Back
[00:57:17] <Roguish> fenn: thanks for the help. made progress. happy. ttfn.
[01:50:19] <Ziegler> http://images.myonlinesite.com/cnc_mill/20071014/videos.html
[01:50:29] <Ziegler> attempt at a thread mill...
[01:51:00] <Ziegler> without the thread mill :\
[01:53:43] <fenn> well.. make one
[01:53:58] <fenn> only need one tooth
[01:54:09] <Ziegler> yeah... just was to impatient
[01:54:20] <fenn> btw usually they start at the bottom for some reason
[01:55:01] <fenn> chip clearance maybe
[01:55:12] <Ziegler> Not always but they will if chips are going to build up
[01:55:16] <Ziegler> yeah
[01:57:41] <toast> it's hard to crash a thread mill if you come out of the hole.
[01:57:59] <toast> most thread mill cycles can't be paused, so do it in a direction that will never crash.
[01:58:43] <fenn> that's dumb
[01:58:56] <toast> same with tapping cycles on a lot of machines
[01:58:56] <toast> etc
[01:59:23] <fenn> i could understand lathe threading, but not mill threading
[01:59:37] <toast> we don't have a single mill that you can feed hold
[01:59:44] <toast> while it is tapping or thread milling
[01:59:58] <fenn> there's nothing special about thread milling. it's just a helix
[01:59:59] <toast> not that we are a super representative sample
[02:00:13] <toast> we have like 9 mills, not all that many
[02:00:52] <toast> even so, it's still easier to go out than in
[02:01:04] <fenn> now that i finally got this point_at tranform to work, i cant figure out how to squeeze it into my program
[02:18:05] <fenn> oo it works
[02:26:18] <fenn> http://fenn.dyndns.org/pub/irc/hexapod_sim.png
[02:26:45] <fenn> still some scaling issues with the model but it still looks really cool
[02:27:56] <jepler> it does look good
[02:28:26] <jepler> I am sure you had to fight quite a bit with vismach.py to get what you wanted, since it's so centered around the idea of stacked rotary or linear axes
[02:29:19] <fenn> not too bad. i wish i opengl had a way to rotate around a point
[02:30:03] <jepler> you mean a shorthand for translate point to origin -- rotate -- translate point back?
[02:30:19] <fenn> also, i cheated. i get the platform position from emc so i dont have to do inverse kinematics
[02:31:02] <jepler> "know when to cheat" is good advice
[02:32:41] <fenn> one annoying thing in vismach is that you have no control over the order the tree is traversed, so you end up trying to invert empty matrices (empty lists)
[02:33:06] <fenn> and then it spews tracebacks for "cant unpack with zero values" or somesuch
[02:36:25] <jepler> hm
[02:36:45] <jepler> if a Collection's parts are a list or tuple then the traversal order should be defined
[02:36:58] <jepler> if its argument is a dict or set that won't be the case though
[02:52:17] <cradek> can windows people generally view/print a ps file?
[02:52:48] <tomp> i think not, cus adobe reader doesnt
[02:53:07] <cradek> hmmmm
[02:53:13] <cradek> how do I convert ps to pdf?
[02:53:22] <cradek> oh, ps2pdf
[02:53:24] <tomp> maybe if it's a type for 'word' or 'msdco'
[02:53:27] <cradek> sorry
[02:53:38] <tomp> msdoc
[02:54:29] <cradek> anyone here currently on windows?
[02:55:11] <cradek> if so can you visit
http://timeguy.com and see if sfm.pdf looks right?
[02:56:24] <DanielFalck> cradek: I'm not on windows, but that looks nice
[02:56:31] <cradek> thanks
[02:58:01] <SWPadnos> looks good
[02:58:06] <SWPadnos> if the top is radius, it's not labeled
[02:58:19] <SWPadnos> nevermind - I see decimal/fracional
[02:58:20] <cradek> the top is also diameter
[03:01:15] <SWPadnos> I do wonder if there's somewhere to stick S<SFM> into G-code
[03:01:31] <SWPadnos> also feed per tooth
[03:02:08] <cradek> there's already both
[03:02:19] <cradek> well, almost
[03:02:33] <cradek> surface speed yes, feed per rev yes. feed per tooth no
[03:03:06] <cradek> and it does use the S word for surface speed like you propose
[03:04:12] <CIA-8> 03fenn 07TRUNK * 10emc2/src/emc/usr_intf/axis/Submakefile: added hexagui
[03:04:13] <CIA-8> 03fenn 07TRUNK * 10emc2/src/emc/usr_intf/halui.cc: added hexagui
[03:04:13] <CIA-8> 03fenn 07TRUNK * 10emc2/configs/hexapod-sim/minitetra.ini: added hexagui
[03:05:55] <CIA-8> 03fenn 07TRUNK * 10emc2/lib/python/vismach.py: added Track transform
[03:06:04] <CIA-8> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build FAILED ; see
http://linuxcnc.org/compile_farm/emc2head_slot1_log.txt
[03:07:11] <CIA-8> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build FAILED ; see
http://linuxcnc.org/compile_farm/emc2head_slot1_log.txt
[03:07:19] <cradek> oops
[03:07:33] <CIA-8> 03compile-farm 07BDI-4.51 (2.6.16.20-rtai) * 10emc2head/: build FAILED ; see
http://linuxcnc.org/compile_farm/emc2head_slot6_log.txt
[03:07:36] <CIA-8> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build FAILED ; see
http://linuxcnc.org/compile_farm/emc2head_slot2_log.txt
[03:08:11] <cradek> I think he forgot a file
[03:08:26] <SWPadnos> perhaps a forgotten cvs add
[03:08:40] <SWPadnos> does CSS work for mills?
[03:09:05] <cradek> I don't see why not
[03:09:06] <SWPadnos> or FPR, for that matter
[03:09:17] <SWPadnos> well, it needs to know the radius, for one thing
[03:09:23] <cradek> hmm
[03:09:33] <cradek> ok maybe not.
