#emc | Logs for 2006-12-07

[01:26:13] <tomp> hello
[01:33:11] <tomp> 64
[01:39:05] <SWPadnos> 128
[01:40:22] <tomp> sorry fumbled on kb :-(
[01:40:34] <SWPadnos> heh
[01:40:42] <SWPadnos> was that meant to be 53?
[01:41:02] <tomp> no, stood kb on edge makin room for soup
[01:41:08] <SWPadnos> heh
[01:42:31] <tomp> lookin at new spinbutton widget, for analog setpoint of process
[01:42:40] <SWPadnos> cool
[01:42:41] <tomp> thanks jmk
[01:43:18] <SWPadnos> did he add spinbuttons, or are you doing that?
[01:43:34] <tomp> but need analog out.. maybe the progress bar? to stay inside 'normal widgets' of gtk+
[01:43:43] <tomp> no, john did it
[01:43:57] <SWPadnos> ah right - spinedit, not a knob - duh
[01:44:38] <tomp> knob not necc, but some output for an analog, tho a label can even be used...
[01:44:44] <SWPadnos> it should be very easy to make a "digital display" - all you do make a label widget and sprintf
[01:44:46] <SWPadnos> yep
[01:45:35] <tomp> the jitter of a changing signal is why meters are good, lsd is jiggly on some signals
[01:46:01] <SWPadnos> you can filter signals before you pass them to the display
[01:46:22] <tomp> i want to see the jiggly bits...
[01:46:22] <SWPadnos> there's a component that's basically an integrator
[01:46:24] <tomp> :-)
[01:46:28] <SWPadnos> oh - well in that case ... ;)
[01:46:47] <SWPadnos> this isn't really analog - don't forget that
[01:46:58] <tomp> what isnt...
[01:47:24] <SWPadnos> you won't see a wiggling meter needle (drawn with X) much better than a toggling digit (drawn by X)
[01:47:55] <tomp> hmm, vtk was good at it.. and some other toolkits...
[01:48:01] <SWPadnos> the UI only updates ~10 times per second as well (user settable, I think, but dependent on system load)
[01:48:15] <tomp> oh yes, it's 'snapshit-ed' yes
[01:48:24] <SWPadnos> yep ;)]
[01:48:24] <tomp> snapshot-ed
[01:48:28] <SWPadnos> close enough
[01:48:36] <SWPadnos> I like snapshit, actually
[01:48:41] <tomp> but so are my eyes at some freq
[01:49:12] <tomp> magicians rely on it
[01:49:15] <SWPadnos> hmmm - it would be interesting to have a "peak hold with decay" HAL component
[01:49:38] <tomp> some linux meter widgets have the peak nedle
[01:49:50] <SWPadnos> feed in signal, and on a fast thread, update max and min, but decay those toward the input value at some rate
[01:50:26] <SWPadnos> like level meters on old cassette decks
[01:50:50] <tomp> i saw beautiful svg meter somewhere, i saved 'em but svg spec and browsers change too much
[01:50:51] <SWPadnos> for display, you can have error bars around the needle
[01:51:01] <tomp> yeh, level meters are nice tools
[01:51:11] <SWPadnos> yep
[01:51:27] <tomp> and there were meter you could set bands and get alarms
[01:51:49] <SWPadnos> hmmm - never seen those (but I'm a relative newbie at the automation game)
[01:52:06] <tomp> oh, very old stuff
[01:53:11] <tomp> maybe a flash2emc interface
[01:53:30] <SWPadnos> not macromedia flash?
[01:53:47] <SWPadnos> (ewwww)
[01:54:13] <tomp> i think it could be a great gui builder
[01:54:24] <SWPadnos> * SWPadnos slaps tomp
[01:54:33] <cradek> * cradek slaps tomp
[01:54:39] <SWPadnos> if only they had god Linux support ...
[01:54:44] <SWPadnos> or even good
[01:54:46] <tomp> http://www.eclipse.org/tptp/platform/documents/design/ChartingService/concepts/meterchart.html
[01:54:49] <cradek> if only it were free software
[01:54:59] <SWPadnos> if only it were good software
[01:54:59] <tomp> yeh, thatd be nice
[01:55:22] <cradek> I'd much rather have free/open than good (from some company)
[01:56:11] <tomp> like, just now, ffox 1.4 blew up (dissapeared, when i left an svg page
[01:56:16] <tomp> 1.5
[01:56:59] <skunkworks> internet explorer is free
[01:57:10] <SWPadnos> free, not free
[01:57:10] <skunkworks> buha ha ha ha ha
[01:57:12] <SWPadnos> * SWPadnos slaps skunkworks
[01:57:13] <cradek> different free
[01:57:25] <tomp> adobe bought mmedia and will drop the svg plugin
[01:57:44] <SWPadnos> of course - it competes with postscript
[01:58:14] <tomp> competes with flash
[01:58:32] <SWPadnos> I suppose, but SVG has no provision for user interaction, AFAIK
[01:59:27] <tomp> hmm? i used it wit DOM and had btns, meters, lotsa widgets ... www.kevlindev
[01:59:58] <SWPadnos> ok - with other programming means, you can do it, but SVG is just a graphics format, like jpg (though vector)
[02:00:09] <SWPadnos> unless they've added to it
[02:00:30] <tomp> yeh, just the nice little unix tool to be used with the others...
[02:00:38] <tomp> unix like
[02:02:29] <tomp> why gnome (gtk+) widgets chosen ? and not qt?
[02:08:39] <SWPadnos> most of the code was gtk based (like classicladder, halscope, etc)
[02:09:08] <SWPadnos> that may hae to do with the C-like nature of it, rather than the c++-like nature of qt (from what I recall anyway - I could be wrong about that)
[02:13:38] <SWPadnos> http://www.mumblelina.com/index.php?gtkvumeter=1
[02:14:51] <tomp> looks nice.. lemm e see if it takes floats as input
[02:15:11] <SWPadnos> probably 16-bit ints, but that should be easy to change
[02:16:21] <tomp> and it exists as xmms plugin for easy testing ( with dynamic input )
[02:16:28] <SWPadnos> heh
[02:16:46] <tomp> hmm, so where it says download, what do you click on?
