#emc | Logs for 2007-01-05

Back
[00:01:11] <jepler> or get 2 of these + 4 of the biggest CF cards you can afford: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=140070487817
[00:01:23] <jepler> (and a mobo that still has 2xPATA interface)
[00:01:56] <jepler> or maybe I'm wrong that CF is cheaper than a "thumb drive" these days
[00:02:38] <alex_joni> I think it's more expensive
[00:02:56] <alex_joni> seen 8G pen drives
[00:03:00] <SWPadnos> thumb drive = $22-ish for 2G
[00:03:48] <SWPadnos> I like the flash hard drives though - they're about the size of a double-high IDE connector (but a bit more expensive than the other options)
[00:04:30] <SWPadnos> http://www.globalamericaninc.com/other/dom.php
[00:04:50] <SWPadnos> oops - bbl
[00:08:59] <goslowjimbo> Several people have mentioned CVS head in their emails. I tried to access it, but the "cvs" isn't recognized by my computer. I was in what I call my home file (jsr). This version of linux was created from the EMC2 live install disc.
[00:09:13] <alex_joni> sudo apt-get install cvs
[00:10:37] <alex_joni> goslowjimbo: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?CVS
[00:12:02] <goslowjimbo> OK I tried the apt-get and it is loading now. Am I downloading a newer version of EMC2?
[00:12:15] <alex_joni> what is the command you used?
[00:12:41] <goslowjimbo> sudo apt-get install cvs
[00:12:58] <alex_joni> that only gives your system cvs capabilities
[00:13:01] <alex_joni> nothing emc2 yet
[00:14:00] <goslowjimbo> So, I'm loading a CVS routine?
[00:14:22] <alex_joni> it's a program
[00:15:02] <goslowjimbo> I see. It seems to have finished.
[00:15:35] <alex_joni> goslowjimbo: what are you trying to accomplish?
[00:19:00] <A-L-P-H-A> how does one check their ubuntu version?
[00:19:00] <goslowjimbo> Long term, create a DRO with HAL. Short term, get a CVS head update. I just looked up the second comment regarding the EMC website and CVS.
[00:19:11] <A-L-P-H-A> I can't remember if this is 6.06 or 6.10
[00:19:59] <jepler> A-L-P-H-A: cat /etc/lsb-release ?
[00:19:59] <A-L-P-H-A> damn,. this is 6.06
[00:20:08] <A-L-P-H-A> I'll keep it 6.06 for now.
[00:20:13] <alex_joni> A-L-P-H-A: less /etc/lsb-release
[00:20:31] <alex_joni> I heard rumors that 6.10 is a bit worse than 6.06
[00:21:07] <A-L-P-H-A> there are some troubles, yes... I've experienced them.
[00:21:23] <A-L-P-H-A> but I needed xorg 7.0+, cause of some driver issues...
[00:21:35] <A-L-P-H-A> freak'n multiple monitor fun.
[00:24:05] <jepler> goslowjimbo: instructions on how to download the "HEAD" version with cvs, as well as other things to prepare your system to compile emc2, are on our wiki: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_5_10_and_6_06_from_source
[00:24:08] <jepler> bbl
[00:24:46] <goslowjimbo> OK.
[00:24:51] <A-L-P-H-A> alex_joni, that simulator you showed, is that linux or windows?
[00:24:53] <A-L-P-H-A> I didn't check
[00:24:58] <alex_joni> win
[00:25:06] <alex_joni> and.. it's in german
[00:25:42] <A-L-P-H-A> I'm not german... doesn't help. :)
[00:25:46] <A-L-P-H-A> alex_joni, do you speak german as well??
[00:25:52] <A-L-P-H-A> alex_joni, what languages do you know?
[00:27:31] <CIA-8> 03alex_joni 07HEAD * 10emc2/src/hal/drivers/ (hal_stg.c hal_stg.h):
[00:27:31] <CIA-8> First go at adding index pulse homing for the STG boards.
[00:27:31] <CIA-8> This should make it work for the STG1 model, don't attempt to try this with a STG2 yet.
[00:27:31] <CIA-8> The code has been written based on the specs from the manual, without no testing whatsoever yet (except that it compiles).
[00:27:31] <CIA-8> Please let me know of problems/successes with this.
[00:28:07] <alex_joni> A-L-P-H-A: ordering by skills: romanian, german, english, french (a bit), italian (tiny bit)
[00:29:33] <A-L-P-H-A> alex_joni, and you aren't a hit with the academic girls??
[00:30:34] <A-L-P-H-A> order by skill: English (dumb), Cantonese (fluent/conversational, can't read/write), Mandarin (understandable), French (broken)
[00:30:50] <alex_joni> I also know a bit of c, c++, java, lisp, and a few assemblers
[00:30:52] <alex_joni> :-)
[00:30:59] <alex_joni> but chicks don't dig those languages :D
[00:34:44] <A-L-P-H-A> delphi/pascal, c/c++, vb, AVR asm, x86 asm, html/javascript, sql...
[00:35:09] <alex_joni> I'd learn cantonese
[00:35:47] <A-L-P-H-A> For what reason?
[00:36:12] <A-L-P-H-A> I speak it decently enough, with a huge white accent. I'd seriously rather learn Mandarin, as it's more economic potential to do business with.
[00:36:27] <alex_joni> oh, ok.. mandarin then
[00:36:51] <A-L-P-H-A> Cantonese, last 30 years, yeah. for sure... next 50 years, Mandarin.
[00:40:53] <A-L-P-H-A> ahh... that was painless to get mysql on ubuntu to run.
[00:41:01] <A-L-P-H-A> http://ubuntuforums.org/showpost.php?s=3b915fd2adfe3c0b9205ab8eb9417ae6&p=73448&postcount=7
[01:06:07] <mdynac> good evening, well ithink we have a problem with adaptive feedrate....
[01:06:51] <alex_joni> what kind?
[01:07:00] <cradek> uh-oh, I'm not here then
[01:07:05] <cradek> darnit!
[01:07:07] <cradek> I always mess that up
[01:07:14] <A-L-P-H-A> cradek... welcome.
[01:08:34] <mdynac> well when i cut the aluminun, the feedrate didn't vary too much at all, so the part came out fine, however when i cut steel, the feedrate is constantly varying and my did not come out to correct size
[01:09:08] <jepler> is the voltage constantly varying?
[01:09:47] <mdynac> it actually cut about 1.26 inch and the display was reading .8906, it also moves progressively slower and slower.....
[01:10:00] <jepler> if you want it to vary more smoothly you could involve a component like limit2/limit3 or lowpass
[01:10:01] <mdynac> yes the voltage is always varying
[01:10:29] <mdynac> ??
[01:11:26] <alex_joni> this sounds like missing steps
[01:12:13] <mdynac> yes it does, but in a servo system?
[01:12:17] <jepler> maybe I should let you explain the problem more thorougly. but I mean, if you think the problem is that the voltage is noisy then one option is to use a software filter to remove some of the noise. "lowpass" performs a sort of average over the voltage, much like an rc system
[01:12:42] <alex_joni> * alex_joni heads to bed
[01:12:49] <alex_joni> good night all
[01:13:28] <mdynac> it is already filtered, i am using the original cirtcuitry....
[01:13:59] <cradek> maybe you have noise on the encoder lines?
[01:14:30] <mdynac> that is possible
[01:15:17] <cradek> I don't think adaptive feed would make the move short or long. I'd be pretty darn surprised
[01:15:35] <cradek> I'm going to hook up a siggen sine wave to adaptive feed and run a program
[01:15:47] <jepler> cradek: that's an interesting idea indeed
[01:16:06] <cradek> at the end of the program, the gui showed the right location, right?
[01:16:17] <mdynac> hmmm a sinewave is too consistant
[01:16:55] <mdynac> no, i stopped it cause it was so far off
[01:17:27] <cradek> ok
[01:17:53] <cradek> what kind of shape was the gcode?
[01:18:48] <mdynac> an mdi command x -2.0"
[01:19:48] <cradek> ok that's really simple
[01:19:59] <mdynac> sure was
[01:20:01] <skunkworks> mdynac: are you running a motec board?
[01:20:16] <mdynac> yes
[01:20:46] <cradek> you could use halmeter to check your raw encoder counts before/during/after the cut
[01:21:08] <cradek> I suspect you will find incorrect counting
[01:21:30] <mdynac> k
[01:21:32] <cradek> would a cut on steel generate much more RF than alum?
[01:23:09] <mdynac> i would think so, the steel was 1" tall and the AL was about 3/8", so tall part = longer spark....
[01:23:33] <cradek> ok
[01:23:40] <jepler> what code turns on adaptive feed?
[01:23:44] <cradek> I was just going to ask that
[01:24:03] <mdynac> but i must add that the machine hardly ever varied in feedrate while cutting the AL
[01:24:10] <mdynac> M52P1
[01:24:25] <skunkworks> I think it was daveE that had issues with motec and encoder counting... I don't think he ever figured it out. figured it was noise.
[01:25:05] <cradek> I hooked up a 42Hz triangle wave to adaptive feed, and it still goes right to X=1 when I MDI g1x1
[01:25:29] <mdynac> do it for an hour or so......
[01:26:19] <mdynac> and instead of Hz , try Khz......
[01:26:24] <cradek> mdynac: only if you test your encoder counts like I suggested :-)
[01:27:07] <skunkworks> http://sourceforge.net/mailarchive/message.php?msg_id=37458094
[01:27:08] <mdynac> k i will test it tomorrow....i am going to totally reinstall from scratch.....
[01:27:10] <cradek> are they differential and/or shielded?
[01:27:18] <cradek> no, reinstalling is the wrong thing to do
[01:27:31] <cradek> you need to troubleshoot instead -- this isn't windows
[01:27:44] <mdynac> differential encoder outputs, yes
[01:27:49] <jepler> http://emergent.unpy.net/files/sandbox/random-adaptive-feed.png
[01:28:31] <cradek> jepler: that's so beautiful
[01:28:38] <jepler> a component is supplying a random number from 0 to 1 for adaptive feed rate, changing on average every 50ms or so (the squarish wave)
[01:29:09] <jepler> the curving wave is emc staying within the accel constraints (the scales don't quite match up, because one is in % of requested and the other is in inches per second)
[01:29:18] <jepler> this is a small portion of a 1" move at F1
[01:30:17] <mdynac> i realize it is not windows, but why can't i run emc head version out of my home directory if i first remove the ubuntu emc2.0.5
[01:30:47] <cradek> you can, but the 205 package supplies a couple configuration files you would have to reproduce
[01:31:00] <mdynac> i c
[01:31:09] <cradek> so it's easiest to just use the package
[01:31:22] <mdynac> k
[01:31:50] <cradek> but please don't waste time reinstalling. I think you have a hardware problem and I'll help you troubleshoot it.
[01:32:30] <skunkworks> thats so windows ;)
[01:32:34] <mdynac> okay, but it is a little late for not re-installing, hell it's no big deal anyhow, i backed up all my configs....
[01:32:36] <jepler> .22 inches done out of 'g1x10f.1' .. man I am not patient
[01:32:54] <A-L-P-H-A> how does one show the amount of free space left?
[01:32:57] <cradek> df
[01:33:06] <A-L-P-H-A> thanks
[01:33:41] <cradek> skunkworks: vital systems = motenc?
[01:33:47] <cradek> skunkworks: in dave's message I mean
[01:34:02] <jepler> though I'm not sure what this will prove on the simulator
[01:34:33] <cradek> jepler: it was nice to see that feed rate response...
[01:34:58] <jepler> cradek: halscope is so cool
[01:34:59] <cradek> a TP bug could actually cause it to end at the wrong point, or jump at the end of the move - you would see it
[01:35:08] <cradek> jepler: no kidding
[01:35:14] <jepler> (so's python's hal module, for letting me write 'random' so easily)
[01:35:23] <jepler> cradek: do you think I should try a smaller move, rather than a larger one?
[01:35:31] <jepler> and capture the end of it?
[01:36:10] <cradek> his move was twice as long as expected partway through... you will easily see it in the backplot if you have the same problem
[01:36:36] <cradek> mdynac: did you understand what I said about how to read and check your raw encoder counts?
[01:37:10] <skunkworks> cradek: yes http://www.vitalsystem.com/web/motion/motionLite.php?PHPSESSID=7cd075e2c2326310f5215005032739e0
[01:37:11] <mdynac> well i was gonna figure it out, but please enlighten me....
[01:37:49] <cradek> at X0, before you start the move, hook up halscope to motenc.X.encoder-0-rawcounts or whatever it's called, and write down that number
[01:38:21] <mdynac> k
[01:38:22] <cradek> now run your move, and when you measure about 1" of motion on the machine (with your ruler) stop it and get the rawcounts reading.
