#emc | Logs for 2007-10-29

[00:23:02] <CIA-22> 03cradek 07TRUNK * 10emc2/debian/ (drivers.files.in emc2.files.in): fix minor packaging bitrot
[02:08:52] <robin_sz> meep?
[02:09:02] <robin_sz> * robin_sz bounces like a bouncy thing
[02:15:48] <jmkasunich_> why is robin bouncing?
[02:18:18] <ds2> * ds2 turns down the gain
[02:18:32] <ds2> stable loops better then bouncing ones ;P
[02:18:50] <SWPadnos> unless he's some cute chick in a bikini, I don't want to know about his bouncing ;)
[02:19:10] <SWPadnos> or his bouncy things, for that matter
[02:19:50] <CIA-22> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py:
[02:19:50] <CIA-22> fix display of error dialog when saving a gcode file to a read-only directory
[02:19:50] <CIA-22> thanks to cradek for catching this
[02:25:54] <CIA-22> 03jepler 07TRUNK * 10emc2/bin/.cvsignore: new items to ignore; sort
[02:26:24] <CIA-22> 03jepler 07TRUNK * 10emc2/docs/man/man1/.cvsignore: it made no sense to ignore these files; they aren't generated files
[02:28:03] <Jymmm> SWPadnos: Per your request... http://gfx.download-by.net/screen/20/20617-catherine-bell-sex-e-screensaver.jpg
[02:28:19] <SWPadnos> I
[02:28:26] <SWPadnos> I'm pretty sure that's not Robin
[02:28:41] <SWPadnos> so it's acceptable for me to know about her bouncy things
[02:28:44] <CIA-22> 03cradek 07TRUNK * 10emc2/configs/sim/axis_9axis.ini: fix up homing
[02:29:38] <Jymmm> lo
[02:29:40] <Jymmm> lol
[02:48:38] <CIA-22> 03jepler 07TRUNK * 10emc2/src/ (configure.in configure): use hal.so from the correct location (lib, not share)
[02:48:38] <CIA-22> 03jepler 07TRUNK * 10emc2/tcl/emc.tcl.in: use hal.so from the correct location (lib, not share)
[02:49:30] <CIA-22> 03jepler 07TRUNK * 10emc2/debian/emc2.files.in: additional file for package
[02:55:09] <CIA-22> 03cradek 07TRUNK * 10emc2/configs/common/core_sim9.hal: I want all these controls in AXIS for testing, and fix rotary jog speed
[02:55:10] <CIA-22> 03cradek 07TRUNK * 10emc2/configs/sim/axis_9axis.ini: I want all these controls in AXIS for testing, and fix rotary jog speed
[03:01:20] <CIA-22> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py:
[03:01:20] <CIA-22> * Fix shift-home (touch-off axis to zero)
[03:01:20] <CIA-22> * Document 4..8 shortcut keys to select axes
[03:04:09] <CIA-22> 03cradek 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: remove debug output
[05:14:58] <CIA-22> 03fenn 07TRUNK * 10emc2/src/hal/utils/meter.c: make single clicks (and dragging) select the signal to probe. double click closes the window. keyboard arrows dont work for some reason
[05:16:53] <SWPadnos> thanks!
[05:35:46] <CIA-22> 03fenn 07TRUNK * 10emc2/src/hal/utils/meter.c: get rid of apply and ok buttons since they're redundant. apply button didnt work with keyboard selection anyway. changed cancel to close since it doesnt keep your old value
[05:36:04] <fenn> i wonder why my close button looks different from cancel in halscope's signal picker
[05:41:25] <SWPadnos> do you mean close in halmeter vs. cancel in halscope?
[05:41:44] <fenn> yes
[05:41:49] <SWPadnos> it could be a default button thing
[05:42:04] <fenn> its made using the same calls
[05:42:12] <fenn> will change the name to cancel and see
[05:42:53] <SWPadnos> hmm. I'm not sure what ways the default button, selected button and any "cancel" button are hilighted, but that's where I'd look
[05:43:04] <SWPadnos> (based on Windows UI experience, not so much gtk)
[05:45:13] <SWPadnos> ok - time for bed. good night
[05:46:03] <fenn> night
[05:48:32] <fenn> well, i'm not that interested in it i guess..
[06:05:45] <Rabbitbunny> so is 0 on ecah axis supposed to be thie middle of the table? I had visualized it being one corner so I could work off a known point versus an imaginary one..
[06:05:50] <Rabbitbunny> each*
[06:09:15] <skinnypuppy34> For a mill likely center, for a router/plasma type system where home is at the two corners and you likely want to program all + XY values
[06:09:50] <Rabbitbunny> Is there a reason it's center on a mill? some usage i can't think of?
[06:10:35] <toastydeath> what?
[06:10:39] <skinnypuppy34> Just according to how you want to program.
