#emc | Logs for 2010-04-19

[00:02:00] <MarkusBec> MarkusBec is now known as MarkusBec_away
[00:04:08] <cliffb> Yeah I think I'll use a couple of those suggestions, hard stops with the servo overload wired to the e-stop, large difference between 2 joints will kick an e-stop... that should work.
[00:10:47] <cliffb> is the homeing issue not a problem for any other machine types, like hexpods where things might jam if joints arn't coordinated?
[00:17:22] <pfred1> cliffb get some pigeons I hear they're excellent at homing
[00:17:44] <cliffb> :) just a question..
[00:18:47] <pfred1> the hexapod screwballs probably use cloded loop systems
[00:18:52] <pfred1> closed even
[00:20:43] <cliffb> Ok I'll live with it...
[00:22:06] <jimbo> When I enable TOOL_CHANGE_POSITION = 002 under [EMCIO] I received an error on startup of EMC stating that there is no iocontrol.0.user-enable-out? Do I have to have this output pin connected? It does not make sense as this is an output.
[00:22:09] <calcite> There must be a reason that "gantry" designs are not popular as industrial machine tools ;-) Difficult to keep aligned may be one/
[00:30:18] <andypugh> calcite: Well, in most situations it makes sense to only have one motor moving your gantry, and to link the two drives mechanically.
[00:31:05] <andypugh> Though I can see that the shaft will mess up the workpiece much of the time.
[00:31:10] <pfred1> andypugh why's that?
[00:32:06] <andypugh> Just less to worry about. I am thinking of conventional machine tools mainly.
[00:32:41] <calcite> Bridge mills with moving tables (long axis) are popular with large tools
[00:32:44] <pfred1> andypugh personaly I prefer engine troubles to transmission troubles
[00:38:00] <andypugh> Dual-engine setups in cars are fairly unusual though.
[00:38:47] <pfred1> even in electric cars?
[00:39:06] <andypugh> A fair point.
[00:39:19] <andypugh> Anyway, it's that time again.
[00:39:53] <pfred1> though i have seen some quad engine dragsters do some bad things
[01:01:33] <Jymmm> HAHAHAHA http://sfbay.craigslist.org/sby/zip/1698611079.html
[01:15:22] <Tonok2_> Tonok2_ is now known as Tonok2
[01:15:40] <calcite> Jymmm The other evening you mentioned a source for Lenovo r51 laptop batteries. Can you advise?
[01:32:14] <Jymmm> T4x actually http://dealextreme.com
[01:36:16] <calcite> ?
[01:42:32] <atmega> if you were using normally open switches for limit switches (because I have 4 with only NO contacts), would you use a pull-up on the port pin and then ground it via the switch, or use a pull-down and +5v?
[01:51:27] <ries_> ries_ is now known as ries
[02:01:45] <pfred1> atmega d all of the above
[02:03:07] <pfred1> atmega personally I plan to have the switch turn on a photodiode in an opto and have that trip a schmitt trigger but I'm just complicated like that I guess
[02:03:37] <Tonok2_> Tonok2_ is now known as Tonok2
[02:04:42] <pfred1> atmega if I do it right with a slow enough opto i may even eliminate debounce
[02:04:56] <atmega> too many wires/connections/etc.
[02:06:00] <pfred1> atmega oh thats nothing compared to when I go for optical switches later on I'm sure
[02:06:54] <atmega> I'm not really worried about debounce
[02:06:59] <pfred1> atmega initially I'm going with plain old microswitches though I've a pile of them
[02:07:21] <atmega> I have a bunch, but the ones with arms suitable for limits are NO only
[02:08:28] <pfred1> NO is a great way to go
[02:08:32] <atmega> so, I ran a 10k pullup from +5 to pin15 for estop-in and the HAL monitor thing said it was on
[02:08:56] <atmega> then I ran it to GND instead, and it said it was on.
[02:09:14] <pfred1> gotta hate when you get results like that
[02:09:45] <pfred1> though estop is an oddball you have to reset it
[02:09:55] <atmega> e-stop in is?
[02:09:55] <pfred1> estop is a latch
[02:10:05] <pfred1> once its tripped you have to reset it
[02:10:31] <pfred1> I think it latches in emc itself
[02:10:42] <atmega> I could see estop-out doing that
[02:11:10] <pfred1> I'm telling you estop is an oddball once you're in estop you're in it until you get out of it
[02:11:37] <pfred1> try your switch thing on another limit switch input and i think it'll work how you think it will
[02:11:39] <atmega> any idea where that is documented so I can read it?