[01:38:49] <cradek> now figure out how many counts it SHOULD have had for an inch move (this is your input scale)
[01:39:26] <cradek> wait now I'm confused :-)
[01:39:40] <cradek> let me think
[01:42:07] <cradek> I think you will have counts missing (the difference in your two readings will be less than your input scale by a substantial amount)
[01:42:25] <cradek> and the display on emc will show the move has gone much less than an inch
[01:42:50] <cradek> skunkworks, jepler: am I right?
[01:43:22] <jepler> cradek: yes, if it is noise then the encoder "raw counts" will agree with the DRO but not with the ruler
[01:43:35] <cradek> right
[01:43:39] <jepler> because "raw counts" and the DRO are both on the same side of the hypothetical encoder noise
[01:43:47] <jepler> that is, the DRO in your emc gui (tkemc, axis)
[01:44:23] <cradek> and he will know the counting is wrong if he sees it move a (real) inch without an inch's worth of counts
[01:44:33] <cradek> that's the basic test we're doing
[01:45:04] <mdynac> okay, got it....
[01:45:16] <cradek> good, I barely do :-)
[01:45:23] <A-L-P-H-A> damn this server is loud as hell
[01:46:03] <cradek> that would tell us it's noise, or a motenc bug, or a motenc driver bug.
[01:46:25] <jepler> yes, that result could have any of those as its cause
[01:46:25] <cradek> or heck it could be a gnat's whisker in one slot in your encoder
[01:46:36] <mdynac> that email from dave does not look too good for me.....however the machine runs just fine when i am in manual mode and execute g-code
[01:47:14] <skunkworks> is it cutting in manual mode?
[01:47:17] <mdynac> i can draw very complex parts all day with the drawing table...
[01:47:19] <cradek> MDI works exactly like a one-line gcode program. I don't think that's a relevant difference
[01:47:35] <cradek> with no electrical noise going on?
[01:47:39] <jepler> do gnats have whiskers?
[01:47:40] <jepler> ??
[01:47:41] <jepler> ???
[01:47:41] <mdynac> correct...
[01:47:55] <cradek> jepler: yes
[01:48:16] <cradek> mdynac: to test for a bad encoder you could do the same test in Y only
[01:48:29] <mdynac> k
[01:48:42] <jepler> gnat's piss, Noun. A particularly weak and tasteless drink.
[01:48:47] <jepler> huh didn't know about that either
[01:48:51] <cradek> ha
[01:49:28] <mdynac> encoders, no prob, i have 20 brand new ones...the ones on the machine are original......maybe i will change them just for gp.
[01:49:58] <cradek> please troubleshoot first!
[01:50:13] <cradek> we should do this in a systematic way since there are so many variables
[01:50:24] <mdynac> i will.....
[01:50:47] <mdynac> i'll do the count thingee first...
[01:50:52] <cradek> ok good luck :-) we'll nail it no matter where it is
[01:51:15] <mdynac> kewl
[01:54:17] <jepler> mdynac: your following error is much less than 1.26-.89 inches, right?
[01:54:40] <wb9mjn> Yea...Cradek is right, the encoder may be fried, but might not be the cause of being fried...Thus re-
[01:54:49] <wb9mjn> placing it will just fry the new good one...
[01:55:13] <jepler> you might want to switch between "commanded" and "actual" position in the gui too, to make sure it's not that pid's taken the day off, and the discrepancy is just ferror
[01:55:33] <cradek> what are the ferror limits?
[01:55:41] <jmkasunich> put halmeters on command and feedback
[01:55:51] <cradek> hi jmk
[01:55:57] <jmkasunich> hi
[01:58:47] <jmkasunich> wow, lots of people must be looking at my website
[01:59:05] <jmkasunich> 950 meg sent, only 768 received
[01:59:16] <jmkasunich> usually its 10:1 the other way
[01:59:23] <cradek> your images are largeish
[01:59:36] <jmkasunich> yeah, right out of the camera
[01:59:50] <jmkasunich> bout 400K, but only if people click on them
[01:59:56] <jmkasunich> the thumbs are 20K or so
[02:00:08] <jmkasunich> I should probably make the thumbs smaller yet)
[02:01:03] <cradek> dave's message about the motenc makes me feel a little uneasy for mdynac
[02:03:08] <jmkasunich> I don't recall any problems with the motenc on the mazak
[02:03:20] <jmkasunich> but thats not an EDM with sparkin' all around
[02:03:25] <cradek> is that's what's counting encoders?
[02:03:27] <cradek> that
[02:03:29] <jmkasunich> yep
[02:04:12] <cradek> motenc lite?
[02:04:24] <jmkasunich> I think its a lite
[02:04:28] <jmkasunich> ray probably remembers
[02:04:39] <jmkasunich> could be the full size motenc
[02:04:45] <cradek> I haven't seen ray around much lately
[02:07:04] <cradek> jepler: did your move ever finish?
[02:07:24] <jepler> cradek: the 10" one? I stopped it after awhile
[02:08:05] <jmkasunich> It takes a special kind of patience to do edm I think
[02:08:11] <cradek> oh I figured it was 1"
[02:08:28] <jepler> no I did a f.1 x10 move -- it would be 100 minutes or something
[02:08:46] <jepler> I'll start another one, but with servo_sim instead
[02:08:49] <jepler> f.1x1 or something
[02:10:11] <cradek> that's probably a pretty decent test
[02:11:55] <jepler> starting a new one with servo_sim
[02:11:57] <jepler> g1x1f.1
[02:12:12] <cradek> remember the M5mumble Pmumble
[02:13:43] <jepler> yes, I did
[02:13:43] <cradek> wow, 581KB/s downloading the firefox update
[02:23:53] <jepler> my 1" move is at .6
[02:26:07] <jepler> looks like the hal-level encoder stuff agrees with the onscreen
[02:27:06] <cradek> good
[02:30:57] <cradek> jmkasunich: you should redirect your / to your blog
[02:31:12] <cradek> (that's what I did)
[02:39:51] <A-L-P-H-A> I think I need a new mouse... it keeps dying on me like it's dead from the hub.
[02:39:53] <A-L-P-H-A> maybe it's hub.
[02:40:08] <jmkasunich> cradek: how do I do that?
[02:40:09] <A-L-P-H-A> cradek, a meta refresh would do the trick.
[02:40:17] <A-L-P-H-A> jmkasunich, a meta refresh would do the trick
[02:40:40] <cradek> Redirect /index.html http://timeguy.com/cradek/
[02:40:51] <cradek> I have this in a .htaccess in the top directory
[02:40:55] <A-L-P-H-A> that's a .htaccess.
[02:40:58] <A-L-P-H-A> but another would be.
[02:40:58] <A-L-P-H-A> <META HTTP-EQUIV=Refresh CONTENT="10; URL=http://www.htmlhelp.com/">
[02:41:47] <A-L-P-H-A> inside a html file.
[02:42:04] <A-L-P-H-A> don't forget to turn on directives allow all in your httpd.conf if .htacess isn't working
[02:42:22] <cradek> jmkasunich: I'd ask A-L-P-H-A, I think he knows about this stuff more than I do
[02:47:09] <jmkasunich> I'm lazy
[02:47:23] <jmkasunich> there's a link.... let them click
[02:48:59] <A-L-P-H-A> haha
[02:49:18] <A-L-P-H-A> it's not hard at all.
[02:49:20] <jepler> 32777 s32 RO 6000 encoder.0.rawcounts
[02:49:25] <jepler> the encoder ended up at the right spot
[02:49:54] <jmkasunich> I'd be shocked if it didn't
[02:51:27] <jepler> well me too really
[02:51:34] <jepler> now we just need sim-digital-noise
[02:53:38] <cradek> sim-mdynac
[02:54:53] <SWPadnos> sim-analog-noise-coupled-into-digital-lines
[02:57:42] <mdynac> ??
[02:58:05] <SWPadnos> everything appears to work fine with simulated hardware :)
[02:58:29] <A-L-P-H-A> :)
[02:58:42] <mdynac> but no motenc board.......
[02:58:48] <A-L-P-H-A> I haven't eaten in aobut 20hrs.
[02:58:55] <A-L-P-H-A> probably not good for me at all.
[02:59:22] <SWPadnos> no, but what that says is that it appears that the motion controller and adaptive feed system work correctly, which tends to point further down the chain for the error
[02:59:45] <SWPadnos> so it could be the driver, the motenc board, the wiring, the encoder, or the environment
[02:59:50] <SWPadnos> or some conbination
[02:59:56] <SWPadnos> combination, too
[03:00:32] <jepler> http://www.youtube.com/watch?v=GQ3AcPEPbH0&NR
[03:01:07] <jepler> (youtube needs a "play at 2x" mode, but that video is still cool even though it gets long by the end)
[03:01:25] <SWPadnos> it seems a bit long ~30 seconds in :)
[03:01:47] <mdynac> well the environment is pretty harsh, rfwise......an edm will totally wipe out the AM broadcast band within say 50 feet, thats how i can tell when my machine stops....the radio clears up!!!
[03:02:42] <mdynac> no problem o fm tho....
[03:03:08] <cradek> thanks mr. armstrong
[03:04:27] <A-L-P-H-A> watching
[03:06:12] <jepler> SWPadnos: for you, it needs a x16 mode then
[03:06:48] <A-L-P-H-A> damn.
[03:07:41] <SWPadnos> heh
[03:07:53] <SWPadnos> it got better at around the 40 second mark (after the guy stopped pushing buttons)
[03:08:07] <A-L-P-H-A> yeah... I speed up till near the end,
[03:08:10] <cradek> that's pretty cool.
[03:08:10] <A-L-P-H-A> when it was putting on tires.
[03:08:11] <jmkasunich> thats gotta be thousands of dollars worth of mindstorm stuff
[03:08:16] <A-L-P-H-A> probably more
[03:08:20] <A-L-P-H-A> lego is EXPENSIVE.
[03:09:30] <jepler> yeah a small fortune
[03:10:46] <jepler> 'night guys
[03:10:56] <jepler> wow I am tired early tonight
[03:10:57] <A-L-P-H-A> http://copyfight.corante.com/archives/2005/11/19/canada_lets_mega_bloks_build_on_expired_lego_patents.php
[03:11:01] <A-L-P-H-A> that's an interesting read.
[03:11:06] <A-L-P-H-A> horray for some Sanity in Canada.
[03:13:30] <SWPadnos> yay
[03:13:58] <A-L-P-H-A> so I'll buy my kids, when/if I have them, mega bloks.
[03:14:05] <A-L-P-H-A> cheaper ver of lego!
[03:14:12] <A-L-P-H-A> cross compatability.
[03:16:28] <skunkworks> I had a lego mind storm set - pretty damn cool.
[03:16:38] <A-L-P-H-A> lucky bastard.
[03:16:48] <A-L-P-H-A> mindstorm came out when I was in university... I think.
[03:17:02] <A-L-P-H-A> for $400, I could have spent on other stuff... which I did.
[03:17:10] <skunkworks> I someone made a C like language interface for it. That is what I used
[03:17:11] <A-L-P-H-A> now I play with real robots... cnc style!
[03:17:17] <skunkworks> 200 dollars
[03:17:25] <A-L-P-H-A> not when it came out.
[03:17:29] <A-L-P-H-A> and 200 was the tiny kit.
[03:17:42] <A-L-P-H-A> not bad with all the sensors and stuff I guess.
[03:18:00] <A-L-P-H-A> but IR sensors... common, you can rip open a $5 mouse to get those.
[03:18:29] <A-L-P-H-A> kinda want to build a silly IR robot...
[03:18:50] <A-L-P-H-A> 4 wheels, two sensors, travelling a circle drawn with a thick marker.
[03:18:55] <A-L-P-H-A> add an atmel.
[03:19:08] <skunkworks> :)
[03:19:13] <A-L-P-H-A> one sensor on either side of the marker.
[03:19:32] <A-L-P-H-A> if the ir sensor hits the marker line, the other wheel would speed up.
[03:19:55] <A-L-P-H-A> causing it to follow the line...
[03:19:58] <A-L-P-H-A> at least that's the logic...
[03:20:10] <A-L-P-H-A> wonder how hard it would be to get it to work.
[03:32:32] <fenn> what a waste of legos
[03:33:28] <ejholmgren> yeah, I'd rather build a spaceship
[03:34:01] <ejholmgren> ... with CO2 lasers
[03:34:09] <SWPadnos> I don't think legos are spaceworthy
[03:34:16] <fenn> sure they are
[03:34:24] <SWPadnos> as a ship, anyway ...
[03:34:28] <fenn> there's tons of lego spaceship designs
[03:34:35] <A-L-P-H-A> and theyfly too!
[03:34:42] <A-L-P-H-A> I remember making helicopters when I was a kid.
[03:34:45] <SWPadnos> oh, you mean a spaceship *model* ;)
[03:34:49] <A-L-P-H-A> with the mechanized lego stuff.
