#emc-devel | Logs for 2008-02-03

[00:12:33] <jepler> cradek: isn't resetting the interpreter the only thing that makes the offsets be changed in the var file?
[00:12:39] <jepler> does touch off still work right?
[00:14:16] <cradek> it does still work right
[00:14:34] <cradek> switching modes also makes the var file get written out
[00:14:59] <jepler> oh really
[00:15:36] <cradek> I will test more after dinner, bbl
[00:20:06] <CIA-21> EMC: 03tissf 07TRUNK * 10emc2/src/po/fr_axis.po: French translation update
[00:27:29] <CIA-21> EMC: 03tissf 07TRUNK * 10emc2/docs/src/quickstart/stepper_quickstart_fr.lyx: French translation update
[00:27:31] <CIA-21> EMC: 03tissf 07TRUNK * 10emc2/docs/src/gui/image-to-gcode_fr.lyx: French translation update
[01:03:48] <fenn> uh, why not explicitly write the var file if that's what you want to happen?
[01:36:48] <CIA-21> EMC: 03cradek 07v2_2_branch * 10emc2/src/emc/usr_intf/axis/scripts/axis.py:
[01:36:48] <CIA-21> EMC: don't reset the interp before when reloading the file. This fixes
[01:36:48] <CIA-21> EMC: SF#1825638 as well as other things like loss of the G20/G21 setting
[01:36:48] <CIA-21> EMC: and other modal things from MDI.
[01:38:04] <jmkasunich> I guess that means you have some confidence it doesn't break other things?
[01:38:15] <cradek> yeah I'm fairly sure it's better this way
[01:39:10] <jmkasunich> I'm getting braver with the shoptask
[01:39:20] <jmkasunich> 1/4", 0.040 DOC, 2400 RPM, 10ipm
[01:39:29] <cradek> you've only braved yourself out of one endmill so far :-)
[01:39:38] <jmkasunich> no, that was stupided out
[01:39:48] <cradek> ah, that's different
[01:40:33] <cradek> is that in steel?
[01:40:45] <jmkasunich> lol
[01:40:45] <jmkasunich> no, aluminum
[01:40:51] <cradek> it's only 150sfm
[01:41:00] <jmkasunich> yeah, but the belts don't go any faster
[01:41:09] <cradek> oh right
[01:41:12] <jmkasunich> the 10 ipm feed is that gets the part done in less time
[01:41:14] <cradek> I know the feeling
[01:41:18] <jmkasunich> s/that/what
[01:41:42] <jmkasunich> a month ago I was cutting aluminum at 1 ipm
[01:41:50] <jmkasunich> and a week ago it was 2-3 ipm
[01:41:52] <cradek> 2 flute? that's .002/tooth, seems fairly heavy
[01:42:08] <jmkasunich> like I said, I'm getting braver
[01:42:14] <cradek> cool
[01:42:28] <cradek> you should get 3 flute mills if you can find them
[01:42:43] <jmkasunich> lots of tooling I should get
[01:42:51] <jmkasunich> for now I'm running with what I've got around
[01:43:04] <jmkasunich> I used to have some high helix 3 flute 1/2" mills
[01:43:08] <jmkasunich> really liked those
[01:43:20] <cradek> yeah they're nice
[01:43:58] <cradek> I don't know why people don't like the 3 flutes...
[01:44:07] <cradek> can't measure the diameter I guess, but who cares?
[01:44:29] <jmkasunich> well, if you ever have them resharpened, you'll care
[01:44:52] <cradek> I have some resharpened ones - they came with a sticker on them stating the diameter :-)
[01:46:43] <jmkasunich> for most hobby projects, its "I want to do this NOW, so I will use whatever I've got"
[01:46:56] <CIA-21> EMC: 03cradek 07v2_2_branch * 10emc2/debian/changelog: .
[01:46:58] <jmkasunich> if I know I'm going to be doing a lot of something, then I get the proper tools
[01:48:32] <cradek> what are you making anyway?
[01:48:45] <jmkasunich> pieces of an engineering week project
[01:48:51] <cradek> fun
[01:48:59] <jmkasunich> bearing mounts in this case - "honey, I shrunk the pillowblocks"
[01:49:00] <cradek> helping a student?
[01:49:09] <jmkasunich> no - competition at work
[01:49:33] <cradek> huh, sounds fun, I work with a bunch of programmers
[01:49:43] <cradek> no offense jepler
[01:50:01] <jmkasunich> theres gonna be a programming aspect to this as well
[01:50:10] <jmkasunich> I picked up an AVR board
[01:55:03] <jmkasunich> http://jmkasunich.com/pics/pilloblocks.jpg
[01:55:17] <jmkasunich> the large bearing is 13mm OD, small ones is 3/8" OD
[01:55:24] <cradek> nice!
