#emc | Logs for 2008-11-24

[01:04:55] <JymmmEMC> jmkasunich: Sorry, throwing a tv dinner in the microwave is NOT cooking ;)
[01:05:09] <JymmmEMC> alex_joni: ping?
[01:05:16] <JymmmEMC> alex_joni: you use rsync?
[01:05:30] <JymmmEMC> oh, nm
[01:36:45] <eric_u1> I can't believe I'm having heat related problems with my router in the middle of winter
[01:38:33] <jmkasunich> JymmmEMC: Chicken Stroganoff, from scratch!
[01:38:42] <skunkworks> Mmmmm
[01:38:43] <jmkasunich> and it was damned good too
[01:39:04] <SWPadnos> we made an interesting pie today
[01:39:11] <SWPadnos> dark chocolate pecan pie
[01:39:23] <jmkasunich> ohhhh
[01:39:33] <eric_u1> "interesting pie" is generally not considered a compliment
[01:39:47] <SWPadnos> it was excellent actuayyl - the recipe is from Lake Champlain Chocolates
[01:39:51] <SWPadnos> actualy
[01:39:54] <SWPadnos> +l
[01:39:59] <SWPadnos> damn. too much pie ;)
[01:40:12] <JymmmEMC> jmkasunich: And you dind't share with the rest of us, eeeeeeesh
[01:40:24] <eric_u1> "email this dinner"
[01:40:25] <SWPadnos> it was so dense I only had about a 1/10 pie serving, and couldn't eat any mroe
[01:40:26] <JymmmEMC> wait... chicken?!
[01:40:27] <SWPadnos> more
[01:40:35] <jmkasunich> yes, chicken
[01:40:37] <SWPadnos> like beef, only smaller
[01:40:49] <jmkasunich> feathers instead of furr
[01:40:52] <eric_u1> better than dog, I suppose
[01:40:55] <JymmmEMC> SWPadnos: and no horns
[01:41:00] <SWPadnos> beaks though
[01:41:04] <SWPadnos> pointy little things
[01:41:18] <jmkasunich> SWPadnos: that's the problem with things like pecan pie
[01:41:24] <SWPadnos> heh
[01:41:34] <jmkasunich> unless its the holidays or something, and you have a thundering herd to feed
[01:41:35] <SWPadnos> I stuck a lot of pecans in it too - not just a layer on top like usual
[01:41:47] <SWPadnos> I'm usually the equivalnet of a thundering herd :)
[01:41:48] <SWPadnos> gah
[01:42:39] <SWPadnos> I had very little of the vegetable leek quiche too
[01:42:58] <SWPadnos> (yes, leek is a vegetable ...)
[01:43:02] <ds3> wheeee 3D scanner works
[01:43:21] <SWPadnos> cool
[01:44:01] <ds3> emc needs more 3D scanner support ;)
[01:44:20] <SWPadnos> patches gratefully accepted
[01:44:22] <skunkworks> our favorite dish at the moment is 3 cheese penne with sweet turkey sausage
[01:44:31] <LawrenceG> ds3... tell more! pictures?
[01:44:31] <SWPadnos> sounds good to me
[01:44:50] <SWPadnos> we've been staying away from cheesy stuff since my wife is lactose intolerant
[01:44:56] <SWPadnos> so creamy stuff is also out
[01:45:02] <SWPadnos> which is a big bummer
[01:45:02] <skunkworks> aww
[01:45:22] <skunkworks> it has red pepper in it. has just a little bite
[01:45:23] <SWPadnos> $x alfredi is/was one of her favorite dishes too :(
[01:45:26] <SWPadnos> heh
[01:45:41] <skunkworks> jmkasunich: :) http://imagebin.ca/img/eERZslj.png
[01:45:43] <SWPadnos> I did that - made some garlic clam linguine, threw in a bit of crushed red pepper
[01:46:14] <ds3> LawrenceG: let me find a place to push them onto... the first one has some holes so I am playing with parameters
[01:46:45] <LawrenceG> ds3, homemade or purchased..... I would like to build one
[01:47:29] <ds3> LawrenceG: home made
[01:47:41] <ds3> most sites seem to talk about mach3
[01:48:00] <SWPadnos> advertising helps with that :)
[01:49:49] <ds3> LawerenceG: http://www.ugcs.caltech.edu/~huny/tmp/run2a.bmp is what the point cloud looks like... it is missing stuff but it suggests it works
[01:50:10] <eric_u1> I suppose I should figure out how to make a copper pour like that skunkworks
[01:50:49] <jmkasunich> skunkworks: what are the funny horizontal lines around the fets?
