#emc | Logs for 2007-10-04

[00:05:53] <Ziegler> jlmjvm: worked just fine...
[00:05:58] <Ziegler> now to invert my spindle
[00:06:10] <jlmjvm> Ziegler:great
[00:06:26] <Ziegler> newsig spindle-on bit
[00:06:28] <Ziegler> linkps motion.spindle-on => spindle-on
[00:06:28] <Ziegler> linksp spindle-on => parport.0.pin-01-out
[00:06:42] <Ziegler> I tried parport.0.pin-01-out-not
[00:06:44] <Ziegler> and got errors
[00:07:27] <jepler> for outputs, you "setp parport.0.pin-01-out-invert 1" to invert it
[00:07:34] <jepler> (check the exact pin name using 'halcmd show pin parport')
[00:07:38] <Ziegler> ahh... forget to add the number 1
[00:07:47] <jlmjvm> http://pastebin.ca/724956
[00:07:53] <jlmjvm> try those
[00:07:58] <Ziegler> looking
[00:08:05] <jlmjvm> k
[00:09:25] <Ziegler> I dont actually have the ability to control direction at this time... right now I am looking to just turn it off and on... and it does the inverse of what I want. Ill give jeplers hint a try and see how it works
[00:09:54] <Ziegler> jepler: does it matter if it goes before or after the pin assignment?
[00:11:02] <Ziegler> be back in a bit
[00:29:54] <jlmjvm> cradek:i never could get that brake signal inverted,just gonna run it thru a relay
[00:31:31] <jlmjvm> prolly be better anyway,the ssr will operate a relay instead of a 110v air solenoid
[00:32:22] <SWPadnos> jlmjvm, the parameter uses a . instead of a - before the word invert. that should have been easy to spot using halcmd or halshow
[00:33:15] <SWPadnos> so the parameter is ppmc.0.dout.02.out.invert
[00:33:36] <Ziegler> worked
[00:34:28] <Ziegler> thanks guys... another question...
[00:34:38] <Ziegler> I was attempting to run many of the example programs...
[00:34:58] <jlmjvm> is that all that was stopping that
[00:35:17] <SWPadnos> yes
[00:35:39] <Ziegler> I keep getting.. "program will run outside of machine boundaries" but my ini file I have set way beyond those limits. Sometimes the program will run... sometimes not.
[00:35:49] <SWPadnos> tha's why cradek suggested that you comment that line, run EMC, and look at the HAL using one of the available tools :)
[00:37:17] <jlmjvm> thats what i did,but it had been a heck of a day already and i musta looked over it
[00:37:52] <SWPadnos> easy to do, and those things should be standardized anyway
[00:37:55] <SWPadnos> so it's nobody's fault :)
[00:39:32] <jlmjvm> im gonna try that tomorrow and if it works im gonna email it to Jon as an example for the files on his website
[00:40:13] <jlmjvm> it could be in the hal file commented out,but there if a guy needs it
[00:42:16] <SWPadnos> makes sense. also, maybe remind me (or Jon) that the names should be changed to match all the other drivers (pin.nn-invert)
[00:42:30] <jlmjvm> the bad thing was i had to get the it dept to make me a long cat5 cable,and then convince them that the irc room was safe to access
[00:42:41] <fenn> then you get to change it back next time :)
[00:43:08] <jlmjvm> took several hours just to get internet access
[00:43:20] <fenn> dont you know irc is full of hackers and teenage boys
[00:43:26] <fenn> and other undesirables
[00:43:27] <jlmjvm> but its all downhill now
[00:43:40] <jlmjvm> only irc ive ever been in
[00:46:50] <jlmjvm> SWPadnos:when you say halshow do you mean in terminal,or start emc and go to show hal configuration?
[00:47:10] <fenn> in a terminal type halcmd -fk
[00:47:16] <SWPadnos> no
[00:47:35] <SWPadnos> that's OK, but I did mean "show HAL configuration" when I said "halshow"
[00:47:45] <jlmjvm> k
[00:52:45] <Ziegler> I keep getting.. "program will run outside of machine boundaries" but my ini file I have set way beyond those limits. Sometimes the program will run... sometimes not.
[00:52:56] <Ziegler> Any thoughts or suggestions?
