#emc-devel | Logs for 2010-08-11

[00:11:16] <cradek> oooh, it looks really promising
[00:13:13] <jepler> cradek: fake tach>?
[00:13:17] <cradek> yes
[00:13:27] <cradek> it moves smoothly - needs tuning
[00:13:44] <cradek> hope the tach input on the amp adjusts far enough - this voltage is a lot lower
[00:21:48] <skunkworks> cool
[00:43:50] <Jymmm> Jymmm is now known as Red70sShow
[00:45:04] <Red70sShow> Red70sShow is now known as Jymmm
[00:51:53] <skunkworks> wow - I think I have a shift component that I can test..
[00:52:12] <skunkworks> rough around the edges - but a good starting point.
[00:53:52] <skunkworks> cradek: are they 60v tachs?
[00:54:49] <Jymmm> skunkworks: woohoo!
[00:54:49] <skunkworks> heh - that probably doesn't make sense.
[00:55:04] <skunkworks> are they 60v at top rpm?
[00:55:42] <skunkworks> Jymmm: thanks :) I think with the right hal component - we could take over the world!
[00:56:25] <Jymmm> skunkworks: Jsut keep saying "I think it can! I think it can!"
[00:57:05] <cradek> hmm, it works just as well with the tach unhooked
[00:57:13] <cradek> hooking it up stomps on the dac output
[00:58:41] <cradek> I think it's driving 6kohms so it seems like it should be ok
[01:00:09] <cradek> I'm going to try hooking up 5v directly to the 7i48
[01:13:39] <cradek> hmm no....
[01:17:42] <cradek> the measured load is 6k and it completely flattens the dac output
[01:17:42] <cradek> something is not right here
[01:19:58] <cradek> oh - I bet it's wired backward - duh
[01:19:58] <cradek> grounded out
[01:21:06] <skunkworks> http://www.youtube.com/watch?v=o70K3mvduqw
[01:21:46] <skunkworks> (running with a 12v charger ;))
[01:23:12] <cradek> haha I'm an idiot
[01:23:22] <cradek> how long did that take me to figure out!?
[01:23:41] <skunkworks> heh - try not to think about it to hard..
[01:25:41] <cradek> no, that was my problem in the first place.
[01:26:33] <cradek> hey now the knobs work!
[01:27:16] <skunkworks> heh
[01:32:51] <cradek> holy crap it works
[01:33:01] <cradek> it sounds smoother than it ever has
[01:35:09] <cradek> wheee
[01:35:11] <skunkworks> cool beans :0
[01:35:24] <skunkworks> so - are you going to convert all the other axis?
[01:36:00] <cradek> I'm now wondering if X would sound better if I did
[01:38:15] <cradek> little tiny blips of .0005 ferror at the beginning and end of accel, on 450ipm rapids
[01:40:21] <jepler> sounds nice
[01:44:14] <cradek> the real thanks goes to seb_kuzminsky for all the work he did on the high quality encoder velocity
[01:44:45] <jepler> and of course hostmot2 for having the timestamps that make it possible
[01:45:01] <cradek> yep
[02:33:32] <skunkworks> very cool!
[02:35:18] <cradek> skunkworks: you moved your machine!?
[02:35:56] <cradek> haha the world went upside-down at the end there as you got more and more breathless :-)
[02:38:03] <skunkworks> that was dad - I have been sick :)
[02:38:23] <skunkworks> he was running it with a 6/12v charger
[02:38:41] <cradek> cool
[02:38:57] <cradek> so not quite closed loop control yet :-)
[02:39:09] <cradek> do you have the amps that you'll use?
[02:44:59] <skunkworks> yes
[02:45:06] <skunkworks> 40amp amc
[02:47:21] <cradek> closed loop soon then...
[04:04:35] <CIA-5> EMC: 03cmorley 07v2.4_branch * r526292394aa6 10/src/emc/usr_intf/pncconf/pncconf.py: change XML entry from stepconf to pncconf
[04:04:36] <CIA-5> EMC: 03cmorley 07v2.4_branch * r920af707dc64 10/src/emc/usr_intf/pncconf/pncconf.py: Fix - spaces are not allowed in machine names
[04:04:37] <CIA-5> EMC: 03cmorley 07v2.4_branch * r642dd1d57d8e 10/src/emc/usr_intf/pncconf/pncconf.py: fix mesa pinnames when using a different locale
[04:04:50] <CIA-5> EMC: 03cmorley 07v2.4_branch * rc2463dc28e20 10/src/emc/usr_intf/pncconf/pncconf.py: ready pncconf for parport and spindle helpfiles
[12:59:35] <SteveStallings> SteveStallings is now known as steves_logging
[12:59:57] <moopy> i am having a stupid newbie problem, i am running the livecd installed and upgraded to compile emc2.2.5.0pre. I run an emc config .ini file and if fails with error "waiting for s.axes"
[13:01:15] <moopy> the config runs fine in the default emc2.3.0 version, anyone know what is happening?
