#emc-devel | Logs for 2009-11-11

[00:26:12] <cradek> machine is joint space, gcode is world/axis space
[00:26:25] <cradek> it makes no sense to read joint limits into gcode
[00:26:48] <cradek> what is the problem you're trying to solve that led you to think about adding this capability?
[03:58:35] <dgarr> cradek: for no tool, i considered making all items 0 -- is that better?
[03:59:32] <cradek> yes I think they should always read as zero. I am not sure but maybe writing to them should be an error
[03:59:52] <cradek> the write has no effect, so I think the user should know that he's doing something unexpected
[04:01:07] <cradek> my tool table is in inches. I programmed G21 and then #5403=1. I get 1 inch, not 0.03937 inch in the tool table. I think this is right but I was a little surprised.
[04:01:28] <cradek> G21 G10 L1 P1 Z1 gives me 0.03937 in the tool table
[04:02:35] <cradek> G21 G10 L2 ... reads mm values for offsets but #5223=1 writes directly always in native units. you have done the same thing for tools which I think is right.
[04:05:33] <cradek> hm there is some incorrect interaction here
[04:06:08] <cradek> T1 M6 => Tool 1, Offset 1, diameter 0.125
[04:06:24] <cradek> G10 L1 P1 Z2 => Tool 1, Offset 2, diameter 0.125
[04:06:45] <cradek> #5410=0.1 => Tool 1, Offset 1, diameter 0.1
[04:06:56] <cradek> expected it to remain at Offset 2
[04:07:23] <cradek> G10 L1 P1 Z3 => Tool 1, Offset 3, diameter 0.125
[04:07:27] <cradek> expected it to remain at diameter 0.1
[04:20:06] <cradek> dgarr: ignore my rambling above "some incorrect interaction"
[04:28:50] <dgarr> the values different from the expectations seem like real issues? so now i'm not sure what you are indicating should be ignored?
[04:29:34] <cradek> yes the values different from expected are bugs IMO
[04:29:42] <cradek> not sure what you are asking about ignored?
[04:32:55] <dgarr> oh, now i see the "above" -- are the only issues you are remarking on are the values different from expectations after g10?
[04:33:09] <cradek> yes
[04:33:21] <cradek> ignore my talk about units because I think it works right
[04:33:46] <dgarr> ok, i'll study that
[04:33:54] <cradek> the thing I think is wrong is how #5410=0.1 changed both offset and diameter in that situation
[04:34:18] <cradek> also wrong that G10 L1 P1 Z3 changed both offset and diameter in that situation
[04:34:39] <cradek> the two methods are fighting somehow
[17:35:05] <CIA-40> EMC: 03cradek 07master * r2b2ac14fdd9a 10/src/emc/usr_intf/touchy/ (emc_interface.py touchy.glade touchy.py): fill out dro with abcuvw
[17:35:08] <CIA-40> EMC: 03cradek 07master * ra87ab2420b12 10/src/emc/usr_intf/touchy/ (emc_interface.py touchy.glade touchy.py): support various configurations of axes in touchy
[17:39:03] <cradek> ugh, there's so much more to that
[17:39:17] <cradek> maybe someone else will finish it (ha)
[17:42:17] <Jymmm> cradek: "Step towards the light" ;)
[17:42:58] <alex_joni> cradek: and some of it needs redoing for ja4
[17:43:35] <cradek> yeah it doesn't do kins
[17:44:17] <cradek> when (if) we fix jogging/teleop all the guis need a rework anyway
[18:04:11] <Jymmm> cradek: I just found the PERFECT clock for you... http://shop.tuxgraphics.org/electronic/detail_scope_clock.html
[18:08:14] <cradek> nah, I made my own crt clock
[18:08:37] <Jymmm> cradek: It's a o'scope clock
[18:09:03] <cradek> yeah, that's the easy way out
[18:11:40] <Jymmm> lol
[20:33:25] <christel> [Global Notice] Hi all, due to the memleak issues we are experiencing with hyperion (our ircd) we will be taking a handful of servers out of rotation for the production network in order to utilise them to better test our new ircd; ircd-seven. Affected users will be notified seperately. Apologies for the inconvenience.
[20:47:50] <CIA-40> EMC: 03cradek 07master * r2c75d7295464 10/src/emc/usr_intf/touchy/ (hal_interface.py touchy.glade touchy.py): Add jogwheel buttons and hal interface for the other axes
[21:40:35] <jt-plasma> in dev is there any hal pin that would have the current programmed feed rate?
