#emc-devel | Logs for 2009-06-08

[01:08:29] <CIA-48> EMC: 03chris * rb7bc9daddccc 10/cia: testing cia from git
[01:08:38] <CIA-48> EMC: 03chris * r7f456be3b9a6 10/cia: delete test file
[01:32:19] <CIA-48> EMC: 03chris * rba7ffd4b93a0 10/cia: remove test file again
[01:32:19] <CIA-48> EMC: 03chris * r7a2c71e40739 10/VERSION: mark version specially
[01:35:18] <cradek> is it really true that you can't find a certain commit by id in gitweb?
[01:51:34] <jepler> cradek: I think you can form the URL but I don't think there's a way to create the link using the gitweb search functionality
[01:52:27] <jepler> http://emc2.timeguy.com/gitweb?p=emc2-experimental.git;a=commit;hb=b7bc9daddccc "testing cia from git"
[12:34:31] <christel> [Global Notice] Hi all, it appears we're experiencing some issues at the data centre of one of our sponsors. We're waiting on further information and apologise for any inconvenience caused. Thank you for using freenode and have a good day.
[15:02:50] <skunkworks_> http://www.machsupport.com/forum/index.php/topic,11267.0.html
[15:34:51] <jepler> ahaha
[15:34:51] <jepler> http://www.machsupport.com/MachCustomizeWiki/index.php?title=Rs274ngc.h
[15:34:57] <jepler> #define ISNT !=
[15:35:51] <cradek> #define SET_TO =
[15:35:53] <cradek> I love it
[15:36:10] <cradek> * cradek waves at his old friend, the emc1 code
[15:47:08] <skunkworks_> heh
[15:48:31] <cradek> it was so otherworldly
[15:49:09] <skunkworks_> I thought this was interesting http://www.cnczone.com/forums/showthread.php?t=69469&page=2
[15:49:33] <skunkworks_> no floating point math - if jon is right
[15:50:42] <skunkworks_> come a long way - huh :)
[15:51:03] <skunkworks_> I am still tempted to buy some of those servos.
[15:52:25] <cradek> well I have played with git merges and rebases, and I think I'm getting comfortable with it
[15:52:38] <cradek> still trying to fight my way through jeff's three documentation URLs
[15:53:09] <cradek> what servos?
[15:54:11] <jepler> I assume he means those itty bitty servos I got
[15:58:23] <skunkworks_> yes
[16:00:13] <cradek> yeah, those are neat
[16:01:28] <jepler> with any luck, Chris is just about done designing me motor mount variations
[16:01:52] <jepler> I'll get the encoder cables wednesday, then I can make up a connector for the mesa servo boards
[16:02:33] <cradek> haha, I've only drawn about 5 variants so far, I'm just starting
[16:03:19] <cradek> jepler: if we end up with a version that takes a 1/4 ID bearing, I've got a ton of them
[16:03:48] <cradek> jepler: the big expense will be the flex couplers at $20+ each
[16:03:49] <jepler> those are the size of the bearings that the far-end of the screws sit in, too
[16:04:08] <cradek> I wish I had a feel for how it would work without them.
[16:04:13] <jepler> I still haven't measured the ODs of those
[16:04:34] <cradek> whatever mine are might be different anyway.
[16:04:44] <jepler> yeah
[16:04:50] <jepler> Is there a common OD for each ID?
[16:05:04] <cradek> I don't know
[16:05:16] <jepler> I assume a typical OD is way under 1"
[16:05:25] <jepler> 1/2 or 5/8 maybe?
[16:05:40] <cradek> the ones I have may be 1" OD
[16:05:51] <jepler> that is too big, the hole circle for the motors is 1"
[16:06:04] <cradek> hm.
[16:07:33] <skunkworks_> your thinking belt reduction?
[16:07:43] <cradek> hey the clamp-on couplings would work well on dowel pins
[16:07:53] <cradek> skunkworks_: hoping for direct drive.
[16:08:04] <skunkworks_> oh - cool
[16:08:20] <cradek> but we don't have a good feel for whether it would work ok.
[16:08:48] <jepler> skunkworks_: if you press on the end of the shaft, there's a fair bit of motion. chris wants to fix that by having a thrust bearing
[16:09:41] <skunkworks_> neat. usually they have a spring behind the bearing race to take up slack (flat spring) I would think your steppers where the same way
[16:10:28] <cradek> there's some feeling of spring, but not much at all
[16:11:00] <cradek> jepler: how many inch-pounds are the motors?
