an edge finder is just a fancy dowel contraption with a spring in it, and costs about $10. this looks like generic info:
http://www.youtube.com/watch?v=f0od-cp_9dg i have never seen one smaller than 3/8" diameter, but it would be easy to make one somewhat smaller. a plain dowel and a piece of shim material (like paper) as a feeler works too. the point of a center finding macro is to circumvent human based steps in the setup process, like finding a calculator, correctly typing in a number and dividing by two, and then correctly entering the result into the machine control. conductive probing of conductive objects works, and i've used such a simple setup as watching an ohmmeter connected between workpiece and tooltip to find surfaces before. but i've also seen 100% failure of conductive probing for tool measurement in a production environment because of the swarfy, oily water environment in the business area of the machine tools. the results are automated mechanical destruction of things. the initial results for setting up tool offsets are great, but the scheme seems to fail rather quickly. --- On Sat, 3/24/12, gene heskett <ghesk...@wdtv.com> wrote: > From: gene heskett <ghesk...@wdtv.com> > Subject: Re: [Emc-users] Tool change question > To: emc-users@lists.sourceforge.net > Date: Saturday, March 24, 2012, 8:39 AM > On Saturday, March 24, 2012 11:27:23 > AM charles green did opine: > > > looks like there are some numbered params listed in the > master doc also: > > > > #<_x> - Return absolute machine X coordinate. > Same as #5420. > > #<_y> - Return absolute machine Y coordinate. > Same as #5421. > > #<_z> - Return absolute machine Z coordinate. > Same as #5422. > > #<_a> - Return absolute machine A coordinate. > Same as #5423. > > #<_b> - Return absolute machine B coordinate. > Same as #5424. > > #<_c> - Return absolute machine C coordinate. > Same as #5425. > > #<_u> - Return absolute machine U coordinate. > Same as #5426. > > #<_v> - Return absolute machine V coordinate. > Same as #5427. > > #<_w> - Return absolute machine W coordinate. > Same as #5428. > > #<_current_tool> - Return number of the current > tool in spindle. Same as > > #5400. > > > > i wonder if these are also effected in the released > versions? this is > > more like what i was imagining might be the access to > various states of > > machine and loaded table values like tool offset > values. > > > > on many milling machine controllers, G43 Hxxx requires > a movement of the > > z axis to'take up' the tool length offset, like G0 G43 > Z-3.0 H5. at > > the end of the move (which differs from the commanded > -3.0 units by the > > length offset value in offset table position number 5) > the end of the > > tool is aligned with the currnt coordinate > system. if you know the > > parameter locations of the current tool and its length > offset, the code > > for taking up the current tool offset becomes G0 G43 > Z-#[<current tool > > number> + <tool length table parameter > offset>] H#[<current tool > > number> + <tool length table parameter > offset>]. so there is no axis > > movement, and that coding step in a tool introduction > is uniform for > > all tools, regardless of length. (the tool change > is often near the > > travel limit of the z axis, so a large tool offset > might result in an > > overtravel if the axis movement is not large enough. so > in the first > > example, G0 G43 Z-3.0 H5, if H5=10.00 and there is only > another +1.0 to > > go in the z axis beyond the tool change position, the > limit will be > > reached 6.0 short of completing this step.) > > > > at the place i have been working, all the mill > controllers have been set > > up with a macro override of the T word that > encapsulates all the > > discreet coding steps required for MDI tool changing, > as well as part > > programming. the result is improved setup time on > the machine, and > > faster part operation coding which is portable between > the various > > different machines. > > > > another handy macro (call it G192 for example) finds > the center of two > > touch off locations along an axis. the procedure > goes like, > > > > touch the -x edge of the part with edge finder > > G92 X0 > > touch the +x edge of part > > G192 X0 (<- "X0" is just convenient, the X-word > value is not used in the > > calculation.) > > > > where the G192 macro uses the machine coorinate > parameter values of the > > selected axis name (X in this example) to set the > origin at the center > > of the part edges. > > > > it would be nice to adapt such tactics in linuxcnc, > using the custom M > > codes perhaps? will the custom codes work from > MDI? they would at > > least be valuable for modularization of part > programming, like for the > > tool change routine. command M106 Txx Sxx: > > > > (M106 TOOL CHANGE MACRO) > > M9 > > G91 G30 Z0 M5 > > T#5400 M6 > > S#???? M3 > > G0 G43 Z-[H#5400] H#5400 > > G17 G90 M8 > > M99 > > > > ..or something like that. > > > G38.2 doesn't demand a Z probe, it can do any axis, so > LinuxCNC is quite > able to do that. > > I did something similar in tholefinder.ngc where I probe the > open end of a > small piece of brass tubing let into the pcb pallet to find > all 4 contact > points and declare the center, with all co-ordinates offset > from that for > the rest of a pcb etch & drill run. That worked > well but needed a very > fine wire, in the spindle with the spindle running, and a > smallish > capacitor to capture the first contact. I actually > used a sewing machine > needle threader wire for that. Repeatability seems to > be about a thou. > Often less. I don't own an edge finder, so I haven't a > reference I can > compare my method to. It works for me & cost $2. > :) > > > --- On Fri, 3/23/12, Sebastian Kuzminsky <s...@highlab.com> > wrote: > > > From: Sebastian Kuzminsky <s...@highlab.com> > > > Subject: Re: [Emc-users] Tool change question > > > To: emc-users@lists.sourceforge.net > > > Date: Friday, March 23, 2012, 9:29 PM > > > On 03/23/2012 11:42 AM, charles green > > > > > > wrote: > > > > nicer, but still only a handful of > documented > > > > > > params. i was looking at a document dated a > little > > > over a year ago (11jan2011). > > > > > > > do you know if the params #31 to #5000 > contain useful > > > > > > info? like g90/g91 state, for example, or > any > > > registration of the other groups apart from group > 12? > > > > > > > "you started this. show me everything." > - ellen > > > > > > ripley > > > > > > Oh... I don't think you want to mess with > > > that... - Hicks > > > > > > Parameters 31-5000 are deliberately unused by > linuxcnc, and > > > thus > > > available for the g-code programmer to use. > > > > > > Ah yes, that's even what the docs say (somewhat > cryptically, > > > focus on > > > > > > the "user parameters" part): > > > > * > > > > > > > > /1-5000/ - > G-Code user > > > > > > parameters. These parameters are global > > > > > > > in the G Code > file. > > > > > > The (slightly) improved docs in the master branch > > > (http://linuxcnc.org/docs/devel/html/gcode/overview.html#sec:parameter > > > s) > > > > > > say this: > > > > 31-5000 > > > > > > > > G-Code user > parameters. These > > > > > > parameters are global in the G Code > > > > > > > file, and available > for general > > > > > > use. Volatile. > > > > > > > > > > > > Oh hey look! In master (aka devel), there > are some > > > predefined named > > > parameters that carry the g90/g91 state: > > > > > > http://linuxcnc.org/docs/devel/html/gcode/overview.html#_predefined_na > > > > med_parameters_a_id_sec_predefined_named_parameters_a > > > > > > I think those variables are new in the master > branch, and > > > not available > > > in the 2.5 branch. > > > > > ------------------------------------------------------------------------ > > ------ This SF email is sponsosred by: > > Try Windows Azure free for 90 days Click Here > > http://p.sf.net/sfu/sfd2d-msazure > > _______________________________________________ > > Emc-users mailing list > > Emc-users@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > Cheers, Gene > -- > "There are four boxes to be used in defense of liberty: > soap, ballot, jury, and ammo. Please use in that order." > -Ed Howdershelt (Author) > My web page: <http://coyoteden.dyndns-free.com:85/gene> > Hardware, n.: > The parts of a computer system that can > be kicked. > > ------------------------------------------------------------------------------ > This SF email is sponsosred by: > Try Windows Azure free for 90 days Click Here > http://p.sf.net/sfu/sfd2d-msazure > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