[06:10:49] <toastydeath> i've never seen machine 0,0 anywhere but the corner
[06:10:55] <toastydeath> is that what you are discussing?
[06:11:08] <toastydeath> or are we talking work offset
[06:11:29] <Rabbitbunny> a 3axis mill, i've got some programs i'm running in sim mode that have 0,0,0 at the center of the table.
[06:11:31] <skinnypuppy34> I was thinking work offset
[06:11:47] <Rabbitbunny> i don't know what a work offest is..
[06:11:59] <toastydeath> work offset is where the program wants 0,0 to be
[06:12:39] <toastydeath> and like skinnypuppy34 said, you can put that anywhere on the table
[06:12:50] <Rabbitbunny> interesting.
[06:13:00] <toastydeath> and change it from program to program
[06:13:07] <toastydeath> or even within the program.
[06:13:17] <toastydeath> g54/55/56/etc
[06:13:29] <Rabbitbunny> so I could clamp something dow anywhere, measure where that is and set my work offest to line it up perfectly?
[06:13:35] <toastydeath> yes
[06:13:36] <Rabbitbunny> down*
[06:13:42] <Rabbitbunny> that's pretty cool.
[06:13:44] <toastydeath> you still have to align it to the two axes
[06:13:47] <skinnypuppy34> ;o) awfully handy if you keep a vice, rotary table and angle plate bolted down
[06:14:04] <toastydeath> but you can set one point as 0,0,0
[06:15:35] <Rabbitbunny> wow, this 'make anything' part is getting alot easier the mor I learn.
[06:15:42] <toastydeath> 0,0,0 is usually set to some feature that can be picked up again later
[06:15:48] <toastydeath> if you have to take it off and set it down again
[06:15:58] <toastydeath> you do not often put it off in space floating somewhere
[06:16:48] <skinnypuppy34> I'm about finished making my first stepper bracket for my mill, cut a .2xl pulley blank yesterday...
[06:17:13] <Rabbitbunny> i'm still learning to cast Al to make my mill...
[06:17:22] <skinnypuppy34> Cool !
[06:18:02] <skinnypuppy34> I was wondering what kind of foam cuts pretty clean for making lost foam core castings
[06:18:03] <Rabbitbunny> but, I've drawn it out in 3ds, and have EMC running in a vm. I'll be milling cool stuff soon enough
[06:18:55] <Rabbitbunny> lost foam is easy. I've got a full sheet of the good stuff, let me look at it to get a name for it, it's always been 'house foam' to me.
[06:19:07] <skinnypuppy34> Yeah I keep one in a VM ware machine to sim with
[06:19:59] <skinnypuppy34> I figured it was , we got an old cafeteria mixxer like 30 gallon to mull greensand... need to order greensand
[06:21:25] <Rabbitbunny> it says 'PERMA "R"' all over it with a penguin picture and a blue facing on the printed side. 3/4" thick stuff.
[06:21:37] <Rabbitbunny> I know a lot of guys like the 4" foam but they don't have it here
[06:22:00] <Rabbitbunny> yeah, I use a washing wachin for my sand.
[06:22:10] <Rabbitbunny> s/wachin/machine
[06:22:19] <skinnypuppy34> figured it would eat one of those
[06:22:38] <skinnypuppy34> did you fill the basket holes
[06:22:38] <skinnypuppy34> ?
[06:22:48] <Rabbitbunny> it's awesome how nice the castings come out. I have pictures in a photobucket somewhere.
[06:23:05] <Rabbitbunny> oh, no I took the basket with holes out to use for fishing.
[06:23:17] <skinnypuppy34> If you can dig em up I'd like to see
[06:25:08] <Rabbitbunny> let's see, the 3DS cnc: http://i232.photobucket.com/albums/ee146/Rabbitbunnyrandtan/butt2.jpg
[06:25:25] <Rabbitbunny> my first pour (dry sand): http://i232.photobucket.com/albums/ee146/Rabbitbunnyrandtan/DSC01199.jpg
[06:25:47] <Rabbitbunny> my second pour (damp packed sand): http://i232.photobucket.com/albums/ee146/Rabbitbunnyrandtan/DSC01204.jpg
[06:26:11] <skinnypuppy34> Just something I had been kicking in my head, is there a way to keep greatstuff expanding insullation from sticking to things?
[06:26:19] <Rabbitbunny> magic?
[06:26:35] <Rabbitbunny> saran wrap...
[06:27:11] <Rabbitbunny> but i don't think it burns very well, although it would be useful to move patterns from plaster to foam.
[06:27:21] <toastydeath> mold release wax
[06:27:27] <skinnypuppy34> magic more like voodoo ...
