#linuxcnc | Logs for 2013-06-06

[00:00:11] <Valen> (they also have 4 legs and tails ;->)
[00:00:52] <PetefromTn> Wow I am not sure we are talking about the same thing anymore.
[00:01:48] <skunkworks> I used a carbide wood round over bit to shape a turbo exauste housing...
[00:02:22] * Valen has 2 dogs ;->
[00:02:37] <skunkworks> http://electronicsam.com/images/turbo/DSC02740.JPG
[00:02:43] <PetefromTn> skunkworks Yeah I've done that too..
[00:03:06] <skunkworks> http://electronicsam.com/images/turbo/DSC02764.JPG
[00:03:08] <PetefromTn> not the housing part.
[00:03:38] <skunkworks> http://electronicsam.com/images/turbo/DSC02768.JPG
[00:04:38] <skunkworks> worked for as long as I had the car...
[00:05:52] <skunkworks> that was when the K&T had the old control.. I think that was the last thing done on it. (it smoked then)
[00:06:12] <PetefromTn> Oh well Gotta get to bed. Night fellows.
[00:06:18] <skunkworks> same here.
[02:04:50] <DJ9DJ> moin
[02:05:11] <Loetmichel> mornin'
[05:06:17] <L84Supper> evenin'
[05:09:04] <L84Supper> pcw_home: way overpriced ($129) but the quad core version of the wandboard is out http://linuxgizmos.com/wandboard-gains-quad-core-cpu-and-enhanced-gpu/
[05:10:02] <L84Supper> the price difference between the single core i.mx6 SOC and the 4 core is only $20 max
[05:11:39] <L84Supper> http://www.wandboard.org/ but they do bump it up from 512MB to 2GB DDR3
[06:08:17] <Andy1978> is it possible to have a XZ configuration and on that top an arm with a fixel length which can be rotated? Imaginde 1 rotational axis SCARA on a XZ Table
[06:13:15] <jthornton> I'm pretty sure you can define just about any kind of kinematic
[06:14:34] <Andy1978> jthornton: Ah, nice. And LinuxCNC does the kinematic even when there are several possibilities to reach a point?
[06:16:07] <jthornton> you would have to either modify a kins or make one from scratch
[06:37:56] <L84Supper> Andy1978: what are you making with that setup?
[06:38:26] <Andy1978> I have seen this setup in the metal trash
[06:38:44] <Andy1978> L84Supper: I think it was used for pick and place tasks (robotics)
[06:39:23] <Andy1978> And I wonder if I could replace the clamp with a spindel and use it
[06:40:05] <L84Supper> Andy1978: interesting, any idea what the range of travel was for the Z that the SCARA was mounted on?
[06:40:15] <Andy1978> 100mm
[06:40:28] <L84Supper> SCARA's usually have their own Z actuator built in
[06:40:49] <Andy1978> L84Supper: Its not a real SCARA
[06:40:55] <Andy1978> hm, should I make a picture?
[06:40:57] <L84Supper> http://www.robotmatrix.org/images/ScaraRobotWorkSpace.gif
[06:56:55] <Andy1978> L84Supper: http://tech-chat.de/images/kin1.jpg and http://tech-chat.de/images/kin2.jpg
[06:57:52] <Andy1978> linear X, linear Z and a rotational head. The "arm/cantilever" is removed
[07:25:02] <L84Supper> http://machinedesign.com/bearings/new-ways-damp-vibration-linear-axes
[10:23:37] <IchGuckLive> hi all from summer hot Germany
[10:28:20] <IchGuckLive> hi tjb1
[10:39:23] <IchGuckLive> till later By
[10:49:45] <DJ9DJ> re
[11:47:10] <Loetmichel> re @ home
[11:48:05] <Jymmm> re @ ISS
[11:54:08] <JT-Shop> this is funny "The reason for the analog speedometer going to 200 is because it was built in Canada and we go by kph. So, we have a speedometer that will go up to 200km/hr (125 mph). If it only went up to 140 then Canadians would only have a speedometer good for 140kph (87mph)."