[03:35:38] <ejholmgren> we made ships that slid down the railing by the stairs
[03:35:54] <ejholmgren> the one that exploded into the most individual pieces at the bottom won
[03:37:23] <fenn> in college we had to assemble a whole mars colony from scratch with nothing but 4 rcx bricks and some kit
[03:38:00] <fenn> the delivery rockets would just drop stuff at random
[03:41:29] <jmkasunich> wow... peak inrush current for this 48V 10A power supply is 80A
[03:41:57] <SWPadnos> that's a lot
[03:43:08] <jmkasunich> combination of transformer inrush and cap charging...
[03:45:30] <SWPadnos> oh - I stuck a couple of 4-ohm 23A ICLs in your box
[03:45:35] <SWPadnos> I hope you don't mind :)
[03:46:05] <jmkasunich> why thank you
[03:47:34] <SWPadnos> 0.065 ohms at 1/2 current (11A)
[03:47:48] <SWPadnos> http://www.ametherm.com/Data%20Sheets/SL32%204R023.pdf
[03:47:51] <SWPadnos> it seemed like a good one
[03:48:10] <jmkasunich> 7.8W at 11A
[03:48:25] <SWPadnos> yeah
[03:48:26] <jmkasunich> not bad
[03:48:52] <SWPadnos> I figured 2 in series (one on each leg) would be about right for the inrush - 15A until it heats up
[03:49:12] <jmkasunich> even 1 per leg should be OK
[03:49:14] <jmkasunich> 30A
[03:49:30] <SWPadnos> you mean one on the primary and one on the secondary?
[03:49:44] <jmkasunich> one on each side of the 240V line
[03:49:49] <SWPadnos> I was thinking one on each side of the primary
[03:49:55] <SWPadnos> ah - I'm going to be on 120V :)
[03:50:08] <jmkasunich> I am too, for now
[03:50:08] <SWPadnos> though I should think about using the 220 line
[03:50:18] <jmkasunich> so I'll use one in the L1 leg, and nothing in the neutral
[03:50:35] <jmkasunich> but eventually I'll move to 240V, then I'll use one in L1 and 1 in L2
[03:50:58] <jmkasunich> the cap on this supply charges to about half voltage in one half cycle
[03:51:14] <jmkasunich> and is about 90% charged in 4 half-cycles
[03:51:40] <SWPadnos> ok - that's not bad. the ICL is only rated for 23A though (and it's very low resistance when hot - 20 mOhm
[03:51:42] <SWPadnos> )
[03:52:23] <jmkasunich> peak primary current is about 80-90A, so the peak secondary current is about 220-250A
[03:52:27] <SWPadnos> eek
[03:52:49] <SWPadnos> what portion of that inrush is just saturating the coil?
[03:53:03] <jmkasunich> its hard to say
[03:53:13] <jmkasunich> I suppose I could disconnect the cap
[03:54:12] <jmkasunich> but I'm lazy
[03:54:18] <SWPadnos> heh
[03:54:31] <jmkasunich> actually I probably should do that
[03:54:41] <SWPadnos> so - do C-rings always deform when you remove them, or is it just me?
[03:54:57] <jmkasunich> it might make more sense to do inrush limiting between the bridge and the cap
[03:55:16] <jmkasunich> I sometimes get them off without damaging them, but not always
[03:55:18] <SWPadnos> well, Robin pointed out that the saturation current of the primary can also be a significant factor
[03:55:25] <SWPadnos> ok
[04:01:01] <jmkasunich> transformer inrush is significantly less than cap charging inrush
[04:01:16] <jmkasunich> maybe 5A peak on the first cycle
[04:01:23] <jmkasunich> and assymetrical
[04:01:25] <SWPadnos> ok - that's not bad
[04:01:51] <jmkasunich> each subsequent cycle gets 20% lower, takes 20 cycles or more to die out
[04:02:30] <SWPadnos> I bet that limiting on the primary side is better though - the size ICL you'd need on the secondary would have higher hot resistance, and would dissipate loads more power
[04:06:16] <SWPadnos> well, if you don't feel like building your own, here's a small supply that should provide for a lot of torque: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=320044377286
[04:06:36] <SWPadnos> at HGR even :)
[04:14:40] <ejholmgren> _only_ $7k?
[04:15:17] <SWPadnos> cheap - some of the ones I've worked on for them are $1000000 plus
[04:15:24] <SWPadnos> actually, lots plus
[04:15:39] <skunkworks> we had quite a few of these... Nice parts http://cgi.ebay.com/Parametrics-Parajust-AC-Motor-Speed-Control-Model-6010_W0QQitemZ200064705874QQihZ010QQcategoryZ71393QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[04:15:43] <skunkworks> for tinkering
[04:16:11] <skunkworks> most of them had holes in a few of the transisters.
[04:16:51] <skunkworks> why we have so many 600v 1200uf caps - or there abouts
[04:17:26] <SWPadnos> how much do those weigh?
[04:18:50] <skunkworks> no clue - got them from the local junk yard iirc. used most of the boxes for controls for other machines/.
[04:18:55] <jmkasunich> 600V?
[04:19:04] <jmkasunich> thats very high
[04:19:13] <jmkasunich> most electrolytics top out at 450-500V
[04:19:31] <skunkworks> I don't remember - could be 300v - the ones I am using for the amp
[04:19:51] <skunkworks> h-bridge
[04:20:00] <jmkasunich> I think you said those were 300V
[04:20:05] <skunkworks> ok
[04:26:15] <skunkworks> could have swore I had a picture of one on my site.. guess not
[04:28:34] <skunkworks> Time for bed - night guys.
[04:28:39] <jmkasunich> goodnight
[04:29:12] <jmkasunich> SWPadnos: I was thinking about DC side inrush limiting if I use a resistor and bypass contactor instead of a PTC
[04:29:42] <SWPadnos> sure - that would work too, of course
[04:29:48] <jmkasunich> you can arrange it so that the same resistor that charges the bus at power up discharges it on power down
[04:30:13] <SWPadnos> depends on how ou want to control the contactor
[04:30:18] <SWPadnos> s/ou/you/
[04:30:18] <jmkasunich> yeah
[04:30:32] <jmkasunich> I'm probably over-engineering things as usual
[04:31:55] <jmkasunich> I intend to interrupt power to the 48V DC bus (steppers) and the 330V DC bus (spindle) on an estop
[04:32:19] <jmkasunich> it would be nice if both buses very quickly got discharged in that case
[04:33:38] <SWPadnos> plus the added benefit of absorbing returned energy
[04:33:44] <jmkasunich> right
[04:33:52] <jmkasunich> the 48V bus has 22000uF
[04:34:00] <jmkasunich> the 330V bus will be somewhere between 2000 and 4000uF
[04:35:06] <jmkasunich> the 330V bus will have a pair of 100ohm 200W resistors that can be dropped across it by a 30A darlington for dynamic breaking
[04:35:28] <jmkasunich> so I guess I could just arrange for the darlington to turn on when the estop opens
[04:35:42] <SWPadnos> for the 330V bus: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=270075265897
[04:35:55] <SWPadnos> for the 48V bus: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=130065451755
[04:36:13] <jmkasunich> he doesn't list the ohms - dumbass
[04:36:19] <SWPadnos> it's 70 ohms
[04:36:32] <jmkasunich> yeah, but why should you have to look that up in the pic?
[04:36:37] <SWPadnos> or possibly 7, if that's a unmber and exponent (in one of the photos)
[04:36:38] <SWPadnos> heh
[04:36:46] <jmkasunich> its 70 ohms
[04:37:08] <SWPadnos> I had some of the 6.8ohm ones at CNC workshop
[04:37:10] <jmkasunich> (we have simular (slightly smaller) resistors at work, they don't mess with the exponent thing
[04:37:16] <SWPadnos> I'm thinking about buying more - they sold pretty well :)
[04:37:42] <SWPadnos> hdb Electronics also has some 95W ones, 10 ohm I think
[04:37:53] <jmkasunich> 150mS time constant across my 22000uF cap.... not bad
[04:39:23] <SWPadnos> the 70-ohm one has a peak power of 1.55kW at 330V, so that should be fine
[04:40:24] <jmkasunich> yeah, but the two 100 ohm ones I already have
[04:40:37] <SWPadnos> ah
[04:41:30] <jmkasunich> I also have a bunch of 50W metal cased wirewounds (the old mil style), 33 ohms
[04:41:58] <jmkasunich> 4 or 5 of those in parallel will discharge the 48V supply right quick, without taking up too much panel space
[04:42:23] <jmkasunich> only 2" long not counting leads
[04:43:28] <SWPadnos> oh - that's not bad
[04:43:46] <SWPadnos> the 220W one is ~8 or 9 inches long and an inch or so in diameter
[04:44:02] <SWPadnos> though it could be mounted vertically if the cabinet is deep enough
[04:44:21] <jmkasunich> that sounds about right - 300W is a foot ling and one inch diameter
[04:44:32] <jmkasunich> those are a pain to mount too
[04:44:43] <SWPadnos> heh
[04:44:50] <jmkasunich> these just take two screws
[04:44:56] <SWPadnos> right
[04:45:19] <jmkasunich> I'm trying to figure out the overall power distribution
[04:45:35] <jmkasunich> I'll have an unswitched (except for main disconnect) 120/240V bus
[04:45:43] <jmkasunich> that will run the PC, monitor, etc
[04:45:58] <jmkasunich> then there will be the main estop contactor, feeding the switched 120/240V bus
[04:46:05] <jmkasunich> rectify that to make the 330V dc bus
[04:46:19] <jmkasunich> tranform and rectify to make the 48V dc bus
[04:46:35] <jmkasunich> I wonder what if any other power supplies I'll need
[04:47:32] <SWPadnos> do you expect to run all contactors (coolant, cother cooling, etc) off the 48VDC bus, or do you need 24V?
[04:47:54] <jmkasunich> probably need 24V
[04:48:18] <jmkasunich> 24Vdc and 120Vac are the most common coil voltages
[04:48:22] <SWPadnos> right
[04:48:25] <jmkasunich> 12vdc is more common than 48vdc
[04:48:36] <SWPadnos> your servos will be at 48V?
[04:48:44] <jmkasunich> steppers, but yes
[04:48:50] <SWPadnos> right - ok
[04:48:53] <jmkasunich> its actually more like 54 unloaded
[04:49:08] <SWPadnos> with steppers, it's unlikely to be unloaded much :)
[04:49:14] <jmkasunich> not true
[04:49:28] <jmkasunich> they run constant current, but its PWMed
[04:49:33] <SWPadnos> isn't idle current reduction usually down to 1/2 current or so?
[04:49:40] <SWPadnos> true
[04:49:41] <jmkasunich> when they're not spinning the motor voltage is low
[04:49:58] <jmkasunich> 6A x 2.1V = 12 watts idle dissipation, plus iron loss from the PWM
[04:50:47] <jmkasunich> I should figure out how much power the PC uses - and how much 12V will be left over from that supply
[04:51:28] <SWPadnos> what size PC supply?
[04:51:55] <jmkasunich> 200W
[04:52:04] <SWPadnos> oh, then you may not have lots to spare :)
[04:52:19] <SWPadnos> anything in the 350W and up category would
[04:52:27] <jmkasunich> I'm only running a P3 mobo, matrox video card, and a single hard drive
[04:52:38] <jmkasunich> not some tricked out gaming box
[04:53:20] <SWPadnos> you may be OK. the total power draw for my touch panel PC (celeron 500, single HD, 512M) is ~56W
[04:53:23] <SWPadnos> including the LCD
[04:56:10] <jmkasunich> hard disk draws 1.75A of 12V during spin up
[04:56:20] <jmkasunich> (about 1.5 seconds)
[04:56:33] <SWPadnos> are you thinking of getting 12V relays instead of 24V?
[04:56:44] <jmkasunich> dunno yet
[04:56:45] <SWPadnos> (or - what goes on the external 12V bus)
[04:56:55] <jmkasunich> 12V is good for fans and such
[04:57:08] <SWPadnos> true - PC fans and the like inside the cabinet
[04:57:15] <jmkasunich> right
[04:57:25] <jmkasunich> and any analog stuff I want for the spindle drive
[04:57:41] <SWPadnos> I'd keep machine related stuff off the PC supply - the PC isn't meant to deal with inductive loads on its supply rails
[04:57:47] <SWPadnos> (even a contactor)
[04:58:25] <jmkasunich> good point
[04:58:37] <jmkasunich> I have a pile of various linear supplies around
[04:58:47] <jmkasunich> lamba, power-one, etc
[04:58:59] <SWPadnos> yeah - a little DIN supply or whatever seems like a much cleaner solution
[04:59:12] <SWPadnos> or an open-frame supply like those
[04:59:18] <A-L-P-H-A> huh?
[04:59:23] <A-L-P-H-A> what are jabbering about?
