#linuxcnc | Logs for 2013-01-06

[02:30:46] <DJ9DJ> moin
[06:35:53] <yskapell> hello
[06:36:04] <yskapell> where can I download a trial for emc powerpath?
[06:39:30] <cncbasher> yskapell: sorry your in the wrong irc channel for that , nothing to do with linuxcnc
[06:41:31] <ReadError> scumbag emc
[06:42:25] <yskapell> sorry
[06:56:50] <IchGuckLive> hi all B)
[09:43:58] <skunkworks> Mmmm blueberry pancakes
[11:12:58] <mikeggg> haha. we just finished blueberry pancakes ourselves
[11:13:38] * JT-Shop tries real hard to not post this response "Best hint I have is RTFM then ask a real question..."
[11:14:02] <archivist> go on post it I dare you
[11:14:15] * archivist ducks
[11:14:54] <archivist> and dreams of pancakes with a sprinkle of sugar and some lemon
[11:15:08] <ReadError> http://www.cncrouterparts.com/crp2448-2-x-4-cnc-router-kit-p-164.html
[11:15:13] <ReadError> opinion? yay or nay ?
[11:18:02] <archivist> nay for me, price and it is rack and pinion therefore can have backlash
[11:18:33] <ReadError> whats a better option you are aware of?
[11:19:07] * JT-Shop tosses a herring to the east
[11:20:11] <archivist> for what definition of better :)
[11:20:30] <ReadError> a decent router kit
[11:20:37] <ReadError> i mean i could source everything myself i suppose
[11:20:57] <JT-Shop> good, fast, cheap never go in the same sentance :)
[11:21:34] <Tom_itx> pick 2
[11:24:15] <JT-Shop> we seem to be getting a rash of strange replies in the forum
[11:29:50] <ReadError> im looking for a decent sized cutting area
[11:30:00] <ReadError> the backlash that comes from the rack/pinion
[11:30:09] <ReadError> is it consistent to where it can be dialed out via software?
[11:30:30] <L84Supper> by use of an encoder
[11:31:26] <archivist> backlash can be the cause of cutter breakage
[11:31:39] <andypugh> I think that you can arrange for the rack-pinion to have fairly low backlash by running deeper mesh, for example.
[11:31:43] <archivist> especially if climb milling
[11:31:51] <L84Supper> http://www.automationtechnologiesinc.com/products-page/cnc-router/kl6045-desktop-cnc-router-with-3-stepper-motors
[11:31:57] <ReadError> well it has a tensioning spring
[11:32:06] <ReadError> i figured that would dial out a good portion of it
[11:32:12] <JT-Shop> isn't there some scheme to have two gears on the rack to reduce backlash
[11:32:45] <andypugh> Anti-backlash gears might work. Two gears with a spring to twist one relative to the toher.
[11:33:15] <andypugh> Double helical gears would let you eliminate backlash simply by biasing the gear sideways.
[11:33:41] <andypugh> Not that I have ever seen doubel-helical rack.
[11:33:45] <IchGuckLive> hi all
[11:34:01] <IchGuckLive> i got realy to mutch time ! Sunday -> http://www.youtube.com/watch?v=QOqn0pIVZeo
[11:34:05] <andypugh> I like the Bell-Everman Servobelt idea, and that is what I would use.
[11:34:21] <IchGuckLive> Spring analysis on mittelpoint behevior of ProE
[11:34:24] <L84Supper> that's the fun of engineering, how to do it, comparing all the variables
[11:34:41] <L84Supper> and whatever you value your own time to be
[11:34:42] <JT-Shop> is that the one that has a fixed belt and the second belt rides in the teeth of the other?
[11:35:40] <ReadError> http://www.cncrouterparts.com/images/R&P.jpg
[11:36:34] <L84Supper> giant spring on pinion, just keep it heavily loaded to one direction
[11:36:58] <L84Supper> replace gear and motors as required :)
[11:37:01] <andypugh> JT-Shop: That's the one.