[01:51:28] <skunkworks> sorry - I had the top and bottom restricts turned on.
[01:51:44] <jmkasunich> the ground connection looks good. maybe want a wee bit more clearance between the red floods (motor leads) and the red ground traces
[01:52:22] <jmkasunich> and you should do something so that the other ground connection (to FET source) can be disconnected
[01:52:28] <jmkasunich> (I guess you could just cut the trace)
[01:52:39] <LawrenceG> ds3, link seems borked, looks like tmp is the file not a directory
[01:52:55] <jmkasunich> I got a 404 on that one too
[01:53:07] <LawrenceG> http://www.ugcs.caltech.edu/~huny/tmp/
[01:53:57] <LawrenceG> light blue background with some white stuff
[01:53:58] <ds3> let me check
[01:55:21] <ds3> fixed... mv from the wrong place
[01:56:15] <LawrenceG> is that a depth map of a statue?
[01:56:31] <ds3> the white stuff is the decoded point... the shadow is right but the reflection from the object seems to be very much reduced in amplitude
[01:56:45] <ds3> LawrenceG: effectively yes
[01:56:52] <LawrenceG> cool
[01:57:09] <ds3> Let me upload this other run... tweaked some values, should be a bit better except for some ghost objects
[01:58:39] <ds3> LawerenceG: http://www.ugcs.caltech.edu/~huny/tmp/run3.bmp
[01:58:56] <LawrenceG> better
[02:00:01] <LawrenceG> ds3, what does scanning hardware look like?
[02:01:41] <ds3> LawerenceG: one hacked up camera + a projector
[02:02:02] <ds3> let me find a suitable camera to take a snapshot
[02:04:01] <skunkworks> jmkasunich: yes - I was planning on cutting the trace.
[02:12:09] <ds3> LawerenceG: http://www.ugcs.caltech.edu/~huny/tmp/setup.jpg
[02:13:26] <LawrenceG> cool.... camera is stationary and the laser line projector sweeps across object?
[02:13:34] <ds3> yep
[02:13:57] <ds3> once it works better, I'll optimize it a bit; the whole mess is held together with shell scripts
[02:14:08] <LawrenceG> are you suing something like the david software or all homebrew?
[02:14:23] <ds3> all home brew; math figured out from first priciples
[02:14:53] <LawrenceG> good work.... a school project?
[02:14:58] <ds3> all the point cloud collection stuff is done entirely in linux; it is the point cloud display and cad stuff that will be nonLinux
[02:15:02] <SWPadnos> what's the angle between the camera nad laser?
[02:15:08] <SWPadnos> and
[02:15:32] <ds3> no... just a fun project... maybe someone will find it interesting and hire me ;)
[02:15:51] <ds3> SWPadnos: I donno... I compute that at run time; I give the location and distances
[02:15:57] <SWPadnos> ok
[02:16:03] <ds3> and it is not a laser
[02:16:12] <SWPadnos> do you rotate the part or sweep the laser?
[02:16:19] <SWPadnos> looks like a laser line generator
[02:16:21] <ds3> sweep the red line
[02:16:28] <SWPadnos> ok, interesting
[02:16:38] <ds3> SWPadnos: this is actually the PicoDLP projector thing from TI
[02:16:54] <SWPadnos> oh - they have a DLP dev kit?
[02:17:05] <ds3> yeah...
