Back
[11:44:30] <skunkworks> log
[11:44:35] <skunkworks> zlog
[11:53:29] <JT-Shop> the one time you use log and Tom has that PC offline for some other work
[11:55:36] <skunkworks> :)
[11:57:50] <pcw_home> Can LinuxCNC do 5 Axis tool length compensation?
[11:59:15] <seb_kuzminsky> pcw_home: yes
[11:59:47] <seb_kuzminsky> some docs on it:
[11:59:50] <seb_kuzminsky> http://linuxcnc.org/docs/devel/html/gcode/g-code.html#gcode:g10-l1
[11:59:57] <seb_kuzminsky> http://linuxcnc.org/docs/devel/html/gcode/tool-compensation.html
[12:00:03] <cradek> beware people seem to mean different things when they ask that
[12:00:13] <seb_kuzminsky> a wild cradek appears!
[12:00:22] <cradek> yeah I'm back now
[12:00:28] <seb_kuzminsky> did they let you back in the country, ya hippie?
[12:00:37] <cradek> yeah it was easy, I'm very white
[12:01:20] <pcw_home> No political test yet?
[12:01:49] <seb_kuzminsky> or religious?
[12:02:34] <cradek> I came back through customs wednesday, before the biggest bombshell
[12:02:44] <cradek> but again, I'm very white
[12:02:56] <pcw_home> lucky timing
[12:04:15] <seb_kuzminsky> lucky coloring
[12:04:25] <cradek> yeah
[12:04:34] <cradek> siiigh
[12:05:05] <pcw_home> Trying to come up with a flexible strategy for "data painting"
[12:05:06] <pcw_home> I suspect I would want the non-compensated waypoints but the position control would need the compensated ones
[12:06:26] <pcw_home> maybe this thing with getting rid of 1/2 the regulations will drop the travel ban
[12:30:47] <skunkworks> cradek, did you see
https://www.youtube.com/watch?v=lTNzBLNBNiE
[12:31:10] <skunkworks> got it down to less than .00008"
[12:31:38] <cradek> neat
[12:32:00] <skunkworks> they seem like a good fit. won't turn the ball screws inside out I think
[12:45:42] <skunkworks> cradek, where did you go?
[12:55:12] <skunkworks> pcw_home, what are you working on?
[12:56:38] <pcw_home> I have some custom work plus a low cost Ethernet analog servo card (7I97 = like the 7I96 but analog servo)
[13:01:12] <skunkworks> ooh... when is that coming out? I think dragon would be interesting fromt hat
[13:03:56] <pcw_home> maybe mid year (still routing it as I have time)
[13:05:04] <skunkworks> awesome
[13:06:50] <skunkworks> you mean 7i76?
[13:09:55] <skunkworks> e
[13:18:03] <pcw_home> 7i96
[13:18:17] <cradek> skunkworks: southern india
[13:18:35] <skunkworks> oh - the wedding that jepler talked about?
[13:18:41] <skunkworks> obviously ;)
[13:18:54] <skunkworks> no- I just happened to go to india at the same time..
[13:19:00] <cradek> heh
[13:19:25] <skunkworks> I don't know the 7i96
[13:19:36] <cradek> jeff is still gallivanting, a few more days I think
[13:20:04] <skunkworks> I don't find it on your site
[13:23:37] <pcw_home> I think its in the store
[13:23:48] <skunkworks> http://store.mesanet.com/index.php?route=product/search&search=7i96
[13:24:08] <pcw_home> Hmm maybe not
[13:24:12] <skunkworks> heh
[13:25:36] <pcw_home> I conjured it into existence using bad memory
[13:25:49] * seb_kuzminsky hands pcw_home a stick of ECC RAM
[13:26:52] <pcw_home> :-)
[13:27:05] <pcw_home> I think Kim K has one
[14:08:35] <pcw_mesa> unless I conjured that up also
[14:24:58] <skunkworks> heh
[14:46:57] <stustev> pcw_home: yes LinuxCNC does 5 axis tool length compensation - the Cinci at MPM has used it for years.
