#emc | Logs for 2010-04-03

[00:34:19] <tom3p> the position capture is likely more accurate with slower velocities, and with agreeing repetition ( and stiffer machines) so, the hardness of the surface isnt very important, and being able to re-dress it may be valuable ( esp if it doesnt work as planned :)
[00:38:18] <tom3p> we use this voltage drop method in sink edm all the time. i use a dowel pins & ball bearing for the 'touch plate' , very very accurate diameters and cheap & easy to get.
[00:39:05] <tom3p> dowel pin for smallish work, ball bearing for large ( exaggerates dip in plane )
[00:40:15] <tom3p> got a 123 block with tapped & thru holes? there's your plate
[01:59:12] <MattyMatt> variation on EDM for wood. run a termite along a surface and anything he can reach, he can eat
[02:04:06] <skunkworks> people for the ethical treatment of termites
[02:05:14] <MattyMatt> so you won't like my idea for doing complex internal cuts
[02:05:50] <tom3p> no termites have been harmed in the manufacture of this product ( they were really ants dress up like termites and had small dremels )
[02:07:10] <clytle374> save the ants.
[02:07:43] <Tome> #nick tome
[02:08:03] <Tome> Tome is now known as Guest76369
[02:09:15] <Guest76369> How does one identify to the NickServ? I try but keep getting -NickServ- You failed to identify in time for the nickname...
[02:11:51] <whimsical> Your name choice might be in use?
[02:11:52] <SWPadnos> /msg nickserv identify password
[02:12:05] <SWPadnos> make sure there are no spaces before the /
[02:12:30] <SWPadnos> and you should be using your registered nick when you do that
[02:13:18] <Guest76369> what is the password?
[02:13:25] <Guest76369> I don't think I have one...?
[02:13:42] <SWPadnos> then you don't have a registered nickname, and you can't identify
[02:13:48] <Guest76369> OH
[02:13:53] <whimsical> It's a secret!
[02:13:59] <Guest76369> Ok, let me back up then.
[02:14:08] <SWPadnos> you can register by changing your nick, then /msg nickserv register password
[02:14:08] <Guest76369> How do I register a nickname
[02:14:10] <SWPadnos> (I think)
[02:14:16] <SWPadnos> /msg nickserv help :)
[02:14:18] <whimsical> Unregistered pass words are first come etc.
[02:14:33] <SWPadnos> your client may allow you to abbreviate that with /ns
[02:14:34] <Guest76369> cool, help worked thanks!
[02:15:02] <SWPadnos> you should do all this in the freenode tab, not in a channel (like this) - in case you mistype something, you don't want it broadcast and archived
[02:15:30] <Guest76369> ok
[02:34:31] <Guest489> Ok, I give up on trying to register a nickname.
[02:35:32] <SWPadnos> what name are you trying to register?
[02:35:41] <Guest489> Any
[02:35:53] <SWPadnos> well, if someone else has it, you can't register it again
[02:35:57] <SWPadnos> so your choice matters
[02:37:11] <MattyMatt> MattyMatt is now known as Any
[02:37:14] <Guest489> It says I am trying to regsiter Guest489
[02:37:20] <Guest489> which I am not
[02:37:24] <SWPadnos> you need to change your name first
[02:37:30] <SWPadnos> /nick newname
[02:37:31] <Any> it's free :)
[02:37:43] <Any> Any is now known as MattyMatt
[02:37:44] <SWPadnos> then /msg nickserv register password
[02:37:53] <SWPadnos> (or whatever the help said)
[02:37:58] <Guest489> Guest489 is now known as grommit
[02:38:45] <SWPadnos> I don't know if you noticed, but you can also ask for extra help on specific commands, like /msg nickserv help register
[02:38:58] <SWPadnos> (which tells you that you also need to provide your email address)
[02:39:18] <SWPadnos> so it's /msg nickserv register password email@address
[02:39:40] <grommit> ok
[02:39:46] <grommit> finally
[02:39:54] <SWPadnos> welcome, grommit ;)
[02:40:00] <SWPadnos> how's Wallace?
[02:40:09] <clytle374> hehe
[02:40:11] <grommit> He's full of cheese
[02:40:20] <SWPadnos> figures
[02:40:33] <EricKeller> I'm going to guess that grommit has been registered previously
[02:40:36] <SWPadnos> ok, bedtime for me. have fun with your new nick
[02:40:41] <clytle374> Or rabbit is named shawn after the sheep
[02:40:49] <grommit> oh hoy
[02:40:58] <grommit> joy, that is
[02:41:30] <SWPadnos> huh. grommit hadn't been registered
[02:41:33] <SWPadnos> amazing
[02:41:37] <SWPadnos> night
[02:41:40] <grommit> night
[02:41:57] <EricKeller> it lets me use various nicks that it says have been registered
[03:03:54] <whimsical> I think a nick can be used even if registered, but if the Owner shows up, the nick can be reclaimed and the unregistered user booted
[03:05:18] <toastydeath> it changes it for you if you don't identify.