[00:53:29] <SWPadnos> none here, sorry. I'm not so knowledgeable about the UIs and actual use of EMC :)
[00:54:27] <jlmjvm> so then when you find what your looking for,say e-stop for example,check the syntax in halshow and compare it against what you added for typo,s
[00:55:22] <fenn> hal components are supposed to have man pages too
[01:00:38] <Guest987> test
[01:00:55] <Ziegler> yeah we read you
[01:00:58] <Ziegler> oops
[01:20:42] <Roguish> SWPadnos: axis checks the limits of motion in the .ini file against the loaded gcode.
[01:21:14] <Roguish> if the gcode exceeds the machine boundaries axis complains.
[01:21:44] <Roguish> been there before.
[01:22:08] <Ziegler> Roguish: my ini file I put at -5 to 12
[01:22:38] <Ziegler> nc file limits in axis... shows limits between -2.4 and 5
[01:23:08] <Roguish> Ziegler: check all the defined axes.
[01:23:38] <Ziegler> it complains about the specific axis... but all of them give plenty of room
[01:23:50] <Ziegler> without changing anything
[01:24:00] <Ziegler> besidedes reloading emc
[01:24:07] <Ziegler> and reloading the nc file
[01:24:10] <Ziegler> sometimes it will work
[01:24:47] <Ziegler> no limit switches
[01:25:10] <Roguish> that's all i know about the limits. sorry.
[01:25:42] <Roguish> try another gui?
[01:25:56] <Roguish> tkemc? mini? just for test.
[01:26:54] <Ziegler> I will
[01:29:07] <Ziegler> knowing my track record the past couple days... im probably missing something obvious
[01:53:42] <fenn> make sure you're loading the right .ini file
[01:56:11] <Ziegler> yeah I am
[01:56:14] <Ziegler> go check tho
[01:56:18] <Ziegler> good*
[01:57:29] <fenn> does the program say g20/g21 at the top?
[01:57:59] <fenn> and does it do any offset stuff
[01:58:27] <fenn> or tool length compensation
[02:01:21] <Ziegler> they are the example programs that come with the install
[02:01:23] <Ziegler> like chips
[02:01:41] <Ziegler> looking
[02:08:38] <cradek> axis/emc2.1 can be slightly wrong about exceeding limits if using tool length comp, but is right otherwise
[02:09:07] <cradek> emc2.2 does not have this problem
[02:10:45] <Ziegler> 2.2 still cvs?
[02:10:54] <cradek> yes
[02:11:04] <fenn> how is it slightly wrong?
[02:11:05] <Ziegler> 2.2 has the speed increase?
[02:11:07] <cradek> but unless you are using large tool len offsets, this is not your problem
[02:12:07] <cradek> fenn: a tool len offset can move the controlled point outside of the limits, and axis/emc2.1 does not compensate
[02:13:00] <cradek> for example if the tool is short, you can maybe get to Z+4 in machine coords, but if it's long, you can only get to Z+3
[02:13:47] <cradek> I'm going to go make some Al chips before it gets any later.
[02:33:05] <tomp> fenn: mazak png looks good
[02:43:32] <CIA-8> 03jepler 07TRUNK * 10emc2/docs/src/hal/drivers.lyx: trivial typo
[02:49:00] <fenn> tomp: right after i drew that someone asked on my LUG mailing list if someone could give a presentation on inkscape (what a coincidence)
[02:49:15] <fenn> so i said yeah sure i guess
[02:49:32] <Jymmm> presentation on inkscape?!
[02:49:36] <tomp> what do you want the presentation to do?
[02:50:09] <tomp> of to demonstarte what inkscape can do?
[02:50:14] <fenn> i'm just going to show how you work with inkscape.. all the little tricks that take a while to figure out (like clones and blur-clipping)
[02:51:12] <tomp> oh to demonstrate inkscape... yes you did a nice job, and i bet you could write the user's guide ;)
[02:51:44] <tomp> could you post the svg somewhere?