[11:54:50] <JT-Shop> anyone want a sammich while I'm up
[11:56:02] <archivist> cheeeeeeeze :)
[11:56:38] <Jymmm> JT-Shop: I'd prefer babyback rib and an ice cold beer
[12:06:26] <IchGuckLive> Hi all B)
[12:21:14] <Loetmichel> JT-Shop: better than the germans
[12:22:09] <Loetmichel> BMW: X1 car has 100HP, vmax of 215kmh but the speedo goes all the way up to 270 :)
[12:51:36] <generic_nick|2> cradek: JT-Shop: the issue with the tool offset seemingly being ignored on my lathe was solved by removing the G0 in front of the t5m6
[12:51:46] <generic_nick|2> for some reason mstercam put one there
[12:53:27] <IchGuckLive> generic you need to G40 G49 bevor tool change
[12:54:45] <IchGuckLive> and after toolchange reset all your ofsets on the given H parameters
[12:55:11] <IchGuckLive> aölso D R and so one
[12:57:02] <Tom_itx> why do you cancel comp before a tool change?
[12:59:12] <IchGuckLive> to free all parameters inside the control #
[12:59:30] <JT-Shop> generic_nick|2: my DX-32 control (386 vintage) uses a G00 on every line except G01 G02 etc.
[13:00:15] <generic_nick|2> i dont know what to tell you, but i tried everything and that's what worked
[13:00:59] <cradek> my lathe's original control didn't recognize G0. You had to program G01 F80000 to get a rapid
[13:01:33] <ssi> the hnc?
[13:01:37] <cradek> yeah
[13:01:45] <ssi> I never actually wrote any code for my original control :P
[13:01:50] <ssi> didn't have a tape punch :D
[13:02:01] <generic_nick|2> thr g0 is the only difference between the tool crashing into the spindle or going to the correct tool offset.
[13:02:02] <cradek> I have a punch but mine wouldn't read the tape right
[13:02:29] <cradek> generic_nick|2: that's fascinating
[13:03:14] <generic_nick|2> yea i dont get it. i know i have other programs where it has a G0 before a tool change and it's fine
[13:03:16] <cradek> generic_nick|2: I'd love to see that reproduced with a simple program in sim
[13:03:34] <cradek> hm, sounds like that might not be the full pattern then
[13:03:55] <generic_nick|2> cradek: i'd be happy to link you to the program in pastebin and you can try it yourself
[13:05:24] <generic_nick|2> perhaps it has something to do with the canned threading cycle
[13:05:31] <generic_nick|2> i never threaded on this lathe before
[13:05:41] <generic_nick|2> and that's the tool it is effecting
[13:08:14] <Crashdemon> hey folks
[13:08:29] <IchGuckLive> B)
[13:08:37] <IchGuckLive> some crash today O.O
[13:10:05] <Crashdemon> i've got a problem with the actual version of linuxcnc, homing is only possible on z-axis
[13:10:13] <Crashdemon> is there an known bug?
[13:10:23] <cradek> elaborate on "only possible"
[13:10:36] <IchGuckLive> 2.5.2
[13:12:54] <Crashdemon> in stepconf moving on all axis working, if doing a manual homing, only z-axis is moving
[13:13:21] <Crashdemon> also if i'm choose x or y-axis
[13:13:36] <IchGuckLive> is there a switch on etch axis
[13:13:49] <IchGuckLive> or all switches to one port
[13:14:35] <IchGuckLive> Crashdemon: is it posible that you can piost the ini file and the hal from your mashine to pastebin
[13:15:25] <andypugh> Crashdemon: I suspect that you haven't wired the amp-enable pins, and just got lucky with Z.
[13:16:56] <IchGuckLive> Crashdemon: what country are you from
[13:17:41] <Crashdemon> germany
[13:17:54] <IchGuckLive> Ich bin in Kaiuserslautern
[13:18:15] <IchGuckLive> wir können auch in deutshc wenns so besser geht
[13:18:34] <Crashdemon> ja, wäre nicht schlecht
[13:19:59] <tjtr33> Control Engineering mag lists a new reactable project at CMU http://www.youtube.com/watch?v=BBQPA5fsLTA and a pickNplace robot from IAI http://www.youtube.com/watch?v=uwqfgpBc69E
[13:20:01] <tjtr33> projected touchy?