[11:37:03] <ReadError> http://www.cncrouterparts.com/images/IMG_0747_.JPG
[11:37:19] <ReadError> thats the spring to keep the mesh tight
[11:37:22] <JT-Shop> that is a pretty neat system and fast IIRC
[11:39:12] <L84Supper> http://shop70018708.taobao.com/ you're making a cheap router for soft parts, not a precision mill
[11:40:16] <archivist> squiggles and no pics here
[11:40:25] <andypugh> It's very clever. Pretty much as rigid as a rack and pinion, but no backlash. Because the static belt is bonded to the rails it's like having a very short belt, and so there is not much spring.
[11:40:32] <andypugh> http://www.bell-evermannews.com/wp-content/gallery/biomedical-servobelt/servobelt-mini-gantry-assembly-closeup.jpg
[11:40:55] <ReadError> interesting andypugh
[11:41:11] <L84Supper> just more flex to the belt
[11:41:33] <andypugh> It looks to be successfully making the transition from "interesting hobby idea" to "profitable commercial product"
[11:41:52] <L84Supper> just needs a nice composite belt
[11:42:09] <andypugh> One cool thing about the servobelt is that it can go round curves. Possibly variable-radius curves.
[11:43:49] <andypugh> Also, because there is a belt in the belt, it should be pretty resistant to contamination.
[11:47:21] <JT-Shop> http://bbs.homeshopmachinist.net/threads/57407-Manual-Bridgeport-Series-One-2J-Head-CNC-Retrofit
[11:49:29] <L84Supper> http://www.bell-everman.com/whats-new/bell-everman-news/item/72-servonut-replaces-linear-motor-on-large-format-printer servo-nut vs servo-belt
[11:50:34] <ReadError> http://www.ebay.com/itm/1-5KW-WATER-COOLE-MILLING-AND-GRINDING-SPINDLE-MOTOR-INVERTER-DRIVE-VFD-f6-/230715508514?_trksid=p2047675.m2109&_trkparms=aid%3D555003%26algo%3DPW.CAT%26ao%3D1%26asc%3D154%26meid%3D4687955620156391538%26pid%3D100010%26prg%3D1091%26rk%3D4%26sd%3D321049786432%26
[11:50:45] <ReadError> seems nice
[11:51:07] <IchGuckLive> finish for the polasma construct today first Freecad use
[11:51:35] <L84Supper> from China, haven't heard any complaints yet from anyone that's used them
[11:52:09] <L84Supper> other than they overrate the capacity, so get one oversized
[11:52:22] <ReadError> L84Supper: on that spindle ?
[11:53:28] <L84Supper> ReadError: you found a US supplier of those Chinese spindles
[11:53:57] <Tom_itx> how are the bearings in them?
[11:54:12] <L84Supper> scroll down the page for the source
[11:54:16] <andypugh> They claim that the bearings are from Germany.
[11:54:28] <IchGuckLive> http://mechmo.de/plasma.jpg my plasma drawing
[11:54:43] <IchGuckLive> german bearing is good
[11:54:44] <andypugh> Which is interesting, as I didn't know the Germans were big in bearings. Sweden, France, UK and Japan, yes.
[11:54:50] <IchGuckLive> comes all from china
[11:54:52] <skunkworks> Gene45: we picked the seros/amps/powersupplies by what was the least expensive. The servos came out of a 80's vintage HMC - 8 brush high torque. (currently with the drives we have - 40ft-lbs peak) Amc drives off ebay.
[11:54:55] <L84Supper> ShanHai Hangzhou Medical Equipment Plant
[11:55:29] <L84Supper> they are about a hours drive away
[11:55:33] <Tom_itx> german bearings made in china?
[11:55:34] <Tom_itx> :)
[11:55:52] <L84Supper> I will get around to visiting them by the spring
[11:56:10] <IchGuckLive> Madler the bigest suplier of Norms wait every day for shipment of out stock parts
[11:56:25] <IchGuckLive> telling costemer its on the ship !
[11:57:22] <andypugh> It's hard to say where bearings are made, as the big manufaturers (SKF, RHP, NTN, FAG) all make bearings for each other. It works out better to make 4 x the number of 1/4 the range each.
[11:57:24] <L84Supper> "This spindle motor adopts Germany imported bearing" so what adopts actually means i don't know
[11:57:29] <Tom_itx> 8k-24krpm is a bit much for some milling jobs
[11:58:31] <L84Supper> could be a German bearing or maybe they had Weisebeer for lunch when they made them
[12:01:11] <L84Supper> http://www.soigeneris.com/smoothstepper-details.aspx why would anyone prefer this to Linuxcnc + Mesa?