[03:08:58] <whimsical> of course
[03:43:25] <whimsical> SPL311
[04:53:39] <jimbo> How do you use net to link multiple outputs from mux2 to an input.?
[05:26:37] <TD--Linux> TD--Linux is now known as TD-Linux
[05:35:51] <elmo40> http://www.opennurbs.org/
[05:36:06] <elmo40> I wonder if heeksCAD can implement some of this
[06:49:49] <Dave911> logger_emc:bookmark
[06:49:49] <Dave911> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2010-04-03.txt
[06:56:46] <awallin_> http://www.youtube.com/watch?v=LytH7eCuvDk
[07:31:50] <alex_joni> awallin_: very nice
[08:40:42] <morfic> awallin_: nice, but with that music you would think they cut a little faster
[08:53:44] <morfic> awallin_: thanks for making me look on youtube for more examples
[09:10:22] <awallin_> morfic: need to buy a Syil X-7 or similar and outfit it with an ATC and AC-servos, that would give speed
[09:10:53] <awallin_> this is a poor old Optimum BF-20 outfitted with linears, ballscrews, servos. no atc :(
[09:19:15] <morfic> i amuse myself on the "crashes" right now
[09:19:22] <morfic> don't ask why i am not sleeping
[09:19:24] <morfic> i dunno
[09:20:05] <morfic> http://www.youtube.com/watch?v=Rjm90eJCjO4&feature=related what is "too much material on your part"?
[10:46:26] <archivist> interesting kludge http://www.youtube.com/watch?v=3psLvh5ssPQ&feature=related
[11:20:41] <MattyMatt> awallin_ does anyone ever fit those IOM with gps and send them across the Pacific by themselves?
[11:30:29] <MattyMatt> archivist: what's the kludge? the motor/gearbox/chain?
[11:31:01] <MattyMatt> isn't that how all rotaries were driven before steppers?
[11:32:10] <MattyMatt> * MattyMatt bends nail a lttle more. there ya go, it's a fly cutter now
[11:35:05] <MattyMatt> hi Any
[11:36:40] <grommit> grommit is now known as grommmit
[11:37:14] <MattyMatt> :)
[11:39:16] <grommmit> I thought you were Any :-)
[11:43:45] <MattyMatt> only after a few drinks
[11:44:11] <MattyMatt> Yesh, I'm the Prime Minishter!
[13:14:27] <MattyMatt> ouch. very little survived blender->DXF->heeks
[13:15:26] <grommmit> what is heeks?
[13:16:16] <MattyMatt> heekscad
[13:16:25] <grommmit> Have you played with the Blender to gcode script?
[13:16:49] <MattyMatt> not yet
[13:17:11] <grommmit> It looks like it might actually work
[13:19:38] <MattyMatt> HeeksCAD does the same thing, in python too, so I'm figuring on using both
[13:21:08] <MattyMatt> ah blender->STL->heeks worked OK, and shows me I only exported what was selected
[13:24:27] <grommmit> Good to know
[13:25:07] <MattyMatt> STL has no concept of solids, but then neither does blender :p
[13:26:05] <MattyMatt> heeks could make objects out of closed meshes tho
[13:29:18] <MattyMatt> it makes 1 object per import
[13:30:41] <grommmit> http://blenderart.org/issues/ Issue 11 has an article on solid modeling with blender
[13:30:48] <MattyMatt> aw no, step/igis export
[13:31:10] <MattyMatt> yeah I do solid modeling all the time
[13:32:00] <MattyMatt> and you can assign mass and CoG to an object, for physics
[13:32:47] <MattyMatt> and you can even use 3d textures, and the knife tool, but it's still not a solid
[13:34:00] <MattyMatt> getting close tho :) it just needs a property called PhysMat, because Material is all about surface appearance
[13:37:10] <MattyMatt> Blender could be turned into a fine CAD/CAM program, but it'll probably cause a schism with the Freeform Organicists
[13:37:40] <grommmit> It could be both, selectively...