[04:59:24] <jmkasunich> hard disk seems to draw about 100mA steady state from the 12V rail, with peaks of 3/4A during seeks
[04:59:37] <SWPadnos> A-L-P-H-A, we're jabbering about CNC control cabinet design
[04:59:40] <A-L-P-H-A> oh
[05:00:36] <jmkasunich> and about 500-600mA of 5V
[05:00:48] <SWPadnos> surprising
[05:00:52] <jmkasunich> so thats 3-4 watts steady state
[05:01:05] <jmkasunich> why?
[05:01:11] <SWPadnos> funny that there are HD-based USB drives that are powered entirely from the bus
[05:01:19] <SWPadnos> it just seems like a lot for the logic
[05:01:28] <SWPadnos> USB 160G drives ...
[05:01:56] <SWPadnos> (I got one at Costco, but returned it because it didn't work with my main machine)
[05:01:58] <A-L-P-H-A> :) yummy... made a ubuntu LAMP box... unloading some services from my workstation to that ubuntu box.
[05:02:05] <A-L-P-H-A> hopefully it'll become a myth box as well.
[05:02:33] <jmkasunich> SWPadnos: yellow is +12 and red is +5, right?
[05:02:39] <SWPadnos> yep
[05:02:46] <jmkasunich> just makin' sure
[05:02:49] <SWPadnos> I think so anyway ;)
[05:02:55] <A-L-P-H-A> jmkasunich, get a volt meter!
[05:02:58] <jmkasunich> the mobo seems to draw about 150mS at +12
[05:03:12] <SWPadnos> with the CPU doing what?
[05:03:29] <SWPadnos> I'm not sure if the P3 has the CPU running off the 12V rail
[05:03:50] <A-L-P-H-A> How does one remotely access xserver? Is it possible with win32 connecting to a linux server?
[05:04:16] <SWPadnos> is xserver something, or are you talking about a remote X client?
[05:04:20] <SWPadnos> err - server
[05:04:22] <jmkasunich> I don't think so, there is only one yellow wire in the bundle, and no separate 12V connector on the mobo
[05:04:29] <SWPadnos> ok
[05:04:37] <A-L-P-H-A> SWPadnos, I want my win32 to be a remote X client.
[05:05:01] <SWPadnos> hmmm - it may use 12V - I think the PPro did (I remember the voltage regulator modules)
[05:05:13] <SWPadnos> A-L-P-H-A, actually, it's a server :)
[05:05:21] <A-L-P-H-A> you fucker.
[05:05:22] <SWPadnos> but cygwin/x works well for me
[05:05:22] <A-L-P-H-A> :P :)
[05:06:05] <SWPadnos> it's backwards - the display device is the server, the program (possibly running on a separate machine) is the client (to the display)
[05:06:27] <A-L-P-H-A> lets speak of this as VNC.
[05:06:30] <A-L-P-H-A> VNC server, is the host machine
[05:06:36] <A-L-P-H-A> VNC Viewer is the client machine
[05:06:37] <SWPadnos> I've never tried VNC, so let's not :
[05:06:39] <SWPadnos> )
[05:06:41] <SWPadnos> :)
[05:06:42] <jmkasunich> SWPadnos: running a make clean; make, no change in the 12V current
[05:06:48] <SWPadnos> ok
[05:06:57] <A-L-P-H-A> jmkasunich, did you already get your MESA?
[05:07:08] <SWPadnos> the P3 probably doesn't have too many effective power reduction schemes anyway
[05:07:23] <jmkasunich> A-L-P-H-A: no, probalby either saturday or monday
[05:08:47] <jmkasunich> there are 4 red (5V) wires, I can only put two at a time in the current probe
[05:09:11] <jmkasunich> running about 2A in those two, so probably 4A total
[05:09:46] <jmkasunich> interesting, it drops from 2A to about 0.7A for 5-10-20mS at a time, then jumps back up
[05:09:53] <jmkasunich> must be the CPU going idle
[05:10:59] <jmkasunich> about an amp and a half in the brown (3.3V) wires
[05:11:12] <jmkasunich> there are two of those, I'm measuring both
[05:13:05] <SWPadnos> well, it sounds like 9 + 2.5 + 20 + 5 + 2 W, or ~30W on the DC side
[05:13:29] <SWPadnos> I measured the 56W my machine took with an AC power meter, so it included losses and the LCD/backlight
[05:14:03] <jmkasunich> I'm skinning the outer jacked off of a spare IEC power cord so I can clamp the probe around an AC conductor
[05:14:22] <jmkasunich> not gonna be able to accuratly compute watts tho, because of the nonlinear waveform
[05:14:23] <SWPadnos> heh. there's also the kill-a-watt meter and friends :)
[05:15:18] <jmkasunich> ain't got one of them
[05:15:25] <jmkasunich> do got a cord and a pocket knife
[05:15:37] <SWPadnos> ok. I think I'd like to try going to sleep before 3 AM tonight, so I think I'll give it a try now :)
[05:16:09] <SWPadnos> see you later. good luck with the power measurements (htough I'd say that even a 200W supply has ample margin at this point :) )
[05:16:29] <jmkasunich> yeah
[05:16:47] <SWPadnos> wow - still 48 degrees here - unbelievable
[05:16:50] <jmkasunich> specifically, even during spinup the 12V load never goes over 2A, and the supply is rated for 6A on that rail
[05:19:29] <jmkasunich> 15A inrush peaks when the PC supply is plugged in
[05:32:01] <ejholmgren> ouch ...
[05:32:14] <ejholmgren> that would explain the price of the kit I got
[05:32:25] <ejholmgren> ssr's are pricey
[08:18:27] <big_chris> Can anyone advise how to remove the 'real-time' kernal so that I can get back to the kernel I was using before EMC installation?
[08:52:26] <Dallur> big_chris: did you compile a new kernel with RT for EMC ? What distro are you running ?
[09:36:08] <paragon36> Morning All
[09:36:31] <Dallur> morning
[09:43:52] <paragon36> Dallur is there a way to display optical encoder counts by number? Reason being I suspect that the etch servo is missing some counts as it servo apears to not alway end up where I expect it to?
[09:53:57] <Dallur> paragon36: hmm isn't there a signal you can watch with halmeter ?
[09:54:24] <Vq^_> Vq^_ is now known as Vq^
[10:43:11] <paragon36> Does the ubuntu Live EMC2 V2.0.5 have the Pluto-P driver?
[10:54:18] <lerneaen_hydra> 'LO
[10:54:22] <lerneaen_hydra> err
[10:54:24] <lerneaen_hydra> 'lo
[10:56:16] <fenn> 'hi
[11:01:27] <paragon36> Hello Lerneaen! Does the ubuntu Live EMC2 V2.0.5 have the Pluto-P driver?
[11:02:31] <A-L-P-H-A> lo
[11:04:49] <fenn> doubt it
[11:04:55] <fenn> pluto seems kinda dumb to me
[11:05:40] <fenn> why not use an fpga that has programming support under linux, and whose company isnt as much of a bunch of dickwads who hate hobbyists/students/amateurs/small businesses
[11:06:31] <paragon36> I was recommended to try pluto last night so I just commited to buying it
[11:07:05] <paragon36> This whole fpga is new to me!
[11:07:08] <fenn> well, you probably wont be doing the development on it, so it shouldnt matter to you
[11:07:53] <paragon36> Can you recommend a Linux supported fpga?
[11:08:43] <fenn> well, they all suck really
[11:08:55] <fenn> but just recently xilinx released webpack under linux
[11:09:05] <fenn> and supposedly they aren't as assy
[11:10:27] <paragon36> $60+ = Pluto not the cheapest chip on the block
[11:12:27] <lerneaen_hydra> paragon36: you'll probably need to run off of emc head to run the pluto
[11:13:10] <paragon36> emc head is that the latest cvs version?
[11:13:14] <lerneaen_hydra> yeah
[11:14:16] <paragon36> Can it be installed along side the ubuntu live version (2.0.5)?
[11:14:23] <lerneaen_hydra> yeah
[11:14:25] <lerneaen_hydra> that's not a problem
[11:15:47] <paragon36> Thanks lerneaen Ill go get'em now :-)
[11:25:32] <paragon36> what does the --enable-run-in-place directive represent in ./configure ??
[11:29:04] <paragon36> get this error is this because I have installed from the live cd? (configure: error: version.h not found - Is the kernel headers package installed ?)
[11:29:27] <fenn> it means you dont have to install it to be able to run it
[11:30:04] <fenn> paragon36: sudo apt-get build-dep emc2-axis
[11:30:19] <paragon36> ok
[11:30:31] <fenn> then sudo apt-get install libpth
[11:30:51] <fenn> er, libpth-dev
[11:31:19] <lerneaen_hydra> fenn: libpth?
[11:31:33] <lerneaen_hydra> since when did you need that and what does it do?
[11:32:07] <fenn> oh, it might be for the simulator
[11:32:13] <lerneaen_hydra> paragon36: run in place makes the emc2 cvs version run from that current folder rather than installing itself like a normal program
[11:32:34] <paragon36> apt-get install libpth-dev
[11:32:36] <paragon36> <lerneaen_hydra>fenn: libpth?
[11:32:54] <paragon36> ooops cut n paste muck up ;-)
[11:32:54] <fenn> portable threads library
[11:34:01] <paragon36> OK they are both installed run the ./configure again?
[11:34:18] <lerneaen_hydra> paragon36: you'll probably want run-in-place
[11:34:58] <paragon36> OK lerneaen ... I was just being lazy .... ;-)
[11:35:55] <lerneaen_hydra> right ;)
[11:36:58] <paragon36> That looks better! currently running the make .... thanks again Chaps!
[11:37:38] <paragon36> Much nicer then the emc1 install :-)
[11:38:58] <fenn> no shit
[11:39:19] <fenn> tell that to paul corner
[11:40:14] <paragon36> Why did it require the build-dep emc2-axis?
[11:41:51] <fenn> build dep gets all the source files and dev libraries (header files etc)
[11:42:15] <fenn> its like, the future and stuff, man
[11:43:01] <paragon36> Cheers Fen .... ok it has compiled sucessfully I ran make setuid how do I run it from the command line?
[11:43:50] <fenn> ../scripts/emc
[11:43:55] <paragon36> oh need to run scripts right?
[11:43:58] <fenn> if you are in src/
[11:44:23] <paragon36> Just saw that in the README .... I should know better!
[11:47:17] <paragon36> Its fired up!!! This is the future Man ....Far Out ....
[11:48:31] <paragon36> Cant see anything in the start up menu that refers to pluto though?
[11:50:06] <fenn> its just a hal component
[11:50:15] <fenn> you'll have to read the docs, i have no idea how it works
[11:50:55] <fenn> read the manual section on hal first
[11:52:50] <paragon36> Ok No problem Fenn .... Thanks for your assistance Fenn / Lerneaen ... Off to read all about HAL !
[11:55:28] <lerneaen_hydra> paragon36: you could also do as I do, which is stay here and bug/whine/poke the devs that know about it (jepler mostly, maybe some more), usually after enough whining they'll tell you what you're wondering ;)
[11:55:46] <lerneaen_hydra> (sorry devs ;) )
[13:28:47] <skunkworks> probably recycled http://www.zeronews-fr.com:80/flash/the-incredible-machine.php
[13:31:10] <jepler> paragon36: there aren't any sample configurations for the pluto-servo yet. skunkworks may have a test .hal file that he can share with you.
[13:33:01] <jepler> paragon36: you can start with one of the similar servo board configurations (like configs/stg, and specifically stg_motion.hal in that directory) and change the pin names to the pluto equivalent. for example, stg.0.position-scale is pluto-servo.encoder-0-scale (I think) and stg.0.dac-value is pluto-servo.pwm-0-value (I think)
[13:33:56] <jepler> one of the documentation files (pluto_servo.9.html) gives the names of all the pins and parameters that are available
[13:34:50] <skunkworks> Yes - my hal file is pretty simple. Will post it later today.
[13:46:12] <Guest166> good day gents..
[13:46:21] <skunkworks> Hello
[13:48:35] <Guest166> does anyone know where I can get indepth docs on classicladder / toolchanger info
[13:53:14] <skunkworks> hmm - you could look in the mazak setup
[13:53:19] <skunkworks> ok maybe not
[13:53:24] <skunkworks> :)
[14:02:42] <paragon36> Cheers Jepler / Skunkworks just got back from lunch and read your replies.
[15:29:54] <goslowjimbo> I am trying to load EMC-HEAD from CVS. WHen I do a build, I get this message: Linking python module emc.so
[15:29:54] <goslowjimbo> /usr/bin/ld: cannot find -lGL
[15:29:54] <goslowjimbo> collect2: ld returned 1 exit status
[15:29:54] <goslowjimbo> make: *** [../lib/python/emc.so] Error 1
[15:30:22] <goslowjimbo> What do I need to do to resolve this?
[15:30:33] <jepler> try installing libglu1-mesa-dev and libgl1-mesa-dev
[15:30:47] <goslowjimbo> OK
[15:38:23] <alex_joni> hi guys
[15:39:33] <goslowjimbo> Now I get can't find -lXmu. What is that? Did I skip a step somewhere? I'm following the instructions on the 'net.