[03:09:37] <SWPadnos> heh
[03:09:44] <cradek> I bet fpr works as-is
[03:10:02] <SWPadnos> also, you don't necessarily want spindle-synched motion, you just want to calculate the feed rate
[03:10:58] <SWPadnos> it probably does work, but probably not in exactly a way that would make a lot of sense to a user :)
[03:11:22] <cradek> nobody expects the spanish inquisition
[03:12:49] <SWPadnos> crash!
[03:13:27] <SWPadnos> interesting. I got two of the fail messages from the compile farm before the commit messages came in
[03:13:35] <SWPadnos> it's much faster now ;)
[03:13:49] <cradek> that is strange, they both go through sf
[03:14:18] <SWPadnos> fail messages at 11:01, 11:03, commits at 11:04 ...
[03:19:08] <cradek> now I'm trying to picture the graph that correlates rpm, ipm, feed per tooth, number of teeth
[03:19:25] <cradek> but that's 4 things so I don't think I can do the same kind of thing
[03:23:47] <cradek> one axis could be teeth/min (RPM * teeth)
[03:24:49] <fenn> yarr. i didnt forget anything, i'm just slow
[03:33:30] <CIA-8> 03fenn 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/tracking-test.py: add tracking test for new vismach.py code
[03:33:31] <CIA-8> 03fenn 07TRUNK * 10emc2/configs/sim/ (tracking-test.hal README): add tracking test for new vismach.py code
[03:34:25] <CIA-8> 03compile-farm 07BDI-4.51 (2.6.16.20-rtai) * 10emc2head/: build FAILED ; see
http://linuxcnc.org/compile_farm/emc2head_slot6_log.txt
[03:34:28] <CIA-8> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build FAILED ; see
http://linuxcnc.org/compile_farm/emc2head_slot2_log.txt
[03:36:38] <CIA-8> 03fenn 07TRUNK * 10emc2/configs/hexapod-sim/hexagui.hal: add hexapod visualization
[03:36:38] <CIA-8> 03fenn 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/hexagui.py: add hexapod visualization
[03:38:21] <CIA-8> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build PASSED
[03:40:48] <fenn> i'm hoping to find a way to run halui from a .hal file without having to be in the same directory as the .ini file
[03:40:59] <CIA-8> 03compile-farm 07BDI-4.51 (2.6.16.20-rtai) * 10emc2head/: build PASSED
[03:41:01] <CIA-8> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build PASSED
[03:42:43] <CIA-8> 03fenn 07TRUNK * 10emc2/src/emc/usr_intf/axis/Submakefile: tracking test
[03:48:27] <fenn> nice graph
[03:49:52] <cradek> thanks
[03:49:59] <cradek> I'm almost done with its mate
[03:50:36] <fenn> feed per minute?
[03:51:08] <cradek> yes, teeth/min (RPM*teeth) + feed/tooth => feed (ipm)
[03:56:21] <cradek> done
[03:56:30] <cradek> http://timeguy.com
[03:56:44] <cradek> whee
[03:57:22] <fenn> teeth/min! you mean i have to do MATH!!!
[03:57:43] <cradek> haha
[03:57:49] <cradek> it gives you the "formula" right there...
[03:58:00] <cradek> is there a simpler way? I couldn't come up with one
[03:58:40] <fenn> no probably not
[03:58:43] <SWPadnos> slide rule
[03:58:59] <SWPadnos> hmmm. maybe that's waht we should make for CNC workshop next year
[03:59:16] <cradek> :-)
[03:59:42] <SWPadnos> at least that wouldn't go out of date as quickly as a CD or g-code reference ;)
[03:59:42] <fenn> do you know how to use a slide rule?
[03:59:49] <SWPadnos> yes. I used to have on
[03:59:50] <SWPadnos> e
[04:00:02] <fenn> i imagine its something like that chart
[04:00:08] <cradek> seriously, you think they're useful? I thought I was the only one who has to stop and think about these things
[04:00:26] <SWPadnos> well, some of can do this in our heads, but they're still useful ;)
[04:00:29] <fenn> the chart or a slide rule?
[04:00:39] <SWPadnos> both
[04:00:59] <fenn> think of the charts as a charm to ward off murphy's minions
[04:02:02] <cradek> I have a nice slide rule ... somewhere
[04:02:53] <cradek> that would actually be a good application for a slide rule. it's exactly what the ipm graph is
[04:03:06] <SWPadnos> right
[04:03:13] <cradek> you'd slide it to set feed/tooth and directly read it for any rpm
[04:03:19] <SWPadnos> you do a ddouble-sided wheel. one for SFM, the other for IPM
[04:03:43] <SWPadnos> rings are the number of teeth
[04:03:48] <cradek> that would be neat.
[04:03:56] <cradek> I bet they exist already
[04:04:01] <SWPadnos> I'm sure they do
[04:04:09] <cradek> but ... I don't have one
[04:04:29] <SWPadnos> heh
[04:04:31] <fenn> this sounds like a job for postscript
[04:04:43] <cradek> hey! it could be turned automatically by the varispeed readout
[04:04:54] <cradek> I'd have to think about that...
[04:05:23] <cradek> I don't "blog" much. the front page goes back 2 years.
[04:07:38] <cradek> wow, I just looked at
http://jmkasunich.com/pics/wood-duck-flapping-1136.JPG - that's a cool picture.
[04:10:15] <jmkasunich> you "just" looked at it?
[04:10:24] <jmkasunich> cause I think I removed it a while ago
[04:10:47] <jmkasunich> (gonna put a slightly smaller version of it back in a bit)
[04:11:22] <steves_logging> steves_logging is now known as stevestallings
[04:12:06] <stevestallings> .. OK so I am not having problems with your server 8-)
[04:12:50] <stevestallings> whew... spent the last 2 hours quickly scanning IRC to catch up, lots going on with new release
[04:13:13] <stevestallings> thanks to everyone who is working on this
[04:13:55] <SWPadnos> it's all for you
[04:13:58] <SWPadnos> or
[04:14:00] <SWPadnos> not! ;)
[04:14:06] <SWPadnos> hi Steve
[04:14:13] <fenn> it's all for steve
[04:14:18] <SWPadnos> I am steve!