[02:11:57] <pfred1> I just noticed messing around in simulation how estop was like that
[02:12:30] <pfred1> state had to be right at the pin then you had to reset it in order to unestop
[02:12:59] <atmega> my wiring is getting messy as is, adding discrete pullups/etc isn't helping.
[02:13:06] <pfred1> by default with nothing connected the sim is in estop
[02:13:12] <SWPadnos> atmega, were you using halshow, or halmeter/halscope?
[02:13:29] <SWPadnos> halshow doesn't always update - it will continue to show you the state the pin was in when you selected it
[02:14:08] <atmega> SWPadnos: I'm not sure which which one it was. BUt... I had it look at pin14 (spindle) and the screen updated when I changed it via axis
[02:14:14] <pfred1> SWPadnos now thats a handy feature
[02:14:21] <SWPadnos> hmmm. ok
[02:14:31] <SWPadnos> the best thing to use is probably halmeter
[02:15:18] <SWPadnos> pfred1, there are a couple of modes for halshow - one is "watch mode" which is what you need for things to update
[02:15:50] <SWPadnos> I think there's another mode, where things you click on will be shown as you click them, but it's more for exploration, not for monitoring
[02:16:20] <pfred1> SWPadnos I'm in such a mood with inconsistancies with all this CNC stuff lately I'm ready to chuck it all for a hobby that is more based in logic like shooting craps
[02:16:21] <Jymmm> SWPadnos: welcome home!
[02:16:41] <SWPadnos> Jymmm, thanks. I'm only headed to New York for the day tomorrow :)
[02:17:02] <atmega> SWP: any preference on pull-up/down for inputs?
[02:17:08] <SWPadnos> pfred1, you could do math puzzles or something. those are rooted in logjc
[02:17:10] <Jymmm> SWPadnos: Isn't 3D pr0n a hobby with consistancy?
[02:17:32] <Jymmm> SWPadnos: OH!!! Pr0n 360!!!!!
[02:17:46] <Tonok2_> Tonok2_ is now known as Tonok2
[02:17:48] <SWPadnos> atmega, it's slightly more fault tolerant for the input to be pulled low by the switch
[02:18:03] <Jymmm> SWPadnos: put that ANGLE button on the DVD player to good use!
[02:18:17] <pfred1> SWPadnos believe me whne I say I've done a lot of math puzzles lately in connection with EMC
[02:18:33] <SWPadnos> only because most inputs float high, so if the switch (and resistor) get disconnected, the input should float to the "unsafe" state
[02:18:48] <SWPadnos> Jymmm, that's why it was added :)
[02:19:14] <pfred1> SWPadnos that is how mine is rigged
[02:19:22] <atmega> I need a pulldown on my spindle relay, it comes on when the comptuer turns on until EMC runs
[02:19:44] <pfred1> atmega thats what a watchdog timer is for
[02:20:11] <atmega> I'm also tempted to ditch this xylotex and get a G540 just to make the wiring neater
[02:20:20] <pfred1> disables everything until the watchdog signal is generated
[02:21:01] <pfred1> me I'm just gonna have a big bad toggle switch on my controller for a while ;)
[02:21:12] <atmega> that's a future plan (disable A/C to drives/spindle/etc) until it is told to turn on.
[02:21:36] <atmega> I have all the parts, but the wiring is just too messy.
[02:21:47] <pfred1> but eventually I'll have to implment the automatic watchdog circuitry
[02:23:12] <pfred1> atmega yeah Mariss needs a new pair of shoes!
[02:23:14] <Calfin> Calfin is now known as Calcite
[02:23:55] <atmega> never actually seen one (or used one), but the 540 package looks good for clean wiring
[02:24:15] <SWPadnos> it's surprisingly small
[02:24:31] <pfred1> SWPadnos stuff costs money you know?