[06:27:30] <toastydeath> mold release wax
[06:27:42] <toastydeath> works on fiberglass, expanding foam, etc
[06:27:54] <skinnypuppy34> thanks toast
[06:27:59] <toastydeath> nm
[06:27:59] <toastydeath> *np
[06:28:15] <toastydeath> you usually put down something like aluminum foil
[06:28:21] <toastydeath> and put the mold release on top of that
[06:28:31] <toastydeath> should release really clean from the foil
[06:28:33] <toastydeath> and no mess
[06:29:15] <skinnypuppy34> yeah I had been wondering how to make lots of foam cores from plaster,
[06:29:40] <skinnypuppy34> Cool thanks for the tips again
[06:29:51] <toastydeath> used to use it doing car audio stuff
[06:29:57] <toastydeath> and they use a lot of fiberglass and expanding foam
[06:29:56] <toastydeath> so
[06:30:03] <toastydeath> i assume it will work a+ for your application
[06:30:04] <toastydeath> if not, sorry
[06:30:37] <skinnypuppy34> Hmmm I did a lot of interesting car audio shape blending with spandex and fiberglass
[06:31:10] <skinnypuppy34> And another amplifier showed up to be fixed last week
[06:31:28] <toastydeath> owned
[06:31:34] <Rabbitbunny> heh
[06:31:36] <toastydeath> spandex?
[06:31:46] <toastydeath> we used regular polyester fleece
[06:31:55] <toastydeath> not that i was involved for very long
[06:32:41] <Rabbitbunny> i only did it twice, but that's what I used too
[06:32:56] <skinnypuppy34> Yeah we'd cut up tupperware bowls glue them to a base form , spray glue and pull spandex across the form to the bowl part where the speaker was going to center
[06:33:11] <toastydeath> i saw some molds for sand casting that were fiberglassed
[06:33:15] <toastydeath> i thought that was pretty cool
[06:33:19] <toastydeath> finished really nice
[06:33:29] <toastydeath> painted red and everything
[06:33:34] <Rabbitbunny> cool.
[06:34:13] <Rabbitbunny> i liked the look of the ones lionel oliver did for his split molds, I'm using a 5 gallon can now.
[06:34:44] <skinnypuppy34> Sweet, btw I mounted my geckos on an old Rockford Fosgate Power 300 heatsink ... someone used tinfoil for fuses and the circuit board had a hole burned through it
[06:35:46] <toastydeath> lol
[06:39:57] <skinnypuppy34> http://imagebin.org/11378
[06:40:00] <skinnypuppy34> http://imagebin.org/11379
[07:13:44] <fenn> aw damn i missed a lost foam conversation
[07:15:00] <Rabbitbunny> i'm still here, but now EMC is laughing at my g-code usage.
[07:15:57] <fenn> the great stuff foam releases cyanide when you burn it. and it doesn't burn out too well either
[07:17:26] <alex_joni> :)
[07:18:10] <fenn> i would like to find something like the blue or pink stuff but without fire retardants
[07:19:18] <Rabbitbunny> cyanide... i know tha name for some reason, i'm thinking 'bad'. does it put bubbles in the pour or something?
[07:19:39] <fenn> puts bubbles in your blood, matey
[07:19:49] <Rabbitbunny> ...oh.
[07:20:03] <Rabbitbunny> * Rabbitbunny notes to not burn 'great stuf' foam.
[07:20:27] <fenn> apparently it doesnt give off too much since i havent heard of anyone dying or getting sick
[07:21:56] <alex_joni> cyanide is what they used on executions too.. right?
[07:22:12] <fenn> " It has also been used for suicide. Some notable cases are Erwin Rommel, Eva Braun, Adolf Hitler, Wallace Carothers, Hermann Göring, Heinrich Himmler, Alan Turing, and Odilo Globocnik."
[07:22:42] <fenn> or, basically everyone of importance in WWII that killed themself
[07:23:23] <Rabbitbunny> ...it stops your nerves from working. including your heartbeat.
[07:23:37] <Rabbitbunny> that's definatly in the 'bad' column.
[07:24:08] <fenn> er, where did you read that? that's not how it works
[07:24:15] <Rabbitbunny> wikipedia
[07:24:42] <Rabbitbunny> http://en.wikipedia.org/wiki/Cyanide under toxicity
[07:25:12] <fenn> ok, well, it stops everything else from working too
[07:25:57] <Rabbitbunny> in the scope of what it stops, i really care about my heartbeat the most, I could type with a thingy like Christopher Reeves had if anything.
[07:26:11] <fenn> no, you would die. FOOLS!
[07:26:22] <Rabbitbunny> lol
[07:27:31] <alex_joni> fenn: noticed I fixed the sitesearch?
[07:27:50] <alex_joni> (that was some really poor php programming :/)
[07:28:05] <anonimasu> 4hm
[07:28:06] <anonimasu> ^_^
[07:28:22] <anonimasu> alex_joni: how close is 2.2.x?