[12:02:10] <archivist> they use windows on a laptop
[12:02:16] <ReadError> Tom_itx
[12:02:23] <ReadError> im going to be cutting CF/G10 mostly
[12:02:39] <ReadError> so i think a higher rpm would be better?
[12:03:03] <L84Supper> ReadError: making PCB's?
[12:03:08] <skunkworks> L84Supper: because it isn't linux
[12:03:15] <ReadError> L84Supper: ive done some yea
[12:03:17] <L84Supper> heh
[12:03:22] <ReadError> but i would rather just do that on my taig
[12:03:34] <ReadError> maybe some aluminum work
[12:03:45] <andypugh> It might be useful to re-structure LinuxCNC to allow buffering of motion commands to those devices.
[12:04:48] <skunkworks> andypugh: yeck
[12:04:50] <skunkworks> ;)
[12:04:53] <andypugh> The only practcal drawback is that you only get the features included in the board firmware. (so, for example, probing and threading need to be controlled by the board, not the PC)
[12:05:18] <L84Supper> something like a single axis SLA printer doesn't require any realtime, just short slow moves
[12:05:24] <skunkworks> plus feed rate override, feed hold and such need to be re-thought
[12:06:35] <L84Supper> so just STEP, Direction and Enable for 1 axis
[12:06:59] <L84Supper> but the smoothstepper is ~$160 ea
[12:08:30] <L84Supper> just wondering, I see strange comments in #reprap
[12:08:31] <IchGuckLive> hi i did it (view on freecad ) http://www.youtube.com/watch?v=V6kllGspehQ
[12:08:51] <IchGuckLive> plasma freecad turning
[12:13:29] <L84Supper> how to get Freecad running on a Mac?
[12:13:50] <L84Supper> already does, nevermind
[12:15:02] <DJ9DJ> lol "ich finds so schön wenn'er rumdreht" :D
[12:15:32] <DJ9DJ> freecad looks nice
[12:16:19] <pcw_home> Probably the cleanest buffered solution would be to move all of realtime = HAL/motion/some of task) to the remote device. This isgetting more practical all the time. This would require some untangling of RT/non RT tasks
[12:26:23] <L84Supper> I'm just thinking ahead for LinuxCNC controlled 3d printers that combine tech such as laser, inkjet and spindles
[12:26:52] <skunkworks> why not?
[12:28:28] <L84Supper> real time + non-realtime motion and process control via comedi
[12:29:33] <L84Supper> rather than heading down some new odd path as some reprap applications have done
[12:29:35] <IchGuckLive> DJ9DJ: B)
[12:29:51] <DJ9DJ> :)
[12:30:12] <IchGuckLive> the timelaps between press and turning and then not fix it maes moves in betrwen
[12:34:34] <IchGuckLive> soomeone here with portugise language Brazil ?
[12:39:16] <L84Supper> who makes nice small <1HP planetary gearboxes?
[12:40:29] <L84Supper> http://www.ccmmade.com/pic/jufei2/zhuhe/14zuhe.jpg
[12:42:33] <L84Supper> http://www.kollmorgen.com/en-us/products/gearheads/planetary/nematrue/nematrue/
[12:43:10] <L84Supper> Precision: 13 arc-minutes
[12:43:45] <archivist> I hate sites like that that redirect to the home page
[12:43:54] <L33TG33KG34R> so, I have ordered the following:
[12:44:06] <L33TG33KG34R> http://www.newegg.ca/Product/Product.aspx?Item=N82E16813131878
[12:44:16] <L33TG33KG34R> http://www.newegg.ca/Product/Product.aspx?Item=N82E16819116408
[12:44:27] <L33TG33KG34R> http://www.newegg.ca/Product/Product.aspx?Item=N82E16823205018
[12:44:41] <L33TG33KG34R> http://www.newegg.ca/Product/Product.aspx?Item=N82E16820148513
[12:44:49] <L33TG33KG34R> http://www.newegg.ca/Product/Product.aspx?Item=N82E16826105447
[12:45:36] <awallin> why celeron and not i3 or i5 ?
[12:45:45] <p0st4L> whats that for?