[02:24:44] <SWPadnos> it's about the same size as the face plate to a 1/2 height 5.25" floppy drive
[02:24:49] <SWPadnos> and about an inch thick
[02:24:52] <SWPadnos> yeah, I know
[02:25:05] <atmega> money vs. time
[02:25:22] <pfred1> no i meant Mariss buying the parts
[02:26:02] <pfred1> he's got to be having his stuff made in China by now there's no way he's still assembling it at home
[02:26:46] <SWPadnos> nope
[02:26:57] <SWPadnos> all done at his plant
[02:27:07] <pfred1> damn and i always thought he was smart
[02:27:21] <SWPadnos> he is. he has absolute control over quality this way
[02:27:32] <SWPadnos> and absolute responsibility
[02:27:54] <pfred1> give the Chinese the right specs they can deliver the goods
[02:28:18] <SWPadnos> that'is one choice
[02:28:21] <SWPadnos> -i
[02:29:12] <pfred1> heck they build every PC on the planet now i can't imagine they can't put together a motor controller
[02:30:15] <SWPadnos> the only reason to do offshore assembly of a product is to save money. if you have a large enough markup, and a large enough population is willing to buy your products with that markup, then there's no need to outsource
[02:30:20] <pfred1> I can't remember the last time I saw a piece of electronics that wasn't assembled in China
[02:30:53] <atmega> I got a fishfinder from .nz
[02:31:26] <pfred1> atmega are you sure they didn't slap the board in the case in nz?
[02:31:40] <atmega> no, but it said made in .nz
[02:31:53] <pfred1> made is a vague term
[02:31:54] <L84Supper> EMC2 for Ubuntu 10.04 should be fun. Canonical went with a kernel with broken DRM
[02:32:04] <atmega> they aren't exactly known for electronics
[02:32:59] <SWPadnos> night folks. got an early flight tomorrow morning
[02:34:45] <pfred1> I've worked with asians assembling electronics and i have to admit they are gifted at the task
[02:34:52] <atmega> I have a bunch of DIN mountable optos, but they seem to need 20ish mA
[02:36:12] <pfred1> I just got my TLP521s the other day and initially I drove them at 16 ma and they made nice waveform but a bit time shifted
[02:36:16] <L84Supper> well it's getting harder to find US PCB assemblers that can handle fine pitch BGA's
[02:36:32] <pfred1> so I just bumped up on the transistor side but i haven't measured or calculated the draw going on there yet
[02:37:27] <pfred1> L84Supper the little bit of sufrace mount we did we had to do all by hand which is why I detest surface mount today
[02:37:58] <pfred1> and at that the worst package we dealt with was the transistors
[02:38:38] <L84Supper> heh, I just redesigned a board for customer in the US that was last designed in 1985
[02:39:04] <pfred1> that was around the time I did that sort of work maybe a bit later
[02:39:08] <L84Supper> it's all through hole except for 4 out of 600 parts
[02:39:19] <pfred1> yeah we did all through hole
[02:39:23] <pfred1> wave soldered
[02:39:48] <L84Supper> I had already forgotten the layout rules for through hole
[02:40:03] <pfred1> I throw our surface mount stuff
[02:40:09] <L84Supper> 10+ years since the last one
[02:40:13] <pfred1> right into the trash
[02:40:54] <L84Supper> I had a hard time finding through hole parts to do the job
[02:41:09] <pfred1> oh its an issue for me building motor drivers
[02:41:14] <L84Supper> 2-3X the price of smt as well
[02:41:24] <pfred1> though none of the surface mount stuff handles any kind of power
[02:41:39] <atmega> I wanted some DIP i2c A/D chips, they don't seem to make them
[02:42:16] <pfred1> atmega for logic lots of times you can get away with using chip carriers
[02:42:39] <pfred1> but power stuff needs heat vias and whatnot
[02:42:41] <L84Supper> I was happy to find I2C GPIO expanders in TH
[02:42:46] <atmega> yeah, I got some SOT23 to DIP boards and hand soldered them
[02:43:46] <pfred1> I think half the rush to SMT is to totally kill the hobbyiest end of electronics
[02:44:07] <L84Supper> evil communist plot? :)
[02:44:09] <atmega> easier to automate SMT
[02:44:50] <pfred1> atmega well then what are we going to do with a billion unemployed chinese?
[02:45:04] <L84Supper> http://www.theonion.com/articles/intel-unveils-oversized-novelty-processor,9790/
[02:45:11] <atmega> create govt jobs programs!
[02:45:27] <pfred1> atmega that hasn't worked out to owell here really
[02:45:40] <atmega> heh, really?