[04:14:24] <SWPadnos> or Steve!
[04:14:27] <stevestallings> hi Steve - or do I repeat...
[04:14:33] <SWPadnos> yes, yes you do
[04:15:32] <cradek> jmkasunich: it was in a tab...
[04:15:38] <stevestallings> lots of catch-up going on in my life, office is finally happening
[04:15:46] <cradek> goodnight guys
[04:15:53] <stevestallings> and I just got this home -
http://www.metalworking.com/personal/ITSHOME.JPG
[04:15:57] <cradek> stevestallings: congrats on the successful lathe transport
[04:16:17] <jmkasunich> stevestallings: I just put the pic back on the server
[04:16:38] <jmkasunich> append an m to the filename (for medium size - the other one was kinda bloaty)
[04:16:59] <jmkasunich> http://jmkasunich.com/pics/wood-duck-flapping-1136m.JPG
[04:19:29] <SWPadnos> ooooh - nice trailer
[04:19:45] <stevestallings> great pic, technical quality excellent, fantastic luck catching that!
[04:19:51] <SWPadnos> I'm not sure my Jeep could tow that machine on that trailer ;)
[04:20:23] <fenn> its clever how the ramp has a support beam right next to the hinge
[04:21:02] <stevestallings> trailer was acquired specifically for the job, 14K lbs rated, just barely enough, truck pushed 1k beyond rating but preformed very well
[04:21:27] <SWPadnos> hmmm. what does the trailer weigh unloaded?
[04:21:55] <stevestallings> about 3K lbs
[04:22:04] <jmkasunich> that's a lot of trailer
[04:22:04] <stevestallings> lathe was 11K lbs
[04:22:06] <SWPadnos> ok, that's ha'f the rating for the Jeep
[04:22:12] <SWPadnos> half, too
[04:23:53] <stevestallings> hopefully I will get into the new office/warehouse by December and have enough power to test the lathe
[04:24:16] <fenn> what kind of lathe is that?
[04:24:22] <stevestallings> if it proves to be a keeper, then I get to put into the shop I haven't built yet 8-)
[04:24:49] <stevestallings> 1984 vintage Hardinge SuperSlant
[04:25:00] <SWPadnos> Hardinge SuperBigAssLathe
[04:25:38] <stevestallings> it was Hardinge's first CNC from the ground up, previously they had sold converted HC chuckers
[04:25:54] <stevestallings> they sort of went overboard making it heavy and rigid
[04:26:29] <SWPadnos> that's not exactly a bad thing
[04:26:32] <SWPadnos> unless you're the one moving it
[04:26:41] <stevestallings> 3 axis machine, one side turrent with X and Z, plus one end turret with Z only, twin Fanuc 6T controls
[04:26:47] <fenn> at 11klb its too much of a good thing
[04:27:12] <jmkasunich> good for hurricane country - you can use it to hold down the shop
[04:27:14] <SWPadnos> dump the Fanucs, that'll bring it down to 8k or so
[04:27:33] <fenn> bad for sand country - it'll make the shop sink
[04:27:33] <stevestallings> dollars or lbs ? 8-)
[04:27:38] <SWPadnos> heh
[04:28:03] <stevestallings> I wish for sand, here we only have clay
[04:29:27] <SWPadnos> braking must have been fun while towing that
[04:29:57] <stevestallings> the clay moves and causes foundation problems, I had to pay for soil engineering study as part of permit process
[04:30:33] <jmkasunich> how about clay with rocks in it?
[04:30:33] <stevestallings> fortunately I did not have any incidents that tested the breaking significantly
[04:30:44] <SWPadnos> beaking, braking
[04:30:50] <SWPadnos> you don't want good breaking ;)
[04:31:17] <stevestallings> trailer has brakes on both axels and I did make sure trailer brakes alone would stop the rig by testing at 30 mph
[04:32:19] <SWPadnos> that had to be a slightly - err, unsettling - test
[04:32:25] <stevestallings> sudden lane changes to avoid idiots were my main worry
[04:32:34] <SWPadnos> heh
[04:32:58] <stevestallings> no, the 30 mph test was on an empty industrial access road
[04:33:13] <jmkasunich> and trailer brakes only means no jacknife risk
[04:33:22] <jmkasunich> truck brakes only would be scary
[04:33:49] <stevestallings> truck brakes only are illegal over 3K lbs, and not even close to adequate
[04:35:51] <SWPadnos> it's not just illegal, it's really really stupid!
[04:35:56] <SWPadnos> triple the fun!
[04:36:31] <stevestallings> I'm getting too old for "fun"
[04:37:20] <stevestallings> at least that type of fun, putting power on the lathe for the first time should be enough to get my juices flowing
[04:38:12] <stevestallings> Hardinge specifies 460 VAC 3ph 60 amp service, which I conveniently contracted for in the new warehouse
[04:39:11] <stevestallings> stupid amount of electric capacity really, but they had an old fashion 13 HP spindle with DC motor intended to go 0 to 4K rpm in a couple of seconds
[04:39:51] <stevestallings> power factor correction is not present in the motor controller, so mains load is horrible
[04:40:15] <jmkasunich> only 60A? ;-)
[04:40:25] <SWPadnos> makes me wonder what the service was for tha 30HP (or was it 50HP) lathe they had at NIST 3 years ago
[04:40:48] <stevestallings> probably similar, considering it was modern
[04:41:00] <jmkasunich> if there's a VFD in the picture, 60A will do 50HP, or close anyway
[04:41:07] <jmkasunich> at 480V, its about 1.2A per HP
[04:41:20] <jmkasunich> and the VFD eliminates the starting current requirement
[04:43:02] <stevestallings> if the Hardinge works out OK, there may be a 15 HP induction motor and VFD in its future, I don't need jack rabbit starts
[05:01:04] <stevestallings> stevestallings is now known as steves_logging
[06:33:13] <user_> user_ is now known as SkinnyPuppy34
[07:12:31] <lerman__> lerman__ is now known as lerman
[10:17:55] <_epineh> Hey all, I have a quick question about the Pluto-p if any 1 has a min...
