#emc | Logs for 2008-05-28

[00:00:43] <JymmmEMC> SWPadnos: They made the conf artigo kits special (no cables). And no serial cables are available from VIA directly. His suggestion was to buy a pico board from fry's and that will come with the cables.
[00:07:44] <Hugomatic> jepler: Thanks a lot! my wifi is working now... I've been trying for hours
[00:13:48] <Hugomatic> Now for the fun stuff. I'm going to try to make PCBs with EMC and eagle, using: http://unpy.net/cgi-bin/viewcvs.cgi/eagle/ulp/gcode.ulp
[01:03:13] <jepler> Hugomatic: that's great news .. good luck with that ulp
[01:03:19] <jepler> you'll probably have some customizing to do ..
[01:23:54] <toastydeath> fff
[02:30:17] <tomp2> "The impact of jitter on stepper motor control is analyzed." Will Shackleford & Fred Proctor title: "Real-time Operating System Timing Jitter and its Impact on Motor Control" http://www.isd.mel.nist.gov/projects/rtlinux/motor-jitter.pdf
[02:36:48] <fenn> phoenix descends to the martian surface: http://hirise.lpl.arizona.edu/HiBlog/wp-content/uploads/PSP_008579_9020_descent.jpg
[02:38:44] <eric_U> don't look like it's gonna get out of that hole
[02:39:00] <fenn> it landed in front of the crater
[02:39:25] <fenn> super-duper telephoto lens :)
[02:40:57] <eric_U> this one is for TomP: http://www.rcuniverse.com/forum/fb.asp?m=7541729
[02:44:10] <eric_U> tomp2, where do you get the motors for your electric bicycles?
[02:58:35] <tomp2> eric_U: nice link, thx . i dumped the cyclone crap and got a unit from walmart, cheap and it actually works.
[02:58:56] <eric_U> cyclone was the batteries?
[02:59:00] <eric_U> or the motors?
[02:59:48] <tomp2> cyclone taiwan was the motor lifepo4 batteries chargers pas system, no bike
[03:00:23] <tomp2> brushless dc 500W 24V
[03:00:25] <eric_U> there is an alternative to a123 called K2, seem to be very good cells
[03:01:18] <tomp2> i'll be lookin, the walmart is cheap but sla battereies
[03:02:16] <fenn> i'm thinking d-cell ni-mh
[03:02:30] <eric_U> the a123's are so much better
[03:02:49] <eric_U> fast charging, high discharge rate with no cell degredation
[03:03:02] <fenn> what's the $/kWh or your preferred units
[03:03:31] <tomp2> http://zeva.com.au/tech/K2/
[03:03:36] <eric_U> what's a kWh?
[03:03:43] <fenn> kilowatt*hour
[03:03:52] <eric_U> never heard of it
[03:04:00] <eric_U> :)
[03:04:42] <eric_U> k2 is currently going fore 3.2v * 3200 mAh for $12
[03:04:49] <eric_U> I think
[03:06:03] <eric_U> that's a temporary price I think, normal price is 2500 mAh * 3.2v for $12
[03:06:34] <eric_U> a123 goes for $16.50
[03:06:49] <tomp2> 200wH/mile
[03:07:30] <tomp2> but k2 from Oz or Shanghai might add to cost
[03:07:50] <eric_U> you can get them from www.battlepack.com
[03:09:13] <eric_U> they are about the same pricing as a quality NiMH without the problems
[03:09:23] <tomp2> very cool, thx for that link :) ( now wth is a cnc battery? i imagine its for portable cnc's? )
[03:10:37] <eric_U> I think it is for memory backup, they are gel-cell bats
[03:10:47] <eric_U> http://www.youtube.com/watch?v=xfVX71YqPuQ a123 are very dangerous :)
[03:15:43] <fenn> ack.. didn't expect to see him get killed
[03:15:49] <eric_U> he's ok
[03:16:15] <tomp2> wow, i hope he's allright, the damn bike looked like it stuck itself right into the side of that van
[03:16:42] <eric_U> I'm pretty sure that was just precautionary, and he's ok