[02:49:31] <pfred1> L84Supper thats for those old folks that need big phones and remotes you know?
[02:50:12] <pfred1> L84Supper just imagine how they feel when it comes time for them to upgrade their CPU!
[02:50:49] <L84Supper> it would be about the size of a integrated single chip Xeon server
[04:55:56] <Tonok2__> Tonok2__ is now known as Tonok2
[08:11:51] <herron_> herron_ is now known as herron
[08:24:12] <sealive> hi someone ever did pdf to g-Code?
[08:27:49] <herron> never heard of such madness
[08:35:52] <sealive> thanks
[08:36:19] <sealive> irecived a shape PEN Drawing via pdf
[08:38:06] <herron> here I downloaded a PDF editor and extracted screen grabs
[08:38:42] <sealive> i use Gimp
[08:59:12] <mrsun_> sealive, just export as a jpeg or something ? :)
[08:59:22] <sealive> yes
[09:02:51] <piasdom> g'mornin all
[09:05:15] <sealive> hi
[09:08:39] <piasdom> what would i use to convert a jpeg to g-code ?
[09:09:44] <herron> there is something on the wiki image2gcode or something like that
[09:10:21] <piasdom> i'll look...thanks herron
[09:11:00] <sealive> i use freemill http://www.sammellothar.de/mount_m.png
[09:13:34] <piasdom> sealive; that looks cool
[09:15:30] <sealive> http://www.sammellothar.de/beet_m.png better then obove
[09:36:11] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[09:50:09] <piasdom> i tried image2gcode and all i get is a bunch of this 'Q|KÉøþCµé|{3<íÁã
[09:56:44] <PCW_> PCW_ is now known as PCW
[11:37:51] <MarkusBec> MarkusBec is now known as MarkusBec_away
[11:42:50] <skunkworks> JT-Dev: overtravel limits.. That is how the machine is setup (all in series) maybe when we get all done we will have some extra wires we can use.
[11:43:10] <JT-Dev> cool
[11:43:49] <Jymmm> skunkworks: With PANDUIT, You Can Do It!!!
[11:47:32] <Mortez> I tried mecsoft freemill but I can't save the output. Am i missing something...
[11:53:42] <Mortez> hmm it saves the gcode to a temp.acl no prompts, well I quess it works
[12:08:24] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[12:51:33] <frallzor> http://pici.se/p/large/SOBlbCaNU/ progress =)
[12:53:50] <mrsun_> frallzor, oo nice =)
[12:53:59] <mrsun_> är du nöjd med fräsen då ? :)
[12:54:06] <mrsun_> för det priset borde du fan vara det :P
[12:54:14] <frallzor> yes I am
[12:54:38] <mrsun_> ops, was emc not elektronik :P
[13:04:55] <elmo401> frallzor: good job :) What do you use for making the gcode?
[13:05:10] <frallzor> artcam
[13:05:15] <elmo401> I see
[13:05:30] <frallzor> tweaked step over too so its spot on
[13:05:47] <elmo401> looks very smooth
[13:05:51] <elmo401> not much sanding required ;)
[13:06:03] <frallzor> doing both roughing and final run with same toom
[13:06:05] <frallzor> *tool
[13:06:27] <elmo401> 1/8" ballnose?
[13:06:46] <frallzor> 6mm
[13:06:51] <frallzor> 1/4 aprox
[13:06:56] <elmo401> I am trying to figure out this inkscape gcode plugin.
[13:07:26] <frallzor> i made a small mistake though
[13:07:42] <frallzor> didnt realize tool settings would be the same for both roughing and final run
[13:07:50] <frallzor> so it does rough a bit too good =)
[13:09:51] <Dave911> Jymmm.... are you a Panduit salesman, or do you just like wireduct .. .... ;-)
[13:10:51] <Dave911> OK, off the subject of Panduit ...
[13:11:17] <Jymmm> It's good shit. Makes troubleshooting easy, and gives a professional and clean appearance to the job.
[13:11:43] <Jymmm> It's the bling bling of the electrical world =)
[13:12:15] <Dave911> You are an easy guy to please then ... :-) electrically speaking ...
[13:12:55] <Dave911> wire duct is nice..... I have been using the AD wire duct but Panduit brand is not that much more expensive
[13:13:45] <Dave911> Actually I think it a lot harder to wire stuff without it than with it .... and still make it look neat .
