LIMIT = -10.25000000 > MAX_LIMIT = 0.36900000 > > Which I've always been told is in machine units. That MAX_LIMIT is > about 10 thou from crashing the nut bracket into the top of the slot > in the machines post. And running it by hand, thats where it stops, > still sitting on the switch. But thats close to 5" above the red > machine limits shown in the backplot when I do. > > I'm running out of hair! > > Most recent, as of about 15 hours ago version of 2.8.0-pre. > > Have I hit a bug because its not 2.7.4-something? IDK, only that its > hugely frustrating and seems to be a quite recurrent problem, > occasionally fixable by shutting down, nuking linuxcnc.var and > var.bak, and restarting and rehoming. But thats not fixing it, I have > chased that rabbit around the briarpatch 4 times in the last 4 hours. > > FWIW, I just copied that tool.tbl to sim.table here, and loaded the > code which ran, not right because a conditional is bass ackwards, but > it ran. > > This is 2.7.4-sim.axis > > So what do I do now? > > If its a bug, thats fine, and maybe I've headed it off from getting > into the mainline code most run, I am cool with that. Frustrated yes, > but if I can bleed & fuss a little and make the stable code better, I > am happy. > > Thanks everybody. > > Cheers, Gene Heskett
It occurs to me that I might be able to diddle the Z home data in the ini file, possibly the offset value, in such a way as to place the freshly homed z axis 0.000- point nominally 5" below the switch, and then ajdust the MIN_LIMIT and MAX_LIMIT settings to restore the limits at the same place as they are now. So, what should I do to this section of the ini do do that? HOME = 0.00000 HOME_OFFSET = 0.25000 MIN_LIMIT = -10.25000000 MAX_LIMIT = 0.36900000 Everytime I try to do something with those HOME lines, it seems to move it backwards. With 10.369 + 0.250= 10.619" of total travel, what do I do to the HOME* data to arrive at a symatrical up-down set of limits, eg -5.45 and 5.45. But I just padded out in my socks and trued again. About 10 or 12 restarts later, I think I have that. But with my vise height thrown in, Z=0.0000 is about .800" below the top of the workpiece. I'll go rebuild the tool.tbl tomorrow useing g38.2, and it looks as if the limit lines in the backplot are now correctly drawn. Who knows, maybe I can get lucky. I also found that an t0m6g43 does NOT cancel any TLO in effect, I must use a G49 to do that. Maybe that was my problem all along? Because what I was doing was cumulative? Which begs the question: Was it intended that way? For the time being, I will do that before I call the tool change because that, it stands to reason, would standardize the height of a G0 Z4 prior to the tool change call too. I'll get this basket of rattlesnakes sorted yet. :) 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> ------------------------------------------------------------------------------ Transform Data into Opportunity. Accelerate data analysis in your applications with Intel Data Analytics Acceleration Library. Click to learn more. http://pubads.g.doubleclick.net/gampad/clk?id=278785111&iu=/4140 _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