[10:24:47] <_epineh> will it handle power for 4 encoders directly or should I provide seperate power and opto's ?
[11:29:06] <Paragon37> Hello Everyone
[11:45:05] <Paragon37> I have a strange situation occuring :-
[11:45:05] <Paragon37> I have set max feed rate velocity at 900mm/min
[11:45:08] <Paragon37> When I execute G0 X100 Y100 it moves without error but when I try to return to Y0 I keep getting a joint following error?
[11:45:08] <Paragon37> If I hit F2 then enter G0 Y0 I get the error again.
[11:45:11] <Paragon37> Additionaly I get Z: 0.049 from 0.000 and the velocity shows as 7.500 ??
[11:45:12] <Paragon37> If I then hit F2 then G0 X0 it moves OK
[11:45:14] <Paragon37> I am using pluto-step with the following setup in the ini file...
[11:45:14] <Paragon37> Motion Control setup
[11:45:15] <Paragon37> BASE_PERIOD = 100000
[11:45:18] <Paragon37> SERVO_PERIOD = 1000000
[11:45:18] <Paragon37> TRAJ_PERIOD = 10000000
[11:45:20] <Paragon37> Traj setup
[11:45:21] <Paragon37> DEFAULT_VELOCITY = 4
[11:45:23] <Paragon37> MAX_VELOCITY = 15
[11:45:24] <Paragon37> DEFAULT_ACCELERATION = 5.0
[11:45:26] <Paragon37> MAX_ACCELERATION = 20.0
[11:45:29] <Paragon37> Axis Setup for all
[11:45:29] <Paragon37> MAX_VELOCITY = 15
[11:45:30] <Paragon37> MAX_ACCELERATION = 5.0
[11:45:34] <Paragon37> INPUT_SCALE = 1600
[11:45:36] <Paragon37> FERROR = 0.150
[11:45:38] <Paragon37> MIN_FERROR = 0.01
[11:49:52] <Paragon37> Additionaly if I just move with G0 Y100 it gets there but issuing G0 Y0 gives me the same error 'joint 2 following error'
[11:49:54] <Paragon37> Strange that it shows an error for joint 2 as this would pertain to Axis 2 (Z) right?
[11:56:53] <Paragon37> I think there is a bug within emc where the Y axis negative move shows as a Z axis positve move...
[12:13:37] <Paragon37> Just ran HAL scope and monitored Zpos-cmd, Ypos-cmd, Xpos-cmd triggering on 'pluto-step.stepgen.0.counts'
[12:13:40] <Paragon37> I can see Zpos-cmd and Xpos-cmd at the scope move as commanded but Ypos-cmd is not even displayed!!
[12:13:40] <Paragon37> Also when commanding Y in the negative plane I see Zpos-cmd changing with a raised step from HAL Scope!!
[12:42:35] <jepler> Paragon37: I don't spot any particular problems with the inifile snippets you've shown. I don't doubt that there are big bugs in pluto-step. However, it's simply not getting any of my time or attention right now. If I don't have time to find and fix these bugs, the driver will be dropped before emc 2.2.0.
[12:45:38] <Rok> hello. i need help with a newbie problem: i am using emcsh console to control a machine. when i run a program, no other commands can be issued until the program finishes. if i use 2 emcsh shells, the program can be aborted from the second one. why is this so, and can i abort a program from the emcsh that is used to run it. tnx in advance.. :)
[12:46:39] <Paragon37> Jepler where is the source code located for pluto-step I don't think this is a problem with the pluo-p verilog but rather the EMC driver?
[12:47:09] <jepler> Paragon37: it's all with the regular emc source distribution, src/hal/drivers/pluto_step.comp src/hal/drivers/pluto_step_firmware/
[12:48:19] <Paragon37> Thanks I'll take a look ... Not sure if I can help though.
[12:50:27] <jepler> Rok: Which command do you issue in emcsh? each one I've tried returns to the emcsh prompt right away. In particular, "emc_run 0" did, and I could type "emc_abort" to make the g-code stop executing.
[12:59:07] <Rok> each one returns for me too, except emc_run
[12:59:34] <Rok> if i write emc_abort nothing happens until the program finishes, only then it is exeuted
[13:00:42] <jepler> are you running interactively? If so, is the "%" prompt shown?
[13:01:17] <jepler> what version of emc are you using? I was testing with the development version, but I'll test on 2.1.7 in just a moment
[13:01:20] <Rok> yes, i'm running it interactivey, but the % shows only after the program finishes
[13:01:57] <Rok> it's 2.1.6 or 2.1.7., how do i check that fast (told you i was a newbie :))
[13:03:24] <jepler> I *do* get the behavior you describe when I run 2.1.7
[13:03:31] <jepler> when I type "emc_run 0" I don't get the "%" prompt back..
[13:04:12] <Rok> you'll get it when the program finishes.. :)
[13:06:01] <jepler> I didn't let it go that long..
[13:06:16] <jepler> but tkemc does exactly that to run a program, and it doesn't stop responding during that time ... !
[13:07:27] <Rok> yes, i looked at the tkemc.tcl file already. it's weird, right? another thing is: if the program is started from tkemc gui it can be aborted from the shell at any time
[13:11:48] <jepler> Rok: Issue this command: % emc_set_wait received
[13:11:58] <jepler> otherwise, emcsh is doing the right thing: waiting for the command to complete
[13:12:01] <jepler> it's just not the thing you want
[13:12:44] <Rok> yes, that did it
[13:12:54] <Rok> tnx for your help!
[13:13:10] <jepler> in the development version, it looks like the default has been changed to 'received', instead of 'done'
[13:13:14] <jepler> I'm glad I could puzzle it out for you
[13:13:29] <jepler> bbl
[13:14:40] <Rok> thanks again. bye
[13:19:16] <Paragon37> Hello Jepler could you point me in the right direction of where I should start to look with regards to the pluto issue?
[13:19:18] <Paragon37> I have dissable the hal commands for axis.2 (Z) and now axis.1 (Y) is working and shown in HAL Scope...