[13:17:18] <frallzor> http://pici.se/p/large/lOhKsnJrQ/ current status elmo401 =)
[13:17:43] <elmo401> pretty shallow cuts. what is your IPM feed?
[13:18:33] <elmo401> nice wires :P can't put them in a loom?
[13:19:08] <frallzor> the cuts are around 5mm but step over = 0.3mm
[13:20:37] <frallzor> 0.012" i believe is the step over
[13:21:21] <herron> rats nest wiring rulz
[13:22:06] <Dave911> I'm surprised that Jymmm isn't suggesting Panduit .. ;-)
[13:22:40] <Dave911> frallzor.. that looks really nice.... I see you got a dust skirt made
[13:23:10] <frallzor> temp one
[13:23:24] <Dave911> OK ..
[13:23:30] <frallzor> having trouble design the new one for the proper dust collector =)
[13:23:34] <frallzor> *designing
[13:23:42] <Dave911> What are you using for CAM to do those designs
[13:23:46] <frallzor> artcam
[13:23:58] <Dave911> Very nice
[13:24:38] <Dave911> Machine Referencing: Is there a way to "home reference" a machine via Gcode (set machine coordinates G53, back to all zeros via Gcode? Or via a custom M code? I don't want to actually have the machine to actually search for a home switch, I just want the machine G53 coordinates to be reset to zero at the end of a long program full of incremental moves..
[13:24:54] <frallzor> * frallzor loves his network camera
[13:25:20] <Dave911> network camera ... how does that work?
[13:26:18] <frallzor> monitoring the work from inside
[13:26:40] <frallzor> I visit the garage from time to time though
[13:26:48] <Dave911> oh.... I got it .... for long runs that would be nice
[13:27:06] <frallzor> this will take 307 minutes =)
[13:27:12] <frallzor> according to emc
[13:27:41] <Dave911> Hmmmm... don't feel like watching the machine for 5+ hours!
[13:28:27] <Dave911> Good thing that machines don't need sleep and food... and coffee breaks and .....
[13:28:30] <frallzor> thats why im inside doing other stuff :P
[13:28:46] <Dave911> makes sense
[13:32:42] <elmo401> hrmm, I downloaded, extracted to the /plugins directory yet the gcode plugin for Inkscape is not showing...
[13:37:41] <elmo401> ahh, supposed to be in /share/extensions directory... why so many different locations :P
[15:19:59] <JT-Work> mighty quiet around here
[15:21:18] <frallzor> Scchhhh
[15:21:25] <frallzor> Im hunting wabbit
[15:21:37] <JT-Work> ok
[15:22:35] <frallzor> bah I messed up X now :P
[15:22:47] <frallzor> just because I wanted to stop the material from lifting :P
[15:24:14] <skunkworks_> JT-Work: how is the hardage?
[15:24:49] <JT-Work> Doing good, I'm lacking a few things to finish up
[15:25:22] <JT-Work> waiting on the lube tank replacement, need to change the Z belt and get my FOK working
[15:28:15] <skunkworks_> neat
[15:29:05] <JT-Work> got the garden in yesterday so not much time to work on the Hardinge after serving a rat writ to a rat in the garden shed
[15:34:29] <Dave911> logger_emc:bookmark
[15:34:29] <Dave911> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2010-04-19.txt
[16:05:24] <terrylm> Hi all
[16:05:49] <terrylm> Has anyone here ever used the auto tuning PID?
[16:06:08] <terrylm> at_pid
[16:06:37] <cradek> I've tried it approximately twice
[16:07:58] <JT-Work> auto tune in EMC?
[16:08:43] <pcw_home> And it worked perfectly each time?
[16:09:16] <cradek> it didn't give very usable results either time
[16:09:37] <terrylm> How do I interface with it to trigger it? That is, it uses bi-directional pins, pyvcp only has uni-directional pins, the ladder also only has uni-directional pins?
[16:10:13] <cradek> halcmd setp always works
[16:10:31] <terrylm> Oh, ok, thanks.
[16:24:31] <cradek> seb_kuzminsky: how's the machine coming?
[16:24:57] <seb_kuzminsky> it's making parts :-)
[16:25:01] <seb_kuzminsky> i made a t-slot cleaner :-)
[16:25:15] <cradek> cool! how did you hold it to cut that odd shape?