[13:20:02] <Tecan> (uwqfgpBc69E) "Cma - Flodyne Hydradyne - Iai Ct4 HighSpeed Cartesian Robot" by "pixel3gp" is "People" - Length: 0:00:39
[13:20:27] <tjtr33> hah Tecan you only catch the last url
[13:20:53] <Tecan> hmm
[13:20:59] <tjtr33> np:
[13:22:16] <tjtr33> and better docs on the CMU worldkit http://chrisharrison.net/projects/worldkit/worldkit.pdf
[14:32:05] <GammaX> Hey guys
[14:37:53] <gene78> Hey Guys, update-manager is refusing to update linuxcnc. I'm on the 2.6.0-pre branch, problems?
[14:38:25] <GammaX> prolly the error between the keyboard and the seat ;)
[14:39:56] <gene78> It shows it as available, but will not take a check in the box.
[14:44:23] <gene78> I wonder if its been disabled because of the md5sum bug reports i saw earlier today
[17:08:19] <gene78> I have inkscape displaying the .svg file from Rockhopper, big enough I can read it on screen, so it should print.
[17:09:28] <DJ9DJ> gn8
[17:09:43] <gene78> But how do I get it, its about 5 feet wide ATM, to a printable, as in paste/tape all the pages together, file?
[17:28:42] <Tom_itx> use a billboard printer
[17:37:53] <gene78> :)
[18:45:04] <JT-Shop> nice first question on the forum
[18:45:04] <JT-Shop> how do I set up the machine for cutting foam wings and how do i create a file for cutting?
[18:49:38] <Tom_itx> hot wire
[18:49:46] <Tom_itx> 2 forms, one for each end
[18:50:00] <Tom_itx> :)
[18:50:20] <Tom_itx> my friend did alot of pylon racing and made his own wings
[18:50:26] <JT-Shop> well you forgot the bow and the spring and the power supply but close
[18:50:34] <Tom_itx> ya ya
[18:50:39] <JT-Shop> I used to cut them too
[18:51:20] <Tom_itx> i had a 6' foam cube at one point that got used for them
[18:51:32] <Tom_itx> not sure where i got it now
[18:52:06] <Tom_itx> ok, time to walk the puppies
[18:52:14] <JT-Shop> I sold everything I had for RC a few years back
[18:52:26] <Tom_itx> i never flew much mostly rc boats
[18:52:28] <JT-Shop> ok, don't step in any or on any
[18:52:43] <Tom_itx> got a box full of trophies for that, got tired of the politics and quit
[18:52:56] <JT-Shop> yea, that part is a pain
[18:53:15] <Tom_itx> it was fun though
[18:53:20] <JT-Shop> petty politics will make me walk faster than anything
[18:53:48] <Tom_itx> i still think there's a fairly decent flying club around here
[18:54:28] <JT-Shop> none around here, I was very active when I lived in LaPlace La.
[18:54:40] <Tom_itx> alot of them used to meet at the Beechcraft activity center until it got sold and turned into business's
[18:54:42] <JT-Shop> we had two nice clubs
[18:55:41] <Tom_itx> my friend owned a hobby shop so i got stuff pretty cheap
[18:55:54] <Tom_itx> wrote him a point of sale for his shop so that kinda helped
[18:56:25] <Tom_itx> ok puppies are impatient...
[18:57:19] <JT-Shop> time for me to start dinner too
[19:50:07] <ssi> I tried to get into rc flying, and the club politics kept me out
[20:01:43] <jdh> there is a lot of that around
[20:01:46] <RyanS> I crashed a few......
[20:02:02] <jdh> me too
[20:02:21] <jdh> last time was 1980 though
[20:02:44] <RyanS> stupid I should have stuck with the trainer for longer
[20:02:44] <Valen> by default does emc still use its pid loop when running steppers?
[20:06:24] <ssi> which pid loop would that be
[20:07:53] <Valen> I'm not sure....