[02:17:01] <SWPadnos> the revision numbers
[02:17:05] <tomp> doh!
[02:24:13] <tomp> it requires automake1.7
[02:25:03] <SWPadnos> bummer
[02:25:08] <SWPadnos> or - no problem
[02:25:12] <SWPadnos> I don't know which :)
[02:26:39] <tomp> heh
[02:33:46] <tomp> ok, it seems to take floats as the min/max and the level, now into the code to see how it aligns to vcp_widget rules
[02:34:33] <tomp> oh, and it scales itself to the window it's in
[02:35:44] <tomp> can be ine or log
[02:35:47] <tomp> lin or log
[02:36:57] <SWPadnos> sounds good so far ..
[02:40:07] <tomp> whats anjunta?
[02:40:18] <tomp> some ide?
[02:40:30] <SWPadnos> yes
[02:40:35] <SWPadnos> Anjuta, I think
[02:40:41] <SWPadnos> (is there a second N?)
[02:41:07] <SWPadnos> ahhhh - the smell of SOOW wire
[02:41:40] <tomp> no 2nd n, it is an ide
[02:42:38] <SWPadnos> yes - I have it installed, but I'm not at that machine
[02:43:56] <tomp> looks like too many deps... it already compiles w/o it, so i'll just read, hope it can be simpler ( no gradient... more like a meter)
[02:44:30] <SWPadnos> yeah, though having green/yellow/red thresholds would still be good
[02:49:30] <tomp> unreadable when min/max less than 1, but ok at typical transducer levels of +/- 10
[02:49:56] <SWPadnos> there's probably some int remnant in there
[02:50:05] <tomp> and i dont get the use of green yellow red on bipolar signals
[02:50:32] <SWPadnos> actually, magnitude > threashold should trigger a color change
[02:50:48] <SWPadnos> motor drive strength, for example
[02:52:43] <tomp> so, the 'level' is a set point, not the immediate value?
[02:52:54] <SWPadnos> I ahve no idea - I'm not looking at the code ;)
[02:53:48] <tomp> i'm looking at the sample app, it shows a vumeter with dialog for min/max/level, radiobtns for log/lin
[02:54:43] <SWPadnos> I'm looking at an IRC window and a transformer that I'm adding wires to :)
[02:55:26] <tomp> ok, watch your fingers ( keep one thumb hooked in the center rear belt loop :-)
[02:56:25] <SWPadnos> the transformer has never been hooked to power here - I just pulled the CT wire and am separating the two secondaries so I have - err, two secondaries ;)
[02:57:01] <SWPadnos> at which point I'll have a nice 44V / 40A toroidal transformer
[03:05:29] <tomp> 1.6kva near 70v dc... for stepper drives?
[03:08:38] <SWPadnos> servos
[03:09:04] <SWPadnos> closer to 60V after the bridge
[03:09:44] <SWPadnos> the servos are rated to 100V (actually 150, but 100 gives max speed of 2500 RPM), and the continuous current is ~10A, peaks near 40A each
[03:10:21] <SWPadnos> hey - do you think nail polish remover would remove enamel coating from transformer wire?
[03:11:09] <tomp> dunno, i use flyback dope on tiny electrodes that cut deep, i remove the coating with edm :-)
[03:11:26] <SWPadnos> heh
[03:11:36] <tomp> flyback dope aka trafo enamel
[03:12:14] <SWPadnos> well - nail polish is enamel as well, so I'm trying it
[03:12:23] <tomp> 400 watt motors?
[03:12:46] <SWPadnos> closer to 1KW at full power
[03:12:55] <SWPadnos> or 6 KW at peak
[03:12:55] <tomp> yeh 100x10
[03:13:44] <tomp> if you just want to strip the coating of a length, try scotchbright
[03:13:49] <SWPadnos> they're nice Baldors - the voltage rating is actually higher than the sum of (full current * winding resistance) + back EMF at full speed
[03:14:09] <SWPadnos> so you can still get full peak torque, even at full speed
[03:14:11] <tomp> i use baldors, m55 series 175V 40amp
[03:14:35] <SWPadnos> yeah - I was using sandpaper, but got lazy ;)
[03:14:48] <SWPadnos> these are MTE4070-BLBCE, or something like that
[03:14:57] <tomp> brushed dc? or..
[03:15:07] <SWPadnos> yep
[03:15:31] <tomp> 1 year later make sure you check the brushes ( esp on the tachos )
[03:16:10] <tomp> at least open 1 brush cap and apply dry air
[03:16:37] <SWPadnos> I'm not using the tachs, but I will check the motor brushes once I actually mount them to the mill ;)
[03:17:13] <tomp> i used copley amps with 'em
[03:18:10] <SWPadnos> geckos here, but I may actually go to something like skunkworks' driver with a mesa card
[03:18:34] <SWPadnos> I would definitely want current limiting and short-circuit protection though
[03:19:36] <tomp> i see skunkworks here sometimes, and you say he has a driver... is he a developer?
[03:19:47] <SWPadnos> he's here a lot
[03:19:59] <SWPadnos> he was palying with a design, and JMK helped him finish it off
[03:20:14] <SWPadnos> or did it, depending on your perspective ;)
[03:20:23] <tomp> yeh, it's me that is snapshotting :-)
[03:20:47] <tomp> i found another vumeter.. http://vumeterplugin.sourceforge.net/ i think i'll look into that if it's gtk+ able
[03:21:00] <tomp> cuz it'
[03:21:08] <tomp> cuz it's more meter like
[03:21:37] <tomp> skinable
[03:21:38] <SWPadnos> drawing a VU meter is pretty easy
[03:21:46] <SWPadnos> I don't know how hard it is to make gtk do it though ;)
[03:22:12] <tomp> i can read the code so it must not be dificult
[03:22:18] <SWPadnos> I think the last time I did that was using the Turbo Graphics toolkit for Turbo Pascal (circa 1987)
[03:22:21] <tomp> the gtk code
[03:22:31] <SWPadnos> reading != writing :)
[03:23:12] <tomp> was pascal the := language that told you all the things you did wrong?