[13:38:30] <MattyMatt> it could indeed, but there's politics if you want to add e.g. dimensioning to the core
[13:38:31] <grommmit> The article above concludes with: Solid modeling is becoming possible in blender but the development needs per- haps to consider the option to create solid objects (not meshes) to achieve mechanical models. This represents new opportunities because management for solid objects is little different than mesh- es.
[13:40:34] <MattyMatt> a PhysMat property would do. set an object to Phy:Steel_Mild and it'll calc its own weight, surface textures etc
[13:42:12] <MattyMatt> defining a shape in terms of cutting ops could be done at a higher level with scripts
[13:42:58] <MattyMatt> the core GUI is mostly scriptable in 2.5 :) I should try and build that
[13:45:39] <grommmit> You go.
[13:45:48] <MattyMatt> ideally Heeks & Blender would use the same python for toolpaths
[13:45:50] <Jymmm> girl!
[13:46:18] <MattyMatt> * MattyMatt is doing it for himselves
[13:48:04] <MattyMatt> constraints are weird, there's no simple pin/sliding/ball joint choice
[13:48:57] <MattyMatt> they can all be made, but by messing about with local axes
[13:56:06] <grommit> foo
[14:12:26] <skunkworks> alex_joni: the cc on you tube is hilarious! (never played with it before.)
[14:16:19] <elmo40> MattyMatt: have you toyed with this? http://www.opennurbs.org/
[14:25:08] <roh> MattyMatt have you seen heeksart?
[14:25:18] <roh> thats for nurbs-mesh stuff afaik
[14:38:49] <MattyMatt> I wouldn't use a lib for nurbs, it's just bezier patches
[14:39:29] <MattyMatt> but for Heeks it'd make sense, as it's well-defined and has a Somebody Else's Problem field erected
[14:39:49] <MattyMatt> 2.5 alpha 2 is nice :) no building required
[14:39:59] <MattyMatt> ^blender 2.5a2
[14:40:29] <MattyMatt> I haven't seen heeksart
[14:41:27] <MattyMatt> oh sweet. I have now
[16:21:31] <grommit> two questions on configuring 7i43 to run spindle motor...
[16:21:58] <grommit> P4-39 19 PWMGen 2 PWM/Up (out)
[16:22:21] <grommit> Is Up and Down equiv to Clockwise and Counter-CW?
[16:23:28] <grommit> If I have my wiring on P3 instead of P4, can I edit the .PIN file and make P4 (above) into P3 and expect it to work?
[16:27:40] <elmo77> try it ;)
[16:29:07] <elmo77> morfic: how goes the EMC2? have you installed and tested it on anything yet?
[16:30:03] <grommit> I would try it, but am not near the machine until monday. I am playing with it in simulation mode
[16:31:10] <grommit> EMC is going fine. I was able to move a stepper using the default file. Peter Wallace is to get me a firmware build for the 7i47 on Monday which will allow me to (hopefully) get it all working
[16:31:16] <pcw_home> The /PIN file reports the capabilites/pinout of a particular .BIT file, thats all, it make no sense to edit it
[16:31:17] <pcw_home> in up/doen PWM mode one PWM output is active for positive input commands and the other for negative input commands
[16:31:27] <pcw_home> .PIN
[16:32:21] <grommit> So, just to be clear, I need to wire my hardware to the pins that the firmware is expecting the signals on (and not think I can reconfig it in software)?
[16:32:32] <elmo77> :P you think?
[16:33:08] <elmo77> why be difficult and wire it incorrectly... just like someone putting the yellow Video RCA into the white Audio location.
[16:33:32] <grommit> i guess that is one way to think about it.
[16:34:16] <grommit> I didn't think of it as "wiring incorrectly", I wired it logically (to me) and figured I would just reconfig the FPGA to know where things are.
[16:36:08] <pcw_home> "So, just to be clear, I need to wire my hardware to the pins that the firmware is expecting the signals on (and not think I can reconfig it in software)?"
[16:36:09] <pcw_home> Yes the firmware bitfiles have fixed pinouts, this doent prevent you from makin a new pinout but that involves re-compiling the FPGA config
[16:38:10] <pcw_home> Recompiling for a new pinout is relatively easy if you have and are used to the tools (only the pindesc file needs to be changed)
[16:48:19] <pcw_home> So if theres a custom pinout that makes sense I dont maind making a bitfile for that config, its usually just a matter of some cut+pasted editing and existing
[16:48:21] <pcw_home> pin description file to create a new one (BTW the reason that the pinouts are not dynamically changeble is that this uses a _lot_ of FPGA resources)
[16:53:52] <grommit> Ok, now that I understand it is not a runtime configuration feature, I can play by the rules. I had just misunderstood that. OTOH, it would be nice to have the tools (and instructions) available to build a cutom FW load...