[03:17:27] <tomp2> now emc folks, that is acceleration
[03:17:36] <fenn> hope he learns his fucking lesson before he does get killed then
[03:17:56] <fenn> 1) wear a helmet
[03:18:02] <fenn> 2) make sure you have a helmet on
[03:18:33] <eric_U> 3) let the female assistant do the burnouts
[03:18:44] <tomp2> i bought a stupid helmet for my stupid 450W walmart after 1st ride (maybe not so stupid)
[03:18:52] <fenn> burnouts are dumb anyway
[03:18:54] <eric_U> gotta have a helmet
[03:19:10] <eric_U> burnouts serve a purpose on a drag vehicle
[03:19:26] <fenn> electric tire heater
[03:19:52] <eric_U> it is just promotion, no harm to it
[03:20:12] <eric_U> bill dube is posting on the killacycle blog, so he must be ok
[03:28:10] <tomp2> jmkasunich: thx, an lm318 fixed my board
[03:28:26] <jmkasunich> cool
[03:51:53] <toastydeath> oh no
[03:51:58] <toastydeath> this beer found its way to my desk
[03:52:02] <toastydeath> and is forcing me to drink it
[03:52:04] <toastydeath> help help
[03:52:15] <K`zan> NO!
[03:53:37] <K`zan> :-)
[05:29:10] <JymmmEMC> * JymmmEMC grabs the bottle from toastydeath and fills it with prune juice
[05:32:15] <K`zan> Well, I think I have stiffended up the Z axis stepper all that is possible with this arrangement, or very close to it and I still have about +/- 0.01 slop in the mechanism on stepper movement :-/.
[07:16:02] <ChanServ> [#emc] "This is the #emc channel - talk related to the Enhanced Machine Controller and general machining. Website: http://www.linuxcnc.org/, wiki at http://wiki.linuxcnc.org/"
[08:09:19] <pjm_> good morning
[08:12:25] <alex_joni> 'lo
[09:57:16] <acemi> was the RTAI kernel in 8.04 compiled with the kernel source from Ubuntu repo or from vanilla repo?
[10:01:28] <acemi> I'm trying to compile RTAI in Debian Lenny but the kernel source 2.4.24 from Debian repo is incompatible with RTAI. Many files are not in the expected places
[10:09:36] <fenn_> fenn_ is now known as fenn
[10:23:28] <acemi> I found that my problem is related with the kernel-headers prepared by make-kpkg. I can compile RTAI modules with kernel-source but not with kernel-headers.
[10:28:06] <alex_joni> acemi: I had a problem compiling RTAI with the headers generated initially
[10:28:17] <alex_joni> so I used an additional small patch to ipipe.h
[10:28:54] <alex_joni> (I used the Ubuntu kernel sources + the ADEOS/IPIPE patch)
[10:29:12] <alex_joni> hal-linux-2.6.24-x86-2.0-04 I think
[10:29:49] <alex_joni> acemi: also, make-kpkg doesn't seem to produce sensible packages anymore, Ubuntu uses a custom debian/ infrastructure to build packages
[10:29:54] <alex_joni> maybe debian does the same .. ?
[10:31:17] <acemi> there is no /arch/x86/Makefile_32 in the headers directory but there is in the compiled kernel directory
[10:33:04] <acemi> therefore I can't compile RTAI module if I didn't compile the kernel source in this machine
[10:34:59] <acemi> alex_joni: after patching the ipipe.h could you compile RTAI modules using kernel-headers?
[10:37:17] <alex_joni> yes
[10:37:37] <alex_joni> but the kernel headers produced by make-kpkg weren't good
[10:37:56] <alex_joni> only the ones I built with "fakeroot debian/rules custom-binary-flavour=rtai"
[10:37:56] <acemi> what is the alternative
[10:38:08] <alex_joni> (debian/rules is a big script from the Ubuntu kernel source)
[10:39:54] <alex_joni> maybe try getting the debian kernel sources with: apt-get source linux-image-...