[12:45:51] <L33TG33KG34R> because I only had $185 to spend
[12:46:01] <p0st4L> try looking on craigslist
[12:46:02] <L33TG33KG34R> and i3 alone costs $119
[12:46:08] <p0st4L> you might find a better used Dell for that price
[12:46:28] <L33TG33KG34R> also, LinuxCNC doesn't need the power of an i3
[12:46:53] <awallin> are you going to run the new xenomai kernel? would be interesting to hear the latency numbers
[12:46:57] <L33TG33KG34R> this celeron is way better than the atom d525 everybody is so proud of
[12:47:03] <Jymmm> L33TG33KG34R: Um $80 http://www.newegg.ca/Product/Product.aspx?Item=N82E16813121643
[12:47:31] <L33TG33KG34R> I am not getting an atom period.
[12:47:53] <L33TG33KG34R> I refuse
[12:47:58] <awallin> Jymmm: what graphics does that have? the latest Atom boards don't play nice with linux..
[12:49:23] <L33TG33KG34R> I will conjure up a monitor and psu later on
[12:50:06] <L84Supper> http://item.taobao.com/item.htm?spm=a230r.1.10.251.ghHU5Y&id=12722663766&_u=1nmdpnqcc74 no way to know what the lash on these might be without trying
[12:50:08] <Jymmm> awallin: fiik
[12:50:51] <p0st4L> im' using a old ibm thinkpad t61
[12:51:03] <p0st4L> 2.8ghz, 4gb ram, found on CL for $200
[12:51:04] <skunkworks> awallin: I finally found a system that runs xenomai kernel with 'decent' latency. So I think I am installing it right. It just needs some more tweeking...
[12:51:06] <skunkworks> http://imagebin.org/241749
[12:51:46] <awallin> nice GUI there! something recent?
[12:52:01] <L84Supper> German Faulhaber? clone or actual Faulhaber sold out the backdoor? http://item.taobao.com/item.htm?spm=2013.1.48-8524767654.7.XAJ3PC&id=2612247222
[12:52:18] <skunkworks> awallin: yes - let me find the info
[12:52:38] <L84Supper> skunkworks: memleak posted some configs, he'll be back on it tomorrow
[12:53:24] <skunkworks> awallin: http://linuxcnc.mah.priv.at/irc/%23linuxcnc/2013-01-03.html#18:04:27
[12:53:31] <skunkworks> dgar did it
[12:53:40] <skunkworks> dgarr
[12:54:06] <L33TG33KG34R> once I get my CNC running and some machining hours under my belt, I think I'll try my hand at making an embedded system for LinuxCNC. With enough bells and whistles to keep most happy while keeping the project opensource and cheap
[12:54:10] <skunkworks> L84Supper: yes - I have been keeping an eye out
[12:54:33] <awallin> cool. that should be pushed into the repo/master!
[12:55:13] <pcw_home> L33TG33KG34R I would also stay away from Atoms for high performance systems (they barely manage a 1 KHz servo thread) I have a similar system to what you have an it does 6 KHz easy
[12:55:27] <L84Supper> skunkworks: he's currently distracted by a new video game..... kids :)
[12:57:24] <L33TG33KG34R> well, the only reason I am so biased against atom is becaused its already on its way to the silicon graveyard
[12:58:22] <L33TG33KG34R> no point is dishing out $100 for a board which will be obsolete much sooner than the sandy bridge celerons
[12:59:06] <pcw_home> The current D2XXX ones will work +- some video issues
[12:59:40] <L33TG33KG34R> One of the reasons why I am motivated to make an embedded system for LinuxCNC
[12:59:49] <pcw_home> and the new Atoms (end of 2013) should remove the video issues
[13:00:22] <pcw_home> plus be faster
[13:00:26] <L33TG33KG34R> if they survive that long
[13:00:30] <IchGuckLive> ok im off have a nice rest day 1
[13:01:03] <pcw_home> I have little doubt they will survive
[13:07:48] <L84Supper> http://www.newegg.com/Product/Product.aspx?Item=N82E16813157324 no video issues with AMD APU's
[13:09:33] <L84Supper> http://www.newegg.com/Product/Product.aspx?Item=N82E16813130641 down to $45
[13:10:46] <L84Supper> http://www.newegg.com/Product/Product.aspx?Item=N82E16819106013 $40
[13:12:09] <pcw_home> I have one of the ASUS E-350 MBs but its like the atom (ok for 1 KHz servo but not much more)
[13:12:52] <pcw_home> a Intel h41 - 2180 is ~6 time faster
[13:14:19] <pcw_home> Haven't tried the FM1 (or new FM2 trinity)CPUs
[13:30:36] <L33TG33KG34R> woot! got an isp upgrade!