[16:26:05] <seb_kuzminsky> i put an oversized sheet of delrin on top of a piece of wood and put four clamps outside where the part would be cut
[16:26:15] <seb_kuzminsky> it worked great :-)
[16:27:00] <JT-Work> cool! what kind of machine is it?
[16:27:08] <seb_kuzminsky> i wrote the program for it with two tools, a drill and an end mill, but the difference in length was so big i had to move the knee and re-touch-off z in the middle, which was awkward
[16:27:15] <cradek> neat. so you cut all around in one shot? sometimes the part gets nicked (or thrown) when it comes loose...
[16:27:16] <seb_kuzminsky> i think it'd be better to drill in a separate program, or by hand
[16:27:20] <seb_kuzminsky> heh
[16:27:25] <seb_kuzminsky> yes, and yes :-/
[16:27:41] <seb_kuzminsky> JT-Work: it's cradek's old bridgeport :-)
[16:27:59] <JT-Work> cool a series 1?
[16:28:09] <seb_kuzminsky> series 1 r2e3
[16:28:19] <cradek> seb_kuzminsky: with that machine I sometimes had a job that had "short tools" and "long tools" with their lengths measured differently by 1" or 2". then when changing tools, I'd crank the knee up or down by 1" or 2". (did you notice the knee crank has a marked dial? it's 0.1" per turn)
[16:28:55] <seb_kuzminsky> cradek: that sounds like a good way to do it
[16:29:05] <seb_kuzminsky> must be nice on Jr to have all that travel on the head :-)
[16:29:29] <JT-Work> I have a series 1 here in the shop with an Anilam control on it
[16:29:32] <cradek> only in desperation - it's easier to drill by hand first or last
[16:30:12] <seb_kuzminsky> i've been thinking a bit about the "kneekins" we talked about
[16:30:36] <EricKeller> what's a kneekins?
[16:40:38] <seb_kuzminsky> EricKeller: it's a hypothetical way to drive a machine with a knee and a quill, to get the full combined travel of both joints for your Z travel
[16:41:02] <EricKeller> I was wondering. I'd like to do that with my mill
[16:41:08] <seb_kuzminsky> me too :-)
[16:41:19] <cradek> me too (I'd like you to do that with your mill)
[16:41:23] <seb_kuzminsky> heh
[16:41:42] <EricKeller> but it takes so much force to move the knee at this time, gotta work on that
[16:41:44] <seb_kuzminsky> cradek: next time i think i'll drill some holding-holes in the workpiece and screw it to the wood underneath
[16:41:50] <cradek> ideally you could mount a scale and just use an ac motor
[16:42:39] <cradek> seb_kuzminsky: not sure about delrin, but on aluminum I sometimes cut the outline all but 0.005" deep and then whack it with a hammer to get it out, and then file off the little edge
[16:43:05] <cradek> works best with really thick stuff that holds its flatness when you clamp it
[16:43:25] <seb_kuzminsky> cool - i've seen people leave just some little tabs spread around the perimeter, so there's less stuff to file afterwards
[16:43:35] <cradek> yep I bet that works too
[16:43:36] <seb_kuzminsky> heekscad has an option to do that
[16:44:02] <cradek> I did the big plate on the bottom that way: http://timeguy.com/cradek-files/emc/spindle-encoder.jpg
[16:44:12] <seb_kuzminsky> that actually sounds way quicker & easier than mesing with wood screws for more setup
[16:44:27] <JT-Work> If I have at least two holes I'll bolt it down to a piece of aluminum to mill a profile like that
[16:47:23] <cradek> heh, spindle speed a little too high, or spindle running backward? http://www.youtube.com/watch?v=T43DrVzlPnA
[16:49:27] <JT-Work> sounds like both
[16:51:01] <cradek> wow, you can sure learn what not to do, using youtube
[16:51:06] <JT-Work> http://www.youtube.com/watch?v=iQGP97mWWm8&feature=related
[16:51:08] <elmo401> frallzor: did you mess up? :(
[16:52:10] <seb_kuzminsky> cradek: fwoar!
[16:52:13] <JT-Work> like chucking up a long part with a short bit of part in the chuck and running too fast and taking too deep of a cut
[16:52:17] <elmo401> JT-Work: OUCH!