[03:23:37] <tomp> tons of err msgs in your final code?
[03:23:48] <tomp> waitng to be released
[03:24:23] <tomp> vs c which just blew up
[03:24:31] <SWPadnos> yes, that kind of thing
[03:24:52] <SWPadnos> it was still compiled, so it would give you errors at compile-time
[03:25:03] <SWPadnos> but it was more strict, and there were no pointers
[03:25:17] <SWPadnos> (you could "pass by reference", but the compiler took care of the pointer stuff for you)
[03:27:25] <tomp> i thought pointers and structures were the endall of programming, nowadays i fighet with classes ( and typos :-)
[03:27:36] <tomp> fight
[03:28:23] <SWPadnos> pointers are great,m but they also cause all sorts of problems for beginners (and sometimes non-beginners)
[03:28:57] <SWPadnos> classes are excellent in a lot of places, but they're a real bear if you try to use them for other problems, where they're not really suited
[03:35:42] <tomp> more dep problems, lots of things want newer versions than our current base ( these meters want xmms >= 1.2.9) off to marillat!
[03:36:20] <SWPadnos> I'd get rid of everything other than the meter code, and see what's needed to get that into halvcp
[03:36:37] <SWPadnos> the xmms dependency is obviously not a reasonable one for EMC2
[03:37:44] <tomp> you're right, and i see it is already gtk
[03:38:06] <SWPadnos> I searched for "gtk meter widget" ...
[03:38:06] <tomp> i was just looking at stripping the main down
[03:38:10] <SWPadnos> ah
[03:38:27] <SWPadnos> now where did I put that Kapton tape?
[03:38:41] <SWPadnos> ah - on my desk, of course
[03:52:19] <tomp> http://macslow.thepimp.net/projects/cairo-clock/cairo-clock-8.png nice, gtk+ with cairo... could become meters
[03:53:41] <SWPadnos> hmmm - even better - 46V / 22A per coil :)
[03:53:53] <SWPadnos> very nice
[03:54:03] <SWPadnos> we should look at a gtk clock, I guess
[03:54:23] <tomp> how'd you measure current out? carbon pile?
[03:54:34] <SWPadnos> I looked at the spec sheet ;)
[03:54:55] <tomp> that trick never works ( r.squirrel)
[03:54:59] <SWPadnos> heh
[03:55:22] <SWPadnos> well, I don't need more than 60 or so volts, which would be ~150 IPM on my mill
[03:55:31] <SWPadnos> extra current reserve is always useful
[03:55:48] <tomp> yep, for reversals & loads
[03:55:58] <SWPadnos> 150 IPM for a Bridgeport that's on wheels would be a bit scary ;)
[03:56:25] <SWPadnos> but I guess it's the reversals that are scary for wheeled vehicles^Wmachines
[03:56:40] <tomp> mills-on-wheels ( like meals-on-wheels) :-)
[03:56:42] <SWPadnos> heh
[03:57:03] <SWPadnos> ok - I finally got something off my office floor. time to call it a night :)
[03:57:16] <SWPadnos> have fun with the gtk stuff
[03:57:17] <tomp> g'nite & thanks
[05:43:52] <A-L-P-H-A> SWPadnos, you play ut99?
[05:44:09] <SWPadnos> nope - pretty much just Wolf:ET these days
[05:44:23] <A-L-P-H-A> you should come play ut99. :)
[05:44:35] <SWPadnos> maybe one day. is there a Linux version?
[05:44:46] <A-L-P-H-A> yeah, there is. there's a wine version that runs decently enough.
[05:45:00] <SWPadnos> I can't run wine on my big machine
[05:45:07] <A-L-P-H-A> but if you want 1280x1024, you gotta run it in windows... I can't get it to run smooth past 800x600 under wine
[05:45:19] <SWPadnos> atleast, not without a lot of effort I'm not really willing to put in
[05:45:20] <A-L-P-H-A> there's linux version of ut2004
[05:45:38] <SWPadnos> that may work. it would be a shame to waste my 1920x1200 LCDs
[05:46:17] <SWPadnos> of course, wine on a dual dual-core opteron system with 4G RAM and a 7800GT video card may be able to go past 800x600 ;)
[05:46:36] <A-L-P-H-A> maybe.
[05:46:43] <A-L-P-H-A> damn.
[05:46:51] <A-L-P-H-A> my system is old now.
[05:46:59] <A-L-P-H-A> I haven't upgraded anything significant in a long time now
[05:47:16] <SWPadnos> then again, running wine at all on an x86_64 box is not trivial
[05:47:29] <A-L-P-H-A> no, not in 64bit mode
[05:47:34] <A-L-P-H-A> err.
[05:47:38] <A-L-P-H-A> double neg. I hate doing that
[05:47:56] <SWPadnos> hard with 64-bit mode - I've noticed that
[05:50:18] <A-L-P-H-A> yeah... I've read.
[05:50:33] <A-L-P-H-A> I haven't turned on my cnc mill now, for a good 4wks again.
[05:50:39] <A-L-P-H-A> I should probably sell it... if I don't use it.
[05:50:51] <SWPadnos> I haven't made my mill a CNC now, for a good 24 months :(
[05:51:03] <A-L-P-H-A> SWPadnos, say again.
[05:51:18] <SWPadnos> it's been 2 years, and I still haven't done the conversion
[05:51:27] <ejholmgren> what do joint following errors mean?
[05:51:43] <SWPadnos> I finally modified the center tapped transformer I've had for the past year, making it a dual secondary
[05:51:47] <A-L-P-H-A> probably the nodes don't touch.
[05:51:59] <SWPadnos> ejholmgren, servo or stepper system?
[05:52:13] <ejholmgren> not enough ceiling between the stepgen and axis mac ccel and vel?