[07:30:28] <alex_joni> very close :)
[07:30:55] <alex_joni> but very close can mean anything
[07:31:01] <alex_joni> from a wekk to a couple months :D
[07:31:03] <alex_joni> week even
[07:32:24] <fenn> thanks for fixing that alex
[07:40:55] <Jymmm> Yay!
[07:41:34] <Jymmm> only took me a week and a fubar'ed online tutorial ti get my dns server up and runnng!
[07:42:46] <fenn> then you're ahead of the pack
[07:43:13] <Jymmm> fenn: lol, how so?
[07:43:21] <fenn> it only took a week!
[07:43:37] <Jymmm> that seems like a long time to me
[07:43:44] <Jymmm> but I wasn't rushing or anything.
[07:46:35] <Jymmm> fenn: This is a server in a DC, so I had to setup everything from scratch.
[07:57:02] <alex_joni> Jymmm: how's stuff over there?
[07:59:11] <skinnypuppy34> fenn thanks for the cyanide/greatstuff tip.
[08:01:06] <skinnypuppy34> Anyone run any mastercam posts in EMC? My teacher was asking, it may be worth it to fix the boss bp if so he said.
[08:01:34] <fenn> you can usually use a fanuc post
[08:02:14] <skinnypuppy34> Ok, I'll get him to send me something in fanuc post to see how it does.
[08:02:59] <fenn> rayh mentioned he had a mastercam post a long time ago
[08:03:06] <skinnypuppy34> I've got a demo version of X but you can't save or post
[08:03:47] <skinnypuppy34> I'll ask rayh then next time round
[08:03:47] <fenn> i guess this topic comes up every year or so
[08:04:00] <fenn> http://emc.sourceforge.net/NIST-archive/msg01106.html
[08:04:17] <fenn> the .pst file is at the bottom
[08:04:53] <fenn> er, a .bin file whatever that is
[08:05:35] <fenn> ok its just text
[08:06:16] <skinnypuppy34> Yeah just text thanks again fenn that looks useful
[08:07:20] <fenn> actually this seems to be the final version www.linuxcnc.org/Dropbox/EMCMpfan.zip
[08:07:33] <skinnypuppy34> Any new canned lathe cycles in the new EMC?
[08:07:43] <fenn> i know that folder is still kicking around somewhere
[08:08:25] <skinnypuppy34> We're sorry but the page you requested could not be found.
[08:08:27] <skinnypuppy34> www.linuxcnc.org/Dropbox/EMCMpfan.zip
[08:09:09] <alex_joni> here's the old dropbox: http://metalworking.net/Dropbox/
[08:09:09] <fenn> http://www.cnczone.com/forums/attachment.php?attachmentid=28370&d=1167713571
[08:09:13] <alex_joni> but that file isn't there
[08:09:36] <alex_joni> there was a post on our wiki too
[08:10:00] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cam_Post
[08:10:15] <alex_joni> maybe it would be nice if someone would add the mastercam one too
[08:10:20] <Jymmm> alex_joni: which stuff?
[08:11:15] <alex_joni> dunno.. saw some fires on the news in cali
[08:11:29] <Jymmm> alex_joni: Ah, the fires are mostly in SoCal
[08:11:35] <Jymmm> * Jymmm is in NoCal
[08:12:21] <alex_joni> ok, that's good I guess :)
[08:12:47] <skinnypuppy34> Gotta get some Z's it's 4AM here. GNite everyone
[08:13:02] <alex_joni> g'night
[08:13:08] <fenn> skinnypuppy34: i expect a full report tomorrow :)
[08:13:30] <Jymmm> alex_joni: some friends have been evacuated just due to the smoke alone miles away from any fire
[08:13:42] <skinnypuppy34> errrr on the mastercam fenn?
[08:13:43] <fenn> yeah
[08:14:12] <fenn> not for me, just so i can blame you when i say "yeah this file works, skinnypuppy said so"
[08:14:45] <skinnypuppy34> I should be able to get him to e-mail me a post tomorow. I'll be on in another 10 hours or so.
[08:15:01] <Jymmm> alex_joni: I wasn't sure which "stuff" you were tlaking about... my email being out for a week, setting up the new server, fires, etc.
[08:15:20] <alex_joni> Jymmm: something along those lines ;)
[08:15:27] <alex_joni> didn't know about email ..
[08:16:34] <Jymmm> alex_joni: I've resolved that finally. Been working on getting BIND up and going, fubared tutorial I had been reading, just got it functional tonight. Now waiting for the propigation.
[08:23:34] <fenn> that LRK motor is neat
[08:29:39] <alex_joni> Jymmm: cool
[08:29:51] <alex_joni> I usually set the times very low, before DNS changes
[08:30:12] <alex_joni> let those propagate first, then do the changes, then the new stuff will propagate very fast
[08:30:37] <Jymmm> alex_joni: Well, this is a full glorified server... nobody does my dns (upstream), I'm handling it all myself.