[13:31:10] <andypugh> L33TG33KG34R: Does it matter if your controller is obsolete as long as it still controls the machine? It won't stop working just because Intel launch a new CPU
[13:31:24] <L33TG33KG34R> I know that
[13:33:57] <andypugh> I think having no fan on the CPU or the PSU is a big plus for a CNC controller
[13:34:31] <andypugh> But that isn't just Atoms.
[13:35:11] <ReadError> didnt they stop making the d525 ?
[13:46:39] <L84Supper> pcw_home: we had them working much faster a few months ago, we'll post the howtos when we get started again soon
[13:49:23] <pcw_home> Thanks, I would be interested in low power Atom replacement MB
[13:50:42] <pcw_home> andypugh: did you see my suggested BLDC manual notes?
[14:02:54] <mrsun> how can i make program wait for spindle spinup ?
[14:03:05] <mrsun> i need to wait like a sec or two for the spindle to spin up
[14:03:11] <mrsun> but do not have any feedback from spindle
[14:03:40] <mrsun> ahh G4
[14:03:42] <mrsun> G4 P1
[14:07:32] <ReadError> heh so i was looking at this spindle again
[14:07:34] <ReadError> "Adopt Germany Imported Bearing: three times useful time than ordinary spindle motor"
[14:07:48] <ReadError> how long do you suppose that is?
[14:09:32] <pcw_home> 3 times (too short)?
[14:15:27] <skunkworks> it might be better to simulate your spindle in hal - then use spindle-at-speed pin... (if you want extra credit...)
[14:16:17] <andypugh> pcw_home: Yes, I will put it in next update
[14:17:40] <pcw_home> Another thing we noticed in trying to get a customers new motor running is that even if you are going to use qh mode its better to get h mode working first
[14:18:14] <pcw_home> (so all Hall edges are guaranteed to be right as reference points)
[14:21:19] <L84Supper> I have to try one of your servo boards and the calibration
[14:25:32] <L84Supper> I've used your 7i48 but with a servo amp that handled the HALL sensors
[14:27:35] <andypugh> pcw_home: Yes. I really need to finish the Wiki page on how to use the bldc component.
[14:29:14] <pcw_home> BLDC is tricky but there are some shortcuts you learn on playing with it
[14:29:16] <pcw_home>
[14:29:36] <andypugh> The 7i39 manual seems to make no mention of what the SenseA and SenseB pins are for...
[14:30:11] <pcw_home> There are just GPIO normally (7I39 just passes them through)
[14:30:46] <andypugh> OK, in that case my 9-core wire is adequate :-)
[14:31:11] <pcw_home> L84Supper yes all the Hall excotica is for simple minded drives (with the brains in HAL comps)
[14:34:11] <pcw_home> I think with Halls, first wire Halls/select pattern so you get monotonic rotor angle in the proper direction for the axis.
[14:41:49] * skunkworks still drools on mesa hardware
[14:44:13] <pcw_home> Voids the warranty
[14:48:44] <psha> pcw_home: by the power vested in me I voids your warranty! :)
[14:48:48] <psha> ?
[14:52:00] <L84Supper> isn't IP 65 drool proof?
[14:53:15] <L84Supper> oh, IP 68 for full immersion
[14:56:24] <skunkworks> heh
[14:59:46] <skunkworks> pcw_home: did visters get straitend out then?
[15:01:26] <ReadError> is the 2x1 or 2x2 80/20 pretty rigid stuff?