[20:08:10] <Valen> realising I have nfi how emc works with steppers
[20:08:29] <Valen> I always presumed it ran them like servos with stepgen faking it
[20:08:42] <ssi> well I think typically the stepgens get fed position
[20:09:00] <ssi> and the stepgen component itself just knows that to get from the position I'm in to the position i'm commanded to, I have to pulse x times
[20:09:40] <Tom_itx> Valen, i think if you miss steps you just miss them
[20:09:53] <Valen> not talking about that
[20:10:03] <Valen> more how it gets from gcode to steps out
[20:10:17] <Valen> and how that differs from gcode to servo out
[20:10:29] <Tom_itx> probably very similar
[20:10:38] <Tom_itx> so many steps per inch or mm
[20:10:44] <ssi> well in servo mode, you have a position command, and the pid component uses feedback to minimize error
[20:10:48] <Tom_itx> so many ticks of the encoder with a servo
[20:10:50] <ssi> in stepper mode there's no pid
[20:10:58] <ssi> there's a position command, and the stepgen ticks
[20:11:06] <Tom_itx> at least with a servo it has a chance to correct it
[20:11:17] <Valen> my mental model was it still ran through the pid just as a "passthrough"
[20:11:36] <Valen> with the step rate coming from the error between commanded and actual position
[20:11:39] <ssi> I don't think there's a pid component at all
[20:11:45] <ssi> but maybe I'm wrong
[20:11:50] <ssi> it's been awhile since I setup a stepper machine
[20:11:54] <Valen> you're probably right
[20:11:57] <Valen> i've never done one
[20:11:58] <Tom_itx> steppers have no position feedback
[20:12:13] <jdh> unless you add position feedback
[20:12:20] <Tom_itx> righ
[20:12:21] <Tom_itx> t
[20:17:35] <RyanS> the 'DRO' on emc, if your motors miss a step, you would be getting a false reading right?
[20:18:07] <ssi> correct
[20:18:40] <Tom_itx> the dro keeps going even if the stepper doesn't
[20:19:55] <RyanS> hmm, I guess closed loop will deal with problems like that.
[20:20:22] <Tom_itx> yup
[20:21:03] <RyanS> I would have thought machining centres would also have encoders on the ways too..
[20:22:14] <RyanS> or some redundancy in the DRO anyway
[20:28:44] <Valen> they often do use scales on the axies
[20:29:01] <Valen> but they use those for position feedback and the encoders on the motors for velocity
[20:31:56] <RyanS> Makes sense I guess that's the whole point of closed loop
[20:36:41] <ssi> as long as there's no lost motion between the encoder and the joint, you don't need direct feedback from the joint
[20:36:48] <ssi> encoders on the motors is plenty sufficient
[20:37:17] <RyanS> That laser camera measurement technique you were talking about, does it have a technical name? Not having any luck on Google
[20:38:09] <Tom_itx> ssi, where does backlash comp come into play in that equasion?
[20:38:29] <ssi> Tom_itx: ideally in the mechanicals ;)
[20:38:33] <ssi> ie dont' have backlash
[20:38:41] <Tom_itx> but everything does
[20:38:44] <Tom_itx> to some degree
[20:38:50] <Tom_itx> or will with wear
[20:38:51] <Valen> you can have less backlash than you can measure though
[20:38:57] <ssi> yes
[20:39:17] <Valen> we have less than .001mm backlash
[20:39:21] <ssi> my hnc has resolvers that are driven from the leadscrew through a belt reduction
[20:39:29] <ssi> and it still can hold 0.00002" without backlash comp
[20:39:33] <Valen> a pair of ballnuts with a spring between them
[20:41:23] <Valen> well, a set of belville washers with a few hundred KG on them ;->
[20:42:00] <RyanS> does it matter if there is backlash as long as the controller 'knows about it' and its predictable?
[20:42:10] <Valen> backlash is bad
[20:42:17] <Valen> imagine trying to mill a circle
[20:42:30] <Valen> say 10mm in diamiter, with 5mm of backlash in your screw
[20:43:02] <Valen> the controller needs to rip through that 5mm of backlash in zero time as you change direction
[20:43:25] <RyanS> Can't the control software compensate?
[20:43:34] <ssi> with infinite acceleration, yes
[20:43:36] <eric_unterhausen> if a mobo takes atx 24 ps, is there any hope that a standard atx ps will work?
[20:43:49] <Valen> sometimes
[20:44:08] <Valen> most newer mbos will have the 24 pin port on them but may be able to run with 20 pin
[20:44:10] <eric_unterhausen> it also has atx12
[20:44:11] <Valen> check the manual
[20:44:15] <ssi> there's also the issue that with backlash in the system, the cutter can grab and slam the table back and forth the amount of the lash
[20:44:20] <ssi> and no software is compensating for that
[20:44:46] <RyanS> i see
[20:45:07] <RyanS> Manual machines probably not so much an issue?