[05:52:18] <ejholmgren> accel
[05:52:19] <ejholmgren> stepper
[05:52:20] <SWPadnos> stepper them
[05:52:22] <SWPadnos> ok
[05:52:26] <SWPadnos> s/them/then/
[05:52:46] <SWPadnos> yes, it means that stepgen is limited to too low a speed to do what the TP is requesting
[05:52:53] <SWPadnos> that limit can be from any of several places:
[05:53:16] <SWPadnos> 1) stepgen_maxaccel and / or stepgen_maxvel don't have sufficient headroom
[05:53:37] <ejholmgren> doesn't seem to be that
[05:53:52] <ejholmgren> I left them at the default values
[05:53:56] <SWPadnos> 2) the BASE_PERIOD * (steplen + stepspace) takes too much time, and stepgen isn't capable of getting up to the stepgen_maxvel anyway
[05:54:10] <SWPadnos> default has no bearing - you changed the INPUT_SCALE, right?
[05:54:19] <ejholmgren> yes
[05:54:32] <ejholmgren> it's at 20000 right now
[05:54:34] <SWPadnos> what is INPUT_SCALE, and what is the TP MAXVEL?
[05:54:37] <SWPadnos> ok
[05:54:44] <SWPadnos> and maxvel?
[05:55:11] <ejholmgren> defult for emc2 in the stepper_inch
[05:55:25] <tomp> g'nite all
[05:55:41] <SWPadnos> one sec- lemme check the default value
[05:55:51] <ejholmgren> the machine is in the basement atm
[05:55:59] <ejholmgren> I need a wireless card down there
[05:56:20] <ejholmgren> or pop a hole in the nice hardwood floor in this rental house ;)
[05:56:26] <SWPadnos> ok - MAX_VELOCITY is 1.2 IPS or 72 IPM
[05:56:32] <ejholmgren> yes
[05:56:39] <SWPadnos> so - let's look at it, shall we? :)
[05:56:59] <SWPadnos> 1.2 inches /sec * 20000 steps/inch = 24000 steps/sec
[05:57:16] <SWPadnos> steplen and stepspace default to 1, so you need 2 BASE_PERIODs to make one step
[05:57:27] <ejholmgren> ok
[05:57:31] <SWPadnos> each BASE_PERIOD is 50 uS, so you need 100 uS to generate a step
[05:57:38] <SWPadnos> that gives 10000 steps/sec max
[05:57:41] <SWPadnos> you need 24000
[05:57:42] <SWPadnos> oops
[05:57:45] <SWPadnos> :)
[05:58:00] <SWPadnos> what is the CPU on that machine?
[05:58:04] <SWPadnos> (speed)
[05:58:06] <ejholmgren> don't ask
[05:58:15] <ejholmgren> 333mhz
[05:58:16] <SWPadnos> I run a celeron 500 - you can tell me :)
[05:58:23] <ejholmgren> PII
[05:58:26] <SWPadnos> ok - that shouldn't be a big problem
[05:58:46] <SWPadnos> try setting the BASE_PERIOD down to 25000 - that should still run, but the UI may get sluggish
[05:59:01] <ejholmgren> ok ... brb
[05:59:04] <SWPadnos> if the machine locks up, then try 30000
[05:59:12] <SWPadnos> (locks up when you run emc, that is)
[06:00:04] <SWPadnos> if you can get it down to 20000 without it getting unbearably slow or locking up, then the problem is solved (that gives a 25000 steps/sec rate)
[06:01:05] <SWPadnos> if you can't, then you need to reduce the TP MAX_VELOCITY and the STEPGEN_MAXVEL and MAX_VELOCITY in each AXIS section
[06:05:20] <ejholmgren> base period of 25000 isn't too slow in the ui
[06:05:33] <SWPadnos> ok - that's good. did you notice a slowdown?
[06:05:39] <ejholmgren> I can jog each axis @ ~60 IPM
[06:05:49] <ejholmgren> any more than that and I get errors again
[06:05:57] <SWPadnos> yes - 25000 means 20000 steps/second max, which is 60 IPM
[06:06:21] <SWPadnos> can you reduce the step count per inch (by changing microstepping settings)?
[06:06:29] <ejholmgren> I'll try base period of 20000 and see if she'll hold
[06:06:46] <SWPadnos> that would be the best solution, if it works ;)
[06:07:13] <ejholmgren> my drives have full steps / half steps / 1/10th steps / 1/125th steps
[06:07:25] <ejholmgren> full stepping sounds painfull
[06:07:32] <ejholmgren> hald is only slightly better
[06:07:35] <ejholmgren> half
[06:07:36] <SWPadnos> you're at 1/10 now
[06:07:38] <SWPadnos> ?
[06:07:40] <ejholmgren> yes
[06:07:44] <SWPadnos> ok
[06:08:01] <SWPadnos> you may be a good candidate for one of the pluto boards jepler has been working on
[06:08:19] <SWPadnos> with step generation FPGA code instead of PWM
[06:08:51] <ejholmgren> bbiab
[06:08:54] <SWPadnos> ok
[06:08:59] <ejholmgren> down into the cold MN basement
[06:09:17] <SWPadnos> I'm in the pseudo-heated VT basement right now
[06:09:28] <ejholmgren> at least that's where the bar is :p
[06:19:03] <ejholmgren> sweet
[06:19:17] <ejholmgren> 20000 and no crash
[06:19:30] <SWPadnos> cool. that should be good them
[06:19:50] <ejholmgren> and I can actually run the spiral @ 20000 scale
[06:20:15] <SWPadnos> if you think things are too unresponsive in GUI-land, there is a simple hardware solution that doubles the number of steps/sec, so you would be able to increase BASE_PERIOD again
[06:20:54] <ejholmgren> does it involve an axe?
[06:21:02] <SWPadnos> hmmm. no
[06:21:12] <SWPadnos> it could include a saw and/or drill though
[06:21:18] <SWPadnos> (for PCB material)
[06:21:50] <SWPadnos> you can make a quadrature to step/direction converter out of two XOR gates and a few discrete components
[06:22:07] <ejholmgren> like a step multiplier?