[02:51:53] <fenn> its in the same directory, mazak.svg
[02:52:12] <tomp> thanks
[02:52:30] <fenn> i'm going to gpl and submit all the drawings to emc cvs as soon as i get around to it
[02:53:23] <fenn> procrastination is my great motivator, you see
[02:55:04] <tomp> dang, the svg rendering in ffox 1.5 puts a big black rhombus on left side of head. at first it was a pretty believable shadow
[02:55:35] <fenn> yeah i couldnt decide whether to use the shadows or not
[02:55:45] <fenn> it slows everything down
[02:56:00] <fenn> so i just added them at the end. it looks fine without them actually
[02:57:58] <fenn> added mazak-no-shadows.svg
[02:58:02] <tomp> http://imagebin.org/10835
[02:59:33] <fenn> the fanuc arm picture really needs blur-shadows
[02:59:47] <tomp> oh, mazak-no-shadows does load quicker and didnt have that big black area !
[03:00:20] <tomp> and you're showing the intermediate pod position, cool
[03:00:26] <fenn> actually this doesnt look bad at all in firefox http://fenn.dyndns.org/pub/sketches/emc-splash/fanuc.svg
[03:02:27] <tomp> looks real nice, hmm, i want cad rendering in svg
[03:02:44] <fenn> its not an unreasonable idea
[03:03:05] <fenn> i want svg hardware acceleration :)
[03:04:20] <tomp> yeh, and css for svg ... do a 'view source' and see all the repetition, thats why i still do svg by hand and avoid the auto generators
[03:04:58] <fenn> i'm quite annoyed that you can't have a clone on a separate layer in inkscape, otherwise i'd have all the "masters" on their own layer so i can pick-and-place
[03:05:21] <fenn> like on a circuit board, one of those resistors is the master, but which?
[03:05:51] <fenn> a file-based include would be even better
[03:06:17] <tomp> hmmm, in gEDA the master element doesnt have to be visible, just a known template ( like an include )
[03:06:47] <tomp> more like a library reference
[03:07:34] <tomp> but, your drawings are really nice, did you work from photos? use onion skin?
[03:08:03] <fenn> yes a lot of the time i trace a photo on a half-transparent layer
[03:08:14] <tomp> try any svg tracers?
[03:08:40] <fenn> but sometimes you have to construct things geometrically, like the blue spiky wheel thing in the mazak toolchanger
[03:09:29] <fenn> autotrace is built-in to inkscape and is very good, but you end up with one complex blob shape
[03:09:47] <fenn> you can't do stuff like for example rotate the robot arm around
[03:10:14] <fenn> and it usually ends up looking like an auto-traced photo :)
[03:10:23] <tomp> the pod wheel, yes... do you remember the goofy encoder it had? it was holes and prox sensors :)
[03:10:33] <fenn> yeah not-quite binary
[03:11:19] <fenn> your pictures helped a lot in drawing the toolchanger actually
[03:11:29] <fenn> so, thanks for sharing them
[03:11:34] <CIA-8> 03swpadnos 07TRUNK * 10emc2/src/hal/utils/pci_read.c: Utility to read from a PCI Mesa (or other) card
[03:12:03] <CIA-8> 03swpadnos 07TRUNK * 10emc2/src/hal/utils/pci_write.c: Add support for hex and octal numbers on the command line
[03:20:56] <CIA-8> 03swpadnos 07TRUNK * 10emc2/src/hal/utils/pci_write.c: gah! forgot to save without the commented old code
[03:22:09] <CIA-8> 03swpadnos 07TRUNK * 10emc2/src/hal/utils/Submakefile: Of course, it also helps to compile pci_read ...
[03:25:48] <CIA-8> 03swpadnos 07TRUNK * 10emc2/src/Makefile: and it's even more helpful to be able to run the program once it's compiled
[03:26:47] <tomp> haha the claw.svg is great, maybe for a fest you can have a robocrance/claw over a box of emc goodies http://fenn.dyndns.org/pub/sketches/emc-splash/claw.svg
[03:34:09] <fenn> tomp: that was an idea i had for the config-picker
[03:34:19] <tomp> wow, loading motenc.svg actually rebooted me!
[03:34:26] <tomp> still reading logs
[03:34:50] <tomp> what about robocrane fun rides? ( boson's chair :)
[03:35:47] <fenn> heh like that legoland robot that comes 5 inches from slamming your head into a concrete wall
[03:37:59] <tomp> no, hopefully something fun & safe, like one of those doorway baby bouncers
[03:38:25] <tomp> but big enuf to have a little scare
[03:38:37] <tomp> barn door bouncer?