[16:53:04] <seb_kuzminsky> i ran my spindle too slow this weekend, but i just pushed the "spindle faster" switch :-)
[16:54:36] <cradek> hahaha http://www.youtube.com/watch?v=RZB8W81ae_g
[16:56:26] <elmo401> http://www.youtube.com/watch?v=yln_IGDuOCo&feature=related
[16:56:35] <JT-Work> been there done that http://i47.photobucket.com/albums/f163/johnplctech/Machine%20Shop/HPIM0598.jpg
[16:57:28] <elmo401> not in a LONG time.
[16:57:51] <JT-Work> been a while for me too
[16:58:20] <cradek> I don't think I've done exactly that...
[16:58:30] <elmo401> now, at home, it doesn't matter... my machine cost me a total of $500 :P
[16:58:48] <cradek> I've rapided a 3/4 mill down into a (delrin) part that I remember
[16:59:07] <elmo401> tiny aluminum vice is easy to mangle but *shouldn't* break the tool.
[16:59:08] <cradek> offsets helped me to do it
[16:59:23] <sealive> hi do someone now whats thew best to get a g-Code for letter styro cut
[16:59:36] <elmo401> I have seen in my old shop people plunging G0 into the table with 6" face mills... $15,000 later a new spindle
[17:00:51] <JT-Work> that endmill broke because of a -
[17:04:01] <seb_kuzminsky> JT-Work: clearly we should outlaw -'s... for the children
[17:05:00] <JT-Work> yes, the mill should wrap around instead of being able to go back and forth
[17:09:58] <JT-Work> here is one of Dave's http://www.youtube.com/watch?v=Ge9_Qbd09zU&feature=PlayList&p=60288DA968439148&playnext_from=PL&playnext=2&index=63
[17:12:27] <elmo401> interesting playlist...
[17:20:17] <awallin> I wonder if the wing-nut is optimal there...
[17:36:27] <awallin> anyone know why "qtdemo" looks fine in nvidia graphics, but produces rubbish on intel graphics
[17:36:33] <awallin> both are ubuntu 9.04 machines
[18:38:13] <frallzor> finally I gave up
[18:40:33] <frallzor> I need to dust and lube the machine now :P
[18:40:47] <frallzor> its whining after all the dust =P
[19:17:38] <skunkworks_> Lerman: ! how is it going?
[19:40:27] <frallzor> http://pici.se/p/GQiPNneGn/
[19:42:18] <skunkworks_> very nice!
[19:42:37] <mikegg> cool, what is it for?
[19:45:29] <frallzor> nothing :P
[19:45:36] <frallzor> playtime :P
[19:45:39] <frallzor> not much mpre
[19:45:41] <frallzor> *more
[19:48:58] <mikegg> it would make a nice head board on a bed
[19:49:50] <frallzor> I could make a proper one for that
[19:50:02] <frallzor> 16mm mdf aint exactly top notch material :P
[19:50:08] <mikegg> heh
[19:50:11] <alex_joni> skunkworks_: something for you http://www.theregister.co.uk/2010/04/19/google_search_spelling/
[19:50:54] <skunkworks_> heh - I use google a lot to spell.
[19:51:02] <frallzor> I need to plane my workarea though =)
[19:51:19] <frallzor> and make some nice means of attaching pieces properly
[19:53:33] <frallzor> need to make the dust collection working properly too, the machine is whining now with all the dust =)
[19:54:38] <mikegg> yeah, i'm always amazed at how much dust a router can produce. I try to use those things outside if I can
[19:55:53] <frallzor> time to design a simple dust shoe!
[19:57:12] <frallzor> the old dust buster cant take the heat now :P
[19:58:00] <mikegg> ugh I hate solid edge
[19:58:19] <mikegg> revolved cut of 359 deg is fine. but 360? forget it
[20:15:27] <andypugh> Is there a very simple C instruction to increment-and-wrap. I want to count from 0 to 5 then back to 0. Speed matters.
[20:16:20] <bill2or3> if (++foo=6) { foo=0; } ?
[20:16:29] <bill2or3> or simpler?
[20:16:45] <andypugh> That was what I had come up with, but I am allergic to "if" statements
[20:17:27] <cradek> ==
[20:17:33] <andypugh> And shouldn't it be ++foo==6 ? (serious question, I am a C-noob)
[20:18:01] <cradek> you probably won't make a faster version than that.