[06:22:15] <SWPadnos> if you do that, you can use the quadrature output mode of stepgen, which can output a step every BASE_PERIOD
[06:22:23] <ejholmgren> ah
[06:22:40] <SWPadnos> no, it outputs one step per transition, but only one transition is needed instead of two
[06:23:20] <SWPadnos> like I said - very simple, probably $10 or less in parts, and unnecessary unless you want more speed or the GUI seems to get sluggish
[06:23:40] <ejholmgren> do you have a schematic?
[06:23:55] <SWPadnos> no, but there's one on the Geckodrive yahoo group
[06:24:03] <ejholmgren> $10 per axis ... or for all 3?
[06:24:13] <SWPadnos> all 3, I thikn
[06:24:13] <SWPadnos> think
[06:24:40] <SWPadnos> actually, a microcontroller could do it for $2
[06:25:08] <ejholmgren> but then I'd need to buy the programmer
[06:25:09] <SWPadnos> plus the board
[06:25:15] <SWPadnos> true
[06:25:18] <SWPadnos> those are $30
[06:25:18] <A-L-P-H-A> SWPadnos, what are you up to now these days?
[06:25:25] <ejholmgren> which I suppose I'll need eventually anyways
[06:25:26] <SWPadnos> nothing at all, unfortunately
[06:25:51] <A-L-P-H-A> the melatonin I took is making me really drowsy. ciao. I'm going to sleep.
[06:25:51] <SWPadnos> plus our cat is dying
[06:25:55] <A-L-P-H-A> :(
[06:25:58] <A-L-P-H-A> that's not cool at all.
[06:25:59] <SWPadnos> me too - it's late enough
[06:26:07] <SWPadnos> no - not cool in the slightest
[06:26:12] <SWPadnos> see you later
[06:26:26] <ejholmgren> night both of you
[06:26:28] <A-L-P-H-A> yeah. c ya.
[06:26:34] <SWPadnos> ejholmgren, I'll see if I can find a schematic for you tomorrow, or some microcontroller code
[06:27:13] <ejholmgren> thanks
[06:27:26] <ejholmgren> does that list require membership?
[06:27:31] <SWPadnos> yes
[06:27:39] <SWPadnos> if you have a yahoo signup, then it's easy
[06:38:06] <ejholmgren> geez ... yahoo's captchas must be effective
[06:38:15] <ejholmgren> I can barely decipher them
[06:39:32] <SWPadnos> I get very little junk main from that address - it's all from yahoo themselves
[06:43:27] <ejholmgren> I'm in
[06:43:43] <ejholmgren> have any idea what the title of the post was?
[06:43:55] <SWPadnos> it's in the files section
[06:43:58] <SWPadnos> one sec
[06:46:41] <SWPadnos> ok - I guess I was wrong - it's in the files section of the CAD_CAM_EDM_DRO group ;)
[06:49:10] <SWPadnos> CCED group/files/CIRCUITS/Pulse Gen And Timing/Quad2StepDir.pdf
[06:56:06] <ejholmgren> founf it
[06:56:09] <ejholmgren> found
[06:56:12] <ejholmgren> thanks
[07:02:14] <ejholmgren> 'night, off to bed
[08:39:39] <alex_joni> morning Jymmmm
[09:10:24] <Jymmmm> Hi alex_joni
[09:11:06] <Jymmmm> alex_joni you do Blender type 3d dont you?
[09:47:05] <alex_joni> Jymmmm: not really
[09:47:28] <Jymmmm> ah ok.
[10:55:54] <Guest540> please could explain me as to make the machine home?
[11:11:02] <anonimasu> Guest540: you mean setting up your zero?
[11:11:14] <anonimasu> or do you mean mechanical homing(with sensors and stuff?=
[11:15:16] <Guest540> with sensors and stuff
[11:19:30] <anonimasu> I cant help you with that sorry
[11:20:06] <Guest540> Who can help me?
[11:25:14] <anonimasu> well, you can always begin looking at the wiki
[11:25:19] <anonimasu> and wait for somone that knows it to show up
[11:26:29] <Guest540> in the wiki no information
[11:27:06] <Guest540> I am there!
[11:29:03] <anonimasu> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Home_Switches_With_EMC
[11:29:32] <anonimasu> atleast it's something..
[11:29:53] <anonimasu> you should wait for somone that knows to show up :)
[11:31:48] <Guest540> Who is Somone?
[11:32:40] <anonimasu> some person that knows.
[11:33:01] <Guest540> ok
[11:33:29] <Guest540> it goes to be here today?
[11:34:56] <anonimasu> probably
[11:41:02] <alex_joni> homing is described in the User Manual
[11:41:05] <alex_joni> http://www.linuxcnc.org/docs/devel/html/config/ini_homing/index.html
[11:41:37] <anonimasu> hey alex
[11:41:46] <alex_joni> Guest540: how do you drive your machine? stepper or servo?
[11:41:49] <alex_joni> hey anonimasu
[11:43:07] <Guest540> my machine is stepper
[11:43:15] <alex_joni> ok, then you don't have index
[11:43:38] <alex_joni> USE_INDEX = false
[11:43:46] <alex_joni> HOME_SEARCH_VEL = 10
[11:43:54] <alex_joni> HOME_LATCH_VEL = 2
[11:44:18] <Guest540> as to place the sensors
[11:44:33] <alex_joni> that will begin to move the axis in POSITIVE direction until it finds the home switch
[11:44:54] <alex_joni> once it's found it goes back (NEGATIVE) direction and searches again with low speed (2 mm/sec)
[11:45:11] <anonimasu> alex_joni: short question, when you have home switches how do you set up parts?
[11:45:30] <anonimasu> do you zero at your _home_ and add a offset for your fixture?
[11:45:39] <alex_joni> usually like that
[11:45:51] <anonimasu> how do you do that in axis?
[11:45:56] <alex_joni> touch-off
[11:46:04] <alex_joni> there's a button for that
[11:46:41] <anonimasu> I know, but is that for machine/other offsets?