[08:30:47] <alex_joni> coo
[08:30:59] <Jymmm> big learning exp, I'll say that much
[08:32:39] <fenn> your dns entry should still propagate to various caches around the world
[08:33:48] <Jymmm> they do, but a WHOLE LOT DIFFERENT than asking youtrr upstream to do it for you.
[08:35:07] <fenn> yep
[08:35:49] <Jymmm> ORA DNS & BINS 5th Edition is worht it's weight in gold =)
[08:39:13] <fenn> i wonder how they came up with the power rating on the LRK
[13:00:18] <xemet_> hello
[13:00:42] <xemet_> still trying to get emc2 working on 7.04 (maybe impossible...)
[13:01:20] <xemet_> situation: I've the RTAI kernel working (i think) I can run all the testsuite scripts without problems
[13:01:27] <xemet_> latency seems to be ok
[13:01:44] <xemet_> compiled EMC2-TRUNK version without problems
[13:02:52] <xemet_> when I try to start EMC2 after picking the configuration I get a lot of scrolling vertical dots and after, the message:
[13:03:04] <xemet_> .
[13:03:04] <xemet_> Can't write to /dev/rtai_shm - aborting
[13:03:04] <xemet_> Realtime system did not load
[13:03:04] <xemet_> Shutting down and cleaning up EMC2...
[13:03:04] <xemet_> RTAPI: ERROR: could not open shared memory (errno=2)
[13:03:04] <xemet_> HAL: ERROR: rtapi init failed
[13:03:06] <xemet_> halcmd: hal_init() failed: -9
[13:03:08] <xemet_> NOTE: 'rtapi' kernel module must be loaded
[13:03:12] <xemet_> RTAPI: ERROR: could not open shared memory (errno=2)
[13:03:14] <xemet_> HAL: ERROR: rtapi init failed
[13:03:14] <xemet_> halcmd: hal_init() failed: -9
[13:03:16] <xemet_> NOTE: 'rtapi' kernel module must be loaded
[13:03:18] <xemet_> Cleanup done
[13:03:20] <xemet_> EMC terminated with an error. You can find more information in the log files
[13:03:22] <xemet_> /home/xemet/emc_debug.txt
[13:03:25] <xemet_> and
[13:03:26] <xemet_> /home/xemet/emc_print.txt
[13:03:28] <xemet_> as well as in the output of the shell command 'dmesg' and in the terminal
[13:03:47] <xemet_> if someone has suggestions they are welcome...
[13:04:02] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting#RT_device_files
[13:05:51] <xemet_> thanks!
[13:05:55] <SWPadnos> sure
[13:06:12] <xemet_> so, if I type: ls -l /dev/rt*
[13:06:15] <xemet_> I get:
[13:06:28] <SWPadnos> there's a rules file as well "emc2.rules" I think
[13:06:32] <xemet_> crw-rw---- 1 root audio 10, 135 2007-10-29 12:33 /dev/rtc
[13:06:34] <SWPadnos> that goes in the udev director
[13:06:40] <SWPadnos> y
[13:08:22] <xemet_> so, what should I do? In the wiki I read: If these files do not exist, you can create them, for example: sudo mknod /dev/RTAI_SHM c 10, 254;
[13:08:59] <xemet_> I tried, but if i type : ls -l /dev/rt* again, I get the same result
[13:09:10] <SWPadnos> rt is lower case
[13:09:14] <SWPadnos> RT is upper case ;)
[13:09:28] <SWPadnos> ls rt* won't show RTAI_SHM
[13:09:40] <xemet_> ops
[13:09:43] <SWPadnos> heh
[13:10:00] <SWPadnos> I've got to run. check this wiki page as well: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?RtaiSteps
[13:10:16] <acemi> ttp://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Debian_Etch_Compile_RTAI#Udev_configuration
[13:10:20] <acemi> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Debian_Etch_Compile_RTAI#Udev_configuration
[13:10:57] <SWPadnos> I don't remember where it is, but there is a file that's supposed to be put in the /etc/udev/rules (or rules/d) directory
[13:11:37] <xemet_> ok, thanks, now I read that page (I know it, but I've used another page to installa RTAI in 7.04)
[13:15:15] <xemet_> GREAT
[13:15:19] <xemet_> now it works!
[13:15:24] <xemet_> thank you very much!
[13:15:39] <xemet_> now I've only get axis work...
[13:15:48] <xemet_> I think I forgot to install tkinter
[13:20:32] <xemet_> ok, now I think I have a new problem!
[13:20:33] <xemet_> :)
[13:20:47] <xemet_> if i try to start the stepper_mm configuration
[13:20:48] <xemet_> I get:
[13:21:03] <xemet_> Starting EMC2...