[01:55:53] <jmkasunich> the fits aren't good, but thats OK
[01:56:03] <jmkasunich> I intend to slit them where the screws are
[01:56:07] <cradek> I was going to ask if you bored them
[01:56:21] <jmkasunich> that would take too long
[01:56:46] <jmkasunich> bad ehough that it took me all of last night and much of today to get this far
[01:56:58] <jmkasunich> actual machining time was less than a half hour
[01:58:15] <cradek> are they .25 thick?
[01:58:25] <jmkasunich> about 0.3
[01:58:36] <jmkasunich> made from scraps
[01:58:55] <jmkasunich> the blank had some kind of glue on one side, so I milled it down
[01:59:07] <jmkasunich> probably 0.330 or something when I started, maybe something metric
[02:01:03] <jmkasunich> oh, the arc feedrate reminded me of something I noticed today
[02:01:24] <jmkasunich> if you cut an arc using tool diameter comp (or any arc for that matter), the feed you get isn't the feed you programmed
[02:01:38] <jmkasunich> imagine a 1.000" inside arc, and a 0.750" cutter
[02:01:45] <jmkasunich> the tool moves in a 0.25" circle
[02:01:55] <jmkasunich> it moves along _that_ circle at the programmed feedrate
[02:01:57] <cradek> afaik the interp gets that right
[02:02:07] <jmkasunich> but the point that is cutting moves 4 times faster
[02:02:22] <jmkasunich> (around the 1" radius)
[02:02:38] <cradek> oh I get what you mean
[02:02:58] <cradek> you're not talking about diameter comp
[02:03:26] <jmkasunich> if you helical interp a 1.0 hole with a 0.95" cutter, the circle is only 0.050 in diameter, and cuts very quickly
[02:03:48] <fenn> yuck
[02:04:34] <cradek> but you programmed the .050 arc right?
[02:04:51] <jmkasunich> in the case I observed, yes - so maybe using radius comp corrects for that effect
[02:05:02] <cradek> I actually think it does
[02:05:26] <fenn> you were complaining earlier bout not being able to cut a hole with radius less than tool diameter using tool comp
[02:05:30] <cradek> otherwise you can't expect emc to guess what feed you mean
[02:06:03] <jmkasunich> cradek: of course in the case where I program the 0.050 I have to figure out the feed
[02:06:14] <cradek> right
[02:06:24] <fenn> you could do inverse time mode, might be easier to understand
[02:06:38] <cradek> also true
[02:06:58] <jmkasunich> well, its done now - I just hit FO during the small bores
[02:07:31] <jmkasunich> I think I did the bores at 6 ipm anyway - didn't go up to 10 till I did the outside
[02:10:06] <cradek> I was wrong - radius comp does not change the feed
[02:10:21] <cradek> only if in inverse time mode, it recalcs the feed because the arc length changes
[02:10:41] <jmkasunich> so you get the programmed feed along the tool centerline path, not the programmed path, right?
[02:10:48] <cradek> I think so
[02:11:21] <cradek> beware I only looked at the source :-)
[02:13:39] <cradek> an interested person could certainly fix this - it would be easyish
[02:14:02] <cradek> you'd have to remember the requested feed and put it back after each arc
[02:14:30] <jmkasunich> lowering the feed on an inside curve is kind of a no brainer
[02:14:45] <jmkasunich> I don't think raising it on an outside curve is a good choice though
[02:14:58] <jmkasunich> at least not always
[02:15:05] <cradek> I'm sort of uncomfortable with that position
[02:15:28] <jmkasunich> if you are milling from the solid, the center of the tool is cutting - if you raise the feed, you increase the chipload on the tool
[02:15:29] <cradek> did you see http://pastebin.ca/889665
[02:15:50] <jmkasunich> otoh, if you are cleaning up the side of something that is already cut out, you are only cutting along the programmed line
[02:15:51] <cradek> it warns about what you just said
[02:16:20] <cradek> I'm beginning to think the current behavior is right
[02:16:20] <jmkasunich> zactly
[02:16:32] <jmkasunich> even for inside curves?