[11:46:52] <anonimasu> err work/machine or what you would call them
[11:46:55] <alex_joni> Set the proper offsets for each axis by jogging and using the ``Touch Off'' button.
[11:47:17] <alex_joni> no, it's for a part
[11:47:22] <anonimasu> yeah, but with home switches how does that change?
[11:47:22] <alex_joni> http://www.linuxcnc.org/docs/devel/html/gui/axis/index.html
[11:47:28] <alex_joni> anonimasu: not at all
[11:47:38] <alex_joni> unless it's a lathe
[11:47:43] <anonimasu> ok, you just have your machine always zero?
[11:47:47] <alex_joni> right
[11:47:49] <Guest540> necessary to place the sensors in standart_pinoult?
[11:47:54] <alex_joni> Guest540: yes
[11:47:55] <anonimasu> so you home it again, and start a part and it goes to the standard place..
[11:48:03] <anonimasu> without setting up your fixture/zero again..
[11:48:41] <Guest540> how to place the sensors in standart_pinoult ?
[11:49:40] <anonimasu> alex_joni: also, does emc save the offsets when you close it?
[11:55:24] <anonimasu> ah well.. nm
[11:58:48] <Guest540> necessary to place the sensors in standart_pinoult?
[12:09:53] <Guest540> necessary to place the sensors in standart_pinoult?
[12:09:54] <Guest540> necessary to place the sensors in standart_pinoult?
[12:09:55] <Guest540> necessary to place the sensors in standart_pinoult?
[12:09:55] <Guest540> necessary to place the sensors in standart_pinoult?
[12:09:57] <Guest540> necessary to place the sensors in standart_pinoult?
[12:09:59] <Guest540> necessary to place the sensors in standart_pinoult?
[12:10:01] <Guest540> necessary to place the sensors in standart_pinoult?
[12:11:24] <anonimasu> Guest540: STOP REPEATING!
[12:12:19] <Guest540> please i need for informantion
[12:12:21] <anonimasu> Guest540: no one active knows, it's not going to help you if you spam the channel with your question 7 times, just wait until somone shows up.
[12:13:53] <Guest540> ok sorry -me
[12:13:57] <Guest540> but
[12:14:13] <anonimasu> Guest540: no one active knows, it's not going to help you if you spam the channel with your question 7 times, just wait until somone shows up.
[12:14:16] <anonimasu> whoops..
[12:14:19] <anonimasu> sorry :D
[12:14:41] <Guest540> ok ok
[12:14:48] <Guest540> Good bless you
[12:15:30] <Guest540> God bless you
[12:19:20] <Guest744> heloo
[12:19:27] <Guest744> sory form my stress
[12:20:21] <Guest744> I only want to know as to place the sensors in the standart_pinoult.hal
[12:29:55] <alex_joni> Guest744: there are other sample configs for emc2, some of them already have limit switches in them
[12:31:36] <alex_joni> newsig Xhome bit
[12:31:49] <alex_joni> linksp Xhome <= parport.1.xxxx
[12:31:54] <alex_joni> linksp Xhome => axis.0.home-sw-in
[12:31:57] <alex_joni> the same for Y and Z
[12:32:03] <Guest744> very tnhanks
[12:32:30] <Guest744> I obtained the information
[12:32:45] <Guest744> in the folder demo_mazak
[12:33:50] <alex_joni> good
[12:33:57] <alex_joni> that means you are starting to understand these things :)
[12:44:38] <Guest744> I obtained!
[12:45:37] <alex_joni> nice :)
[13:02:11] <Guest744> Very thanks
[13:40:05] <Guest744> please 1 more information
[13:40:10] <Guest744> Its possible?
[13:41:44] <Guest744> please one information ..
[13:41:46] <Guest744> my engine makes 5 revolves for 1 milimeter
[13:41:53] <alex_joni> ok
[13:42:00] <alex_joni> how many pulses / revolve?
[13:42:02] <Guest744> its correct?
[13:42:03] <alex_joni> 200?
[13:42:14] <alex_joni> Guest744: depends on your motors / screws
[13:42:31] <Guest744> 200
[13:43:37] <Guest744> 200 pulses for 1 revolve
[13:43:48] <alex_joni> ok, and 5 revolves for 1 mm
[13:43:58] <alex_joni> so you need 1000 pulses for 1 mm
[13:44:02] <alex_joni> INPUT_SCALE = 1000
[13:45:01] <Guest744> Which the acceleration?
[13:46:11] <Guest744> Which MAX_VELOCITY and
[13:46:28] <Guest744> Which MAX_ACELARATION
[13:46:47] <Guest744> in the my ini file is:
[13:47:06] <Guest744> MAX_VELOCITY: 1.2
[13:47:20] <Guest744> MAX_ACELARATION 20.0
[13:47:30] <alex_joni> that is mm / sec
[13:47:34] <alex_joni> 1.2 mm / sec is very slow
[13:47:57] <alex_joni> you can increase max_velocity up to a certain point
[13:48:06] <alex_joni> if it's too high, you'll get following errors
[13:51:36] <Guest744> if increase appears menssagem of error
[13:52:44] <Guest744> Which the speed that I can place?
[13:52:52] <alex_joni> Guest744: what is your BASE_PERIOD ?
[13:53:07] <Guest744> one moment..
[13:53:41] <Guest744> its 50000
[13:53:54] <alex_joni> ok, that means one cycle takes 50 microseconds
[13:54:19] <Guest744> its correct?
[13:55:01] <alex_joni> if you have BASE_PERIOD at 50 microseconds, you can output max 10 kHz pulses
[13:55:18] <alex_joni> with 1000 pulses / mm it means you can go at fastest with 10 mm / sec
[14:03:30] <Guest744> I have 40 puses/mm, how make this?
[14:03:39] <alex_joni> 40 ?
[14:03:57] <alex_joni> you mean one turn of the screw is 5 mm ?
[14:04:12] <Guest744> yes...
[14:04:19] <alex_joni> INPUT_SCALE = 40
[14:05:09] <mdynac> i have an adaptive feedrate question.....
[14:06:50] <Guest744> I I go to make the tests, grateful for the time being.