[03:38:54] <fenn> redneck rollercoaster?
[03:40:03] <tomp> yeh, when foreigner come heree to the muidwest, and we pass one of those mega-graineries, i tell 'em thats an agricultural fun ride :)
[03:40:27] <fenn> tomp: http://www.youtube.com/watch?v=RRuskNIgBoA
[03:41:36] <tomp> thats great! call up red green
[03:42:20] <tomp> still no clue about the crash, tho Xlog shows all clients being rejected
[06:11:36] <ds2> okay, is 0.010inch of backlash ever normal on any ball screw?
[06:12:06] <SWPadnos> maybe one that's 37" diameter and 1/2 mile long
[06:12:46] <ds2> heh okay
[06:13:07] <ds2> do ball nuts typically have a backlash adjustment?
[06:13:23] <SWPadnos> mine don't, but there are that do
[06:13:43] <SWPadnos> the design of the ballscrew is supposed to eliminate the need for that though
[06:13:44] <ds2> guess I am not out of luck then
[06:14:15] <ds2> that's what I thought but I put a dial indicator on the cross slide and jogged it back and forth and found 0.010 of play
[06:14:33] <SWPadnos> it could be anything though - the mounts for example
[06:14:50] <SWPadnos> or bad bearings
[06:15:06] <ds2> hmmmm so I can't conclude anything til I did a tear down of the entire cross slide?
[06:15:10] <SWPadnos> or play in the drive system
[06:15:37] <SWPadnos> you can only conclude that between the motors and the table, there's something (or some things) that causes some error
[06:15:45] <ds2> the drive is through a timing belt to a stepper so i wouldn't think that can do anything bad
[06:16:21] <SWPadnos> unless the belt isn't tight, or the pulleys are slipping on the shafts, or ...
[06:16:26] <fenn> ds2: there are ballscrews that don't have any preload out there... but it's probably just crap
[06:16:44] <ds2> fenn: but what would be a typical backlash on those?
[06:16:43] <SWPadnos> if you have a keyway on wither uolley, that could be worn
[06:16:58] <fenn> i dont know, probably less than 0.01"
[06:17:08] <ds2> this is on a CNC trainer that came out of a school...suppose to be very lightly used
[06:17:07] <SWPadnos> argh
[06:17:21] <SWPadnos> that was supposed to say "either pulley"
[06:17:32] <fenn> woother wooley
[06:17:48] <SWPadnos> what would elmew fudd say?
[06:19:06] <fenn> if anyone's into graphics and non-realistic rendering (hand-drawn style) this guy's page is amazing: http://graphics.uni-konstanz.de/publikationen/
[06:19:13] <ds2> what about 0.002 of 'lash? is that likely to be typical? (this is on the Z axis
[06:19:51] <SWPadnos> 0.002 isn't that great for a ballscrew, especially if it's over a short travel
[06:20:00] <SWPadnos> if it's good wnough for your purposes though, it's good enough
[06:20:02] <SWPadnos> enough
[06:20:34] <ds2> i just got it, trying to see what are likely trouble spots I need to investigate before replacing the controls
[06:20:49] <fenn> why replace the control?
[06:21:04] <ds2> it is a 6502 thing that doesn't use G code (uses a series of menus to program)
[06:21:12] <fenn> oh, rip 'er out :)
[06:21:15] <ds2> i want a G code capable control!
[06:21:36] <fenn> yeah g-code's better than nothin..
[06:22:23] <ds2> the drives, vfd, and power supplies are all functional so I am going through the mechanicals... the spindle has less then 0.0005 runout (limit of my indicator)
[06:57:08] <Jymmm> http://www.supersizedmeals.com/food/article.php?story=20060125050438458
[07:02:40] <fenn> autotrace -centerline seems like it might be good for engraving arbitrary shapes
[08:40:26] <Guest487> Howdy.
[08:44:09] <fenn> mornin
[08:45:53] <fenn> " You reached the limit of 0 lookup queries per day.
[08:45:53] <fenn> Due to some automated bots spidering our
[08:45:55] <fenn> database we had to set a limit."