[13:02:44] <cradek> changes are required for a 2.3 configuration to work with 2.4 and later. check the UPDATING page on the wiki.
[13:28:25] <micges_work1> moopy: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.4
[13:51:25] <moopy> OKay i got the updating thing sorted i think, now i have a problem with simulator loading???
[13:52:41] <moopy> emc says 'Waiting for ./test to become ready.....'
[13:53:01] <moopy> i have a 'loadusr -W ./test' in my hal
[13:53:31] <moopy> the test file is a python script similar to the pumagui one
[13:54:38] <jepler> You probably need to specify the component name in the loadusr command
[13:54:47] <jepler> if it is different than "./test"
[13:55:13] <jepler> example: loadusr -Wn mycomponent ./test
[13:57:06] <moopy> i think its a path problem
[13:58:20] <moopy> yes, it is path, if i copy test to the emc bin dir then it loads
[13:58:34] <moopy> now i got more errors
[13:58:35] <moopy> hehehe
[14:07:45] <CIA-5> EMC: 03jepler 07master * r144bb6f33b47 10/configs/smithy/ (4 files):
[14:07:46] <CIA-5> EMC: Adds the 622keyence configuration for the machine at Champion Spark Plug. An
[14:07:46] <CIA-5> EMC: external digital depth sensor (appearing as an encoder input) is read via M66
[14:07:46] <CIA-5> EMC: and various pyvcp widgets are used to aquire user input and display process
[14:07:46] <CIA-5> EMC: variables using M68.
[14:07:46] <CIA-5> EMC: 03jepler 07master * r7f34a3cc5012 10/src/hal/drivers/serport.comp: serport: improve error on failure to register
[14:07:49] <CIA-5> EMC: 03jepler 07master * r44780f562246 10/configs/smithy/ (12 files):
[14:07:50] <CIA-5> EMC: Improves system behavior by pausing the program when a guard is opened during a
[14:07:50] <CIA-5> EMC: program. When the door is closed the spindle will restart, but the operator will
[14:07:50] <CIA-5> EMC: have to manually resume axis motion. This gives the spindle time to reach the
[14:07:50] <CIA-5> EMC: programmed speed and improves operator safety.
[14:07:51] <CIA-5> EMC: 03jepler 07master * r9a21aa6fa588 10/src/emc/task/emctaskmain.cc: Fix RFL when starting between a sub definition and its call
[14:07:51] <CIA-5> EMC: 03jepler 07master * r66893f0e9cf9 10/docs/src/hal/ (3 files in 2 dirs): add second reader to graphic
[14:07:54] <CIA-5> EMC: 03jepler 07master * r046d44714200 10/ (VERSION debian/changelog): release version 2.4.3
[14:07:55] <CIA-5> EMC: 03jepler 07master * rc3c8d9a74249 10/docs/src/hal/comp.lyx: replace the reference to bool with int
[14:07:56] <CIA-5> EMC: 03jepler 07master * r2728cb469aad 10/src/Makefile: install emctool.h
[14:07:58] <CIA-5> EMC: 03jepler 07master * re1ee4f0a1f1d 10/src/configure.in: configure: detect missing library
[14:07:59] <CIA-5> EMC: 03jepler 07master * r6b46f8ce9fc4 10/src/configure.in: configure: fix truncated warning
[14:08:00] <CIA-5> EMC: 03jepler 07master * r1bd62994ff1f 10/docs/src/hal/comp.lyx: markup fixes
[14:08:02] <CIA-5> EMC: 03jepler 07master * re4612d45f220 10/debian/rules.in: Don't compress sample configurations
[14:08:03] <CIA-5> EMC: 03jepler 07master * r1cf6dc1cad70 10/src/emc/motion/usrmotintf.cc: when sending command to motion fail, print reason in terminal
[14:08:07] <CIA-5> EMC: 03jepler 07master * r06683ff84199 10/src/emc/usr_intf/axis/scripts/axis.py: Improve an error message by making it more vague
[14:08:08] <CIA-5> EMC: 03jepler 07master * r433974fabe5d 10/src/emc/usr_intf/axis/scripts/axis.py: print 's.axes' warning fewer times
[14:08:10] <CIA-5> EMC: 03jepler 07master * rc2463dc28e20 10/src/emc/usr_intf/pncconf/pncconf.py: ready pncconf for parport and spindle helpfiles
[14:08:13] <CIA-5> EMC: 03jepler 07master * r526292394aa6 10/src/emc/usr_intf/pncconf/pncconf.py: change XML entry from stepconf to pncconf
[14:08:14] <CIA-5> EMC: 03jepler 07master * r920af707dc64 10/src/emc/usr_intf/pncconf/pncconf.py: Fix - spaces are not allowed in machine names
[14:08:16] <CIA-5> EMC: 03jepler 07master * rbf0cccc8078c 10/docs/man/man9/.gitignore: add bldc_hall3.9 to .gitignore
[14:08:21] <CIA-5> EMC: 03jepler 07master * rbec24629de85 10/ (27 files in 9 dirs): Merge remote branch 'origin/v2.4_branch'
