#emc-devel | Logs for 2009-08-13

[00:30:07] <tomaw> [Global Notice] In memory of a recently deceased good friend: http://forums.gentoo.org/viewtopic-t-787855-start-0-postdays-0-postorder-asc-highlight-.html
[00:56:55] <jepler> mozmck: no, a signal can only have one writer. motion.spindle-forward and another OUT pin cannot be connected to the same signal. You could use an 'or2' component if you want to have some signal be TRUE when either motion.spindle-forward OR the vcp button's pin is true..
[00:57:59] <mozmck> thanks, I'll play with that. how do the buttons in Axis work? send NML messages?
[01:02:12] <jepler> yes
[01:48:54] <mozmck> in axis.py there is a module "emc" imported. where would I find the source for this module?
[01:57:27] <dgarr> mozmck: src/emc/usr_intf/axis/extensions/emcmodule.cc
[02:00:43] <mozmck> ah, thanks
[06:42:48] <micges_work> hello
[08:24:36] <micges_work> hi alex_joni
[08:25:15] <alex_joni> hi micges_work
[11:58:07] <jepler> micges_work: I thought you might be interested in the work I mentioned on the mailing list a few days ago under the subject "the defer-format branch". Are you planning to comment on it?
[12:00:24] <micges_work> I planned yesterday but I forgot :)
[12:01:35] <jepler> I know how that is
[12:01:35] <jepler> bbl
[13:19:28] <skunkworks_> logger_dev: bookmark
[13:19:28] <skunkworks_> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2009-08-13.txt
[13:48:07] <skunkworks_> cradek: did you get some pictures taken?
[14:54:36] <seb_kuzminsky> yay no more 2.2 cruft! :-)
[14:59:40] <jepler> hi seb_kuzminsky
[14:59:48] <jepler> was I right or wrong about the pci stuff in hm2 not being in 2.2?
[15:00:03] <jepler> or do you not know for sure either?
[15:03:48] <skunkworks_> cradek: found some on bill's site :)
[15:05:58] <jepler> speaking of the pictures on http://geektrap.com/bill/01248711886 anyone know what the function is of the funky curvy bits on the spindle in http://geektrap.com/bill/01248711886 ?
[15:06:30] <jepler> err http://geektrap.com/bill-files/01248711886/dscn7529.resized.jpg
[15:08:34] <skunkworks_> I would guess - maybe cooling
[15:13:33] <cradek> heh, glad he's documenting it
[15:14:44] <skunkworks_> How is it going?
[15:15:30] <cradek> great, got spindle orient and tool change working now
[15:15:46] <cradek> but the spindle drive is flaky, unfortunately
[15:15:55] <skunkworks_> crap
[15:16:05] <skunkworks_> cold solder joint. ;)
[15:16:36] <cradek> my current guess is bogus capacitors
[15:16:43] <cradek> at least that's the first thing to try
[15:16:48] <skunkworks_> better yet
[15:16:49] <cradek> (man it's complicated)
[15:16:51] <skunkworks_> heh
[15:17:06] <skunkworks_> How are you orienting?
[15:17:19] <cradek> it has a mechanical orient pin
[15:17:42] <skunkworks_> oh - nic
[15:17:44] <skunkworks_> nice
[15:17:54] <skunkworks_> (that is how the k&t is)
[15:18:06] <skunkworks_> 'just works'
[15:18:13] <cradek> wait until the spindle is stopped, tell the drive to limit torque to a low value, activate the pin, override the spindle speed to a low value and turn it on, wait for the pin to drop in (switch), stop the spindle drive, change tools, pop pin out, done
[15:18:36] <cradek> yeah it orients in about one second. foolproof.
[15:18:59] <skunkworks_> cool. Like I said - with our vfd set to a low rpm - it doesn't care that it is stalled.