[16:11:06] <jepler> cradek: approximately zero
[16:12:06] <cradek> I'm looking at the flexible shaft couplings. they give a "peak torque" which I bet means "they will bust if you go over this"
[16:12:21] <jepler> 4 lb-in stall, .625 lb-in continuous
[16:12:22] <cradek> I was wondering how stout these would have to be to "probably" not break no matter what you do
[16:12:29] <cradek> whoah
[16:13:22] <cradek> ok, so the ones rated 12-16-28-42-67-84, which are the choices with 1/4 on one end, would be ok.
[16:13:40] <jepler> I know that the first 3 numbers are chest, waist, hips. what are the others?
[16:13:59] <jepler> (I like big couplings and I cannot lie)
[16:14:14] <cradek> there are six couplers with 1/4" on one end; those are the inch-lb ratings for all six models.
[16:14:29] <cradek> what's the motor shaft diameter again?
[16:14:57] <jepler> I think we decided this datasheet says they are within the range .1561 - .1564
[16:15:14] <cradek> = 5/32
[16:15:14] <jepler> .1565 = 5/32
[16:16:53] <jepler> incidentally, enco has 3/8, 1/2, 5/8, 3/4 OD for 1/4 ID ball bearings
[16:17:04] <cradek> http://www.mcmaster.com/#6208k131/=288r8u
[16:17:27] <cradek> hm, 1" length 1" OD
[16:17:41] <cradek> 1" OD is a problem with the 1" bolt circle too
[16:17:59] <jepler> yeah, hm
[16:18:18] <cradek> next one down is 3/4 OD
[16:18:18] <jepler> or else we're gonna end up with two plates or something
[16:19:25] <jepler> the threaded holes are #6. What OD spacers do you end up with #6 screws? Does that leave enough room with a 3/4 OD coupling?
[16:20:57] <jepler> mcmaster has 1/4" OD
[16:21:13] <jepler> (smallest OD)
[16:21:26] <jepler> so you need a coupling less than 3/4 for them to not touch
[16:22:06] <jepler> there's a 5/8, 6208K3
[16:22:59] <cradek> 12 oz-in "peak" is getting down into the worry category for me
[16:23:15] <cradek> (I've seen these things break a few times)
[16:23:31] <cradek> but 12 >> 4 I suppose
[16:23:49] <cradek> I'm thinking about the moving-full-speed-crash-into-something-hard case though
[16:24:00] <jepler> the stall current is 9A and the drivers will be operating with a 3A current limit
[16:24:08] <jepler> so I don't think I can hit 4 lb-in anyway
[16:24:33] <cradek> yeah but you've got something with at least a little mass spinning pretty fast
[16:25:51] <jepler> I see what you mean
[16:26:45] <jepler> if we start with a cheap rigid coupler instead of a flex coupler, what's the downside? have I damaged the motor in the meantime, before I decide it's a bad idea?
[16:30:11] <skunkworks_> some people have had motor shafts actually break if the alignment isn't perfect.
[16:31:43] <cradek> yeah in this case, I bet the weakest part by far will be the motor shaft, and it might flex.
[16:36:33] <skunkworks_> Have you measure the shaft? the ad says 3.5mm-7/64
[16:36:44] <skunkworks_> measured
[16:37:28] <skunkworks_> heh that comes out to 2 different sizes .137 and .109
[16:39:11] <SWPadnos> maybe it's 3.5mm minus 7/64 = 0.028 :)
[16:39:26] <jepler> pretty sure I mic'd it and it came out 5/32
[16:43:16] <jepler> of course, I can barely read micrometers and calipers
[16:43:38] <jepler> skunkworks_: huh, their stock is up to 53
[16:46:23] <skunkworks_> I saw that ;)
[16:47:55] <skunkworks_> jepler: http://www.cnczone.com/forums/showthread.php?t=58546&highlight=snapped+shaft
[16:48:31] <jepler> "oh snap"
[16:53:21] <cradek> "If you get the bellows type, I would suggest the stainless version over the aluminum type, which are really only suited for encoder loads."
[16:54:48] <cradek> hmm.