[15:02:19] <pcw_home> Not sure, but in the mean time ive gotten more familiar with the BLDC comp so can perhaps help more
[15:03:03] <pcw_home> I think his main trouble (at least with speeds) is too slow a servo thread for commutation
[15:04:28] <pcw_home> I also wonder if his problem with jogging is a kins problem
[15:07:15] <DJ9DJ> gn8
[15:08:00] <L84Supper> ReadError: the 40mm metric is pretty stiff, depends on your loads however
[15:08:53] <ReadError> well i was thinking of doing up a cnc router with some
[15:09:09] <ReadError> seems folks have done it in the past
[15:09:20] <L84Supper> ReadError: it's enough for your router, plus they have an extruded table with t-slots
[15:10:05] <ReadError> im looking at linear motion stuff right now
[15:10:12] <skunkworks> pcw_home: would it be possible to put just the mesa and bldc in a faster thread - keeping motion and everthing else in a slower thread?
[15:10:52] <ReadError> http://www.finelineautomation.com/store/show/FLA102-01
[15:11:08] <ReadError> wasnt sure about going with that or the bar+linear bearings
[15:11:19] <L84Supper> http://www.8020cnc.com/home
[15:11:20] <ReadError> seems like the flat kind would be easier to clean and stuff
[15:21:43] <pcw_home> skunkworks probably BLDC doesn't do too much math so it should be OK in a base thread
[15:21:45] <pcw_home> but there may be some issues with some modes
[15:22:56] <skunkworks> I suppose pid also would have to be in faster thread
[15:23:16] <skunkworks> say 10khz while everything else in 1khz
[15:23:46] <andypugh> bldc does floating-point, so can't run in the fast thread.
[15:24:19] <andypugh> (Though I am not sure if the fast-threa always has to be the base thread)
[15:32:02] <pcw_home> I'd forgotton that the base thread cant do floating point
[15:33:02] <andypugh> Which seems a bit arbitrary, really.
[15:34:28] <pcw_home> Found out what the multichannel scaler config is used for (lowres CAT scan of scrap metal for automated sorting)
[15:34:50] <andypugh> Seems exotic
[15:37:54] <pcw_home> I was wondering if it was medical (and hoping it was not)
[15:44:22] <skunkworks> is the servo thread the only one that can do floating point? can you create a faster thread (not as fast as the base thread..) that is floating point?
[15:44:33] <skunkworks> say 10khz
[15:46:14] <micges> all threads created by 'loadrt threads' are fp
[15:46:44] <micges> and they can be 10kHz but you mus be sure that your pc will handle it
[15:46:51] <skunkworks> cool - so I think that may be a good test....
[16:13:53] <mrsun> http://www.youtube.com/watch?v=Kb8dRpLC9Og my little X1 doing some work today! =)
[16:18:44] <skunkworks> so - pid, bldc and mesa in 10khz -> everything else in 1khz thread
[16:19:56] <pcw_home> That may work but I'm not sure what limits the servo thread on Atom MBs
[16:40:15] <mikeggg> we figured out what the problem with the arduino input was.
[16:41:01] <mikeggg> the part of the arduino code that prints the input status to serial. Used Serial.print
[16:41:15] <mikeggg> changed it to Serial.write and bam! success
[16:46:41] <skunkworks> awesome!
[16:49:21] <mikeggg> yeah, I guess they've changed the way that command works since jepler wrote the code
[17:03:56] <L33TG33KG34R> why did purolator ship 2 different packages?
[17:04:01] <L33TG33KG34R> er newegg*
[17:15:37] <pcw_home> Kickback from Fedex?
[17:27:38] <andypugh> The mill is starting to get ther now: https://picasaweb.google.com/lh/photo/bzAH1-jDtgO0YUGhxnSBANMTjNZETYmyPJy0liipFm0?feat=directlink
[17:47:11] <Tom_itx> andypugh, what's the current addition to it?
[17:47:56] <Tom_itx> i wish i had a nice size mill like that and room for it
[17:48:37] <andypugh> side panel on the controller box, connectors for the 7i39, but the main change today was moving it properly in to place, as it was previously on the diagonal so I could get to all sides.
[17:49:22] <Tom_itx> all nice n level now?
[17:49:58] <andypugh> I still haven't actually started on the X axis.
[17:50:05] <Tom_itx> so i see
[17:50:17] <Tom_itx> stepper or servo?
[17:50:45] <andypugh> Not levelled either. That might be tricky as the rear feet are inside a panel I can't get to with it in-situ.
[17:50:58] <andypugh> It's got 700W servos (with resolvers)
[17:51:45] <ReadError> oh cool, theres a site that has a bunch of parts already made up in solidworks, like 80/20
[17:51:48] <ReadError> ;)
[17:51:55] <ReadError> should make things easier to plan
[17:52:23] <ReadError> is there a way to generate a bill of materials from pre-fab'd parts within solidworks?