[13:21:03] <xemet_> insmod: error inserting '/home/xemet/emc2-trunk/rtlib/hal_parport.ko': -1 Device or resource busy
[13:21:03] <xemet_> standard_pinout.hal:4: exit value: 1
[13:21:22] <xemet_> I guess maybe I should do something with the parallel port...
[13:21:27] <xemet_> any suggestion?
[13:23:10] <acemi> did you disable the parallel port module?
[13:23:34] <xemet_> no...
[13:23:41] <xemet_> how to do that?
[13:23:49] <acemi> see this http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Debian_Etch_Compile_EMC2#Disabling_parallel_port_module
[13:24:02] <xemet_> thank again...
[13:24:14] <acemi> or for now, sudp rmmod parport_pc
[13:24:47] <acemi> sudo rmmod parport_pc
[13:24:47] <xemet_> I reboot
[13:29:08] <xemet_> ok, thanks you again! Now everything seems to work
[13:29:28] <xemet_> acemi: are you the one who installed emc on 7.10?
[13:29:48] <acemi> I use debian
[13:29:52] <acemi> etch
[13:30:04] <Guest998> Jepler made the packages for 7.1 I have tested them a bit.
[13:30:15] <Guest998> 7.10
[13:30:22] <Guest998> Guest998 is now known as skunkworks_
[13:32:01] <xemet_> ok, I asked, because I've read on the mailing list today that someone got it works on the 7.10, but I don't remember the name
[13:32:30] <xemet_> now I've to try with the real machine....
[13:32:36] <xemet_> and see how it works
[13:32:46] <xemet_> question
[13:33:29] <xemet_> in the 6.06, with EMC2 installed using the original packages, when I would like to use Halcmd I was used to type: /etc/init.d/realtime start
[13:33:41] <xemet_> and after halcmd -kf
[13:34:02] <xemet_> If I would like to start realtime now in the 7.04 how should I do?
[13:45:48] <alex_joni> the same
[13:46:07] <alex_joni> but if emc is not installed, then you use scripts/realtime start
[13:46:20] <alex_joni> or simply halrun
[13:48:05] <xemet_> thank you
[14:22:00] <anonimasu> hm
[14:23:44] <anonimasu> did anyone cast plastic of yuou?
[14:23:45] <anonimasu> you?
[15:15:08] <xemet> hi
[15:16:02] <xemet> curiosity: http://axis.unpythonic.net/01190912545 where do I find this EMC2 HAL latency test window?
[15:23:57] <skunkworks_> what happens when you type in latency-test?
[15:24:23] <skunkworks_> as far as I know - it is only in trunk
[15:28:21] <xemet> ok, I've the trunk
[15:28:32] <xemet> what should I write and where?
[15:34:40] <jepler> if run-in-place, ". scripts/emc-environment". Then all emc commands such as emc, halrun, and latency-test are available by simply typing the command name at the shell prompt: latency-test
[15:38:49] <xemet> ok thanks, tested
[15:39:03] <xemet> what exactly "max jitter" means?
[15:39:21] <xemet> I think my value is high...but I don't know why
[15:43:48] <jepler> xemet: jitter is the variation in the time between executions of the real-time code. For instance, instead of executing base-thread at t=0, t=50, t=100 (microseconds), jitter means that the actual times might be t=0, t=55, t=100. If a pulse was created by the first period, it's 5uS too long (55 instead of 50); if a pulse is created by the second period, it's 5uS too short (100-55=45 instead of 50)
[15:44:15] <xemet> ok thank
[15:44:47] <xemet> I'm now on the 7.04, I get a max jitter of 20000 ns for the base period that is of 25000
[15:45:51] <xemet> it is really high, isn't it?
[15:46:07] <jepler> imagine that your stepper driver requires that 'dir' be stable for 6uS before 'step' rises. If you get t=0, t=45, t=50 (should be t=0, t=25, t=50 but with 20uS jitter added to the second time) and dir changes at t=45, you have not met the timing requirement of your stepper driver because dir is only stable for 5uS before dir changes.
[15:47:29] <jepler> to get correct behavior, you would either increase BASE_PERIOD or you would increase stepgen.N.dirsetup.
[15:47:34] <xemet> ok, I should try this test on the other system with the 6.06, the one I've always used with my machine and see what happens
[15:47:53] <xemet> are there ways to reduce max jitter?
[15:48:23] <xemet> I've used a 2.6.20 kernel and installed RTAI
[15:48:36] <jepler> this page discusses some things that cause high latency or jitter: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting#RTAI_Latency_test
[15:48:44] <xemet> RTAI latency test said a max latency of about 18000
[15:49:22] <xemet> with other patches I tried and kernel 2.6.19 the max latency was (on the same hardware) 12000
[15:49:27] <jepler> this page gives more details about the relationship between base_period, latency/jitter, and step rate: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TweakingSoftwareStepGeneration
[15:49:56] <jepler> not all versions of the rtai-patched kernel are equally good. It seems you are having first-hand experience of this.