[14:49:25] <pcw_mesa> Thanks, now need to figure out if motion could provide compensated machine waypoints in addition to uncompensated (tool tip) waypoints
[14:52:01] <cradek> stustev: sorry I did not answer your email. I've been away for weeks.
[14:54:55] <cradek> is the path resulting from your algorithm still representable as canonical lines and arcs/helixes in xyzabcuvw space? the current cutter comp algorithm just generates the compensated path in place of the original one, an arc is still an arc, a line is still a line, some additional moves for going around corners are added in between. if your algorithm is like this, you can put it in the same place, in addition. if not, I'm not sure what ...
[14:55:01] <cradek> ... to recommend.
[14:56:21] <stustev> cradek: I thought so. Not a problem. I have 5 axis cutter diameter working in NCL. Looks very good. At this time the application is pointed at machines like the Cinci at MPM. It is not universal.
[14:56:50] <stustev> There are not arcs in the 5 axis path for the Cinci. Arc are not handled at all.
[14:56:59] <stustev> arcs
[14:57:18] <cradek> heh that's a good way to simplify the problem
[14:57:47] <stustev> I am ready to attempt the insertion into LinuxCNC. :)
[14:57:49] <cradek> in this case you would probably want to just generate an error and abort the program if you encounter an unhandled type of move (arc) while comp is on
[14:58:05] <cradek> many things are disallowed like this when comp is on
[14:58:18] <stustev> yes - abort would be the response
[14:59:27] <stustev> When the machine is running with the tool axis at O,O,1 then regular cutter comp would apply.
[15:00:59] <pcw_mesa> In my case just wanted to get surface velocity/ curve position at tool (laser) tip for data painting
[15:01:16] <stustev> Since 5 axis motion on the Cinci type machines is purely point to point then arcs 'should' not be a problem.
[15:01:18] <cradek> am I right that in the general case it's tricky to find the tool vector from a,b,c because of the different rotary configurations?
[15:02:12] <cradek> the interpreter knows which a,b,c axes are defined on the machine but it doesn't know how they are stacked
[15:03:23] <stustev> The finding of a vector from the rotary commands is surprisingly easy. To generalize it would need more information (in possibly the .ini file?).
[15:04:00] <cradek> yeah the ini could say what your rotary stacking is. I think AXIS has this or something similar already to make the preview right...
[15:05:34] <stustev> it gets very simple when you get into it and are able to see it on the screen
[15:05:36] <cradek> I only barely know how that works
[21:33:37] <jepler> hmmm I can't remember my forum login and it doesn't seem to be written down. what an idiot.
[21:35:52] <jepler> aha
[21:36:25] <jepler> .. installing security updates, hope nothing breaks ..
[21:36:42] <cradek> you're dedicated
[21:36:47] <cradek> we appreciate you
[21:38:19] <jepler> you're welcome
[21:38:25] <jepler> I'll do it as long as I can stand it
[21:39:24] <jepler> > CB 2.1.1 further strengthens the legendary security of Community Builder
[21:39:27] <jepler> such confident. so reassure.
[21:39:51] <jepler> > not vulnerable with default or any reasonably possible settings.
[21:49:44] <jepler> ugh an openssl update too !?
[21:50:10] <jepler> .. rebooting because ubuntu says I should ..
[21:52:13] <skunksleep> jepler: I didn't think you really could stand it now... ;)
[21:52:25] <skunksleep> And thank you also
[21:54:50] <Tom_L> can i directly set work offset values (G54) ie the 5221-5230 locations? ... i forgot how
[21:55:15] <Tom_L> other than the axis menu
[21:55:26] <jepler> Tom_L: G10 family
[21:55:33] <Tom_L> ok thanks
[21:55:38] <jepler> .. afk