[10:40:57] <acemi> does EMC have any problem with kernel-headers produced with make-kpkg?
[10:41:14] <acemi> for 2.6.24
[10:44:18] <alex_joni> I don't know..
[10:44:33] <alex_joni> but if rtai doesn't compile/install.. then it's useless to try emc2
[10:46:07] <acemi> I want to prepare linux-image and linux-headers and rtai modules deb packages in a machine and than I using this packages to compile EMC in anaother machine
[10:46:23] <alex_joni> acemi: right
[10:47:04] <alex_joni> I didn't follow debian way of building kernel packages lately
[10:47:27] <acemi> hmm
[10:48:11] <acemi> another question: RTAI is compatible with the standart ADEOS patch?
[10:48:55] <alex_joni> http://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s-common-official
[10:49:02] <alex_joni> acemi: usually yes
[10:49:13] <alex_joni> what RTAI are you trying?
[10:49:25] <acemi> 3.6.1
[10:49:27] <alex_joni> at the moment I would use 3.6.1 (just released a couple days ago), or vulcano CVS
[10:49:41] <alex_joni> ok, there's a 2.6.24 patch in there.. but I haven't tried it yet
[10:51:21] <acemi> why do you prefere adeos patch?
[10:51:57] <alex_joni> I don't prefer adeos patch
[10:52:01] <alex_joni> I used the one from RTAI
[10:52:07] <alex_joni> I think it's the same
[10:52:25] <acemi> I think there are not same
[10:52:57] <acemi> but not sure
[10:58:31] <alex_joni> I see Xenomai uses: adeos-ipipe-2.6.24-x86-2.0-07.patch
[10:58:45] <alex_joni> and for RTAI it is: hal-linux-2.6.24-x86-2.0-07.patch
[10:58:56] <alex_joni> but they differ a bit (content..)
[11:17:55] <micges> hello al
[11:19:49] <BigJohnT> hi
[11:21:37] <BigJohnT> micges: did you see the patch for blending of arcs for high speed cutting
[11:24:12] <micges> no
[11:24:51] <micges> BigJohnT: can you shortly describe bug that you found ?
[11:24:58] <BigJohnT> jepler did it yesterday, it makes G64Pn work on arcs too
[11:25:46] <BigJohnT> at high speed with short moves lines and arcs were not treated the same so the speeds of each were very different
[11:26:39] <BigJohnT> so if you are doing arc line arc line the speeds change so often that the whole machine shook
[11:27:42] <BigJohnT> now it runs very smooth cutting small arcs/lines at high speeds
[11:28:16] <micges> I see
[11:28:24] <micges> cool
[11:28:54] <micges> wonder how it's impact on my laser after update :)
[11:29:07] <BigJohnT> how fast do you cut?
[11:29:21] <micges> 1000 mm/min
[11:29:28] <micges> about
[11:30:00] <BigJohnT> should smooth out the motion when cutting small arcs/lines
[11:30:41] <micges> how small ?
[11:30:44] <BigJohnT> I had to play with the P number a bit and play with my cam software to get the smoothest motion
[11:31:13] <BigJohnT> about < 5mm
[11:33:23] <BigJohnT> do you cut out any art work or letters?
[11:33:47] <micges> yes letters
[11:34:15] <BigJohnT> then you will see an improvement when cutting letters
[11:34:44] <BigJohnT> the speed will be more consistent as you cut
[11:36:41] <micges> ok
[11:39:30] <BigJohnT> If you set the P tolerance on G64 you can control how the blending is done
[11:41:27] <micges> I know
[11:41:40] <micges> my laser run always in g64 p0.01
[11:42:28] <BigJohnT> cool
[11:48:47] <BigJohnT> alex_joni: does the feed override affect the velocity in EMC?