[16:54:52] <grommit> The 7i47 wiring is set due to the pins being labeled for specific purposes (encoder, stepper, etc).
[16:55:26] <pcw_home> On 7I47 only thing that fixed is inputs vs outputs
[16:55:33] <pcw_home> (thats)
[16:55:38] <grommit> The 7i43 (and 7i42 protection) is ultimately generic. So I figured I could wire anything I wanted anywhere...
[16:57:27] <grommit> I can change it, it isn't a big deal as I currently only have the spindle and a couple other things wired to the 7i42 as my limit switches aren't built yet.
[16:57:42] <pcw_home> Well for GPIO thats somewhat true (the HM2 driver has the limitation that special output functions are only enableable on a 0-n basis not individually
[16:57:44] <pcw_home> so that may limit what you can do
[16:58:59] <pcw_home> (in other works if a config has 4 PWMGENS an you want only 2, your only choice is PWMGEN 0 and 1 (driver limit, not firmware limit)
[16:59:08] <pcw_home> (words)
[17:00:17] <grommit> On the 7i47 it shows specificly where to wire each differential encoder input and each step/gen output at least in the instructions. That is what I meant about it being "set".
[17:03:22] <morfic> elmo77: no, not installed anywhere, it'll all be data collecting and presenting to boss only, he will decide which way he wants to go
[17:03:29] <pcw_home> Sure, The maunal shows a stardard pinout (we use this pinout on our test encoders for example)
[17:03:31] <pcw_home> but theres no physical limit on the 7I47 other than inputs and outputs are fixed
[17:03:33] <pcw_home> (you could have a RS-422 serial port on a RX/TX pair if you wished)
[17:03:44] <pcw_home> (manual)
[17:03:56] <grommit> gotcha
[17:05:44] <grommit> Is it just me (it may be!) or is there a lack of documentation [for a beginner] on the use of the Mesa cards?
[17:08:42] <grommit> Once I figure this out perhaps I'll send you a "Mesa cards with EMC for dummies" :-)
[17:13:13] <pcw_home> Well I think the vast majority of people using our cards with EMC use a 5I20, a 7I33 and 1 or 2 7I37s
[17:13:14] <pcw_home> They can use standard bitfiles and start with the sample HAL files so they dont get into much of this exotica
[17:15:32] <pcw_home> Only in that last couple of weeks have people asked to use a 7I47 with a 7I43
[17:17:44] <pcw_home> The good thing is that the Hostmot2 driver and firmware make it easy to support different configurations but all this flexibility
[17:17:46] <pcw_home> can be bewildering at first
[17:20:37] <grommit> I'm not sure if I could tell you how I ended up with the 7i43 (vs a 5i20), but I got the 7i47 after you suggested it for differential encoder inputs. So my exotica is all your fault :-)
[17:22:23] <pcw_home> Well its a good hardware solution, but its first time with 7I43
[17:30:23] <grommit> I am using a DigisSpeed DC-06 board to drive the spindle motor controller. It has a Step input (and Gnd), and two relay inputs (RL1, RL2) which are passed through and used to control a KB electronics KBPB-125
[17:32:36] <sealive> hi i broke my Axis Z at speed test today
[17:32:59] <sealive> someone now if i can messure the L298 if this is broken
[17:33:23] <grommit> I have the DC-06 wired to the KBPB-125 so that RL1 on/off control CW/CCW, and RL2 on/off controls Brake/Run
[17:33:26] <sealive> ther is no power at the Axis
[17:34:00] <sealive> the sterpper works if i conect to X,Y
[17:34:33] <sealive> it has to be the 297 or 298 or both
[17:34:50] <grommit> Are these just wired to GPIO pins, and the Step input wired to a PWM out on the 7i42?
[17:35:42] <sealive> grommit: how you talking to
[17:35:55] <grommit> pcw_home
[17:36:02] <grommit> Or myself :-)
[17:37:29] <sealive> i think everyone is sercing for eggs
[17:37:41] <grommit> I know I am!
[17:39:35] <alex_joni> I sure know where mine are.. no need to search for them
[17:41:23] <sealive> :D
[17:42:20] <pcw_home> If its a step (rate) input I would think you should use a stepgenerator to drive it
[18:14:28] <pcw_home> Yes, the DC-06 uses a frequency (0-25 KHz) input so a PWMGEN is not what you want, a Stepgen in velocity mode will do what you need
[18:15:53] <pcw_home> ( you may be able to use the stepgens DIR output for CW/CCW subject to whatver your VFD reversal limits are)
[18:16:06] <pcw_home> (whatever)
[18:23:43] <JT-Hardinge> * JT-Hardinge has a new rule Put 2 things back for every 1 thing taken out!