[13:22:29] <Paragon37> As soon as I enable this command it fails with the following error 2 for Y (axis.1) neg movement :-
[13:22:29] <Paragon37> net Zpos-fb pluto-step.stepgen.2.position-fb => axis.2.motor-pos-fb
[13:24:14] <fenn> you are using a stepper system right?
[13:24:35] <fenn> and no gantry kinematics or anything
[13:25:19] <Paragon37> Is this for me Fen?
[13:25:22] <fenn> yes
[13:27:02] <Paragon37> Yes I am using pluto-step connected to a microstepping drive (1600 uStep) to a stepper motor unmounted there appears to be a software bug when moving on the axis.1 negativly it interferes and shows up on the Z axis.2
[13:28:03] <Paragon37> I see a spike on the hal scope for Z when hitting the down arrow on the keyboard to move Y axis.1
[13:28:40] <fenn> on which signal?
[13:28:42] <Paragon37> I then get a following error for axis 2 but I have not commanded axis.2 (z) to move
[13:30:13] <Paragon37> Here is what I posted earlier :-
[13:30:15] <Paragon37> I have a strange situation occuring :-
[13:30:18] <Paragon37> I have set max feed rate velocity at 900mm/min
[13:30:19] <Paragon37> When I execute G0 X100 Y100 it moves without error but when I try to return to Y0 I keep getting a joint following error?
[13:30:21] <Paragon37> If I hit F2 then enter G0 Y0 I get the error again.
[13:30:22] <Paragon37> Additionaly I get Z: 0.049 from 0.000 and the velocity shows as 7.500 ??
[13:30:24] <Paragon37> If I then hit F2 then G0 X0 it moves OK
[13:30:25] <Paragon37> I am using pluto-step with the following setup in the ini file...
[13:30:27] <Paragon37> Motion Control setup
[13:30:28] <Paragon37> BASE_PERIOD = 100000
[13:30:30] <Paragon37> SERVO_PERIOD = 1000000
[13:30:31] <Paragon37> TRAJ_PERIOD = 10000000
[13:30:32] <fenn> yes yes i saw all that
[13:30:31] <Paragon37> Traj setup
[13:30:33] <Paragon37> DEFAULT_VELOCITY = 4
[13:30:36] <Paragon37> MAX_VELOCITY = 15
[13:30:36] <Paragon37> DEFAULT_ACCELERATION = 5.0
[13:30:37] <Paragon37> MAX_ACCELERATION = 20.0
[13:30:39] <Paragon37> Axis Setup for all
[13:30:42] <Paragon37> MAX_VELOCITY = 15
[13:30:43] <Paragon37> MAX_ACCELERATION = 5.0
[13:30:46] <Paragon37> INPUT_SCALE = 1600
[13:30:47] <Paragon37> FERROR = 0.150
[13:30:50] <Paragon37> MIN_FERROR = 0.01
[13:30:52] <Paragon37> Additionaly if I just move with G0 Y100 it gets there but issuing G0 Y0 gives me the same error 'joint 2 following error'
[13:30:55] <Paragon37> Strange that it shows an error for joint 2 as this would pertain to Axis 2 (Z) right?
[13:30:57] <Paragon37> I think there is a bug within emc where the Y axis negative move shows as a Z axis positve move...
[13:30:59] <Paragon37> Just ran HAL scope and monitored Zpos-cmd, Ypos-cmd, Xpos-cmd triggering on 'pluto-step.stepgen.0.counts'
[13:31:00] <Paragon37> I can see Zpos-cmd and Xpos-cmd at the scope move as commanded but Ypos-cmd is not even displayed!!
[13:31:03] <Paragon37> Also when commanding Y in the negative plane I see Zpos-cmd changing with a raised step from HAL Scope!!
[13:31:05] <Paragon37> Oh Sorry
[13:32:17] <fenn> i dont think it has anything to do with pluto if you are seeing something on axis.1.*
[13:32:32] <fenn> or any of the axis.* pins
[13:32:46] <Paragon37> I agree I think it is somwhere within EMC2
[13:33:05] <fenn> unless.. pluto is sending a weird feedback value
[13:34:36] <Paragon37> Maybe... it weird though when I dissable axis.2 the Y axis.1 does not error.
[13:34:44] <fenn> try scoping axis.2.motor-pos-fb at the same time
[13:35:56] <Paragon37> I have and I see a spike when i try and move Y(axis.1) in the neg (Down Arror Key) it then errors with the 'joint 2 following error'
[13:37:26] <Paragon37> Sorry I got confused with axis.2.motor-pos-fb I'll try this now!
[13:41:02] <Paragon37> ok just tried that and on moving the Y axis I see the line rise for axis.2.motor-pos-fb
[13:41:23] <Paragon37> That Y axis.1 in the negative plane only
[13:43:58] <fenn> ok so that narrows down the bug
[13:44:31] <fenn> you should submit a bug report on sourceforge so we dont forget
[13:45:41] <Paragon37> Fenn I will do .... :-) but where could I be looking so as to maybe help fix the issue?
[13:47:05] <fenn> http://cvs.linuxcnc.org/cvs/emc2/src/hal/drivers/pluto_step_firmware/
[13:47:38] <fenn> actually, start here
http://cvs.linuxcnc.org/cvs/emc2/src/hal/drivers/pluto_step.comp
[13:47:41] <Paragon37> So it's the verilog stuff at the pluto side sending incorrect feed back
[13:47:49] <fenn> i dont know, that's what you're supposed to find out
[13:48:27] <Paragon37> I'd better start learning then ;-)
[13:51:14] <fenn> look at FUNCTION(read) in pluto_step.comp - will show you what you need to be paying attention to
[13:55:55] <Paragon37> Thanks for the pointer Fenn
[14:09:53] <Guest635> Guest635 is now known as skunkworks
[14:18:00] <CIA-8> 03jepler 07TRUNK * 10emc2/docs/src/hal/tutorial.lyx:
[14:18:00] <CIA-8> Update to latest style: halcmd -> halrun, linkxx -> net
[14:18:00] <CIA-8> fix up some sample outputs to better match what the user will see
[14:35:11] <CIA-8> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/Submakefile: submakefile reorganization makes stepconf strings get into axis.pot
[14:35:12] <CIA-8> 03jepler 07TRUNK * 10emc2/src/po/Submakefile: submakefile reorganization makes stepconf strings get into axis.pot
[14:47:53] <fenn> ohh.. hexagasm
[14:49:22] <skunkworks> umm - what?