[15:39:58] <SWPadnos> did you apt-get build-dep emc2 emc2-axis ?
[15:40:58] <alex_joni> g++,make,libc6-dev,tcl8.4-dev,tk8.4-dev,libgtk2.0-dev,pciutils-dev,libncurses-dev,gettext,libxaw7-dev,libreadline5-dev,lyx-qt,python,tetex-bin,latex2html,yapps2,python2.4-dev,libglu1-mesa-dev,libgl1-mesa-dev,libgnomeprint2.2-dev
[15:41:43] <alex_joni> along with gcc-4.0 for dapper, and linux-kernel-2.6.15-magma and linux-headers-2.6.15-magma
[15:41:53] <goslowjimbo> No, I used sudo apt-get build-dep emc2.
[15:42:42] <mdynac> how does one get these packages WITHOUT using apt-get??? i have no network connection to my emc box.....
[15:42:41] <SWPadnos> I'm not sure what build-dep emc2-axis adds (other than python stuff, probably), but you should probably do that as well
[15:43:36] <SWPadnos> that's a very good question
[15:43:40] <goslowjimbo> OK. alex_joni: are you saying I should make sure I have the list you generated?
[15:43:45] <alex_joni> goslowjimbo: build-dep emc2 helps only partly, because it refers to the 2.0.x version of emc2, I pasted a full list of packages above, yo might want to apt-get install all of those
[15:44:32] <SWPadnos> you can download debs and put them on a CD, but that won't add them to the repository lists without some work (making the CD like a repository)
[15:44:38] <goslowjimbo> Wow, ok. Can I just use 1 apt-get instruction and several of these files at once?
[15:44:53] <SWPadnos> yes - all of them (but no commas)
[15:45:10] <SWPadnos> actually, it may allow commas ...
[15:46:43] <goslowjimbo> I'm sure I have G++, make, and several of the others. I downloaded them on the Synaptic Package Manager. Do I need to get them again?
[15:46:58] <SWPadnos> no. it'll skip any that you already have
[15:49:05] <mdynac> so where can i find these debs, been looking around but i can't seem to locate them
[15:49:38] <SWPadnos> they won't be on your system. where are you looking for them?
[15:50:00] <mdynac> ubuntu site....etc, repos...
[15:50:23] <SWPadnos> can you get network access to the EMC2 machine?
[15:50:43] <mdynac> not here, i gotta take it home with me...
[15:50:47] <SWPadnos> ok
[15:51:16] <SWPadnos> if you run synaptic, you'll be able to search for each of those and see that they're in the repos
[15:52:33] <mdynac> well i was hoping to find a directory somewhere out there that has the debs in it, kinda like what you have at the linuxcnc site....
[15:53:04] <SWPadnos> is your net-connected machine running ubuntu?
[15:53:18] <mdynac> one of 'em
[15:53:36] <SWPadnos> ok. and does it have the emc2 repository listed?
[15:53:42] <mdynac> yep
[15:53:57] <SWPadnos> ok. there's a "download-only" option for apt-get. I don't remember exactly what it is
[15:54:11] <mdynac> okay, i know what ya mean
[15:55:03] <SWPadnos> the problem with it is that the emc2 machine won't necessarily know everything it needs to know about the packages (assuming you burn them to CD and carry them over)
[15:57:21] <mdynac> i'll just drag this box to my house and get them that way.....was hoping to get this done today, but i gotta wait till next week now.....damn i should have kept bdi on this box, easy to update without network...
[15:57:49] <SWPadnos> they're about the same, with the exception that BDI includes the compilers by default
[15:58:12] <SWPadnos> they both use apt, and they both need network connections to apt-get anything that isn't on the CD
[15:58:38] <cradek> neither has anything like a full development environment required for building emc
[15:58:44] <cradek> you simply can't fit that on one CD
[15:58:49] <mdynac> well i did get more stuff off the ubuntu/emc disk when i selected build-essential, but emc wants tcl-dev also
[16:00:23] <cradek> mdynac: any luck checking your encoder noise?
[16:00:44] <mdynac> not quite yet
[16:01:31] <cradek> working on a new machine too?
[16:01:54] <goslowjimbo> Thanks guys. It looks like emc2-axis had everything else I needed to finish the build.
[16:02:33] <mdynac> there was something seperatly wrong with my installation, so i re-installed ubuntu and now i cannot build emc-head because i cannot install packages needed to build without an internet connection, which i do not have at the shop...
[16:03:31] <mdynac> so i must drag the box home, do the updates and drag it back on monday and then test the encoders....
[16:03:50] <cradek> what was wrong?
[16:05:11] <mdynac> not sure, it was running like shit for a while, booting way slow and locking up.....
[16:05:26] <mdynac> runs perfect now...
[16:05:47] <cradek> huh
[16:06:24] <mdynac> memtest ran fine overnight
[16:06:59] <mdynac> well i gotta get back to work, talk with ya on monday about the encoder stuff
[16:07:07] <cradek> ok
[16:07:53] <paragon36> reading the HAL doc and i noticed it mentions Erector set as an example. Out of interest is this what we called Mechano here in the UK it sounds like it?
[16:09:09] <cradek> Like Meccano, it consists of collections of small metal beams with regular holes for nuts, bolts, screws, and mechanical parts such as pulleys, gears, and small electric motors. Erector beams have flanges, which make them more sturdy than those of Meccano.
[16:09:16] <cradek> http://en.wikipedia.org/wiki/Erector_Set
[16:10:03] <paragon36> Thanks Cradek
[16:14:13] <paragon36> Makes interesting reading ..... i guess kids like to play computer games all day ..... they don't now what there missing out on! :-)
[16:28:40] <skunkworks> capsella, erector set, legos and tinker toys where my child hood memorys
[16:29:11] <jepler> mostly legos here
[16:29:19] <jepler> I had some tinker toys but I didn't think they were so great
[16:30:59] <skunkworks> right - I loved the expert builder sets - I always wanted the lego chassis.
[16:32:45] <skunkworks> http://cgi.ebay.com/Technic-set-8860-Auto-Chassis_W0QQitemZ250068174474QQihZ015QQcategoryZ19012QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[16:33:31] <jepler> I could never afford the technic stuff. instead I blew my allowance every time I could afford one of the little $3 (I don't actually remember the price) kits with about 20 pieces inside
[16:33:57] <alex_joni> legos and playmobil here :)
[16:35:23] <skunkworks> I think I only had 1 expert builder set - it was a dragster - had an 4 cylinder engine in it and a differential - rack and pinion stearing
[16:35:56] <skunkworks> when I was older - my girlfriend at the time bought me a mindstorm set.
[16:40:10] <skunkworks> I remember drooling over this also http://cgi.ebay.com/1982-LEGO-ad-Expert-Builder-Dune-Buggy_W0QQitemZ230046992517QQihZ013QQcategoryZ1186QQrdZ1QQssPageNameZWD1VQQcmdZViewItem
[16:42:36] <skunkworks> ok - this is what I had when I was really young. http://www.peeron.com/scans/8847-1/1
[16:43:33] <skunkworks> that concludes the trip down memory lane
[16:43:47] <paragon36> Lego / meccano and one hell of an a imagination so i@m told :::::)
[16:45:58] <paragon36> my ubuntu is stuck in CAPS mode i cant switch it off. i have to hold down the shift key to type lowercase ... weird!
[16:46:16] <paragon36> and annoying
[16:49:52] <jepler> paragon36: bizarre
[16:51:47] <paragon36> I maybe the fact that i have ssh tunneled the xvnc'd back to my work computer. ill check when i get home. :-)
[16:52:40] <paragon36> What did i just say .... lol
[16:53:30] <SWPadnos> I have no idea
[16:53:44] <skunkworks> http://www.cnczone.com/forums/showthread.php?p=238311#post238311
[16:53:46] <skunkworks> make sense?
[16:55:45] <paragon36> Not to newbie ... sorrry.
[16:55:57] <paragon36> Not to this newbie ... sorrry.
[16:57:16] <jepler> it does to me but then cradek's expertise has rubbed off on me
[16:57:31] <SWPadnos> it makes sense to me
[16:58:21] <cradek> great post, with a link to the wiki even
[16:59:31] <cradek> the wiki page explains why you might or might not want that behavior
[17:00:48] <jepler> reportedly, some competing program uses an exact stop mode after a g0; does the wiki page say why we didn't choose to do it that way?
[17:01:00] <jepler> (I don't see where it does)
[17:01:59] <cradek> we could definitely do that, I never considered it
[17:02:14] <jepler> oh I thought I had a memory of discussions about it
[17:02:21] <jepler> I just couldn't remember the way the discussion went
[17:02:29] <cradek> skunkworks: locodave's real problem is that he has [TRAJ]MAX_ACCELERATION=1.0 (inch/s2)
[17:02:43] <jepler> I think it went like this: if you want exact stop mode, ask for it. you can get an exact stop wherever you like by programming a zero-time dwell (g4p0)
[17:03:05] <cradek> skunkworks: so he gets a big blend by default because his accel is so low
[17:03:15] <cradek> jepler: well those are too very good points.
[17:03:19] <cradek> two even
[17:03:28] <cradek> I can see me saying both of them :-)
[17:04:10] <cradek> I suppose the spec doesn't say anything about this.
[17:05:35] <jepler> cradek: I doubt it
[17:05:56] <cradek> skunkworks: if he wants axis accels of 5 in all cases, TRAJ accel should be 8.66 (5 * sqrt 3)
[17:07:13] <skunkworks> really? ok I will add that.
[17:08:02] <cradek> I think he's only seeing this as a problem because accel is (much) lower than he intends
[17:09:26] <jepler> I am trying to figure out whether pluto-servo does the right thing with min-dc and max-dc. suppose min-dc=.3 and max-dc=.8. what is the proper duty cycle for value=-.1? down@30%? What about value=0?
[17:09:45] <cradek> ouch
[17:10:13] <cradek> what's the use of min-dc?
[17:10:26] <cradek> I mean that in the "what is it used for" sense
[17:10:45] <SWPadnos> we had had discussions about using exact stop between G0 and non-G0 moves
[17:10:56] <jepler> I know that max-dc is so that h-bridges with built-in charge pumps don't get their high side drive caps discharged
[17:11:21] <SWPadnos> I think the reason is that it's assumed that aG0 move is a non-cutting move, so blending it with a cutting move may not make sense
[17:11:57] <jepler> it'll discharge completely if you're at 0% DC as well as at 100% DC
[17:12:05] <jepler> .. I think. maybe I'm mistaken.
[17:12:11] <SWPadnos> cradek, some drives won't generate any output unless you apply some minimum voltage (like needing to get above a diode drop or something)
[17:12:21] <cradek> seems that 0 is a value that all H bridges would want to support
[17:12:57] <SWPadnos> consider a srive that takes a 4-20 mA input. <4 means there's a broken wire
[17:13:01] <SWPadnos> s/srive/drive/
[17:13:52] <skunkworks> my drives only discharge the bootstrap cap at 100%duty cycle
[17:16:31] <jepler> so for this hypothetical 4-20mA drive you would want up@30% for value=0
[17:18:37] <skunkworks> I would think for a value of 0 the pwm would be 0 then when any level is commanded - it would jump to 30% initially
[17:23:25] <awallin> do you have some pwm drives that you are going to use with the pluto ?
[17:23:52] <jepler> awallin: I have the l298 that I used with my etch-a-sketch
[17:23:56] <jepler> awallin: cradek's also using pluto with l298
[17:24:18] <jepler> skunkworks has his IR driver/mosfet board
[17:24:40] <awallin> ok, but those are pretty low power amps...
[17:25:20] <skunkworks> awallin: mine should do 20 amps continuous and 44a peak
[17:26:15] <jepler> the pluto board supports the pwm+direction mode which should allow it to work with jon elson's PWM servo amplifier
[17:28:54] <awallin> skunkworks: nice, is your design on the net somewhere?
[17:29:46] <jepler> l298 is 46V/4A continuous per package -- entirely adequate for a sherline servo retrofit, for instance.
[17:30:06] <skunkworks> http://www.cnczone.com/forums/showthread.php?p=238321#post238321
[17:30:13] <skunkworks> awallin: is that you?
[17:30:21] <awallin> yeah ;)
[17:30:34] <skunkworks> I don't think there is a 61.1 any more
[17:30:47] <awallin> really?
[17:31:34] <jepler> I think it's accepted but I don't think it does what the documentation says
[17:31:37] <awallin> I'll have to try G61.1 and see what happens
[17:32:55] <awallin> jepler: is the l298 design on your homepage?
[17:33:41] <jepler> awallin: it's pretty much right out of the datasheet, but jas ..