[14:07:27] <mdynac> linkpp motenc.0.adc-00-value motion.adaptive-feed <-- is this the only line i must add to get it enabled?
[14:08:00] <jepler> does adc-00-value go from 0 (to indicate zero feed) to 1 (to indicate 100% feed)?
[14:08:13] <jepler> I'm not that familiar with adaptive feed, but I think that's the only HAL pin you have to hook up
[14:08:33] <jepler> there's also the G-code you need to issue to make emc respect the adaptive feed value
[14:08:59] <mdynac> i have no idea what the value part is, i am inputting 0 to 5 vdc to the adc-00 pin
[14:09:24] <mdynac> yes the M50P1 correct?
[14:09:55] <mdynac> it wseems to have no effect, the machine moves as normal even with M50P1 in startup code
[14:10:20] <mdynac> i was wondering if my adc hal stuff is incorrect...
[14:10:27] <jepler> the documentation I have says that M52 is the adaptive feed control
[14:10:43] <mdynac> sorry i meant M52P1
[14:10:50] <jepler> M50 is the "feed override" control that is traditionally in the GUIs
[14:11:14] <mdynac> i am doing an M52P1
[14:11:25] <mdynac> in my startup code
[14:12:09] <mdynac> my problem is that i have 0vdc applied to the adc and the machine still moves......
[14:12:35] <jepler> use 'halcmd show' to find out if the adc-00-value is really 0.0, or if it's some other value.
[14:14:24] <jepler> adaptive feed works for me -- I issued 'g52p1' in MDI, then in a halcmd I issued 'setp motion.adaptive-feed 0'. A subsequent 'g1x1f1' didn't move
[14:14:39] <jepler> but when I issued 'setp motion.adaptive-feed .1' to halcmd, the move proceeded slowly
[14:17:04] <mdynac> okay i'll try it...
[14:17:28] <mdynac> so i give it a G52P1 thought it was M52P1
[14:17:53] <jepler> er, maybe I typed the wrong thing into this window
[14:17:56] <mdynac> k
[14:18:19] <jepler> g52p1 gives an error
[14:18:35] <jepler> m52p1 makes "m52" appear among the active modal codes
[14:19:39] <mdynac> k i will go test it right now....brb
[14:28:36] <mdynac> if i set the setp to 0 it still moves and i get an error pin motion.adaptive-feed is connected to a signal
[14:29:23] <jepler> the setp worked for me because I hadn't connected anything else to it
[14:29:35] <jepler> you'd have to 'unlinkp motion.adaptive-feed' to do setp as I did
[14:36:11] <mdynac1> sorry but i keep getting disconnected and then cannot find server....
[14:36:45] <mdynac1> xchat on xp at work.....i have no idea why....
[14:36:51] <jepler> the setp worked for me because I hadn't connected anything else to it
[14:36:53] <jepler> you'd have to 'unlinkp motion.adaptive-feed' to do setp as I did
[14:37:15] <mdynac1> okay
[14:37:21] <mdynac1> brb
[14:38:24] <jepler> did you try using halcmd/halmeter/halshow to see what the value on adc-00-value was?
[14:40:42] <mdynac1> the value is 0.007324
[14:41:09] <mdynac1> unlinkp then setp to 0 and machine still moves
[14:41:41] <jepler> does M52 appear among the active modal codes?
[14:41:46] <jepler> (in AXIS they are shown on the MDI tab)
[14:41:46] <mdynac1> nope
[14:41:55] <jepler> Issue M52 in MDI, then try again
[14:41:57] <jepler> er, M52P1
[14:42:01] <mdynac1> usung tkemc
[14:42:07] <mdynac1> nope
[14:43:33] <mdynac1> i got an M5 and an M53 F1 S0, and the usual G's
[14:43:50] <alex_joni> mdynac1: use M52P1 from MDI
[14:43:58] <alex_joni> the codes at startup might be reset when you load a program
[14:44:01] <mdynac1> okay...
[14:44:21] <jepler> After I issue M52P1 from MDI, I get M52 in the list of active codes in both axis and tkemc
[14:51:08] <mdynac1> okay the MDI code worked like a chrm
[14:51:21] <mdynac1> and the M52 popped up in tkemc\
[14:51:46] <alex_joni> mdynac1: now it should work properly
[14:52:21] <mdynac1> whats up with the RS274NGC_STARTUP_CODE=M52P1 ?
[14:54:37] <jepler> sounds like it doesn't work
[14:55:19] <mdynac1> okay i'll just fire it up on the MDI line or just put it at the top of my g-code...
[14:55:48] <alex_joni> mdynac1: the RS274NGC_STARTUP_CODE gets executed once at startup
[14:56:13] <alex_joni> however, when you load a file, a different reset happens internally, and the RS274NGC_STARTUP_CODE isn't interpreted again
[14:56:21] <alex_joni> I'm not sure if that's a bug or a feature
[14:56:28] <mdynac1> i'll just put it at the top of my g-code
[14:56:37] <jepler> that sounds like a good thing to do
[14:56:41] <alex_joni> mdynac1: that's the sanest thing to do
[14:57:58] <mdynac1> okay i'm going to hook up a voltage to my adc and execute some g-code and see what happens...
[15:10:26] <mdynac> M52P1 at the top of my gcode gives me cannot do a G1 with 0 feedrate.....
[15:10:48] <mdynac> we are getting close.....
[15:12:47] <mdynac> this xchat on xp is awful....keeps disconnecting me......
[15:14:05] <alex_joni> maybe you can try the one at http://www.linuxcnc.org/index.php?option=com_mospjirc&Itemid=8&lang=en
[15:15:46] <Guest369> okay the java version is up...
[15:16:10] <Guest369> so i get this cannot do a g1 with 0 feedrate......
[15:20:07] <Guest369> anyone?
[15:20:57] <Lerneaen_Hydra> 'lo
[15:21:46] <alex_joni> Guest369: maybe put a F10 or something at the top too
[15:21:53] <skunkworks> why would you not have an initial feed rate (assume the adaptive feed will slow it down)
[15:21:56] <alex_joni> Guest369: I'll investigate later
[15:22:21] <alex_joni> bbl
[15:23:59] <Lerneaen_Hydra> what's happening?