[14:52:45] <fenn> i finally got the model coordinates lined up with the kinematics coordinates:
http://fenn.dyndns.org/pub/irc/hexapod_sim1.png
[14:53:37] <fenn> well, almost
[14:55:43] <fenn> enough to trick my primate brain into thinking its real
[14:57:06] <skunkworks> heh - very neat
[14:57:49] <skunkworks> is there any pictures of that style hexipod anywhere?
[14:59:12] <fenn> http://wodan.meb.uni-bonn.de/forsch/gebiete/verschleiss/hexapod.jpg
[14:59:34] <skunkworks> cool - thankyou
[15:00:04] <skunkworks> so - are you making it so - the spindle will be stationary above?
[15:00:20] <skunkworks> or is the spindle going to be mounted on the hexipod?
[15:01:01] <fenn> i'm going to make a prototype/toy that looks like that, and does nothing
[15:01:19] <fenn> and then i'm going to make one with an octahedral frame that's upside down or horizontal
[15:01:30] <fenn> with the spindle in the middle
[15:02:28] <fenn> mounting joints solidly to an octahedron is tricky unless you've got big hunks of steel holding everything together (not what i want)
[15:03:10] <fenn> i've got an idea for a sort of welded pyramid thingy
[15:03:57] <skunkworks> neat
[15:04:13] <fenn> i think there's a flaw in my tracking algorithm
[15:09:05] <fenn> http://fenn.dyndns.org/pub/irc/bad_tracki.png <- the cylinder is supposed to be pointed at the circle
[15:11:51] <fenn> er,
http://fenn.dyndns.org/pub/irc/bad_tracking.png
[15:11:58] <skunkworks> I had found it :)
[15:13:38] <fenn> it also does weird stuff when you move the window around
[15:14:13] <fenn> * fenn tries adding double precision rotation to minigl
[15:31:44] <fenn> nope, that wasnt it
[15:38:47] <Paragon37> Hi All: should I be seeing the led flash on the pluto-p board when I move direction using pluto-step?
[15:39:32] <Paragon37> I am only seeing it flash in one axis X (axis.1)
[15:40:25] <jepler> Paragon37: "When the device is programmed, the LED glows according to the duty cycle of PWM0 (LED = UP0 xor DOWN0) or STEPGEN0 (LED = STEP0 xor DIR0)."
http://linuxcnc.org/docs/devel/html/hal_drivers.html#r1_8_4
[15:41:45] <Paragon37> Thats what I thought but I am only seeing flash on it on the X axis.1 and not when I move Y or Z..
[15:42:24] <lewin1> lewin1 is now known as lewing
[15:42:45] <Paragon37> What should the para port be set up as EPP or standard ?
[15:42:53] <jepler> right, STEPGEN0 is typically the X axis
[15:43:23] <jepler> EPP
[15:44:32] <Paragon37> Cheers Jepler : Pulling my hair out with this apparent bug Y leaks to Z ;-(
[16:42:12] <Paragon37> Jeppler I manged to get it working but with a fudge by setting the high values below :-
[16:42:14] <Paragon37> FERROR = 28.050
[16:42:15] <Paragon37> MIN_FERROR = 20.010
[16:42:17] <Paragon37> What I am seeing when moving the Y axis in minus direction is the Z axis counts up to around +9mm when Y movement has stopped the Z axis count down by aprox 9mm.... Weird
[16:42:18] <Paragon37> I've tried to understand the code but to no avail .... still scratching my head ;-)
[17:12:46] <Guest233> Guest233 is now known as skunkworks
[17:49:10] <skunkworks> cradek: we did something similar when we grabbed a mill that was carbide but was only hhs. It turned into a read glowing mass of mettal.
[17:50:29] <skunkworks> *thought it was carbide
[17:58:06] <JymmmEMC> hand warmer
[18:13:25] <cradek> skunkworks: yes it was very colorful :-)
[18:22:28] <Rabbitbunny> interesting, people are actually here.
[18:22:50] <skunkworks> umm - what makes you think that?
[18:23:09] <Rabbitbunny> So, I grabbed the EMC2 livecd, threw it in a VM and booted it. what's the login?
[18:23:29] <cradek> there isn't a login when you boot live
[18:23:42] <cradek> it boots right to the desktop
[18:24:06] <Rabbitbunny> I know there isn't normally, I run dapper on my toy box. however, this is definatly asking me for a login.
[18:24:30] <cradek> that's very puzzling
[18:25:04] <cradek> that cd is simply a dapper cd with a few packages added. it works exactly the same.
[18:25:25] <Rabbitbunny> downloaded emc2-ubuntu6.06-desktop-i386.iso this morning.
[18:25:39] <Rabbitbunny> Right I booted a normal dapper cd to be sure, it goes right to desktop.
[18:25:44] <skunkworks> could you boot off of it?
[18:25:52] <skunkworks> for shits and grins
[18:26:03] <Rabbitbunny> oh yes. it boots fine.
[18:26:23] <cradek> I'm reaching here - maybe the download or burn was bad. did you check the md5sum?
[18:26:58] <Rabbitbunny> heh, it is too late in the day for me to be forgetting to do that...
[18:27:08] <Rabbitbunny> (i did forget)
[18:27:30] <cradek> worth a try... something is funny. lots of people have used that iso successfully
[18:44:20] <Rabbitbunny> meh it's checking.
[18:44:47] <Rabbitbunny> So, I understand I can create a virtual cnc machine to play around with. this is correct
[18:45:55] <Rabbitbunny> err corrct?
[18:46:03] <SWPadnos> yes
[18:46:42] <Rabbitbunny> awesome, md5 is correct.
[18:46:59] <Rabbitbunny> and, awesome. virtual machine means I don't have to build it first.