[14:08:37] <CIA-5> EMC: 03jepler 07master * r51aeaaaf830e 10/src/emc/usr_intf/axis/scripts/axis.py: axis: gray out 'run from here' on right click menu when emc is not enabled
[14:08:47] <CIA-5> EMC: 03jepler 07master * r642dd1d57d8e 10/src/emc/usr_intf/pncconf/pncconf.py: fix mesa pinnames when using a different locale
[14:09:11] <skunkworks> wow
[14:10:03] <alex_joni> merging 2.4 into master
[14:10:06] <jepler> yeah, it's not as much as it looks like
[14:12:14] <moopy> my configuration system has created its first working conf
[14:12:58] <moopy> but would be nice if i knew how to specify the path of a component in 'loadusr'???
[14:13:57] <jepler> Just like you did: ./test
[14:14:27] <jepler> but when the *component name* is different than the first argument (e.g., "test" vs "./test") then you have to specify the component name
[14:14:31] <jepler> loadusr -Wn test ./test
[14:15:22] <jepler> and for the removal of doubt, remember that "the current directory" is "the directory containing the inifile"
[14:15:27] <moopy> ooh, right i missed that
[14:15:57] <moopy> i will remove the test file from the emc bin dir just to check
[14:16:03] <jepler> but when the command is not found at all there's a different error
[14:16:04] <jepler> $ halcmd loadusr -W this-command-does-not-exist
[14:16:04] <jepler> <commandline>:0: execv(this-command-does-not-exist): No such file or directory
[14:16:07] <jepler> <commandline>:0: this-command-does-not-exist exited without becoming ready
[14:16:58] <moopy> oooooopps, i got it test ./test
[14:17:33] <moopy> as they dont match with the extra path chars
[14:17:47] <moopy> yeah it all works now
[14:17:52] <jepler> right, and yay
[14:19:36] <jepler> there might be a better heuristic for the component name, such as "cut off everything up to and including the last slash" and "cut off everything from the last dot to the end" (so loadusr -W ./test.py would wait for 'test' by default)
[14:19:51] <jepler> I'd be happy to review a patch to halcmd that makes it work that way
[14:20:16] <jepler> probably for master only, due to the small risk of breaking an existing configuration which actually names a component "./test.py" for this very reason..
[14:20:57] <moopy> unfortunatly im working on something else, which maybe ready in a couple of weeks
[14:21:36] <moopy> im too much of a newbie to start hacking halcmd
[15:39:26] <moopy> anyone know about linking homing to pyvcp joints?
[15:44:13] <moopy> seems if i move joint past 180 degrees then home the axis, then jog in reverse the pyvc dial does not move anymore??
[21:27:11] <moopy> are there any docs on how the 3d drawing functions work and how the puma560gui works?
[21:36:32] <alex_joni> moopy: what kind of docs?
[21:38:55] <alex_joni> do you mean vismach?
[21:40:02] <voxadam> I've been looking at EMC2 for awhile now. I've been focused on planning for the conversion of an old VMC to EMC2 from the current archaic that's on it. While reading up on the EMC's current state I've become quite amazed with HAL. Has any thought ever been given to moving HAL to its own git tree? It seems that HAL could potentially see much wider use that just EMC if it became a project unto itself.
[21:45:21] <alex_joni> there was an initial attempt to do that
[21:45:31] <alex_joni> but so far all HAL developers are also emc2 developers
[21:45:32] <cradek> voxadam: I think it's always been part of the design that they're useful separately - but no I don't think anyone's currently serious about splitting them
[21:45:50] <alex_joni> and separating the two just for the fun of it, ain't gonna happen
[21:45:58] <cradek> actually even now, they're quite separate aside from being in the same repository
[21:46:00] <jepler> I think it'd be more useful to talk about changing emc2 so that it can be configured for hal only, or splitting packages so that hal is in a base package that emc depends on
[21:46:08] <alex_joni> especially as that will mean quite a lot of work for packaging and making them work together
[21:46:35] <jepler> as a release manager of emc2 I don't want to complicate things by having another git repository to make releases from
[21:46:54] <voxadam> I can understand that.