[19:13:04] <MercuryRising> so i have a 3d printer set up, and i am using slic3r to turn my stl's into gcode. There are options for speed in there, and if I go too high, my machine will round corners and what not. From this, I am assuming that the travel speeds are set by the slicer, and not linuxcnc?
[19:13:35] <cpresser> MercuryRising: take a look at "G64" in the docu
[19:13:54] <cpresser> you need to understand how the trajectory-planner works
[19:14:09] <MercuryRising> awesome, thanks!
[19:15:36] <cpresser> more than two years ago, exact the same question brought me to this channel :D
[19:16:02] <cpresser> https://ca.rstenpresser.de/~cpresser/tmp/pictures/emc2_error.png (its not an error. its a feature)
[19:16:47] <MercuryRising> yeah i was staring at those, wondering how in the hell linuxcnc knew that my machine didn't actually make the corner when I have steppers without encoders
[19:17:23] <uski> it's fun; a few minutes ago, i was asking stuff about a CNC in #reprap because i didn't know this channel; and now, someone asks 3D printer questions here
[19:18:17] <MercuryRising> just like swapping drill bits
[19:19:16] <uski> :)
[19:20:12] <MercuryRising> thanks cpresser, that helped a lot, and thats an awesome feature!
[19:22:56] <MercuryRising> so i can put G64 Px.xx once in the beginning of a gcode file , and it will 'modify' all the gcodes after that to the tolerance I specify?
[19:25:02] <cpresser> yes
[19:26:31] <cpresser> in fact, it wont modify gcodes. gcodes are absolute. but it will change how the trajectory-planner does motion-blending
[19:27:26] <cpresser> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?TrajectoryControl
[19:28:01] <MercuryRising> okay, so the trajectory planner is like the controller for reading in g-code and making the machine do what the g-code says
[19:29:28] <MercuryRising> and even though it would be difficult, say i wanted one corner out of four (on something like a box) to have no blending, and the other three i didn't really care, i could put a g64 in before i hit the 'good' corner, go around, g61, then g64 before i get back to that corner
[19:29:59] <MercuryRising> opposite of that, but yeah, have some parts in exact stop, and some in continuous
[19:30:09] <cpresser> yes, that would be possible
[19:35:57] <MercuryRising> if you're milling or engraving something, do you have to run it through a program that turns the solid model into g-code, or can linuxcnc read solid models for milling?
[19:43:56] <Tom_itx> model to gcode
[19:44:23] <Tom_itx> you need to have a gcode file to run
[20:02:37] <tjb1> What is the point of a tracking number if after 5 days it doesnt show anything?
[20:06:19] <andypugh> MercuryRising: But.. You can set up LinuxCNC so that if you open an STL file with it then it will automaticallt pass it through a filter file, then open the G-code created.
[20:06:51] <andypugh> This does require that the filter program can take command-line arguments and pass the G-code to stdout though.
[20:07:34] <andypugh> (though you could potentially wrap up another package in a script if the program you wanted to use was tricky that way)
[20:08:07] <MercuryRising> hmmm... does it run a command ( could I run "slic3r STLFILE.stl | cat STLFILE.ngc" ) or something more complex?
[20:08:28] <MercuryRising> errr & instead of pipe
[20:09:18] <MercuryRising> &&
[20:10:37] <andypugh> I am no Linux expert. But have a look here: http://linuxcnc.org/docs/html/config/ini_config.html#_filter_section_a_id_sub_filter_section_a
[20:12:57] <MercuryRising> awesome, i'll have to play around with it, thanks!
[20:35:07] <AR_> so i have this thomson linear shaft
[20:35:16] <AR_> 17 inches long
[20:35:27] <AR_> i want to cut it in half
[20:35:39] <AR_> think this can be done with a dremel and/or hacksaw?
[20:35:40] <AR_> lol
[20:35:52] <AR_> 5/8 diameter
[20:36:12] <MercuryRising> hacksaw yes, dremel maybe but i would mess it up
[20:36:15] <MercuryRising> is it hardened steel?
[20:38:24] <skunkworks> dremel to cut through the hard bearing surface - hacksaw to cut through the aluminum...