[18:47:16] <SWPadnos> not that it will be trivial to do, but it is possible :)
[18:47:29] <SWPadnos> I think the default user name in ubuntu is ubuntu
[18:47:39] <SWPadnos> with no password
[18:47:59] <Rabbitbunny> they really should write that down somewhere...
[18:48:03] <SWPadnos> but it's very strange that you would get a login screen from the liveCD
[18:48:08] <Rabbitbunny> booting the iso again to check
[18:48:10] <SWPadnos> only if you check that it's true first
[18:48:45] <SWPadnos> and you're quite sure it is the liveCD that's booting? (ie, that the VM is set to boot from CD rather than from hard drive?)
[18:49:09] <Rabbitbunny> righto
[18:49:53] <Rabbitbunny> hmm, ubuntu with no pass doesn't work.
[18:50:05] <SWPadnos> I'm probably wrong then ;)
[18:50:52] <skunkworks> Rabbitbunny: you did say that you can boot off of the emc2 livecd normally and it goes right to the desktop?
[18:51:30] <Rabbitbunny> No, I can boot the ubuntu dapper livecd and it goes to the desktop, the emc2 livecd wants a username
[18:52:10] <skunkworks> that is odd.
[18:52:36] <skunkworks> you could just use the ubuntu dapper livecd and install emc2 to it.. if all else fails.
[18:53:06] <Rabbitbunny> right, but I was hoping to play for a bit before getting serious.
[18:53:29] <SWPadnos> I guess I was right:
http://ubuntuforums.org/archive/index.php/t-36690.html
[18:54:02] <SWPadnos> I'm suspicious that something else is booting, not the LiveCD
[18:54:21] <SWPadnos> I don't know how that's being accomplished, but I'm reasonably sure that's what's happening
[18:55:03] <SWPadnos> hmmm - now there's an inopportune spam subject: "Summer is almost here, be ready"
[18:55:15] <Rabbitbunny> they mean next year.
[18:55:20] <SWPadnos> considering that we just got snow in the mountains ;)
[18:56:10] <Rabbitbunny> well, might as well install dapper then emc2, the login thing is annoying.
[18:57:14] <SWPadnos> that's definitely one way to do it, and it's not too much harder
[18:57:54] <SWPadnos> actually, if you plan to only do simulation, that may be a better idea anyway - you can install the sim version of EMC2 on a non-realtime OS
[18:58:09] <SWPadnos> that should run faster if you have it in a VM
[18:58:12] <Rabbitbunny> well, yay.
[18:58:44] <skunkworks> follow these directions
http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_5_10_and_6_06_from_source
[18:58:54] <SWPadnos> I think there are sim packages now
[18:59:10] <SWPadnos> so source son't be necessary until the visualizations need to be compiled
[18:59:14] <skunkworks> then you would do --enable-simulator
[18:59:15] <SWPadnos> won/t
[18:59:17] <SWPadnos> won't
[18:59:19] <SWPadnos> ha!
[18:59:26] <skunkworks> oh - I did not know that.
[18:59:30] <Rabbitbunny> heh
[18:59:42] <SWPadnos> I could be wrong - it happens a lot before I've had enough coffee
[18:59:46] <skunkworks> I have yet to play with the sim
[18:59:53] <Rabbitbunny> righto, all this from a spark.
[19:00:04] <Rabbitbunny> I want a small desktop cnc built from GM IAC's
[19:00:15] <SWPadnos> IAC?
[19:00:23] <skunkworks> gm idle air control?
[19:00:50] <Rabbitbunny> Idle Air Controller, it's a bipoler stepper jammed into the throttle body of early-mid 90's GM trucks.
[19:00:50] <SWPadnos> Generic Modular Integrated Accessory Components?
[19:01:29] <Rabbitbunny> heh
[19:02:04] <Rabbitbunny> They're kinda cool because the shaft is internally threaded. from a glace last year I'd say 6/32" shaft.
[19:02:11] <Rabbitbunny> Have to grab one from a junkyard.
[19:04:23] <Rabbitbunny> I figure 6"x6"x2" should be doable. nice way to practice my AL casting, and if it works a nice way to build controllers for a bigger one.
[19:04:42] <Rabbitbunny> if it doesn't I'm $20 into it. who cares?
[19:05:09] <SWPadnos> that's the spirit!
[19:06:29] <Rabbitbunny> meh, back to the grind
[19:06:30] <jepler> Rabbitbunny: if you use ubuntu dapper on a box already, you can install the "simulator" version of emc. I think this version of the install .sh file will do it:
http://emergent.unpy.net/index.cgi-files/sandbox/emc2-install-sim.sh
[19:06:44] <jepler> you don't have to install a different real-time kernel in this case
[19:07:08] <Rabbitbunny> Yeah, but I break things. I like doing it in a sandbox, then really doing it.
[19:08:53] <SWPadnos> the sim version is strictly userspace - it shouldn't be very easy to break your install with it
[19:09:40] <jepler> compiling the sim is pretty safe -- with --enable-run-in-place, you don't need to run anything as root, and no files are written outside the source tree by the build process
[19:11:16] <jepler> but by all means, do this in whatever way makes you most comfortable
[19:13:06] <ds2> ...unless you do a make install...
[19:13:31] <SWPadnos> I wonder what install does for run-in-place
[19:13:58] <ds2> IIRC, it goes and tries to install stuff all over the place (i.e. /etc/init.d/)
[19:14:13] <ds2> tried it before I got buried in 'real work' :(
[19:14:33] <SWPadnos> are you sure that was with ./configure --enable-run-in-place ?
[19:16:49] <alex_joni> it gives an error
[19:17:08] <alex_joni> Makefile:388: *** You configured run-in-place, but are trying to install.
[19:17:09] <alex_joni> For an installable version, run configure without --enable-run-in-place
[19:17:09] <alex_joni> and rebuild. Stop.
[19:17:19] <SWPadnos> I wonder if install for RIP should just do setuid instead
[19:17:32] <alex_joni> no, install should barf with an error
[19:17:31] <SWPadnos> since that's the RIP equivalent of install
[19:17:35] <SWPadnos> heh
[19:17:42] <SWPadnos> then it's fixed! ;)
[19:17:46] <alex_joni> people who do make install, expect it to install
[19:18:14] <alex_joni> * alex_joni doesn't want a bunch of users asking where did the files go after make install was completed successfully
[19:18:49] <SWPadnos> yeah, and terminals don't have very good support for really big type saying "THIS IS RIP YOU IDIOT!!!"