[15:19:17] <cradek> yeah, cool
[15:19:38] <cradek> it would only be stalled for a few ms before turning off anyway, I bet
[15:19:57] <skunkworks_> right
[15:20:07] <cradek> this one has a switch for the orient pin being out all the way too - so I mask normal spindle operation with that
[15:20:17] <jepler> in that situation I'd expect it to ramp the torque all the way up to the limit. that's what you'd want if you were making chips..
[15:20:47] <skunkworks_> or vfd is dumb - no feedback really
[15:21:09] <cradek> jepler: at only a few rpm, they have almost no power - but I do wonder what the current is doing when it's stalled
[15:21:45] <cradek> I also did the manual drawbar stuff last night
[15:22:14] <skunkworks_> wow - are you going to start retro-fitting for a living? Your progress is awesome.
[15:22:40] <jepler> oh? http://www.yaskawa.com/site/DMSpindle.nsf/(DocID)/TKUR-5EKSYH/$File/SIE-S626-1.2.pdf page 5 shows torque is highest (constant) below 1500rpm and drops at higher speeds
[15:23:10] <jepler> looks a lot like a stepper motor curve, with a constant torque at low speeds and constant power at high speeds
[15:23:14] <cradek> oh maybe I don't know what I'mt alking about then
[15:24:38] <cradek> today I hope to figure out whether the resolver is on the motor or spindle
[15:25:03] <cradek> if it's on the spindle, it can't be for commutation and also lets me use it for tapping if I get a different vfd
[15:27:31] <cradek> also I measured backlash on X,Y again - I can't see any now. I think it's better now that it's lubed up
[15:27:43] <cradek> it will be so slick to be able to make round holes without a boring head
[15:28:25] <jepler> cradek: the pulley is not toothed and not 1:1, right?
[15:28:34] <cradek> both right
[15:29:06] <cradek> hm, if the resolver isn't 1:1 I still can't use it because there would be too many indexes
[15:29:18] <cradek> (even if it's on the spindle, I mean)
[15:35:20] <seb_kuzminsky> about the curvy channel on the piece on the table here: http://geektrap.com/bill-files/01248711886/dscn7529.resized.jpg
[15:35:33] <seb_kuzminsky> it looks like a one-way fluid channel
[15:36:06] <seb_kuzminsky> the structure at the top left has much higher resistance-to-flow in the right-to-left direction than in the left-to-right dir
[15:44:42] <jepler> cradek: I don't understand how it could be on the spindle but not 1:1
[15:45:26] <jepler> I was thinking that you could look at the amplitude of one phase of the resolver while you orient it multiple times. If the amplitude is the same then its on the spindle. If it's not, then it's on the motor.
[15:45:49] <jepler> (isn't it the amplitude that varies as the resolver turns?)
[15:47:05] <jepler> oh, are you saying it could be on neither spindle nor motor (like the resolvers on your lathe axes)?
[15:47:31] <steves_logging> steves_logging is now known as steve_stallings
[15:48:47] <steve_stallings> yes, the amplitude out is the sine or cosine of the shaft angle, detectors use ratio of the two outputs so that common mode amplitude changes do not affect reading
[17:20:45] <cradek> jepler: for instance a belt to the spindle would give velocity (which I think is the reason it's on there) but give you too many indexes
[17:40:06] <steve_stallings> steve_stallings is now known as steves_logging
[20:06:13] <alex_joni> jepler: I think post 2.4.20 or so it got changed to kbuild
[20:55:17] <CIA-41> EMC: 03micges 07task_cleanup * r9fce74ba07e5 10/src/emc/task/emctaskmain.cc: Remove all shorthand NMLmsg pointers variables
[20:55:17] <CIA-41> EMC: 03micges 07task_cleanup * rd936802849f5 10/src/emc/task/emctaskmain.cc: EMC_TASK_PLAN_EXECUTE_TYPE: Remove redundant condition, use constant
[22:01:02] <CIA-41> EMC: 03micges 07joints_axes3 * rb3c8c4471b44 10/src/emc/ (5 files in 2 dirs): Add EMC_AXIS_STAT class