[20:39:47] <AR_> it is case hardened
[20:39:52] <AR_> so yeah
[20:39:54] <AR_> softer core
[20:40:29] <skunkworks> oh - no aluminum
[20:41:08] <Jymmm> dremel cutoff wheels are great, but WEAR SAFETY GOGGLES (not glasses)
[20:41:40] <Jymmm> I dont care if you put your hands into battery acid, wear the goggles!!!!!
[20:42:06] <Jymmm> when those fuckers shatter, THEY GO EVERYWHERE and into a million pieces
[20:46:45] <andypugh> Buy an angle-grinder. You are going to need one for the rest of your life anyway.
[20:47:21] <Jymmm> heh heh
[20:47:41] * Jymmm is ay 24 feet with his angle grinder
[20:47:46] <Jymmm> *at
[20:48:29] <Jymmm> and it's a 4" too!
[20:48:55] <andypugh> I have run out of space for the 500W PSU I got from eBay (http://www.power-one.com/power-solutions/products/embedded-power/ac-dc/din-rail-power-supplies/x-din-rail-series/series ) It looks like a rather high-quality bit of kit. (£375 list price).
[20:49:35] <andypugh> I think I will need to remove my current 24V / 5V PSU (which makes a funny buzzing noise anyway)
[20:49:42] <AR_> Jymmm, I use cutoff wheels all the time without any eye protection
[20:49:43] <AR_> it's just a matter of time
[20:49:47] <andypugh> But that will leave me with no 5V supply.
[20:49:58] <AR_> they always shatter on me too
[20:50:11] <Jymmm> AR_: DREMEL cuttoff wheels?
[20:50:15] <AR_> yes
[20:50:21] <andypugh> I can't decide whether to steal the 5V from the PC PSU, or do something with a regulator.
[20:50:50] <Jymmm> AR_: Yeah, you might just save the time and stab your eye with a red hot poker instead
[20:51:28] <andypugh> I am terrible for not wearing eye protection, I almost never do. I have so far got away with it.
[20:51:50] <andypugh> I rarely weld without protection, mind.
[20:51:59] <Jymmm> andypugh: Same here, BUT.... when using dremel cutoff wheels, I grab the face shield
[20:53:13] <Jymmm> and check the mandrel too, I've found one of mine slightly bent
[20:53:36] <mikegg> so I fried a A3967 step driver earlier. Accidently shorted two of the output leads. Should have been wearing a face shield...
[20:54:06] <andypugh> It seems that the PicoPSU is good for 6A. I wonder how much current the Mesa cards and PC need?
[20:54:23] <AR_> Jymmm, i dont think they are as dangerous as you make them out to be
[20:55:05] <Jymmm> AR_: If I didn't wear glasses, I would have been blind. And I'll leave it at that.
[20:55:59] <Jymmm> $1 to save your eyesight http://www.harborfreight.com/safety-goggles-3-pack-66538.html
[20:58:12] <Tom_itx> do you trust your eyes to something that costs a buck?
[20:58:41] <Jymmm> better than nothing at all
[21:01:22] <Jymmm> Actually, I dont like the goggles personally, so I have this instead; works for me http://www.harborfreight.com/adjustable-face-shield-46526.html
[21:02:13] <Jymmm> goggles fog up and dont fit right with me wearing glasses
[21:03:37] <Jymmm> AR_: Also, realize that the dremel cutoff is creating micro slivers in whatever your cutting too much of the time.
[21:09:20] <AR_> i like it
[21:09:24] <AR_> i like the danger
[21:41:48] <toner> my eyes! the goggles do nothing!
[21:42:29] * ReadError pours toner a pint of homebrew
[21:42:53] <ReadError> damn if roycroft was still here
[21:43:01] <ReadError> i was going got give him 250mL
[21:43:07] <toner> http://www.youtube.com/watch?v=OqfOxm_1BE0
[21:43:08] <toner> haha
[21:43:17] <toner> roy left?
[21:55:19] <p0st4L> if i were to make a partition to install linuxcnc, how big would be ok? 20gb?
[21:55:44] <Jymmm> p0st4L: That would be more than enough
[21:56:00] <Jymmm> and leave you room for adding extras too
[21:56:12] <p0st4L> perfect :)