[08:46:22] <fenn> dont throw the baby out with the bath water, sheesh
[08:53:55] <Jymmm> fenn: what are you looking up?
[08:54:21] <Guest487> how are ya'all?
[08:56:00] <Guest487> wow. must be alseep.
[09:00:25] <alex_joni> hi
[09:01:06] <Guest487> yep, they're asleep alright.
[09:01:15] <Guest487> pauses for breaht.
[09:02:16] <Guest487> taps on glass..
[13:33:07] <lerman_> lerman_ is now known as lerman
[14:18:02] <skunkworks> hmm - I might have tinker time this weekend..
[14:18:14] <skunkworks> don't want to jinx it though
[14:18:42] <skunkworks> so I am not getting my hopes up ;)
[16:21:49] <ADucci> hi guys
[16:22:11] <jepler> welcome ADucci
[16:23:09] <ADucci> i am really struggling to connect my Plasma cutter to the "Spindle On" in HAL... any experience here?
[16:23:38] <ADucci> newsig spindle-on bit
[16:23:39] <ADucci> linkps motion.spindle-on => spindle-on
[16:23:39] <ADucci> linksp spindle-on => parport.0.pin-16-out
[16:23:51] <ADucci> thats the ini file entry.
[16:25:00] <jepler> those are lines for a ".hal" file, not for an ini file
[16:25:12] <ADucci> sorry... true.
[16:25:48] <ADucci> i assume hal is the only place i connect the plasma via parallel pin?
[16:25:57] <jepler> yes
[16:26:05] <jepler> does it give an error when you run emc, or does it just not work as expected?
[16:27:10] <ADucci> no, well in my nc code i give an M3 (spindle on ) and M5 (off ) commands, and it does nothing, the same setup works in windows on other software
[16:27:29] <ADucci> ie, the plasma just wont fire.
[16:27:53] <cradek> you may need an S word, since the default spindle speed is 0
[16:27:55] <cradek> so, try S100 M3
[16:29:02] <jepler> yes, I can confirm that without setting a nonzero spindle speed, the pin stays FALSE when you issue M3
[16:29:20] <cradek> thanks for checking my guess
[16:30:02] <jepler> "It is OK to use M3 or M4 if the spindle speed is set to zero. If this is done (or if the speed override switch is enabled and set to zero), the spindle will not start turning. If, later, the spindle speed is set above zero (or the override switch is turned up), the spindle will start turning." -- documentation
[16:30:09] <ADucci> ok, sounds good, that may work, any idea where i change the default spindle speed,
[16:30:28] <jepler> you can simply put S1 as part of your initialization gcode; I don't know of any other way to set the default speed.
[16:31:29] <cradek> I think you would only need an S word in one place at the beginning of your program
[16:31:46] <SWPadnos> you may be able to put it in RS274NGC_STARTUP_CODE in the ini file
[16:31:53] <SWPadnos> (or whatever that's named)
[16:31:54] <jepler> yes, if you S1 at the top, it is saved when you M5
[16:33:02] <ADucci> is this, generally what you need to do to use M3, M4, M5, oh i guess for a spindle you always need to specify a speed
[16:35:40] <cradek> yes on a mill it would be wrong for the control to guess a speed
[16:38:01] <ADucci> "or if the speed override switch is enabled and set to zero" -- see above, does this mean that i can override the speed requirement
[16:43:38] <ADucci> RS274NGC ? how would i know the commands?
[16:44:05] <jepler> ADucci: http://linuxcnc.org/docs/html/gcode.html (quick reference) http://linuxcnc.org/docs/html/gcode/main/ (more thorough reference)
[16:44:49] <jepler> originally based on this document: http://www.isd.mel.nist.gov/personnel/kramer/pubs/RS274NGC_3.web/RS274NGC_3TOC.html
[16:49:55] <ADucci> great thanks, ill give that a read!
[17:49:55] <Paragon37> Afternoon ....
[17:50:38] <anonimasu> noon
[17:50:45] <Paragon37> what is best parrallel or serial wired bipolar steppers?
[17:56:06] <skunkworks> series - higher top speed lower torque down low. Parrallel - lower top speed - higher torque down low. (that might be a bit generalized) gecko has a pdf about this also iirc
[17:58:07] <Paragon37> Glad you told me that because the guy I bought my stepper motor kit told me the opossite. It was only when I looked at the wiring diagram I thought parallel must be higher torque ... thnaks for for comfirming.