[16:55:16] <jepler> mcmaster's "high precision bellows shaft couplings" have peak torques as low as 0.7 lb-in .. they also cost a bundle
[16:55:51] <BJT-Work> this is the coupling I used on the plasma cutter from the nema 23 stepper to the Z axis shaft 6208K168
[16:56:12] <jepler> busted any of them yet?
[16:56:17] <BJT-Work> kinda pricey but works well
[16:56:19] <BJT-Work> nope
[16:56:44] <BJT-Work> I used them to connect both sides of the Y gantry too
[16:57:09] <BJT-Work> peak torque of 156 in lbs
[16:57:27] <BJT-Work> nema 23 double stack motors
[16:57:54] <jepler> lbs? or ozs?
[16:58:37] <BJT-Work> in-lbs.
[16:59:26] <BJT-Work> the coupling torque that is
[17:02:19] <BJT-Work> you can see one in this picture http://s47.photobucket.com/albums/f163/johnplctech/Plasma%20Cutter/?action=view&current=Plasma012.jpg
[17:02:51] <BJT-Work> these did not work http://s47.photobucket.com/albums/f163/johnplctech/Plasma%20Cutter/?action=view&current=Plasma004.jpg
[17:04:33] <skunkworks_> I don't understand why those are so popular (love joy style)
[17:04:43] <BJT-Work> cheap!
[17:05:01] <skunkworks_> they work great for coupling pumps and such - but servo/steppers
[17:05:04] <skunkworks_> ?
[17:05:20] <jepler> bbl, lunch
[17:12:30] <alex_joni> skunkworks_: we have such on the smaller servos on the bots
[17:12:38] <alex_joni> they work without issues
[17:42:23] <SWPadnos> thanks for noticing
[17:43:22] <alex_joni> skunkworks_: yeah
[17:43:30] <alex_joni> like in BJT's picture
[17:44:56] <alex_joni> we have both this kind: http://www.supermechengineer.net/encoder.jpg
[17:45:11] <alex_joni> and this: http://www.tantexcorp.com/products_limg/20090225203629.gif
[18:18:02] <jepler> this is the style that cradek is suggesting to me : http://www.mcmaster.com/catalog/115/gfx/large/6208kp1l.gif
[18:28:36] <skunkworks_> jepler: neat - should work great.
[18:31:05] <skunkworks_> are you going to have to bore them out to fit - or do they have the sizes you need?
[18:44:54] <skunkworks_> jepler: how much was shipping on the servos you got?
[18:45:42] <jepler> skunkworks_: let me check
[18:50:46] <jepler> skunkworks_: looks like $10 on 4 motors
[18:51:06] <skunkworks_> wow - very nice.
[18:51:29] <jepler> UPS ground
[18:56:24] <jepler> cradek: hm, mcmaster has no pulleys with 5/32 bore size that I can find
[18:58:24] <skunkworks_> do they have any with 1/8 holes?
[19:00:10] <jepler> you're thinking to bore it out? yeah, they do
[19:01:08] <skunkworks_> yes
[19:01:26] <skunkworks_> chris has the technology ;)
[19:02:46] <cradek> I'm not at all sure I have the technology to bore a deep 5/32 hole
[19:02:50] <cradek> more skill than technology
[19:03:19] <skunkworks_> wait - I thought you fixed clocks and such ;)
[19:03:55] <jepler> .563 depth
[19:08:53] <skunkworks_> http://www.use-enco.com/CGI/INSRIT?PMAKA=317-5616&PMPXNO=19499014&PARTPG=INLMK32
[19:12:50] <skunkworks_> I am sure there are less expensive ones.
[19:13:10] <cradek> skunkworks_: no holes on a watch are 1/2" deep...
[19:14:12] <skunkworks_> after seeing you guys bore out a holder on your little lathe at the last cncfest - I don't think you would have a problem.
[19:15:32] <skunkworks_> or make one out of drill rod
[19:15:56] <jepler> the whole pulley? I think I want a timing belt type, that's hard to make on a lathe
[19:16:26] <skunkworks_> I ment the boring tool.
[19:16:38] <jepler> oh duh
[19:17:22] <jepler> Notes: [2] Peak values are theoretical and supplied for reference only
[19:21:10] <skunkworks_> heh
[19:24:25] <mozmck> hey skunkworks
[19:25:03] <skunkworks_> Hello
[19:25:33] <mozmck> how is your servo drive coming? your are sam from the fest right?