[02:17:16] <SWPadnos> ooooh - cool
[02:18:11] <ds3> I wanted to build something like this for a long time... but I never finished my CNC rotary table (2 more operations to go to mount the stepper)
[02:18:17] <SWPadnos> huh - relatively cheap too. I guess that needs to go on my list :)
[02:18:35] <eric_u1> link?
[02:18:39] <ds3> yep... that is a hacked up $11 camera
[02:18:53] <SWPadnos> http://focus.ti.com/dlpdmd/docs/dlpdmdhomepage.tsp?sectionId=597
[02:18:58] <eric_u1> thx
[02:19:07] <SWPadnos> sure - just found it myself ;)
[02:19:33] <SWPadnos> it's lots more expensive for the 1920x1080 kits ;)
[02:19:49] <ds3> the picoDLP kit is cool...smaller then a pack of cards
[02:20:06] <SWPadnos> I didn't see the resolution in 6 seconds of scanning
[02:20:30] <ds3> 640x480 fixed, HDMI input and a built in battery
[02:20:42] <SWPadnos> ok, so VGA was the resolution
[02:21:30] <ds3> so the scans won't be too high res.. can't use it to measure features on a proocessor die ;)
[02:21:40] <SWPadnos> heh
[02:22:15] <ds3> In the last year, did EMC gain any 3D surfacing features?
[02:22:29] <SWPadnos> uh, I don't think so
[02:22:40] <SWPadnos> you mean CAM-like?
[02:22:46] <ds3> yeah
[02:22:54] <SWPadnos> (it's been able to mill 3D surfaces for a ling time)
[02:23:01] <ds3> eventually I want to machine it out
[02:23:16] <SWPadnos> I don't think there's much new in that area
[02:23:27] <ds3> yeah, I think the MadCAM I got a while back can generate the surfacing code
[02:25:50] <ds3> I wonder how bad would a 117deg point surface
[05:20:25] <user__> user__ is now known as SkinnYPuPp_
[09:21:24] <archivist> ds3 look at pycam
[11:22:34] <archivist> hmm ball nuts adjustable to 0 backlash http://www.worldofcnc.com/images/ISEL-1A-1B.pdf
[12:19:37] <alex_joni> JymmmEMC: 'lo
[12:19:39] <alex_joni> sometimes..
[12:23:26] <alex_joni> ds3: too bad your links don't seem to work anymore
[12:23:34] <alex_joni> any chance you could put it on imagebin or the like?
[12:23:50] <alex_joni> btw.. for point cloud, look at meshlab (they have a linux version aswell)
[12:35:30] <micges> alex_joni: have you access to any running fanuc nc controlled machines ?
[12:45:08] <alex_joni> nope
[12:46:51] <stustev> micges: I fanuc controls - what is your question?
[12:46:56] <stustev> have
[12:49:13] <micges> I made someday posprocessor DXF->FANUC NC and want to extend/improve it
[12:49:28] <micges> want some files for fanuc
[12:49:59] <stustev> you want some gcode files for a fanuc control?
[12:50:17] <micges> yup
[12:50:56] <micges> can be one file and dxf for it
[12:52:22] <stustev> the gcode files are interchangeable between EMC2 and fanuc
[12:54:23] <micges> can't be..
[12:54:44] <stustev> for basic motion there is no difference - depending upon the set up parameters in the fanuc control
[12:54:55] <stustev> what difference do you see?
[12:55:33] <micges> I have some files with M900 M801
[12:55:50] <micges> in fanuc dir
[12:56:28] <stustev> are those custom macros for a particular machine? - the builder/integrator wrote those macros and calls?
[12:56:56] <micges> don't know
[12:57:17] <stustev> I don't recognize them as anything I have ever used
[12:57:29] <stustev> what do they do on the control you are talking about
[12:59:25] <micges> stustev: anyway thanks, I must have files from some nc control but as you saing this is not fanuc
[12:59:47] <micges> I'll google for that (I think)
[13:00:13] <stustev> if your goal is basic 3d motion the files are interchangeable - they may be fanuc - just not the usually used codes
[13:01:07] <stustev> if they are custom then google probably will not find an explanation - only if custom writer has published the explanation
[13:03:20] <micges> thanks for info stustev
[15:38:06] <skunkworks> Monday - yay
[15:38:41] <archivist> grmbl...its Monday
[15:39:54] <jymm> archivist: you have permission to slap skunkworks
[15:40:24] <archivist> hmm choice of weapons ?