[20:45:22] <ssi> on manual machines the practice is to always feed one direction
[20:45:23] <Valen> still an issue
[20:45:26] <ssi> so you're always working against the lash
[20:45:34] <ssi> but there's still the issue of cutter grab when climb cutting
[20:45:35] <Valen> at least for the thing vibrating
[20:45:45] <ssi> that's the reason conventional milling is taught
[20:45:54] <ssi> you're feeding into the teeth
[20:45:58] <ssi> rather than with them
[20:46:14] <ssi> so the cutter is trying to push against you, keeping the lash loaded
[20:46:18] <RyanS> ah
[20:46:38] <ssi> most manual machines have a TON of lash
[20:46:49] <ssi> my 9x42 mill has almost 0.100"
[20:47:02] <Tom_itx> that's alot
[20:47:04] <ssi> yep
[20:47:19] <ssi> but it doesn't bother me because I have a DRO with glass scales :P
[20:51:46] <RyanS> I keep hearing the stories about finding old turret mills "everywhere" for ridiculously cheap. They mustn't be talking about Australia?
[20:52:35] <RyanS> $4000-$6000 seems to be the going price around here
[20:53:38] <Valen> not in .au Ryan
[20:53:57] <Valen> the only ones that are cheap are the ones that are too big to move
[20:59:52] <RyanS> http://www.ebay.com.au/itm/Ikegai-AX25N-CNC-lathe-12-tool-turret-/271218127064?pt=LH_DefaultDomain_15&hash=item3f25dc30d8
[21:00:04] <RyanS> The end of the first paragraph is hilarious
[21:00:30] <toastyde2th> Ikeagi, good lathes
[21:00:32] <Valen> rofl
[21:01:06] <RyanS> I can see how the coffee grinder would make a difference.....
[21:01:23] <Valen> i think perhaps there may have been some hyperbole
[21:02:40] <RyanS> It sounds like a tall tale
[21:04:04] <RyanS> The cost of tools would be too much
[21:10:38] <RyanS> Live tooling on that turret?
[21:13:18] <toastyde2th> no
[21:14:02] <toastyde2th> also most residential power can't handle that much current
[21:14:21] <toastyde2th> it's kinda insane how much peak power they pull
[21:15:18] <RyanS> A lot of houses with central air-conditioning have three-phase... Probably only have 30 amp
[21:15:36] <toastyde2th> ya, 240v, 30-40 amp
[21:17:32] <RyanS> Live tooling is more recent technology ?
[21:23:22] <toastyde2th> much
[21:23:24] <toastyde2th> and much more expensive v
[22:01:49] <RyanS> wtf is an oil country lathe?
[22:07:51] <Jymmm> a rock?
[22:08:57] <Valen> guess its from texas?
[22:09:17] <RyanS> lol I was just about to say that
[22:34:45] <skunkworks> probably really really big?
[22:53:23] <qdc> I always assumed it was something only an oil company could afford
[22:54:55] <Tom_itx> http://www.practicalmachinist.com/vb/general-archive/what-makes-oil-country-lathe-74514/
[22:56:12] <Tom_itx> heck, they even have a website
[22:56:13] <Tom_itx> http://www.oil-country-lathes.com/
[23:01:02] <RyanS> yee haw
[23:01:29] <RyanS> Maybe they manufacture oil drills
[23:01:38] <RyanS> rig
[23:07:46] <RyanS> "Billy Bob we done gone have enough room for 'nother oil country lathe in the barn"
[23:18:40] <toastyde2th> RyanS,
[23:18:52] <toastyde2th> oil country lathes are generally:
[23:19:17] <toastyde2th> Big as fuck spindle bore, 25-30" swing or larger, but a relatively short bed _for the spindle size_
[23:19:49] <toastyde2th> they'll also have an elongated spindle housing with a chuck on the other end of the spindle as well
[23:20:15] <toastyde2th> they'll have a bed long enough to do SOME boring, but not a lot - maybe 10-15' long
[23:20:21] <toastyde2th> the idea is to be able to work on drill string
[23:24:55] * roycroft could make some crude, bawdy comment about beds long enough for boring, but decides not to
[23:29:25] <ssi> commenting on the comment is still commenting :D
[23:31:36] <roycroft> of course
[23:31:56] <roycroft> but any interpretation is not my responsibility
[23:33:04] <Valen> you know I hadn't made that interpertation untill you made the comment
[23:34:08] <RyanS> Besides you don't necessarily even need a bed
[23:35:33] <roycroft> true, and a rotary table could add a whole new dimension