[18:06:21] <Paragon37> skunkworks are you sure thats correct? Just got this from the web :
[18:06:22] <Paragon37> There are two ways to connect a stepper motor; in series or in parallel. A series connection provides a high inductance and therefore greater performance at low speeds. A parallel connection will lower the inductance but increase the torque at faster speeds.
[18:06:57] <SWPadnos> series has a lower top speed because you can't get the current to build up as fast with the higher inductance
[18:07:23] <Paragon37> What would be the best setup for a mill?
[18:07:54] <SWPadnos> I thikn Mariss said "use parallel", but you should read the step motor basics PDF from his site - I'm pretty sure it has this in it
[18:08:22] <Paragon37> Thanks SWPadnos
[18:08:46] <SWPadnos> sure
[18:09:11] <skunkworks> boy - I don't know what is happening to my base knowledge.
[18:10:13] <skunkworks> I am not one to make shit up. (I don't think)
[18:30:11] <skunkworks> and now - no one talks to me anymore.
[18:32:32] <skunkworks> Shunned by my peers
[18:34:25] <cradek> skunkworks: um, what are you on about?
[18:35:50] <jepler> skunkworks: I'd talk to you but I don't know that I have much to say at the moment
[18:36:09] <jepler> skunkworks: I saw you ask about zenbot the other day .. I haven't done much with it
[18:36:28] <jepler> it zips around the workarea but I still need to do something about the flatness and alignment of the table
[18:36:47] <cradek> ha, a common problem
[18:37:02] <cradek> want some of the .003 feeler gauge stock I got to put under the right side of my vise?
[18:37:18] <skinnypuppy1334> I've been trollin on and off. But haven't had much to say. So Hi everyone !
[18:38:24] <cradek> jepler: if we could get my probe to fit it somehow, we could probe your table and you could write compensation in kinematics
[18:38:39] <cradek> that would be a fun, if silly, exercise
[18:38:43] <jepler> cradek: that would be interesting to do
[18:38:57] <skunkworks> heh
[18:39:45] <skunkworks> just had 2 thinkos recently that I can remember. stupid mistakes.
[18:40:06] <skunkworks> but the caffine is kicking in and I am already feeling better.
[18:40:07] <cradek> skunkworks: at least your memory is good
[18:40:11] <cradek> :-)
[18:45:34] <anonimasu> hm :/
[18:45:49] <anonimasu> cradek: I like that kind of compensation
[18:51:52] <Paragon37> Would it be OK to use the spindle control sig to signal the +5v enable for the driver I wish to test?
[18:51:55] <Paragon37> linksp spindle-on => parport.0.pin-09-out
[21:03:34] <cradek> Paragon37: I like to use axis.0.amp-enable-out for driver enables
[21:04:49] <cradek> jepler: I bet the probe has a 1/4" shank. I'm not sure how we could rig it up. maybe a replacement for the whole dremel, with a 1/4" hole in it.
[22:12:43] <CIA-8> 03swpadnos 07TRUNK * 10emc2/src/hal/utils/bfload.c: help message text change (possibly from jmkasunich)
[22:49:40] <Ziegler> my boundary issues seem to be that the absolute coords of the machine were different than the coordinates displayed in axis... but the program used the coords displayed in axis
[22:53:05] <jlmjvm> Ziegler:do you have home switches?
[22:54:47] <Ziegler> no I do not
[22:55:25] <Ziegler> maybe its time to make that addition
[22:55:28] <skinnypuppy1334> jlmjvm did you get that card working /
[22:55:49] <jlmjvm> yes,everything on the hurco is working
[22:56:13] <skinnypuppy1334> cool.
[22:56:14] <jlmjvm> got the spindle and brake going today
[22:57:26] <jlmjvm> i think some switches would help you out for sure
[22:58:25] <jlmjvm> you can probably do a g92 and still get it working
[23:01:19] <jlmjvm> how are your limits setup in your ini file?
[23:01:19] <dmessier> hi all
[23:01:42] <jlmjvm> skinnypuppy did you get any motors yet?