[17:34:07] <jepler> http://emergent.unpy.net/projects/01148303608
[17:34:32] <SWPadnos> re 4-20 mA stuff: a command of 0 means 4 mA. those systems are used when you want to detect a wiring fault, so any value below ~3mA or so will be considered a system error
[17:34:49] <jepler> with pluto you would hook UP0 to INPUT1, DOWN0 to INPUT2, UP1 to INPUT3, DOWN1 to INPUT4, and some digital out to ENABLE_A and ENABLE_B
[17:35:24] <lerneaen_hydra> btw, what's the difference between exact stop and exact path mode?
[17:35:48] <jepler> (that digital out hooked to the axis.N.amp-enable-out)
[17:37:20] <jepler> lerneaen_hydra: the theory of exact path mode is that if the two motions are colinear (for instance, you pre-blend by adding arcs between every line--no sharp corners) then you can follow the exact given path without stopping
[17:37:39] <jepler> that's what "exact path mode" would be, but I don't think it works like that in emc2
[17:38:00] <lerneaen_hydra> does exact stop stop between a line and an arc? (if the arc is tangent to the line)
[17:38:16] <SWPadnos> exact stop will stop at every endpoint
[17:38:26] <SWPadnos> even if they're sub-segments of a line
[17:38:40] <cradek> you can get virtually exact path (stop only when necessary) with G64 P[tiny]
[17:38:39] <lerneaen_hydra> oh, so basically it's the most crude type of TP
[17:39:56] <awallin> how can I halscope the axis velocity? do I have to do the speed calculation from motorp-pos-cmd myself?
[17:40:01] <cradek> as far as I know, no version of EMC has ever done what the spec says regarding exact path
[17:40:16] <lerneaen_hydra> haha
[17:40:20] <cradek> awallin: use a derivative (ddt) block, see the sim configuration for an example
[17:40:28] <lerneaen_hydra> sounds perfect
[17:40:35] <awallin> in sim, is ddt.0 the velocity of axis 0
[17:40:44] <lerneaen_hydra> is the current behavior documented somewhere?
[17:40:51] <SWPadnos> it's in the code :)
[17:40:53] <jepler> lerneaen_hydra: you mean, besides the source code? :-P
[17:41:03] <lerneaen_hydra> oh. right
[17:41:06] <jepler> the wiki page is pretty good.
[17:41:10] <lerneaen_hydra> :p
[17:41:31] <jepler> I'll give you some of the leftover ribbon from the christmas packages if you convert that into .lyx documentation
[17:42:40] <lerneaen_hydra> seeing as how I've never done stuff in latex, maybe not. :p
[17:43:19] <jepler> the lyx gui is just like a bad microsoft word clone .. if you can use wordpad.exe you can handle it
[17:43:46] <lerneaen_hydra> wait, lyx isn't latex based?
[17:43:59] <cradek> awallin: probably, but look at the hal file to be sure :-)
[17:45:00] <awallin> jepler: which wiki page are you talking about?
[17:45:15] <cradek> TrajectoryControl
[17:45:18] <lerneaen_hydra> bbl
[17:49:14] <awallin> heh, I do get pretty strange behaviour when a joint accel is set way below the traj accel...
[17:49:42] <cradek> strange right or strange wrong?
[17:50:28] <awallin> cradek: the tool does not reach the programmed position before the next move begins. even in g61 or g61.1 mode
[17:50:55] <cradek> screen shot?
[17:52:06] <awallin> hm, let me explain once more: this is seen when in [TRAJ] ACC=20 and in [AXIS_0] ACCEL = 0.2
[17:52:29] <awallin> so perhaps it's not a bug, it's a feature?
[17:52:48] <cradek> let's talk after lunch...
[17:53:05] <cradek> screen grab of your results would be great
[17:53:06] <cradek> bbl
[17:53:07] <awallin> ok
[17:53:23] <rafa> hello friend
[17:54:42] <rafa> somebody knows as to start emc direct
[17:54:42] <awallin> hi rafa
[17:54:46] <rafa> hi
[17:55:20] <rafa> without needing to choose the configuration?
[17:55:56] <rafa> i have one configuration and a want start with
[17:56:36] <rafa> it could help me?
[17:57:01] <awallin> scripts/emc yourconfig.ini
[17:57:38] <rafa> it could say me as?
[17:58:26] <awallin> i don't understand
[17:58:38] <rafa> ok
[17:58:52] <rafa> em miconfig.ini?
[18:00:28] <tomp> example: (begin in the head of your emc2 ) scripts/emc configs/sim/servo_sim.ini
[18:00:56] <rafa> write this in terminal?
[18:01:06] <tomp> try it
[18:01:25] <rafa> or make one shortcut?
[18:01:41] <tomp> try simple first
[18:01:44] <awallin> if you want an icon to click on, you can make a shortcut on the desktop
[18:06:04] <rafa> ok thanks
[18:06:30] <rafa> someone know to start axis with maximized?
[18:07:42] <awallin> rafa: no, ask jepler or cradek...
[18:08:11] <awallin> anyone know if the amount of samples in halscope can easily be increased ?
[18:08:17] <awallin> or do I just have to lower the sample rate
[18:09:40] <rafa> ok
[18:09:43] <rafa> thanks
[18:12:12] <skunkworks> it depends on which thread you use - and what the period for said thread is .. I think
[18:13:23] <rafa> ?
[18:13:32] <awallin> skunkworks: yes, but the maximum number of samples is still just ca 16k
[18:13:52] <rafa> ?
[18:14:02] <skunkworks> ah - I understand your question. I don't know ;)
[18:14:36] <rafa> ok
[18:14:53] <rafa> thanjs skumkworks
[18:15:06] <rafa> do you know someone?
[18:15:53] <tomp> awallin: in the code, beware of 'sample_len' ...it's really the number of channels , havent found your limit yet
[18:18:06] <rafa> ?
[18:20:38] <awallin> tomp: do you know what halscop does during pre-trig ?
[18:21:17] <awallin> halscope
[18:21:54] <awallin> tomp: ah, I got it, nevermind
[18:24:27] <tomp> awallin: no... but for 1st Q... the limit is shm ("disp->end_sample = ctrl_shm->rec_len - 1" )
[18:26:38] <awallin> tomp: ok, so it's not that easy to change then
[18:26:44] <tomp> nope
[18:26:53] <tomp> ask JMK tho...
[18:30:36] <rafa> ?
[18:30:40] <ve7it> ve7it is now known as LawrenceG
[18:37:07] <tomp> scope_shm_size 65000
[18:51:50] <tomp> state is enum PRE_TRIG if you want to search
[18:53:56] <ejholmgren> I found heaven today ...
[18:53:59] <ejholmgren> http://www.ax-man.com/
[18:54:12] <ejholmgren> they have a location 15 mins from my house
[18:54:53] <ejholmgren> aisles of surplus motors, electronics, etc dirt cheap
[18:56:09] <ejholmgren> even with the flu, I got stuck in there for over an hour
[18:56:35] <Goslowjimbo> I rebuilt emc as emc-head successfully, but I can't get the command "pyvcp /home/jsr/pydro.xml" to work. It keeps not finding pyvcp.
[18:59:31] <Goslowjimbo> Since I'm on Ubuntu, I have to got to root/etc, and type init.d/realtime start. Do I go to a different directory for emc-head?
[19:00:32] <awallin> cradek: if you are interested in the traj stuff at low accel, I just made a wiki page about it: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Behaviour_At_Low_Accel
[19:00:48] <awallin> goslow: run . scripts/emc-environment
[19:01:06] <awallin> goslow: and since yesterday, the pyvcp syntax is: pyvcp -c compname myfile.xml
[19:01:19] <awallin> goslow: the pins created will be named compname.xxx
[19:02:02] <Goslowjimbo> What is compname supposed to represent?
[19:04:32] <awallin> goslow: if you create LEDs, numbers etc, they will be named compname.led, compname.number, etc.
[19:08:04] <awallin> goslow: did you get pyvcp to work ?
[19:10:08] <Goslowjimbo> No, I had to try to start up HAL after I type the scripts instruction, and it isn't working.
[19:11:53] <Goslowjimbo> I keep getting a message that the rtapi module isn't started.
[19:12:14] <awallin> when issuing what command ?
[19:13:15] <jepler> Goslowjimbo: when you are using the version of emc2 in your home directory, use scripts/realtime, not /etc/init.d/realtime
[19:13:19] <Goslowjimbo> init.d/realtime start, halcmd unloadrt all,
[19:13:38] <jepler> similarly you must use scripts/emc not /usr/bin/emc, bin/halcmd, not /usr/bin/halcmd
[19:15:09] <Goslowjimbo> Will this one day make sense to me?
[19:15:36] <awallin> I'll be back later (30-40min). I'd be interested in cradek's thoughts on the traj stuff!
[19:20:46] <paragon36> Hello All
[19:21:29] <skunkworks> Hello
[19:22:42] <paragon36> Hello Skunkworks ... Did you manage to grab the config file for your pluto setup?
[19:22:58] <Goslowjimbo> Now I got the pyvcp command to quit because it couldn't open my xml file!-??
[19:23:18] <skunkworks> not yet - it is on my portable at home.
[19:24:15] <paragon36> No problem!! I popping out tonight for a jar or to anyway ;-)
[19:24:20] <skunkworks> remember it is just a test config. it uses siggen to put a sin signal into the pluto.
[19:24:49] <paragon36> s/to/two/ duh
[19:25:44] <jepler> paragon36: try this: http://emergent.unpy.net/files/sandbox/pluto.hal
[19:25:53] <skunkworks> still early here - 13:22
[19:25:56] <paragon36> At least it would give an idea... Have been reading up on HAL ... I need to read it again some has sunk into the grey matter but not enough .. lol
[19:25:58] <jepler> it will put a sine wave signal on the first pwm output and on the LED
[19:26:31] <jepler> run it with: halrun -I pluto.hal
[19:26:42] <paragon36> Has someone hooked the pluto to there lathe... Chris I recall?
[19:27:08] <jepler> yep it's cradek you're thinking of
[19:27:51] <jepler> oops -- spotted one thing that should be removed from that file. Grab it again if you downloaded it already and it says something about "epp_soft" in the top line
[19:28:43] <paragon36> Has cradek not publically released it yet then?
[19:29:49] <jepler> I don't think he's quite got it all back together
[19:30:48] <paragon36> Oh I see... Ive ordered one the pluto-p's but it could take 10 days to arrive to me here in London anyhow ....
[19:32:40] <skunkworks> I have been to london once. heathrow airport anyways
[19:33:07] <paragon36> Where are you living Skunkworks?
[19:33:09] <skunkworks> Have you ever met tom baker? :)
[19:33:23] <skunkworks> I live in wisconsin usa
[19:33:36] <paragon36> YEAH! and the Queen ... ;-)
[19:34:23] <skunkworks> * skunkworks would be more interested tom baker ;)
[19:34:37] <skunkworks> if he is still alive
[19:34:44] <paragon36> Saw a weird film about wisconsin Mmmm Something like Wisconsin Death Trip ... Real spooky!
[19:35:08] <skunkworks> never heard of it. (doesn't mean anything though)
[19:35:17] <paragon36> My drama teacher was called Tom Baker ...
[19:35:36] <paragon36> but the last I saw of him was in 1984
[19:36:39] <paragon36> Wisconsin Death Trip = Kind of a documentry about the big starvation or something back in the olde days ...
[19:37:17] <paragon36> http://www.wisconsindeathtrip.com/
[19:48:47] <awallin> back.
[19:54:59] <ejholmgren> phewy on WI ;)
[19:55:28] <ejholmgren> I've gone over the MN/WI border once...
[19:55:41] <skunkworks> they let you in?
[19:55:43] <skunkworks> :)
[19:56:01] <ejholmgren> we were driving waaaaaayyy too fast ;)
[19:56:11] <ejholmgren> the guards couldn't stop us
[19:57:24] <ejholmgren> my parent's used to drive to WI because the bars were open later and the drinking age was lower
[19:58:47] <lerneaen_hydra> what's the usual drinking age in the US?
[19:58:52] <ejholmgren> 21
[19:59:06] <ejholmgren> used to be 18 in some states
[19:59:06] <lerneaen_hydra> damn
[19:59:16] <ejholmgren> but that was a long time ago
[19:59:16] <lerneaen_hydra> is that at bars or to buy from stores?
[19:59:21] <ejholmgren> both
[19:59:29] <lerneaen_hydra> hmm, I see
[19:59:38] <lerneaen_hydra> 18 in bars and 21 for stores here
[19:59:46] <ejholmgren> interesting
[20:00:03] <ejholmgren> are you in sweden?
[20:00:07] <lerneaen_hydra> yeah
[20:00:26] <lerneaen_hydra> there are some drinking issues, though not out of the ordinary
[20:00:39] <ejholmgren> same here
[20:00:47] <lerneaen_hydra> those people will get alchohol regardless of whatever age limit they set so that won't make a difference
[20:01:31] <ejholmgren> I'm 25 and got carded today
[20:01:47] <ejholmgren> makes me feel a little younger
[20:01:49] <lerneaen_hydra> carded?