[19:25:36] <jepler> cradek: I see that I was wrong about there being a separate manpage about things to configure with 'git config'
[19:26:05] <skunkworks_> mozmck: yes :) are you the texan?
[19:26:17] <mozmck> :) yes
[19:27:07] <skunkworks_> I have not touched the servo drive since I got back. (busy building a garage atm)
[19:28:11] <skunkworks_> I almost started working on the kerney and trecker but got side tracked putting a air compressor togather. (it was in the way_
[19:28:13] <skunkworks_> )
[19:29:07] <mozmck> I see. I need to fix up a good compressor too. I have an old 2 stage that I think will work if I work on it a little.
[19:29:18] <skunkworks_> http://www.electronicsam.com/images/KandT/DSCCurrent.JPG
[19:29:25] <skunkworks_> it is actually in this picture also
[19:30:25] <mozmck> Are you going to run the k&t with your servo drive?
[19:30:33] <skunkworks_> yes
[19:31:11] <mozmck> that should be interesting.
[19:31:18] <skunkworks_> yes ;)
[19:32:15] <mozmck> cradek: did you get a chance to look at my second Makefile patch?
[19:32:52] <jepler> cradek: git config --global allg "log --stat"
[19:32:57] <jepler> err
[19:33:04] <jepler> git config --global alias.lg "log --stat"
[19:33:11] <jepler> then you can 'git lg' and get --stat specified by default
[19:35:21] <jepler> cradek: http://git.or.cz/gitwiki/Aliases
[19:36:24] <cradek> thanks
[19:36:39] <cradek> I guess they don't want the equivalent of .cvsrc because it might break scripts that use these commands.
[19:36:39] <jepler> it looks like you can't automatically add flags to a builtin command, you have to choose and remember to use a different command name
[19:39:17] <cradek> skunkworks_: that tool is good for a $20 "snap" noise
[19:39:45] <skunkworks_> Heh - only if your not careful. ;)
[19:42:36] <skunkworks_> monday must be mach monday for me. http://www.cnczone.com/forums/showthread.php?t=82753
[19:43:11] <skunkworks_> I like the last post.
[19:43:14] <cradek> mozmck: I see it on the list now. It definitely looks right to me...
[19:43:59] <mozmck> ok. I spent another 15 minutes sat night and figured that out.
[19:44:21] <cradek> you stayed awake later than me then...
[19:46:01] <mozmck> not much. looks like everything else runs fine here.
[19:46:36] <cradek> ok I will get it committed for you, and thanks again
[19:47:10] <micges> http://imagebin.ca/view/mW9Z7mh.html
[19:48:39] <micges> how can it be?
[19:49:25] <alex_joni> is it homed?
[19:49:43] <alex_joni> micges: I assume you want to show that you jogged outside the machine limits
[19:49:44] <micges> http://imagebin.ca/view/SesmPKk7.html
[19:49:50] <micges> yes
[19:50:18] <micges> it's homed
[19:50:58] <micges> I bet that isn't allowed
[19:51:33] <micges> it shows message that limit are exceeded but allow jog further
[19:54:41] <cradek> micges: can you say how to reproduce this problem?
[19:55:52] <micges> run trunk sim_mm
[19:55:56] <micges> home all
[19:56:05] <micges> ang jog in -x direction
[19:56:06] <micges> thats all
[19:56:27] <jepler> happens in configs/sim?
[19:56:41] <micges> yes
[19:57:23] <SWPadnos> interesting. I can jog to -254.00 mm
[19:57:56] <jepler> it stops at -40 inches for me
[19:58:28] <jepler> that looks like it lines up with the working volume
[19:58:38] <alex_joni> MIN_LIMIT = -254
[19:58:44] <SWPadnos> Y stops at -254.00 as well, which is within the volume
[19:59:27] <alex_joni> X & Y are -254 .. 254 mm
[19:59:29] <SWPadnos> ok, interesting
[19:59:31] <jepler> -40 is my inifile minimum
[19:59:35] <jepler> inches here
[19:59:35] <alex_joni> Z is -50.8 .. 101.6
[19:59:38] <jepler> sim/axis.ini
[19:59:41] <micges> ok letme rip again
[19:59:43] <jepler> F1 F2 ctrl-home hold left arrow
[19:59:45] <alex_joni> jepler: he said axis_mm.ini
[19:59:53] <SWPadnos> it looks wrong in the perspective view, but it's correct if I look at a Z plane view
[19:59:59] <jepler> oh, I missed "mm"
[20:00:25] <SWPadnos> ah, and that's because Z is not at the bottom of the work volume
[20:01:07] <jepler> yes, sim/axis_mm.ini also works just fine. same keystrokes
[20:01:48] <SWPadnos> yep. it's working as expected here
[20:01:57] <micges> maybe bugs shows only on dapper ? ;)
[20:02:07] <jepler> dapper here
[20:02:11] <SWPadnos> I was testing on Jaunty, so could be :)
[20:02:14] <SWPadnos> nope
[20:02:42] <jepler> now, I haven't updated recently..
