Thanks for the feedback. Here are the other two screen shots. One showing how the G54 is now applied to the tool position after the touch off so that now the POSition accurately reflects the work coordinate system. The other shows that I want to move in from 0.500" to 0.475" to do a 0.025" pass. This would change the diameter to 0.950".
Haven't looked into how G7 diameter mode is handled yet. My only experience with LinuxCNC at the moment is on the mill and since I didn't want to run into mechanical end stops at full tilt I enabled limit and home switches. So even though it's not easy to see the world coordinates of the machine a G0 G53 X0 does take my mill to the point 0.1" from the limit/home switch. And it's true that once I've done the homing (my LCNC won't do programmed moves without it) I never look at the G54..Gnn values anymore. Don't even use G92. I suspect most of those are targeted more at production environments. However to understand the G54 part workspace coordinates I believe one has to get a feel of what's happening under the covers. For example, show the second DRO screen in AXIS, and it's clear the G54 values never change when the machine is moving around. The G54 causes the DRO actual machine positions to be offset from the G53 machine world coordinates. So do the soft limits use the part coordinate system(G54) or the world coordinate system (G53)? I think G53. @Leonardo. What you've done isn't really much different other than perhaps I'm sharing a limit switch with the home switch. In your case if your limit switch is mid bed somewhere (for Z) then that becomes G53.z == 0.0000. And you're right. It doesn't matter because I'm pretty sure the soft limits are referenced from that G53.z == position and the G54 is referenced from wherever you do the Touch Off sequence using the G53 values to make the DRO POS.xyz == 0.0000. @Andy -- Yes. Your approach for Z as the end of the work for every job is how I designed my ELS too. There the keypad has a BEGIN/END button and generally Z is set to 0.000 after the end of the part is faced off. The BEGIN is slightly to the right of that for ramping up to speed and the END is either travelled to and set or keyed in from the keypad. So turning 2.000" might have BEGIN = 0.200" and END will have -2.000". Therefore, Touch Off at the face and then using G0 Z0.2, and G1 Z-2.0 is essentially the same. (Non synchronized to spindle in this case of course). It's the tool table that ultimately has me interested. You can't change to a different tool and then face off again to find that elusive G54.x == 0. Although I guess one could use a feeler gauge or thin paper to sneak up to the work to touch it and then enter the new, now measured, offset into Tool Touch Off or even just into Touch Off. Sometime today or tomorrow I'll add tool table support to the program and post some more pictures. John > -----Original Message----- > From: John Dammeyer [mailto:jo...@autoartisans.com] > Sent: December-07-22 11:32 PM > To: 'Enhanced Machine Controller (EMC)' > Subject: Re: [Emc-users] Home switches for LinuxCNC on a Lathe > > I've done a bit more work on this. I've made an assumption that the home > switch is about 5" from the > theoretical lathe centerline with a specific tool inserted in a fixed holder. > > I home the cross slide setting 0.000 into the world coordinates G53 location. > Then move it -4.0" inwards > which puts it conveniently at exactly 0.5" from the real lathe center line. > > Now we want the DRO to read 0.5" at this point, not -4.000. So this is where > the G54 comes in and the AXIS > Touch Off Tool dialog. In that dialog we'd enter 0.5" which then ends up > putting a value into the G54 offsets > locations which then is combined with the G53 to fill in the DRO position > fields to show the position relative to > this new 'home' position. > > Since I had trouble getting my head around this I wrote a small program to > simulate some of the AXIS > interface. I haven't done the tool table stuff yet. I want to get this > clarified first. > > What I'm not sure about is whether I'm doing the math (and the sign of the > value or even the value) correctly > in the G54 locations. > > I've attached some screen shots. These shots show the initial move, the > touch and then the change to G54. > It appears to work. > > John > >
_______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