[20:02:07] <ejholmgren> when they look at your ID
[20:02:15] <lerneaen_hydra> oh, haha
[20:02:17] <lerneaen_hydra> nice
[20:04:02] <ejholmgren> got all 4 of my receptacles mounted into my box and the 8 relay parport kit assembled the other day
[20:04:19] <ejholmgren> I should get the flu more often
[20:05:53] <ejholmgren> just have to get emc installed on the controll box so I can start playing with classicladder
[20:14:19] <ejholmgren> I finally figured out what the electrical fire smell in our bedroom was the other day
[20:15:02] <ejholmgren> ... cat was marking his territory
[20:15:17] <ejholmgren> time to get "fixed"
[20:41:44] <Goslowjimbo> awallin: I got pyvcp to at least tell me it can't open my .xml file. What should I look for?
[20:42:29] <awallin> goslow: for some reason, the xml must not start with an empty line
[20:43:34] <Goslowjimbo> I have <pyvcp> on the first line
[20:43:57] <awallin> ok, then pyvcp -c test myfile.xml should work
[20:44:20] <awallin> myfile.xml should be in the direcgtory from where you are running pyvcp
[20:45:08] <Goslowjimbo> I had pyvcp -c ncs /home/jsr/pydro.xml as my command line
[20:45:30] <awallin> and you are sure the file exists ?
[20:45:47] <awallin> and is readable ?
[20:46:35] <Goslowjimbo> It's a good file, but its not where pyvcp is running. Where is pyvcp running?
[20:46:57] <awallin> let me try the same....
[20:47:27] <awallin> anders@el-laser:/$ pyvcp -c t /home/anders/testfile.xml
[20:47:30] <awallin> works
[20:48:08] <awallin> you are logged in as jsr?
[20:48:18] <Goslowjimbo> Do I have to have bin/ in the command line?
[20:48:45] <awallin> try pyvcp by itself. you should get: Usage: pyvcp -c hal_component_name myfile.xml
[20:48:46] <Goslowjimbo> Yes, I'm jsr
[20:49:12] <Goslowjimbo> I DO
[20:49:28] <SWPadnos> thanks, but I'm already married
[20:49:36] <Goslowjimbo> sorry about the caps
[20:49:42] <awallin> can you pastebin the xml file you are trying
[20:49:46] <SWPadnos> :)
[20:49:53] <Goslowjimbo> They are powerful words, arent they?
[20:50:04] <SWPadnos> they seem to have worked for the laast ~10 years ...
[20:50:20] <Goslowjimbo> What is pastebin?
[20:50:30] <SWPadnos> http://pastebin.ca
[20:50:36] <SWPadnos> lets you share text files easily
[20:50:39] <awallin> a way to post files on the web
[20:51:00] <Goslowjimbo> I'll try. Never did it before.
[20:52:57] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/pyvcp_widgets.py: move widget specific info here from vcpparse.py
[20:53:41] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/vcpparse.py: move widget specific stuff to pyvcp_widgets.py
[20:59:21] <SWPadnos> awallin, does it make sense to have both "from pyvcp_widgets import *" and "import pyvcp_widgets" statements?
[20:59:29] <Goslowjimbo> I've posted the file with my nickname as the title.
[20:59:42] <awallin> SWP: no, I'm working on that ;)
[20:59:48] <SWPadnos> heh :)
[20:59:55] <awallin> Goslow: please paste the url here
[21:00:02] <SWPadnos> http://pastebin.ca/307060
[21:00:37] <Goslowjimbo> Yeah, what he said
[21:00:50] <awallin> </pyvcp> at the end
[21:01:18] <Goslowjimbo> Do I remove it?
[21:01:24] <SWPadnos> add the slash to it
[21:02:00] <Goslowjimbo> I see. I missed that distinction. Is that in addition to <pyvcp> ?
[21:02:13] <awallin> <pyvcp> at the start of the file
[21:02:14] <SWPadnos> it's like the <number>blahblah</number> things
[21:02:19] <awallin> </pyvcp> at the end
[21:02:36] <Goslowjimbo> Gotcha. Thanks for your patience.
[21:03:42] <SWPadnos> this is funny. with my hosting service, I could host a pastebin-like site that could take 1 million 2k posts per week, and never use up my storage space
[21:04:27] <anonimasu> hm
[21:04:29] <awallin> SWP: how much do you pay for that?
[21:04:31] <lerneaen_hydra> and that, my friend, is why the earth is shaped as a banana
[21:04:36] <SWPadnos> $16/month
[21:04:36] <anonimasu> SWPadnos: store your posts at google mail
[21:04:55] <SWPadnos> that grows too slowly. DH adds 2GB per week
[21:05:08] <SWPadnos> and 32GB/week of monthly transfer ...
[21:05:29] <anonimasu> dh?
[21:05:36] <SWPadnos> DreamHost
[21:05:38] <anonimasu> ah
[21:05:39] <anonimasu> ok
[21:06:03] <SWPadnos> and remember - if you'd like to support me (and the linuxcnc.org site), use this link if you want to sign up with them: http://www.dreamhost.com/r.cgi?80098
[21:06:06] <SWPadnos> :)
[21:06:28] <cradek> awallin: what version of emc was that, and what are the length units?
[21:06:54] <awallin> cradek: HEAD from last night, sim/axis (inch I guess)
[21:07:20] <Goslowjimbo> OK, ppyvcp works, but it doesn't return control to the terminal - ?
[21:07:50] <jepler> Goslowjimbo: yes, you can use 'halcmd loadusr' or just add a '&' at the end to make it give you the terminal back
[21:07:55] <SWPadnos> add an & at the end
[21:08:46] <Goslowjimbo> OK. I got a control-c to return control. does that mean I need to run the pyvcp command again?
[21:09:02] <SWPadnos> did the DRO remain onscreen?
[21:09:05] <jepler> SWPadnos: do you mention this because there's a problem with pastebin, or just trying to make sense of the amount of storage DH gives you?
[21:09:14] <SWPadnos> the latter
[21:09:25] <Goslowjimbo> No
[21:09:36] <SWPadnos> Goslowjimbo, then you need to run it again :)
[21:09:48] <Goslowjimbo> Ok.
[21:10:46] <jepler> in the terminal, control-c requests the currently running program to exit
[21:11:25] <SWPadnos> and it's now graceful due to the userinterrupt exception being handled
[21:12:27] <awallin> jepler: any idea how to get rid of all the junk I get with pydoc for my modules due to from xxx import * statements ?
[21:12:56] <awallin> import xxx, obviously, but from xxx import * is useful/required sometimes, and it messes up the pydoc
[21:13:42] <jepler> awallin: I think that by defining __all__ in your module, you limit both the things that are imported from it when using 'from ... import *' and the things that are shown in pydoc
[21:14:04] <jepler> a list of strings e.g., __all__ = ["function1", "Class1", "variable1", ...]
[21:14:58] <awallin> ok, but the things that show up in pydoc when I do from Tkinter import * annoy me, any idea how to get rid of that?
[21:15:17] <awallin> i guess i should rtfm
[21:17:36] <cradek> awallin: I'm puzzled, to say the least, by the last plot
[21:17:44] <jepler> awallin: yes, the answer is to put __all__ in your module
[21:17:55] <awallin> jepler: ok, I'll try
[21:18:04] <awallin> cradek: try it for yourself! :)
[21:18:22] <awallin> cradek: the question is I guess if the traj planner should allow for "stupid" users?
[21:18:31] <jepler> awallin: I put the following in a file:
[21:18:33] <jepler> from Tkinter import *
[21:18:33] <jepler> __all__ = ["f"]
[21:18:33] <jepler> def f(): "do something"
[21:18:41] <cradek> what part do you mean is the stupid user?
[21:18:44] <jepler> then "pydoc thatmodule" shows only an entry for 'f', nothing else
[21:19:12] <awallin> cradek: traj_accel=axis_accel=0.2 (a very small number)
[21:19:27] <cradek> sure, it should do what you ask
[21:19:50] <cradek> but there's a problem in that last plot
[21:19:52] <awallin> jepler: right, something that removes all the Tkinter stuff from the default would be more useful
[21:21:11] <A-L-P-H-A> SWPadnos, going to the bank, and post office now, after I get changed.
[21:21:32] <SWPadnos> ok
[21:22:55] <awallin> what's the 'officia' tab-length for the emc sources? 4 or 5 both seem to be in use??
[21:22:59] <awallin> official
[21:23:21] <jepler> for python code I prefer 4 space indentation
[21:23:32] <SWPadnos> I think 4 is more common (and preferable to me)
[21:23:50] <awallin> does cvs mess with the tab length ?
[21:23:50] <A-L-P-H-A> 4 is a unlucky number.
[21:23:57] <A-L-P-H-A> 4 means death in Chinese.
[21:24:02] <SWPadnos> but 5 looks odd
[21:24:04] <A-L-P-H-A> 8 is a more lucky number. :)
[21:24:13] <A-L-P-H-A> weren't tabs in MSDOS 8?
[21:24:28] <A-L-P-H-A> someone should just run in through some code beautifiers.
[21:24:46] <cradek> no
[21:24:53] <A-L-P-H-A> yes
[21:25:02] <cradek> you should never reformat code that's under revision control
[21:25:02] <A-L-P-H-A> just cause I want to be different. :)
[21:25:19] <A-L-P-H-A> oh... I didn't know it was under revision.
[21:25:26] <A-L-P-H-A> how about a refactor instead? :D
[21:25:44] <A-L-P-H-A> SWPadnos, how about you come up, and we'll do some extreme programming... you code, I'll watch and critique. :)
[21:25:44] <jepler> pyvcp is young enough that I would instead say: reindent it in a check-in that makes no meaningful changes
[21:26:25] <SWPadnos> A-L-P-H-A, as long as you pay for the gas and the food, I'm game :)
[21:27:04] <jepler> if it's just adjusting whitespace at the left, "cvs diff -b" will even still give useful diffs
[21:27:39] <A-L-P-H-A> hmm... gas + food... how about Air, that's a gas. You pay for lodging, and the strippers, we're GOOD.
[21:28:06] <lerneaen_hydra> * lerneaen_hydra pops into the channel and reads A-L-P-H-A's text
[21:28:09] <lerneaen_hydra> O.o
[21:29:11] <A-L-P-H-A> cheap socks... elastic threads are coming out after one/two washes.
[21:36:01] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/vcpparse.py:
[21:36:01] <CIA-8> prefer 'import module' over 'from module import *'
[21:36:01] <CIA-8> add docstrings
[21:37:47] <cradek> awallin: I think G64 P is off by a factor of 25.4
[21:37:55] <cradek> * cradek grumbles miserably
[21:38:01] <cradek> G64P[.01/25.4]
[21:38:09] <cradek> and you get exactly the behavior you expect
[21:38:12] <lerneaen_hydra> cradek: are you running inch based?
[21:38:23] <cradek> yes it's an inch config, with inch gcode
[21:38:36] <SWPadnos> some missing {TO,FROM}_USER_UNITS or similar
[21:38:44] <SWPadnos> ?
[21:38:49] <cradek> yep
[21:40:09] <SWPadnos> I guess that was the simple part. the hard part is finding where :)
[21:40:11] <awallin> cradek: ?? if 0.01 was mm, then it should have moved within 0.01mm of 1" which it didnt. If 0.01 was inch, it should have moved within 0.01" of 1" which it didn't - I don't understand
[21:40:34] <cradek> awallin: probably it's a units conversion the wrong way
[21:40:51] <cradek> try P[.01/25.4] and I bet it goes right to .99 inches
[21:41:05] <SWPadnos> unless 0.01 was considered to be mm, but the machine is inch so it gets scaled up by 25.4, then that is used as an inch number
[21:41:29] <SWPadnos> so maybe it's an extra unit conversion somewhere :)
[21:41:33] <awallin> wait... I'll try it.
[21:41:36] <cradek> grrrr
[21:42:09] <cradek> every time the MPG readout on my car says 25.4 I think something must be wrong - it's a phobia now
[21:42:20] <SWPadnos> heh
[21:42:41] <SWPadnos> you start thinking that the 25.4 MPG should be 1 liter/100 km?
[21:43:01] <cradek> yes it goes to 0.990 inch
[21:43:14] <skunkworks> Nice find
[21:43:24] <awallin> yep, G64P[0.01/64] looks much better. btw. I'm in G20 mode, and the .ngc file has no G21/G20
[21:43:36] <awallin> skunkworks: thanks ;)
[21:43:46] <skunkworks> :)
[21:44:42] <awallin> didn't both USA and the UK move officially to the SI system in the 1970ies ?? ;)
[21:44:53] <cradek> troll!!