[21:42:25] <cradek> 32770 float OUT 0 motion.current-vel
[21:42:42] <cradek> that is the current ACTUAL feed rate
[21:42:48] <cradek> current programmed, no
[21:43:37] <jt-plasma> I'm needing to see if I'm within a percentage of programmed feed that that would be 1/2 of what I need
[21:44:35] <cradek> not sure how you would get that - F word is only a request - lots of things can slow you down
[21:45:14] <jt-plasma> yes, and I only want make Z corrections if I'm close to the request
[21:45:28] <jt-plasma> for the THC
[21:46:11] <cradek> would you want feed override included in the requested value?
[21:46:26] <cradek> and adaptive feed?
[21:46:36] <jt-plasma> no
[21:46:52] <cradek> *that* number you could get with a few changes to motion - F word, no
[21:48:42] <jt-plasma> I could get feed override from halui if needed :)
[21:54:45] <alex_joni> jt-plasma: I think you want what cradek suggests
[21:54:53] <skunkworks> for now - couldn't you add a something to halui that you could manually set a number that is the feedrate in the program?
[21:54:56] <alex_joni> the requested value wtih all overrides applied
[21:55:14] <alex_joni> skunkworks: the feed in the program is pretty irrelevant
[21:55:20] <alex_joni> say I program F9999999999999
[21:55:28] <alex_joni> my machine only does 10 on X, and 20 on Y
[21:55:33] <jt-plasma> alex_joni: I don't think I understand what cradek is suggesting :/
[21:55:41] <alex_joni> I will be never close to the programmed F word
[21:55:47] <alex_joni> jt-plasma: let me explain
[21:56:01] <alex_joni> you have the F-word in the g-code (the F999999999999 above)
[21:56:02] <jt-plasma> ok
[21:56:21] <alex_joni> that gets scaled based on what the machine could possibly do (XYZ vels and TRAJ vel)
[21:56:29] <alex_joni> that new scaled velocity is sent to motion
[21:56:47] <alex_joni> it's still a requested feedrate, not sure the machine is really doing it
[21:57:24] <alex_joni> that requested feedrate then gets scaled with feed overrides (AF - which is probably 1 for you, and FO - which you can change)
[21:57:53] <alex_joni> a new requested feedrate is computed, and this one then gets used by the TP to generate the periodic waypoints
[21:58:17] <jt-plasma> ok, but then that is a programming error on the cam side and in my case there would not be any correction and the cut would be bad
[21:58:22] <alex_joni> (now.. based on blending and whatnot, the real velocity will be similar to the requested, but not always equal)
[21:58:35] <jt-plasma> because the speed is too fast to cut
[21:59:12] <alex_joni> jt-plasma: maybe it's best if you describe what you want to do
[21:59:25] <jt-plasma> right, and if it is close enough I want to apply Z corrections as needed to maintain tip voltage
[22:00:04] <jt-plasma> I need a couple of things for the THC comp, the requested feed rate and the requested tip voltage.
[22:00:33] <alex_joni> I think requested feed rate is the clamped velocity as explained above
[22:00:46] <jt-plasma> if the actual feed rate is close enough then apply Z offsets to get the tip voltage within the tolerance window
[22:00:49] <alex_joni> if you programmed carefully then the clamped velocity == F-word in the g-code
[22:01:03] <alex_joni> carefully meaning under the machine constraints
[22:01:08] <jt-plasma> yes
[22:01:21] <alex_joni> go with cradek's suggestion then
[22:01:34] <jt-plasma> which is?
[22:01:51] <alex_joni> 23:44 < cradek> *that* number you could get with a few changes to motion - F word, no
[22:02:06] <alex_joni> jt-plasma: just beg nicely for cradek to add it ;)
[22:02:19] <jt-plasma> * jt-plasma understands now :)
[22:02:40] <jt-plasma> cradek: pretty please add that when you have some spare time :)
[22:03:47] <jt-plasma> I went looking to see how "I" could change it but got lost in the forest because someone picked up my bread crumbs
[22:04:19] <jt-plasma> alex_joni: thanks for clearing that up for me
[22:04:36] <alex_joni> I think someone ate them
[22:04:49] <jt-plasma> LOL
[22:06:22] <jt-plasma> brb, I need to go and check on a sick neighbor who is by himself to see if he needs anything
[22:06:36] <jt-plasma> before it gets dark...
[22:09:32] <CIA-40> EMC: 03cradek 07master * r19be74094499 10/src/emc/ (5 files in 2 dirs): add motion.requested-vel output for jt's testing
[22:10:45] <skunkworks> heh
[22:10:57] <skunkworks> cradek: you are awesome
[22:15:47] <alex_joni> cradek++
[23:31:47] <jt-plasma> cradek: thanks
[23:50:30] <cradek> :-)
[23:50:39] <cradek> don't thank me until you check and see if it works...