[15:40:25] <Dallur> This is where an emc controlled robotic arm would come in handy
[15:40:52] <jymm> archivist: Preferably non-lethal, but tazer is acceptable
[15:41:00] <eric_u1> but more importantly, what did they do to qtips? The new ones are unusable
[15:41:21] <jymm> eric_u1: plastic stick?
[15:41:41] <eric_u1> yes
[15:41:56] <jymm> eric_u1: yeah, they suck. Gotta find the paper sticks
[15:42:29] <jymm> which is ironic, as I'd think today plastic is more expensive than paper
[15:42:41] <eric_u1> it's not very much plastic, that's the problem
[15:43:07] <jymm> eric_u1: and it metls when dipped in acetone
[15:43:22] <eric_u1> I can imagine
[15:43:33] <jymm> or slides off when in denatured alcohol
[15:43:35] <eric_u1> there probably are expensive alternatives
[15:43:39] <jymm> the cotton that is
[15:43:40] <skunkworks> * skunkworks ducks
[15:44:02] <jymm> * jymm kicks skunkworks in the shins - duck this biotch!
[15:44:45] <jymm> skunkworks: oh yeah.... Goooooooooooooooooooooooooooooooooog Morning!
[15:46:02] <skunkworks> :)
[15:46:17] <jymm> </sarcaism>
[15:47:20] <jymm> It bugs the crap out of me when ppl around the office are saying "Good time-of-day", Can't you just say "Morning" or some such thing, maybe not everyoen is actually having a GOOD moring.
[15:47:49] <cradek> Looks like someone has a case of the Mondays
[15:48:12] <skunkworks> heh
[15:48:23] <jymm> cradek: Nope, not at all actually, though others are. But just in general.
[15:48:58] <jymm> I just think they ppl saying "Good morning" is kinda (unintentionally) rubbing it in.
[15:49:06] <alex_joni> good evening :)
[15:49:10] <alex_joni> * alex_joni runs home..
[15:49:29] <jymm> alex_joni: Evening
[15:49:33] <jymm> =)
[15:49:45] <alex_joni> it's a good one .. as I'm leaving work
[15:50:10] <jymm> hasta
[15:50:22] <alex_joni> luego
[15:52:25] <jymm> But, most have a short week, so this is a good thing =)
[15:54:35] <archivist> * archivist cuts jymm off at the knees to make sure he is short for the week
[15:55:28] <jymm> archivist: It's gonna take much more than that =)
[15:56:16] <archivist> just a flesh wound....(Monty Python Black Knight)
[16:28:13] <jymm> This is totally cool... http://www.cnn.com/video/?/video/world/2008/11/23/lah.jap.monkey.biz.cnn
[16:32:20] <alex_joni> archivist: merely a flesh wound :P
[17:27:08] <cradek> ha
[17:27:14] <cradek> rteyber12 has made a comment on Hardinge HNC conversion running EMC2: holy shit. perfect.
[17:30:00] <alex_joni> heh :)
[17:30:03] <alex_joni> where was that?
[17:31:13] <cradek> I think it's a youtube comment but I'm not sure
[17:35:55] <cradek> yep
[18:06:25] <archivist> cradek, I see you other vid has 15K views!
[18:44:04] <skunkworks> http://www.cnczone.com/forums/showthread.php?t=68650
[18:44:59] <skunkworks> came about from this thread http://www.cnczone.com/forums/showthread.php?t=68155
[19:54:30] <maddash> :(
[19:54:48] <maddash> python: math.sin(math.pi) = 1.2246063538223773e-016
[19:56:17] <alex_joni> http://www.theregister.co.uk/2008/11/21/more_hp_packaging/ ROFL
[19:58:18] <cradek> maddash: that's clearly wrong - my calculator says it's -2.0676e-13
[19:59:09] <fragalot> cradek: rounded pi.