[21:45:49] <SWPadnos> yes, but lots of people here are (a) stupid or (b) too nationalist to allow some other system to take over
[21:45:57] <SWPadnos> (from the system we already borrowed from England)
[21:46:28] <SWPadnos> that's why we call "Aluminium" "Aluminum"
[21:46:29] <cradek> :q
[21:46:33] <SWPadnos> heh
[21:46:33] <cradek> err
[21:46:38] <SWPadnos> ok - I'll stop ;)
[21:48:15] <cradek> http://cvs.linuxcnc.org/cgi-bin/cvsweb.cgi/emc2/src/emc/task/emccanon.cc.diff?r1=1.74;r2=1.75
[21:48:48] <cradek> jepler: we broke it when we fixed the other thing
[21:49:26] <jepler> cradek: pork!
[21:49:50] <lerneaen_hydra> bork!
[21:49:57] <cradek> all these length units should be different C++ types shouldn't they
[21:50:09] <awallin> jepler: now I have __all__=["myclass"], but myclass inherits from Tkinter.Label, which results in about 3 pages of pydoc I don't want...
[21:50:14] <jepler> you'll never really get that to work as you would wish -- but yes that would be grand
[21:50:24] <cradek> grrr
[21:50:41] <cradek> instead I'll just fix this last bug
[21:50:51] <awallin> I'm sorry if this ruined your Friday :)
[21:51:00] <cradek> not at all - thanks for finding it
[21:51:16] <jepler> awallin: python doesn't really have "private inheritance", so in other words I don't know of an easy fix. You can use "has-a" (put a Label as an instance variable of myclass) instead of "is-a" (make myclass inherit from Label) but that can require a lot of changes to your program.
[21:51:25] <cradek> to really ruin my friday, you could tell me it's wrong in 2.0 as well
[21:51:39] <cradek> in fact I'd be thrilled if someone would test it while I'm fixing this
[21:51:55] <jepler> if it really broken in 1.74->1.75 we didn't backport that change to 2.0
[21:52:12] <cradek> yes but that's no reason to not test it :-)
[21:53:46] <awallin> I could probably test the same G-code with 2.0.5 pretty easily
[21:54:14] <cradek> thank you
[21:54:49] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/pyvcp_widgets.py: add docstrings
[21:56:02] <awallin> cradek: except 2.0.5 can't be built with --enable-simulator ?? so I have to reboot the vm with magma?
[21:58:31] <skunkworks> * skunkworks runs away.
[22:02:17] <jepler> awallin: yes, that's right
[22:02:17] <cradek> yeah, there's no simulator
[22:02:28] <jepler> but cradek and I were just testing in 2.0.5 -- it looks like it was right
[22:03:34] <CIA-8> 03cradek 07HEAD * 10emc2/src/emc/task/emccanon.cc: fix tolerance units problem
[22:04:32] <CIA-8> 03cradek 07v2_1_branch * 10emc2/src/emc/task/emccanon.cc: fix tolerance units problem
[22:05:22] <cradek> awallin: thanks again, that would have been an embarassment in 2.1.0
[22:06:15] <awallin> ok, so I'm not testing 2.0.5 today then...
[22:09:43] <SWPadnos> awallin, you still have to specify what the hal pin name will be, do you need <name></name> around it?
[22:10:02] <SWPadnos> (regarding your response to James Rerd)
[22:10:05] <SWPadnos> Reed
[22:10:29] <awallin> SWP <halpin>"myhalpin"</halpin> is correct
[22:10:46] <SWPadnos> ok. that's missing from the snippet in your message :)
[22:11:05] <skunkworks> SWPadnos: says I should recieve the 'package' next tuesday
[22:11:09] <skunkworks> * skunkworks is excited
[22:11:15] <SWPadnos> cool :)
[22:11:24] <awallin> yeah, I'll have checked in the autonaming of pins by the time Reed reads it ;)
[22:11:30] <SWPadnos> heh
[22:11:37] <SWPadnos> shall I post the correction?
[22:11:45] <awallin> SWP: sure go ahead
[22:12:01] <awallin> we need a way to count how many instances of a particular class has been created
[22:12:02] <SWPadnos> ok
[22:12:35] <awallin> that's called a class variable is it ? (one that stays with the class, not with each individual object)
[22:12:43] <SWPadnos> are you thinking of having automatic compname.<type>.<number> naming ?
[22:13:03] <SWPadnos> I guess it would be called that (in C++ anyway)
[22:13:13] <awallin> SWP: yes, if the user doesn't specify anything, at least it shouldn't crash
[22:13:14] <SWPadnos> but it could also be a global int ;)
[22:15:05] <awallin> I think it's a __myvariable__ syntax in Python...
[22:16:15] <SWPadnos> I wonder what these screws on my desk are from?
[22:16:24] <awallin> where did jepler go when I ask python questions - he is so much faster than google...
[22:16:30] <SWPadnos> heh
[22:16:48] <SWPadnos> he could be getting ready to leave work, or driving home or something :)
[22:18:20] <awallin> notice how I said <halpin>"myhalpin"</halpin> just a while ago... you didn't include the quotes in your email...
[22:18:37] <SWPadnos> I guess you should correct me :)
[22:18:52] <awallin> but I can try if it works without them... probably not
[22:19:16] <awallin> nope.
[22:19:46] <SWPadnos> that's strange. I'd expect it to work unless you need <> in the string
[22:20:54] <awallin> <option>value</option> gets mangled into a function call pyvcp_widget(option=value) which is interpreted by Tkinter, and Tkinter wants to see option="something"
[22:21:00] <SWPadnos> ah
[22:21:27] <awallin> But ofcourse pyvcp could add the quotes for you - the current implementation is just a very lazy one.
[22:27:04] <awallin> there. I patched your post.
[22:39:27] <awallin> awallin is now known as etla
[22:39:39] <etla> etla is now known as awallin
[22:41:11] <lerneaen_hydra> having trouble deciding on a name? ;)
[22:41:41] <alex_joni> multiple personality disorder
[22:41:50] <awallin> yeah, some channels require a registered nick...
[22:42:04] <alex_joni> you can have both registered and linked together
[22:42:52] <alex_joni> at one point you'll get a cloak too :)
[22:43:13] <alex_joni> something like awallin@emc/developer/awallin
[22:43:32] <robin_sz> meep?
[22:43:52] <lerneaen_hydra> alex_joni: I was wondering, how do you regiest those?
[22:44:04] <lerneaen_hydra> *register
[22:46:09] <alex_joni> what ? cloaks?
[22:46:18] <lerneaen_hydra> yeah
[22:46:23] <alex_joni> you need superpowers for that
[22:46:43] <alex_joni> first you need a sip from the supercows milk
[22:46:47] <lerneaen_hydra> oh, like invisibility or being able to fly or stuff like that?
[22:47:02] <alex_joni> then you ask a freenode admin to do that
[22:47:15] <lerneaen_hydra> oh, so it's done manually
[22:47:46] <alex_joni> yeah, and only the group contact can request that
[22:48:06] <lerneaen_hydra> oh, I see
[22:48:46] <alex_joni> any reason to ask?
[22:49:11] <lerneaen_hydra> oh, no not really
[22:52:38] <alex_joni> hi robin_sz
[22:55:19] <robin_sz> evening ...
[22:55:22] <awallin> CIA is really slow...
[22:56:00] <alex_joni> * alex_joni kicks CIA-8
[22:56:02] <CIA-8> ow
[22:56:10] <alex_joni> awallin: not that slow :D
[22:56:43] <awallin> it must be 60s since I did a commit
[22:57:44] <alex_joni> sometimes it's hours
[22:58:47] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/pyvcp_widgets.py: remove debug print
[22:59:00] <awallin> ha, that one made it right away
[22:59:04] <awallin> but the previous one was lost
[22:59:36] <alex_joni> that happens sometimes
[22:59:41] <awallin> alex: did you think about the feed aggregator idea?
[23:02:03] <alex_joni> awallin: yes, I will look into it, as soon as I have decent internet connection
[23:03:01] <awallin> alex: ok, still traveling the world? ;)
[23:03:53] <alex_joni> nah, winter holiday
[23:04:49] <daniel_br> hello guys
[23:05:30] <awallin> hi
[23:05:46] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/pyvcp_widgets.py: automatic hal pin naming for checkbutton
[23:06:25] <daniel_br> alex, last time I was here asked about home all and home sequence, remeber?
[23:06:37] <daniel_br> remember?
[23:07:17] <daniel_br> I used a config like that http://cvs.linuxcnc.org/cvs/emc2/configs/sim/axis.ini?rev=1.23;content-type=text%2Fplain
[23:07:58] <daniel_br> but i can not get this working ok
[23:09:16] <cradek> daniel_br: what's the problem exactly?
[23:10:07] <daniel_br> The problem is that config not permit use only one click for home all axis
[23:10:37] <lerneaen_hydra> 'night all
[23:10:46] <daniel_br> I can do that only using classiclader
[23:10:48] <alex_joni> night
[23:11:11] <daniel_br> with a external button
[23:11:22] <cradek> in AXIS you can home all with Control-Home
[23:12:04] <cradek> the keyboard shortcut
[23:12:35] <daniel_br> ok i will try next week
[23:13:04] <cradek> http://linuxcnc.org/docs/2.1/html/config/ini_homing/index.html
[23:13:21] <cradek> you set up HOME_SEQUENCE and HOME_IS_SHARED?
[23:13:35] <daniel_br> Yes
[23:14:20] <daniel_br> and after clicked in the icon home and nothing changed
[23:14:52] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/pyvcp_widgets.py: automatic hal pin naming for button and scale
[23:15:21] <cradek> you should make homing work individually before working on home-all
[23:15:46] <daniel_br> this was working very well
[23:16:18] <daniel_br> and after i used classiclader all is working as i need
[23:16:40] <cradek> ok
[23:16:45] <daniel_br> i push a button and all axis homing in sequence
[23:18:26] <daniel_br> but i can say i very happy with the emc2.1
[23:18:56] <jepler> darn, he's gone
[23:19:08] <jepler> I think that in AXIS there's only a keystroke for "home all", not a button to click
[23:19:31] <cradek> that's right, I looked
[23:25:42] <alex_joni> night guys
[23:28:16] <jepler> see you alex
[23:28:54] <lerman> Hi, guys. Greetings from San Francisco -- I'm here for a week.
[23:28:54] <CIA-8> 03awallin 07HEAD * 10emc2/lib/python/pyvcp_widgets.py:
[23:28:54] <CIA-8> automatic hal pin naming for the rest of the widgets
[23:28:54] <CIA-8> all widgets now support automatic naming
[23:30:00] <lerman> cradek: jepler: in particular. I'm thinking of my prototype GWiz conversational programming code in python.
[23:30:49] <jepler> hi lerman
[23:31:03] <lerman> It will take me a while to become proficient in a new (to me) language and gui environment, but it seems like a reasonable way to go.
[23:31:25] <lerman> And it would probably integrate better with Axis.
[23:31:29] <jepler> I think you'll pick up on it pretty quickly
[23:31:55] <jepler> right now axis can run filters in any language -- only things that are to appear in the main axis window are forced to be in python+tkinter
[23:32:10] <lerman> Also, I'd really like to be able to single step forwards and *backward* in the interpreter.
[23:32:25] <lerman> That would also be nice to integrate.
[23:32:30] <jepler> that's a pretty tall order
[23:32:39] <lerman> Doable, I think.
[23:32:42] <daniel_br> cradek net problem here but i'm here again
[23:34:12] <lerman> In the best of all worlds, you should be able to connect an encoder to another port and run it forwards or backwards, causing the code to run forwards or backwards with the motion along the path syncronized to the motion of the encoder handwheel.
[23:34:25] <lerman> Now, that would be neat.
[23:35:20] <A-L-P-H-A> SWPadnos, the check was mailed... I have enough funds to clear np. :)
[23:35:25] <jepler> daniel_br: in axis there's only a keystroke for home all (shift+home, I think) -- not a button in the GUI
[23:35:47] <lerman> It would need the same type of synchronous motion as is used for rigid tapping or lathe threading. In this case, the axis motion would be synched to the encoded handwheel instead of the spindle.
[23:35:49] <jepler> control-home
[23:36:14] <jepler> I have debated between adding a new button, or making the current button mean "home all" when the homing order is defined, but instead I did nothing
[23:36:34] <daniel_br> Yes i think with that config home_sequence that only button will work for that home all
[23:39:10] <alex_joni> >/away sleep
[23:39:25] <alex_joni> darn keys
[23:40:32] <daniel_br> poor man 3d scanner http://zarria.net/nrmphoto/nrmphoto.html
[23:46:48] <daniel_br> open source python 3d scanner http://www.splinescan.co.uk/
[23:50:36] <A-L-P-H-A> :( my cell phone minutes are being chewed up a shit load this month.
[23:59:28] <anonimasu> lerman: that's cute..
[23:59:34] <anonimasu> lerman: they have that on toolroom lathes(cnc ones)