[19:19:17] <alex_joni> they do have colours and blinking
[19:19:25] <alex_joni> RIP IDIOT !!!
[19:19:28] <SWPadnos> and lots of asterisks
[19:19:32] <alex_joni> that's a great message
[19:20:09] <SWPadnos> I shouldn't write error messages after staying up all night
[19:20:29] <alex_joni> * alex_joni knows that just too well
[19:20:32] <SWPadnos> heh
[19:21:16] <alex_joni> apt just "wrote magic at 66ac"
[19:21:29] <SWPadnos> it's a magic app
[19:21:29] <alex_joni> I *knew* there is vodoo in here
[19:22:14] <alex_joni> didn't do "apt-get moo" in a while though
[19:25:37] <ds2> I wonder if my initial complaining had some effect then
[19:30:09] <fenn> i always wondered what that Super Cow Powers stuff was about
[19:30:42] <alex_joni> me too
[19:32:25] <fenn> http://en.wikipedia.org/wiki/Aptitude_(program)#Easter_egg
[19:37:40] <alex_joni> ROFL
[19:37:50] <alex_joni> * alex_joni also got the elephan eaten by a snake
[19:41:19] <fenn> that mailing list is out of control
[19:41:55] <alex_joni> which one?
[19:41:59] <fenn> the users list
[19:42:28] <cradek> the signal to noise seems pretty high to me
[19:43:06] <fenn> well, yes. but there's a lot of signal
[19:45:40] <skunkworks> I think he got confused with the thread - I am pretty sure the lathe was using 1ppr. It was not however using feedback to the vfd from the ac motor.
[19:45:59] <skunkworks> hence the confusion.
[19:46:03] <cradek> ok that was a little bit noisy
[19:46:18] <cradek> but still not bad
[19:57:00] <CIA-8> 03jepler 07TRUNK * 10emc2/docs/src/Submakefile: quoting in generated index.html manpage links was wrong
[21:41:21] <stustev> http://www.hougen.com/tech_tips/speeds/speed.html
[21:41:38] <stustev> this should give you a speed and feed calculator
[21:45:55] <stustev> cradek: does this look like what you were thinking of for a slide rule type calculator?
[21:45:59] <stustev> http://cgi.ebay.com/Machinists-Speed-Feed-Calculator-CNC-Mill-Tooling_W0QQitemZ230180803131QQihZ013QQcategoryZ104242QQcmdZViewItem
[21:50:02] <JymmmEMC> wth, it doens't have butter or playdo on it - COMPLETELY USELESS!!!!
[21:50:08] <fenn> i think i'd want something that doesnt require me to think so hard
[21:50:32] <fenn> re: the slide rule thingy
[21:51:05] <JymmmEMC> fenn: If slide rule it too much work, you really are a lazy bastard ;)
[21:51:37] <fenn> i am a lazy bastard
[21:51:42] <JymmmEMC> lol
[21:51:47] <fenn> * fenn goes to take a nap
[21:52:07] <JymmmEMC> well, at least you made me laugh, it's been a rough day so far. damn insurance company
[22:00:07] <dmess> JymmmEMC: " Have a Great Day.... Cause Mine has been Completely F#$%ked"
[22:01:01] <IceGuest_7> Hi cradek, thought u might like 2 know I managed to get the backlash thing sorted. Sherline are in fact selling the old EMC1 with their machines; I managed to get EMC2 installed, use the same INI file as before and got backlash to work properly. Now I just need another monitor which can display 1024x768 ;-)
[22:11:03] <Paragon37> Cradek could you recommend any good resources for learning Verilog?
[22:11:31] <SWPadnos> Paragon37, if you find one, let me know
[22:11:44] <Paragon37> LOL ;-)
[22:11:50] <SWPadnos> there's lots of code you can look at hat http;//www.opencores.org
[22:11:54] <SWPadnos> s/hat/at/
[22:12:02] <SWPadnos> ack
[22:12:22] <Paragon37> Keep up the regex ... :-)
[22:13:02] <SWPadnos> for i in opencores ; do echo
http://www.$i.org/ ; done
[22:13:26] <Paragon37> Thanks for the link...
[22:13:31] <SWPadnos> sure
[22:14:37] <Paragon37> Been having problems with pluto-step whereby the negative movement of the Y (axis.1) bleads into the Z(axis.2) real weird...
[22:14:57] <SWPadnos> I think it's pretty easy to ger your head around either verilog or VHDL, but it's not so easy to (a) really know what the languages can do or (b) have it sink in that you're writing code that will run in parallel (most of the time), not sequentially executed statements
[22:15:11] <SWPadnos> get
[22:17:13] <Paragon37> Sure ..... always @ (posedge_clk) begin .... do this stuff ...
[22:17:21] <Paragon37> ;-)
[22:18:18] <SWPadnos> yeah, but if you do something like "a=6; b=a+1; a=7;" those all execute in parallel - it's hard to know whether a will be 6 or 7, and what b will be :)
[22:19:29] <Paragon37> I reading a little about that .... need to now more ....
[22:20:51] <Paragon37> s/^(w+\s+)/$1was/ .... dam its been a while ... :-)
[22:21:45] <Paragon37> Have good one .... I'm off to me bed ...
[22:21:51] <SWPadnos> sigh
[22:21:53] <SWPadnos> enjoy
[22:22:09] <Paragon37> Thanks SWPadnos ...
[22:22:16] <dmess> inner brackets first..
[22:22:17] <SWPadnos> sure
[22:22:52] <Paragon37> forgot the first escape \ on the w+ ;-)
[22:40:23] <CIA-8> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/stepconf.py: prepare for i18n -- still a bit broken
[22:46:11] <CIA-8> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/ (stepconf.glade stepconf.py): testing on another machine, i18n of the glade file worked fine. huh.