[15:50:55] <xemet> yes...think that it was the first time in my life I compiled the kernel...
[15:52:13] <xemet> but with the 2.6.19 RTAI patched on the 7.04 EMC2 didn't work...so I tried the 2.6.20
[15:52:57] <jepler> that's one of the reasons that we support (pre-build kernel and emc for) so few versions of linux (only one or two versions of one distribution) -- it's a lot of work to build a rtai kernel that has decent performance on a wide range of machines, and with every different kernel version the "tricks" needed vary..
[15:53:23] <xemet> I understand
[15:54:01] <xemet> according on what is in the wiki, whan I run the RTAI latency test, the most important value is "ovl max"
[15:54:12] <xemet> is that the "max jitter"?
[15:54:47] <jepler> yes those measure essentially the same thing. It's a difference in terminology.
[15:57:51] <xemet> according to the RTAI test I have a ovl max of about 33000ns
[15:58:59] <xemet> I've to try with the real machine and see what happens
[15:59:21] <xemet> maybe it could be good because I do not use high feedrates
[16:00:34] <jepler> the spreadsheet file on the Tweaking Software Step Generation page can help you find out what your max step rate is for a given latency/jitter value.
[16:00:57] <xemet> This motherboard has an nvidia integrated video board...so I think I'm in the worst case possible...
[16:16:16] <cradek> best to just disable it and plug in an old pci card
[16:21:12] <xemet> yes the problem is that the motherboard is a small Mini-ITX
[16:21:22] <xemet> the video card is almost big as the mother board
[16:21:46] <cradek> did you try the vesa video driver yet? that's probably the only way it will work, but it's slow
[16:22:02] <xemet> my machine however run very well on this system...probably because I don't need high feedrate
[16:22:22] <xemet> the video card works pretty well
[16:22:33] <xemet> I don't think I need vesa
[16:23:10] <xemet> and I don't get the annoying unexpected realtime error I was always getting with the 6.06
[16:23:35] <cradek> good, if you don't get that error, it's probably ok
[16:24:33] <xemet> maybe it is only a slow computer...it is a PIII 933 MHz
[16:25:28] <xemet> the ovl max is 33000 as measured from RTAI test, but I have a 50000 base-thread and everything seems to woerk well
[16:26:15] <xemet> I will publish a page with a step by step guide to get EMC2 working on the 7.04 so people who want to try can get help
[16:26:50] <xemet> thoug I think it is very similar to the one in the wiki for the 6.10
[16:46:21] <skunkworks_> there is a wiki on 6.10?
[16:50:17] <skinnypuppy34> xemet: I had a MB with onboard nvidia agp, latency went way down just putting in an old pci video.
[18:04:51] <alex_joni> fenn: http://www.etherlab.org/en/success/hexapod.php
[18:41:04] <skunkworks_> Guest570: Hi
[18:42:17] <Guest570> hello, anyone help me with motor tuning ? My table travel only 20 milimetres instead of 25.4 ( 1 inch)
[18:42:25] <alex_joni> Nazdar
[18:42:40] <alex_joni> Guest570: what kind of motors?
[18:42:42] <Guest570> nazdar
[18:42:54] <Guest570> ty ses cech ?
[18:43:41] <alex_joni> Guest570: nope, that's about all I know :)
[18:43:56] <Guest570> alex_joni: sanyo denki stepper motor 1.8 deg/step
[18:44:33] <skunkworks_> any gear reduction between the motor and the leadscrew? How many threads per mm?
[18:45:07] <skunkworks_> how many steps per rev (are you full, half, micro)
[18:47:24] <Guest570> no gear reduction, 1mm = 200 rev full
[18:48:58] <Guest570> and 200steps per rev
[18:50:06] <skunkworks_> soo - your input scale is 40000?
[18:50:31] <skunkworks_> 200rev for 1mm -> 200 steps per rev = 40000
[18:51:10] <cradek> I doubt the leadscrews have 200 threads per mm...
[18:51:33] <skunkworks_> that would be a fine thread :)
[18:51:56] <alex_joni> 1mm = 200 revs?
[18:51:57] <skunkworks_> so most likely his input scale is 200
[18:52:03] <Guest570> i have input scale 4000
[18:52:18] <alex_joni> Guest570: lets start again
[18:52:23] <skunkworks_> heh
[18:52:27] <alex_joni> you said motor has 200 steps/revolution
[18:52:30] <Guest570> yes
[18:52:31] <alex_joni> right?
[18:52:35] <Guest570> yes
[18:52:35] <alex_joni> ok..
[18:52:38] <alex_joni> do you drive microsteps?
[18:52:42] <alex_joni> or full steps?