[19:59:19] <cradek> fragalot: duh.
[19:59:29] <fragalot> gotta love it
[19:59:55] <skunkworks> should be 0?
[20:00:04] <fragalot> my calc tells me .0548036651
[20:00:17] <cradek> fragalot: uh, try radians
[20:00:26] <skunkworks> fragalot: degrees
[20:00:29] <skunkworks> hehe
[20:00:32] <cradek> jeez
[20:00:40] <fragalot> cradek: 0 :D
[20:00:52] <cradek> SWPadnos: you saw peter's message about the HNCs?
[20:00:59] <cradek> kirk's
[20:00:59] <fragalot> silly calc must 've reset itself, it's normally always in rad mode
[20:01:00] <cradek> dangit
[20:03:24] <jepler> maddash: just Python relying on the behavior of the C standard library.
[20:04:00] <jepler> $ gcc -std=gnu99 sinpi.c -lm && ./a.out; cat sinpi.c
[20:04:00] <jepler> 1.2246063538223773e-16
[20:04:00] <jepler> #include <math.h>
[20:04:00] <jepler> #include <stdio.h>
[20:04:00] <jepler> int main(void) { printf("%.17g\n", sin(M_PI)); return 0; }
[20:26:42] <maddash> jepler: out of curiosity, how long did all that take you?
[20:27:20] <jepler> maddash: oh, I think I first looked at the discussion at around 13:59 local time, and I pasted that at 14:04 local time .. so about 4-5 minutes
[20:27:50] <maddash> wow.
[20:48:18] <maddash> rofl http://www.bash.org/?870063
[20:55:11] <jepler> maddash: by the way, if you want to read all about how floating-point math sucks you could do worse than to read this guy's publications: http://www.cs.berkeley.edu/~wkahan/
[20:56:23] <jepler> "Miscalculating Area and Angles of a Needle-like Triangle" and "How JAVA's Floating-Point Hurts Everyone Everywhere" are particularly relevant to your result sin(pi) != 0
[21:04:42] <alex_joni> * alex_joni is now known as alex_joni
[21:23:29] <fragalot> http://failblog.files.wordpress.com/2008/11/fail-owned-accident-kids-card-fail.jpg
[21:59:02] <Gamma-X> hey i forget wich to switch.... if my table is running away after a command is given, would I switch the wires going from the tachs to the amp, or the wires going from the amp to the mesa board?
[22:00:19] <cradek> depends which ones are wrong
[22:00:32] <cradek> or it might be a different problem
[22:00:50] <Gamma-X> not sure.... ive heard of multiple sources that that is most likely the problem..
[22:01:10] <cradek> remember the bottom-up procedure I described before? you're still not doing it.
[22:01:17] <cradek> if you had, you'd know if your tach is wired right
[22:01:40] <cradek> darnit, I have to run, bbl.
[22:01:45] <Gamma-X> thanks cradek
[22:11:00] <Guest978> can anybody shed some light on homing?
[22:11:43] <fragalot> sure
[22:11:47] <fragalot> what part don't you understand?
[22:12:16] <Guest978> my machine uses 2 switches per axis, so on one end I want to share home and limit
[22:12:30] <Guest978> using step config I wind up with a joint error
[22:12:44] <Guest978> as the home sequence ends
[22:13:18] <Guest978> stepconfig is set to both limits and home
[22:13:44] <fragalot> THAT part, i can't help with, my machine doesn't fully run yet, let alone have homing switches, so i can't really check..
[22:14:47] <Guest978> anybody else?
[22:20:44] <reelphoot> #emc-dev
[23:51:45] <skunkworks> jmkasunich: close. http://imagebin.ca/img/A4SFxgvk.png Still want to play with the opto side to get that all on one layer and away from eveything else.