[02:16:56] <cradek> assuming you pick the feed considering what you're cutting, not what is left, it is correct
[02:17:06] <jmkasunich> I guess it again depends on whether you are cutting from the solid, or cleaning something up
[02:17:11] <cradek> right
[02:17:27] <jmkasunich> if you are cutting from the solid and you slow down, you could get in trouble if you are cutting work hardening stuff
[02:17:51] <cradek> right, or burning
[02:30:24] <jmkasunich> damn - broke a tap
[02:30:57] <jmkasunich> fortunately I was able to get it out - otherwise I would have said something much stronger than damn
[02:31:17] <cradek> that's the good kind of tap breakage
[02:31:50] <jmkasunich> I have a foot switch, and I've been meaning for years now to hook it up to reverse the drill press
[02:32:09] <cradek> oh in a tapping head?
[02:32:22] <jmkasunich> as it is, I need three hands to tap on the drill - one to feed the quill, one to flip the switch, and one to hold the work
[02:32:26] <jmkasunich> no, rigid tapping
[02:32:39] <jmkasunich> the tapping head reverse automatically when you pull up on the quill
[02:33:20] <cradek> oh, I've never done it that way I guess
[02:33:37] <cradek> never had a reversing drill press...
[02:33:59] <jmkasunich> I do it a lot, when I either don't have a tapping head collet for the size I need, or when its not enough holes to be worth setting up the tapping head
[02:34:08] <jmkasunich> just chuck it up, dial in a low speed, and go
[02:34:17] <fenn> i had a reversing drill press.. if it was spinning when you hit reverse it kept going the same way
[02:34:47] <jmkasunich> this time I tried to get away without workholding - the parts were in a heavy vise, and I figured it would resist the load of tapping 4-40
[02:35:01] <jmkasunich> once the vise started to turn, that was all she wrote
[02:35:13] <cradek> eek
[02:35:19] <fenn> i push down on the quill feed and turn the belt by hand
[02:35:36] <fenn> the vise spins until the handle hits the column
[02:35:43] <jmkasunich> my drill press was modified before I got it, for very low speed work
[02:35:44] <cradek> I have a tapping handle with a stud out the top
[02:36:04] <jmkasunich> cradek: me too
[02:36:11] <jmkasunich> but power tapping is so tempting
[02:36:17] <cradek> yeah
[02:36:43] <jmkasunich> get too greedy, sometimes you pay the price
[02:36:47] <jmkasunich> I got off easy this time
[02:36:55] <jmkasunich> if that part was scrapped, I'd be really pissed
[02:37:11] <cradek> or you'd be looking for capacitors for your edm setup!
[02:37:34] <jmkasunich> really pissed while looking for capas
[02:37:37] <jmkasunich> caps
[02:37:42] <cradek> heh
[02:40:53] <jmkasunich> oh gawd
[02:41:02] <jmkasunich> #emc has a lunatic
[02:41:12] <cradek> only one?
[02:41:41] <jmkasunich> specifically a lunatic that believes in perpetual motion
[02:46:03] <cradek> meh
[02:47:26] <cradek> good thing it's not my job to convince otherwise everyone who believes silly things
[02:47:37] <cradek> I'd be very busy
[02:48:17] <jmkasunich> I've said all I'm gonna say to him
[02:48:28] <jmkasunich> sad though
[02:48:38] <jmkasunich> he's not a stupid person
[02:50:17] <cradek> bbl
[05:02:41] <cradek> back
[06:06:36] <CIA-21> EMC: 03cmorley 07cl_v7124_branch * 10emc2/src/hal/classicladder/classicladder.c: fix - init data before loading program
[17:17:15] <skunkworks> http://www.cnczone.com/forums/showthread.php?t=51493
[17:56:40] <jmkasunich> gear ratios are always subject to misunderstanding
[17:57:29] <jmkasunich> does 2:1 means two turns of the motor to one turn of the screw (small gear on the motor), or does it mean two turns of the screw to one turn of the motor (small gear on screw)
[17:57:48] <skunkworks> right - try it one way - if it's wrong -try it the other..
[17:58:07] <jmkasunich> that guy has spent more time bitching about it than it would take to try it both ways
[17:58:58] <jmkasunich> I haven't looked at stepconf in a while, so I don't know how jeff does it
[17:59:26] <jmkasunich> I'd be tempted to have it ask for "number of teeth on motor gear or pulley", and "number of teeth on screw gear or pulley"
[17:59:35] <jmkasunich> that is unambiguous
[18:01:59] <skunkworks> http://www.electronicsam.com/images/house/kitchen1.JPG
[18:02:13] <skunkworks> http://www.electronicsam.com/images/house/kitchen2.JPG
[18:02:17] <skunkworks> http://www.electronicsam.com/images/house/kitchen3.JPG
[18:02:26] <skunkworks> getting there.