[18:28:10] <archivist> cant be, you are supposed to have bits left over
[18:30:36] <JT-Hardinge> I mean tools :)
[18:30:44] <archivist> fun part is deciding if you are going to strip a thing to put the extra back in :)
[18:31:02] <grommit> thanks for that Peter!
[18:36:10] <JT-Hardinge> if tool 1 is the master tool in a turret do you touch off Z in G54 then touch off the rest of the tools in the tool table to get the Z offset from tool 1?
[18:41:50] <JT-Hardinge> wow this card had 47 jumpers on it
[18:52:12] <JT-Hardinge> * JT-Hardinge thinks I will take a nap as soon as I can stop the bleeding
[18:53:41] <archivist> bleeding! you been fighting with the machine and lost?
[18:54:32] <JT-Hardinge> naw, just cleaning up a bit and nicked my arm my blood is so thin that it don't take much to start a river of blood coming out
[18:55:55] <archivist> I prescribe a beer with a bit of body to thicken it up
[18:56:13] <JT-Hardinge> just finished the beer but it was a Mexican beer
[18:56:44] <JT-Hardinge> I'm out of good dark ale
[18:57:06] <archivist> I out as well :((
[18:57:52] <JT-Hardinge> drop by and we can make some
[18:58:52] <archivist> I cant afford the beer let alone a flight !
[19:00:19] <JT-Hardinge> yea! the bleeding stopped now I can take a nap
[19:14:17] <cradek> jtr: yes
[19:17:19] <cradek> hm, wrong jt
[19:23:43] <skunkworks_> hmm - I need some more coffee.
[19:26:19] <elmo40> /dcc send skunkworks_ grande_coffee
[19:55:38] <skunkworks_> Thanks!
[20:33:14] <elmo40> sorry, forgot cream and sugar.
[21:19:27] <grommit>
[21:20:08] <MrSunshine_> look grommit knows the keycode for like space or something good boy!
[21:20:09] <MrSunshine_> :P
[21:20:34] <grommit> just checking to see if am here
[21:20:54] <grommit> the java client had apparently crashed
[21:40:58] <frallzor> my simple diy touch plate =P
[21:41:01] <frallzor> cheap as cheap can be
[21:41:39] <MrSunshine_> i wish my gf was like a touch plate, never says no
[22:00:13] <tom3p> OT: how to fix python not finding a file? eg "jppy: Starting..ImportError: No module named jpilot_user" ) i know where the file is, python doesnt, and i dont see how to let python know.
[22:00:31] <tom3p> (seems to be how to edit sys.path dunno how thats done )
[22:07:33] <acemi> tom3p: http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=src/emc/usr_intf/axis/scripts/axis.py;h=371a4363a2ca94ad61966dadf3bc0d46601ddd12;hb=219b79110514b0d01e9d96cc931eac82fb9f0581
[22:07:42] <acemi> line 25-26
[22:14:19] <tom3p> acemi thx
[23:02:35] <frallzor> hmm found some interesting tools for milling glass
[23:02:46] <frallzor> anyone got some experience doing that?
[23:07:37] <VernonM> I've worked with ceramic, but never glass
[23:07:58] <VernonM> I'l bet they're NOTHING alike, too
[23:08:00] <VernonM> haha
[23:09:48] <VernonM> sorry :)
[23:16:05] <frallzor> hmm sleep maybe, need my energy for building the mill today
[23:21:45] <MattyMatt> frallzor, it'll be like drilling glass. very slow and lots of water
[23:22:03] <frallzor> just water?
[23:22:04] <MattyMatt> there are vids on youtube
[23:22:18] <SWPadnos> water actually seems to break bonds in glass
[23:22:29] <frallzor> glycol maybe?
[23:22:46] <MattyMatt> I think sand is involved, but that depends on the tool
[23:22:49] <SWPadnos> if you score a sheet of glass, then drip some water in it and leave it overnight, it's a lot easier to break the next day
[23:23:10] <Valen> I thought they generally ground glass
[23:24:48] <MattyMatt> I'll try it when my mill is waterproofed
[23:25:38] <frallzor> time to go die for a few hours then live again, NN
[23:26:12] <MattyMatt> a few more coats of polyurethane, it'll be good :)
[23:26:38] <MattyMatt> only kidding. I'll gradually replace all the wooden structure