[18:52:57] <Guest570> i drive microsteps
[18:53:02] <alex_joni> how many?
[18:53:38] <alex_joni> (usual values are 2, 5, 8, 10)
[18:53:59] <Guest570> 2
[18:54:21] <alex_joni> ok, so you need 400 steps from the PC in order for the motor to turn one revolution
[18:54:31] <Guest570> yes
[18:54:39] <alex_joni> so far so good
[18:54:56] <alex_joni> (to test this you can put INPUT_SCALE = 400)
[18:55:08] <alex_joni> then if you do G1x1 it should go exactly one turn
[18:57:49] <JymmmEMC> http://www.youtube.com/watch?v=88w9ybua4i0&NR=1
[18:59:19] <skinnypuppy34> BOO ! !
[18:59:36] <skinnypuppy34> Goodone Jymmm
[18:59:40] <skunkworks_> * skunkworks_ thought JymmmEMC didn't like youtube..
[19:01:12] <JymmmEMC> skunkworks I dont, but I trust one or two ppl that send me a link. non of which are in this channel ;)
[19:03:40] <alex_joni> Guest570: still there?
[19:03:45] <Guest570> yes
[19:03:53] <Guest570> testing
[19:31:02] <alex_joni> * alex_joni kicks CIA-22
[19:31:08] <alex_joni> no ow?
[19:39:08] <jepler> he's dead, jim
[19:51:31] <Hugomatic> Hi, I'm having trouble finding info on the EMC tool table. The "tool table page" in the pdf seems to be missing from the user manual, and its not mentioned in the wiki afaik. Can anyone tell me the file and where it is? thanks.
[19:52:20] <cradek> Hugomatic: is this for a lathe or mill?
[19:52:31] <Hugomatic> Its for a mill
[19:53:43] <cradek> http://www.linuxcnc.org/docs/devel/html/gcode_tool_compensation.html
[19:55:00] <Hugomatic> Cradek: thank you
[19:55:40] <cradek> I think that stuff should be in the pdf too...?
[20:04:22] <Hugomatic> Cradek: Maybe the Machine menu could let you know which file is the current tool table file (could be easier than to update the documentation). I'm still a little confused, since my ini file doe not refer to any var file (except for stepper.var).
[20:05:09] <cradek> In emc 2.2, there will be an entry on the File menu to edit the tool table
[20:05:22] <cradek> but I don't understand your comment exactly (var file etc)
[20:05:40] <cradek> the tool table and var file aren't really related
[20:05:53] <Hugomatic> Cradek: From the docs: The EMC uses a tool file that is read in when a machine control is started. In a standard release this file is named emc.var, generic.var, or sim.var and is used by the similarly named run file. The specific name of the file that will be used is set by the ini file that is read at startup.
[20:06:20] <cradek> oh, that's entirely bogus
[20:07:02] <rayh> That description sounds like an EMC(1) rather than EMC2 document
[20:07:11] <cradek> the tool table is something.tbl
[20:07:31] <rayh> However the var and tbl files used are specified in the ini still.
[20:07:43] <cradek> right, those docs are just wrong
[20:07:48] <cradek> Hugomatic: now I understand the confusion!
[20:08:28] <Hugomatic> Cradek: I'm still wondering what file name should I use, and if can I leave it in the same dir as my ini file?
[20:09:46] <Hugomatic> Cradek... I have found a stepper.tbl. I think I have it now. Thanks
[20:09:56] <jepler> The tool table file is specified by [EMCIO]TOOL_TABLE in your inifile
[20:09:57] <rayh> hugomatic.var and hugomatic.tbl
[20:10:30] <Hugomatic> Thanks to all of you
[20:17:54] <jepler> I'll make sure this is improved in the documentation for emc 2.2.
[20:20:00] <cradek> I notice the lathe tool table isn't described there either
[20:23:14] <Hugomatic> Another question about tools: I have a metric machine and I use both g20 and g21, sometimes in the same program. Can I specify the unit in my tool table?
[20:23:44] <alex_joni> the tool table (in emc2.2.x) will be in machine units
[20:23:45] <cradek> you are going to have big problems doing that in emc2.1. In emc2.2 it all works properly.
[20:24:53] <cradek> emc2.2 will be released "any day now"
[20:24:57] <Hugomatic> Cradek: EMC2.2 rocks!
[20:25:24] <cradek> in emc2.1, if you are using the tool table, I recommend not ever switching units
[20:27:19] <Hugomatic> Cradek: all right, I'll keep that in mind while I play with it...
[21:25:30] <skunkworks_> so - has anyone looked at the holmes comet? http://www.spaceweather.com/comets/gallery_holmes_page5.htm?PHPSESSID=3qdppmvk7526t0o05gqk3pl071
[21:25:43] <skunkworks_> should be able to see it with binoculars
[21:37:53] <dmess> hi all