[18:02:46] <jmkasunich> are the cabinets new?
[18:03:10] <skunkworks> new in the mid 90's.. we decided to re-use them. they are hickory
[18:03:35] <jmkasunich> seemed odd that you would have hung new cabinets with other work yet to be done
[18:03:38] <jmkasunich> asking for damage
[18:03:58] <skunkworks> heh. we just didn't take them down.
[18:03:59] <jmkasunich> but if they were already there, might be better to work around them than to remove and replace
[18:04:23] <skunkworks> the house is a bit kittywampus (sp) - you have to go with the angles.
[18:05:09] <skunkworks> biab
[18:05:42] <jmkasunich> never saw a square house yet
[18:05:57] <jmkasunich> if they built cars like they built houses..... we'd all be walking
[18:58:17] <cradek> http://timeguy.com/cradek-files/emc/stepconf.png
[18:58:26] <cradek> the screen clearly documents what it wants
[19:00:10] <cradek> and I think the calcs are right, when I increase the Leadscrew number, the scale goes up since the motor has to turn more
[19:03:46] <cradek> oh I see what he means I think. a larger "Leadscrew" number could mean more teeth on the pulley OR more rotations
[19:04:10] <cradek> I (and jepler) are thinking teeth, he's thinking turns, so the calcs are backward
[19:06:52] <cradek> maybe the word "teeth" could go on the screen somehow to clear it up
[19:07:00] <cradek> or "turns" and reverse it
[19:08:38] <cradek> bbl
[19:11:01] <jmkasunich> teeth are easier to count
[19:12:09] <jmkasunich> although if there is a multi-stage reduction or a closed gearbox, turns can be measured
[19:12:38] <jmkasunich> I doesn't matter which one you pick, as long as they know which one stepconf wants
[19:13:45] <jmkasunich> * jmkasunich states the obvious
[20:11:59] <jmkasunich> the AXIS source is not easy to figure out
[20:24:54] <jmkasunich> cradek: jepler: I'm trying to figure out something that I thought would be simple
[20:25:00] <jmkasunich> (maybe it is, and I'm blind)
[20:25:10] <jmkasunich> where does AXIS set the contents of the titlebar?
[20:26:07] <jmkasunich> it displays "<name-of-loaded-g-code-program> AXIS <emc-version>"
[20:26:29] <jmkasunich> and so far I've tried tracing each chunk of that with no success
[20:27:00] <jmkasunich> root_window.tk.call("set", "version", emc.version) <-- does this result in the version being in the title bar?
[20:33:06] <jmkasunich> thats the only instance of the word "version" in the entire emc/usr_intf/axis tree that isn't obviously something else
[20:34:47] <jmkasunich> likewise, I've looked at every instance of the word AXIS, and can
[20:35:03] <jmkasunich> can't find one that even remotely appears to be setting the application title
[20:41:21] <jmkasunich> _and_ looked at all calls to wm_title
[20:41:49] <jmkasunich> and looked at the "open file" process, hoping to see where the name of the current file winds up in the title bar
[20:42:01] <jmkasunich> now you know why I stick to control.c ;-)
[20:44:41] <jepler> jmkasunich: it's in share/axis/tcl/axis.tcl
[20:45:03] <jmkasunich> thanks
[20:45:39] <jmkasunich> why is there source code that isn't in the src tree?
[20:46:57] <jmkasunich> found it - proc update_title
[20:47:04] <jmkasunich> obvious when you know where to look
[20:48:37] <jmkasunich> well, my hopes of actually contributing a patch to axis are fading
[20:50:06] <jmkasunich> I was hoping to figure out how to make the title bar say '<name-of-loaded-g-code-file> AXIS <emc-version> <infile [EMC]MACHINE>"
[20:52:09] <jepler> ah
[20:52:47] <jepler> I don't think that's available in tcl at this moment, but it could be
[20:52:51] <jmkasunich> (dunno if you saw the mail on emc-users requestiong something like that)
[20:52:57] <jepler> no I haven't read my e-mail yet today
[20:53:04] <jmkasunich> I didn't see anything in axis.tcl that resembeled ini file access
[20:53:23] <jepler> I don't think there is any -- the variables are fetched in axis.py and put in tcl string variables
[20:53:41] <jmkasunich> I figured it was something like that - I saw inifile support in axis.py
[20:54:04] <jmkasunich> but the connections between axis.tcl and axis.py are too much for me - one unfamiliar language is enough
[20:54:30] <jmkasunich> I'm procrastinating anyway, I have parts to make