[20:02:57] <SWPadnos> I have - in the last couple of days
[20:03:13] <micges> I was messing with Exceed soft limit messages, ant on all my rips I suddenly can jog past soft limits
[20:03:21] <SWPadnos> though the version I'm testing with is 2.3.0_~beta1
[20:03:32] <jepler> that's a silly version to test
[20:03:33] <cradek> oh, so you broke it :-)
[20:03:59] <micges> but I have 2 clean installs to test
[20:04:00] <SWPadnos> I'm with cradek on this one :)
[20:04:14] <micges> and screens are from there
[20:04:20] <alex_joni> * alex_joni fires up a fresh dapper
[20:05:07] <jepler> I did cvs up, make, and run; it's still OK here
[20:05:40] <micges> heh ok
[20:05:48] <micges> I'm gonna trace it
[20:06:30] <cradek> micges: about your comment on the helix_length code: that code you copied is right, the theta is wrong
[20:06:48] <cradek> the problem is earlier
[20:07:23] <micges> cradek: let me look
[20:07:33] <cradek> er, circle.angle
[20:08:28] <cradek> it decides to do the full circle instead of the small arc
[20:08:54] <micges> I saw that angle is sometimes different
[20:09:06] <micges> it is 6
[20:09:07] <cradek> notice the DTG is 2 * pi * I word
[20:09:14] <micges> then 7
[20:09:25] <micges> suddenly 15
[20:09:51] <micges> to all: jogging works correnct in 2.2.9
[20:09:54] <cradek> I expect that it's 2pi + a little
[20:10:08] <cradek> 7 or 15 would surprise me
[20:10:45] <micges> maybe all params are connected somehow so then strange numbers
[20:13:56] <micges> cradek: so the problem is in pmCircleInit() ?
[20:14:26] <micges> maybe some overflow with 10^10 numbers ?
[20:18:36] <alex_joni> micges: jogging works fine here
[20:18:50] <alex_joni> dapper, with just updated 2.4.0~pre
[20:21:34] <micges> ok I found it
[20:21:39] <micges> Issuing EMC_TRAJ_SET_TELEOP_VECTOR -- (+231,+84, +25,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,)
[20:21:39] <micges> Issuing EMC_TRAJ_SET_TELEOP_VECTOR -- (+231,+84, +26,0.000000,30.383333,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,0.000000,)
[20:21:39] <micges> I
[20:21:52] <micges> those messages I get while jogging
[20:22:18] <alex_joni> with sim_mm ?
[20:22:33] <micges> yes
[20:23:25] <micges> those I get on 2.2 and they are correct: Issuing EMC_AXIS_JOG -- (+124,+24, +13, +1,30.383333,)
[20:23:27] <alex_joni> I only get jog/abort
[20:24:46] <alex_joni> are you sure you're not running some other kins maybe?
[20:27:52] <micges> searching
[20:29:21] <micges> ok found it, hidden file .axisrc was accidently copied and executed on axis run :|
[20:36:11] <micges> good night all
[20:47:31] <alex_joni> good night all
[20:48:01] <jepler> I wonder what you can write in axisrc to evade soft limits
[20:48:20] <jepler> soft limits are supposed to be enforced in task
[20:50:43] <alex_joni> jepler: right, but if it manages to send TELEOP jog messages, then task will probably do the wrong thing
[20:51:13] <alex_joni> teleop messages aren't checked against axis limits iirc
[20:51:17] <jepler> I see
[20:51:22] <alex_joni> they should be checked against joint limits
[20:51:56] <alex_joni> anyways.. g'night
[20:55:48] <jepler> see you alex_joni
[22:37:36] <CIA-48> EMC: 03cradek 07TRUNK * 10emc2/src/Makefile: two more fixes for "make install", thanks Moses McKnight