[21:47:21] <cradek> I've considered/designed a hal-only (or hal-and-ladder-only) project and always figured I'd install emc but then just not use all of it
[21:47:34] <jepler> but by all means use hal in your own project without emc. I think there are some stuff it would be great for
[21:47:39] <moopy> i need doc to workout how the HalRotate and HalTranslate commands work
[21:47:42] <cradek> realtime start / halcmd whatever
[21:48:02] <voxadam> I've been thinking about using HAL and CL to automate some things that I've been working on in my head.
[21:48:11] <cradek> moopy: there is no doc, except for the many examples
[21:48:22] <alex_joni> HalTranslate is obvious
[21:48:38] <alex_joni> just translate the object with some X.Y.Z offsets
[21:48:47] <jepler> HalTranslate works by getting a value from hal and then performing an OpenGL operation that depends on that value
[21:49:03] <alex_joni> same for HalRotate although the angles might not be obvious if you don't know 3D ops
[21:49:13] <jepler> glPushMatrix and glTranslatef are Python wrappers of standard OpenGL API calls, you'll find them documented online
[21:49:58] <moopy> HalTranslate([tooltip], c, "tool_length", 0,0,-1)
[21:50:12] <moopy> what do the arg represent?
[21:50:51] <moopy> move tooltip -1 in z direction?
[21:50:59] <jepler> c is the hal component
[21:51:06] <jepler> "tool_length" is the name of the pin
[21:51:16] <jepler> 0,0,-1 is the direction of the translation
[21:51:35] <jepler> [tooltip] is the list of items to translate (a single item in this case)
[21:52:07] <moopy> that should be of great help... thanks jepler
[21:52:09] <jepler> so this says: when drawing this HalTranslate object, first translate by tool_length * <0,0,-1>, then draw tooltip, then undo the translation
[21:53:04] <moopy> what.... undo it?
[21:53:53] <jepler> in opengl a transformation stays in effect until it's undone
[21:54:09] <jepler> but objects like HalTranslate that work by applying OpenGL transformations always undo them after drawing the contained objects
[21:55:10] <moopy> so if i make another translation its relative to the original coords?
[21:56:24] <moopy> i think i need to go away work things out and comeback with some more questions
[21:56:29] <cradek> moopy: you should play with it! maybe the best way is to modify some of the existing examples and see what they do.
[21:56:38] <moopy> thanks for the help
[21:57:35] <jepler> http://glprogramming.com/red/chapter03.html#name8
[21:57:48] <jepler> vismach uses opengl so its transformations work like this ^^
[22:01:12] <moopy> just one more question actually..... HalRotate([tool],c,"joint4",1,0,-1,0)
[22:01:50] <jepler> rotate around the axis <0,-1,0> (i.e., the negative Y axis) by 1*joint4 degrees
[22:01:58] <moopy> what are the last 4 arg in hal rotate.... a x,y,z coord + angle?
[22:02:02] <jepler> jkjclass HalRotate(Collection):
[22:02:02] <jepler> def __init__(self, parts, comp, var, th, x, y, z):
[22:02:14] <jepler> I didn't remember that; I looked at the __init__ and saw that they're th (theta), x, y, and z
[22:02:36] <jepler> glRotatef(th * self.comp[self.var], x, y, z)
[22:04:01] <moopy> okay i think i got plenty to be getting on with...... i now need a copy and paste and a few hours to understand things
[22:04:16] <moopy> thanks again for the help
[22:24:10] <jepler> vismach "sun and planet" demo: http://emergent.unpy.net/files/sandbox/solar.hal http://emergent.unpy.net/files/sandbox/solar.py
[22:24:55] <jepler> 'halrun solar.hal'
[22:33:24] <alex_joni> jepler: got a png?
[22:36:47] <jepler> http://emergent.unpy.net/files/sandbox/solar.png
[22:36:55] <jepler> the earth is square so that you can tell it's rotating
[22:37:12] <jepler> the blue spot is where the backtrace comes out of
[22:37:29] <alex_joni> cool
[22:37:42] <alex_joni> the moon is missing :)
[22:37:51] <jepler> backplot, not backtrace
[22:37:59] <jepler> I knew that wasn't the right word
[22:38:07] <alex_joni> gdb hovering around
[23:03:23] <jepler> bbl