[20:18:50] <andypugh> Reminds me of a bug I found in a QBasic dyno control program written by a C-coder. He had written "x = y = 0" to reset x and y to zero...
[20:18:58] <archivist_attic> assembler :)
[20:19:04] <cradek> although if you mistakenly start with foo equal to 721 this won't work very well. in that case ++foo>5 might be a better test.
[20:19:12] <cradek> (assuming foo is unsigned)
[20:19:28] <andypugh> Well, I wasn't going to use "foo" at all :-)
[20:19:40] <andypugh> Thanks, back to coding.
[20:19:43] <cradek> andypugh: what's that do in qbasic?
[20:20:03] <bill2or3> yeah, ==.
[20:20:06] <bill2or3> <--- also newb.
[20:20:15] <andypugh> it sets x to 1 if y==0 and 0 otherwise
[20:20:42] <andypugh> ie it sets x to the logical value of the comparison y=0
[20:20:46] <cradek> oh, funny
[20:21:32] <andypugh> It took some spotting in someone else's code (and I was there with my mechanical hat on)
[20:24:05] <andypugh> While you are there... a fast "if (index is odd)"? In assembler you would right-shift and look at the carry...
[20:24:53] <cradek> if (index & 1) is roughly the equivalent test in C
[20:25:25] <andypugh> any non-zero is interpreted as logical "true"?
[20:25:31] <cradek> assuming index is unsigned: be careful you know what "odd" means if it is signed and negative
[20:25:41] <cradek> yes
[20:25:56] <andypugh> Thanks.
[20:27:36] <frallzor> thar I go, dustshoe designed
[20:27:45] <atmega> jpegs?
[20:29:18] <cradek> what
[20:29:40] <atmega> dustshow jpegs
[20:30:01] <atmega> assuming a dustshoe is something that sucks up dust
[20:32:14] <frallzor> nah no need for jpegs
[20:32:21] <frallzor> nothing special about it
[20:32:45] <frallzor> just a simple static shoe that doesnt move in Z
[20:33:12] <atmega> mine consists of me holding the shopvac hose.
[20:35:29] <frallzor> I got myself a proper shopvac
[20:35:48] <frallzor> just need to attach it to the gantry in the ceiling and make a shoe for it
[20:37:35] <atmega> I got mine in 1992ish, it wasn't very good then.
[20:39:31] <frallzor> I need to figure out camworks
[20:44:48] <frallzor> dang its complicated
[20:53:58] <Lerman> skunkworks: All is well. Life is good. Busy as usual. And with you?
[20:55:23] <skunkworks_> Lerman: pretty much the same.
[20:55:26] <skunkworks_> :)
[21:27:40] <andypugh> The problem with combination projects like my Arduino thing is that it gets awkward when the computer is downstairs and the Scope, PSU, proto-board etc are upstairs, and you want serial output at the same time as the Scope.
[21:28:28] <andypugh> I think I might bring my HP100 home from work to use as a dumb serial terminal.
[22:47:07] <andypugh> Bother, the compiler has hung. I really should have saved my work at least once in the last 4 hours, shouldn't I?
[23:00:31] <andypugh> Phew! I left it for 20minutes and it recovered.
[23:30:36] <andypugh> I am completely baffled:
[23:31:29] <andypugh> If I have a function int arctan(long V1, long V2) and I don't change the value of V1 or V2
[23:32:50] <andypugh> But I call it with arctan(V[0] - Z[0], V[1] - Z[1]) where V[] is long and Z[] is int. Would you expect the value of Z to change?
[23:33:46] <andypugh> I am guessing that I am falling foul of a call-by-reference?
[23:40:20] <MarkusBec> MarkusBec is now known as MarkusBec_away
[23:40:23] <JT-Hardinge> I'm not even sure I understand what you said lol
[23:44:19] <andypugh> That makes two of us
[23:47:43] <JT-Hardinge> hmm with a 2.4 rip I can't remember how to get past this error Can not find -sec EMC -var NML_FILE -num 1
[23:48:26] <andypugh> Commenting out the function call doesn't remove the problem.
[23:48:35] <andypugh> (My problem that is, not yours)
[23:49:31] <JT-Hardinge> time to strap on the feed bag here anyway