[11:48:59] <BigJohnT> http://www.cnczone.com/forums/showthread.php?p=457118#post457118
[11:51:56] <cradek_> BigJohnT: no, you can neverride past any configured machine constraints (velocity/acceleration)
[11:52:04] <cradek_> ...never override...
[11:52:13] <BigJohnT> that's what I thought
[11:52:16] <BigJohnT> thanks
[11:53:25] <cradek_> cradek_ is now known as cradek
[14:46:43] <skunkworks_> BigJohnT: MAX_FEED_OVERRIDE = 1.2
[14:47:29] <BigJohnT> ok, just when you thought you knew something :)
[14:59:51] <BigJohnT> Stompin' at the Savoy
[15:03:11] <SWPadnos> hey - I know that grapevine
[15:21:26] <BigJohnT> the Ink Spots
[19:45:32] <Hugomatic> I'm having an error with a generated NGC file: "start point too close to probe point". Since my Sherline machine does not have a probe, can I get EMC to ignore it?
[19:46:07] <cradek> have you programmed a G38 code?
[19:47:08] <Hugomatic> Hi cradek... I'm trying to use the gcode.ulp script for eagle
[19:47:37] <cradek> oh, it uses tool length sensing. you will have to take that out and use some other scheme for tool lengths.
[19:48:18] <Hugomatic> cradek: OK, thanks
[19:48:22] <skunkworks_> cradek: can you pull an older verion of the ulp out of your repository? (the older version didn't have tool length)
[19:48:33] <cradek> I'm not sure
[19:48:42] <cradek> should be easy to fix up your copy though.
[19:48:51] <cradek> I would hate to go back because there might be other things fixed.
[19:48:55] <skunkworks_> heh
[19:48:57] <Hugomatic> I'll try to comment out the printfs in the ulp file
[19:49:00] <cradek> you'd have to study the commit messages to figure it out.
[19:49:06] <cradek> Hugomatic: sounds good.
[19:50:22] <Hugomatic> Is the script expecting many tools? I'd figure an endmill for the traces and a drill for the holes, but the script has many of these probes... I wonder why
[19:51:01] <cradek> copper roughing, trace outlines, different drills, board cutout
[19:51:59] <Hugomatic> cradek: this is serious! thanks for the script
[19:52:10] <cradek> hope you get it going!
[20:53:45] <alex_joni> good night all
[20:53:55] <pjm_> gn
[20:57:08] <dmess> hi all..
[22:23:39] <[1]a-l-p-h-a> sup?
[22:34:37] <jepler> [1]a-l-p-h-a: everyone's switched to mach, the channel is closed.
[22:58:04] <SkinnYPup> lol
[23:00:38] <eric_U> it ain't funny at all
[23:01:49] <toastydeath> blebleble
[23:01:55] <toastydeath> POOL'S CLOSED
[23:02:42] <eric_U> that's funny, I'm about to go swimming
[23:12:19] <LawrenceG> anybody using sheetcam?
[23:13:06] <LawrenceG> it seems to run on linux under wine (at least it installs and starts)... I am looking for some sample parts to play with
[23:13:23] <toastydeath> okc needs a "no fatties" filter
[23:13:44] <toastydeath> there should be categories you can click on when you rate someone
[23:13:45] <eric_U> that would empty the place in my experience
[23:13:49] <toastydeath> oops
[23:13:55] <toastydeath> that was totally the wrong channel
[23:13:59] <toastydeath> but since i said it
[23:14:08] <toastydeath> it has been suggested before but the dev team doesn't listen
[23:14:37] <eric_U> I thought you were talking about Oklahoma City
[23:14:42] <toastydeath> oh, no
[23:14:56] <toastydeath> one of my friends works at this internet dating site and made everyone in a channel join
[23:15:37] <toastydeath> trolling it is the national pasttime
[23:45:16] <eric_U> I need to get a better caller ID so the phone doesn't ring for certain numbers