On Tuesday 22 September 2020 02:30:21 dave engvall wrote: > Hi again, > Indeed it is pretty hard to use spindle load if is doesn't change thru > the corner. I suspect it does on smaller machines but that is not a > generalized solution. > Just for experimental purposes it might be interesting to decrease the > velocity thru a corner see if it helps. Lacking the adding another row > to the matrix should control jerk but the may not be the whole > solution when dealing with a change in direction. So much to learn so > little time (and energy). > I suspect the 'big boys' do some pretty fancy adaptation to get good > smoothing thru the corners. Of course it is all propitiatory and those > patents take a long time to run out. Silvia McFarland (sp?) in her > masters thesis (Uni BC) used sin accel rather than trapezoidal.
The proper term for that is sine-squared, as we used it in broadcasting back in Never Twice Same Color days where you start at the peak of the waveform where the velocity change is zero, accelerate in the shape of a sine wave where velocity is greatest going thru the center of the move, and then following the same curve, wind up at max speed and keep on going as the peak of the curve is attained in the new direction. We used a test signal that started and ended at 7.5% black Done right, using a limit3 to try shaping it, the vfd on my rpi4 driven Sheldon lathe, can reverse the spindle direction of a 40 lb 4 jaw chuck with an overtravel of only .25 turns at 100 rpms. Chuck obviously is clamped to the spindle so it can't unscrew itself. It may go faster, but the belts slip some. Since I put the red link belts on it, they slip quietly, the badly worn stock v-belts yelped like Michelin tires on hot blacktop. I seriously need new pulleys, they have a good belly worn into them in their 70+ years of use. > Might > help. I think traj control get much more critical as velocities get > really radical. Some of the new Brothers with 30 tapers absolutely > scream. When the machinist describes stepping back from the machine > when the tool moves toward him then is must be really impressive. > > Dave > > On 9/21/20 4:55 PM, andrew beck wrote: > > hey dave > > > > I hear you. I might get good at coding myself soon.. but would > > love to solve the problem. > > > > I don't think the spindle load could really help much though. > > > > I am machining plastic and aluminium and the spindle load doesn't > > noticeably change in corners. I more need a limit on the initial > > acceleration for a split second I suppose > > > > or go the full way and have something that analyzes the G code path > > and calculates the best speed and path for the machine > > > > regards > > > > Andrew > > > > regards > > > > On Tue, Sep 22, 2020 at 11:35 AM dave engvall <dengv...@charter.net> wrote: > >> HI Andrew, > >> Just because I have ideas does not imply I can implement them. ;-) > >> However, I believe there is a velocity pin that might be used to > >> slow down in corners. Since the spindle loads up going thru corners > >> it may be possible to monitor spindle load and back off velocity > >> thru the corner which should make milling a bit smoother. Adding > >> jerk control involves , as you know, adding another order to > >> motion. (see first sentence). ;-) A lot of bright people are > >> involved in linuxcnc but they all have their own pet projects. Just > >> the way it goes. > >> Hang in there, your heart is in the right place. > >> > >> Dave > >> > >> On 9/20/20 8:19 PM, andrew beck wrote: > >>> Hey guys > >>> > >>> just a few thoughts here. > >>> > >>> I have a high speed cnc machining center (cnc mill) > >>> > >>> it has 16mm pitch ballscrews and weighs 6ton > >>> > >>> linuxcnc is awesome and does a great job controlling it. I make > >>> plastic injection molds and lots of 3d machining all the time. > >>> plus all the > >> > >> little > >> > >>> jobs that a cnc is great for. > >>> > >>> I think the only thing lacking from linuxcnc for my use is some > >>> sort of jerk control for corners. there may already be something > >>> around that solves the issue though without jerk control. so what > >>> happens is when > >> > >> the > >> > >>> machine changes direction the whole machine vibrates and jumps > >>> around. > >> > >> its > >> > >>> quite bad. enough that I may rubber mount the electrical box lol. > >>> > >>> I have accelerations of 750 (i think it is mm/sec2) > >>> > >>> and max rapids of 8m/min > >>> > >>> I would love to turn those up to like 25m/min rapid and > >>> 2000mm/sec2 accelerations > >>> > >>> has any one tried any form of jerk control yet on linuxcnc? or > >>> wants to have a play and code it. I have 3 6 ton cnc machines > >>> that will all use linuxcnc so very keen to solve this jerk control > >>> issue. > >>> > >>> regards > >>> > >>> Andrew > >>> > >>> _______________________________________________ > >>> Emc-users mailing list > >>> Emc-users@lists.sourceforge.net > >>> https://lists.sourceforge.net/lists/listinfo/emc-users > >> > >> _______________________________________________ > >> Emc-users mailing list > >> Emc-users@lists.sourceforge.net > >> https://lists.sourceforge.net/lists/listinfo/emc-users > > > > _______________________________________________ > > Emc-users mailing list > > Emc-users@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/emc-users > > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users 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) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis 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