[15:30:30] <Guest369> k
[15:31:11] <skunkworks> Guest369: I have had issues on windows machines disconnecting from irc. I have pretty much fixed it by also having the ubuntu channel running also.. The higher useage of ubuntu seems to keep me connected
[15:31:29] <skunkworks> odd I know
[15:31:53] <skunkworks> (just here at work - at home I don't have the issue)
[15:36:54] <rayh> I've run irc to keep me connected at home during downloads.
[16:12:16] <Guest369> gentleman.....we are burning parts under emc control!!!!!!!! edm is working....
[16:12:28] <cradek> yay!!
[16:12:42] <alex_joni> Guest369: yay
[16:13:25] <skunkworks> Nice!
[16:20:57] <Guest369> the f1 in the gcode bruoght it up....
[16:21:41] <Guest369> i will post photos of the part here when it is finished....where can i post the pics??? i have not set up a webspace anywhere
[16:21:49] <alex_joni> did you have some G1 at the beginning without F?
[16:22:00] <Guest369> hang on
[16:22:07] <alex_joni> Guest369: there is an upload limit of about 300k on wiki.linuxcnc.org
[16:22:10] <cradek> that makes F1 the maximum speed possible (adaptive only slows down)
[16:22:30] <alex_joni> alternatively you can mail stuff to me, I'll put it somewhere..
[16:23:35] <Guest369> G1 F1 then just plain g's after the first line
[16:24:00] <Guest369> already burned a g2 now back to g1 and moving on to a g3 next line
[16:24:02] <cradek> Guest369: hope your part has an arc on it so you can show off to TomP
[16:24:08] <cradek> cool :-)
[16:24:17] <Guest369> it's a maple leaf.....
[16:24:25] <Guest369> lot's o radii
[16:24:51] <skunkworks> so is it slowing down as it should?
[16:25:13] <cradek> you could plot the actual velocity with halscope - that would be neat to see
[16:25:33] <cradek> also if you use AXIS it will show you the actual resulting velocity
[16:26:38] <skunkworks> if he is using head - it is on by default also.
[16:26:43] <skunkworks> (in axis)
[16:27:12] <cradek> yeah he is using head
[16:27:19] <jepler> The velocity that AXIS shows is estimated, and is not terribly accurate
[16:27:45] <cradek> ok ddt and halscope is the best way then
[16:28:11] <cradek> distance-to-go also could be useful for EDM because it's so slow
[16:29:11] <jepler> the "core_sim.hal" file shows how to use ddt and hypot to compute the XY and XYZ velocity
[16:59:57] <Guest369> edm still burning....althogh the feedrate is pretty steady.....i'm not going to push it till the next part i burn, i just want to prove out the software......
[17:00:26] <Guest369> cutting 1/2" aluminum with .010 brass wire
[17:02:04] <Lerneaen_Hydra> what feedrate is that, roughly?
[17:02:17] <Lerneaen_Hydra> 1mm/min? 10mm/min? 100mm/min?
[17:03:59] <Lerneaen_Hydra> hi chinamill
[17:04:12] <Lerneaen_Hydra> btw, why are you in france?
[17:04:48] <skunkworks> ?
[17:05:45] <Lerneaen_Hydra> do /whois ^Eugenics ;)
[17:05:49] <Lerneaen_Hydra> french ISP or something
[17:05:58] <Lerneaen_Hydra> my client does that whenever someone enters
[17:07:28] <^Eugenics> Greetings Lerneaen_Hydra , yes I'm based in Nice
[17:07:53] <Lerneaen_Hydra> wtf? weren't you in sweden?
[17:07:56] <^Eugenics> My parents live here
[17:07:57] <Lerneaen_Hydra> ah
[17:07:58] <skunkworks> you guys need to quit changing your names :)
[17:08:00] <Lerneaen_Hydra> ok
[17:08:05] <^Eugenics> :)
[17:08:10] <Lerneaen_Hydra> afaik ^Eugenics = chinamill
[17:08:16] <Lerneaen_Hydra> I keep my name :p
[17:08:29] <Lerneaen_Hydra> unless I've started mixing names, again
[17:08:33] <^Eugenics> Sweden = cold :)
[17:08:58] <Lerneaen_Hydra> not this winter
[17:09:02] <simon78> Not anymore. This autumn has been the hottest for centuries..
[17:09:12] <Lerneaen_Hydra> it's still +8° outside here
[17:09:25] <Lerneaen_Hydra> * Lerneaen_Hydra hopes there will be snow this year
[17:09:28] <simon78> Its about the same as a swedish summer.
[17:09:28] <^Eugenics> here is cold to 16 deg
[17:09:38] <Lerneaen_Hydra> haha, yea
[17:09:40] <^Eugenics> :)
[17:09:39] <Lerneaen_Hydra> *yeah
[17:09:40] <jepler> -11C here
[17:09:48] <Lerneaen_Hydra> humid or dry?
[17:09:55] <jepler> dry
[17:10:07] <Lerneaen_Hydra> nice :)
[17:10:08] <^Eugenics> jepler: sounds like a nice winter
[17:10:08] <jepler> clear skies, light winds
[17:10:13] <Lerneaen_Hydra> perfect
[17:10:18] <jepler> the snow will come later
[17:10:29] <Lerneaen_Hydra> here I've got really gusty winds up to 10m/s and rain
[17:10:31] <Lerneaen_Hydra> lots of rain
[17:10:35] <Lerneaen_Hydra> buckets of rain
[17:10:41] <Lerneaen_Hydra> did I mention it's raining?
[17:10:47] <Lerneaen_Hydra> you get the idea ;)
[17:11:06] <^Eugenics> Göööteborg!
[17:11:22] <Lerneaen_Hydra> de ê la gött
[17:11:22] <jepler> winds of 20m/s (40mph) are not uncommon here in winter
[17:11:28] <Lerneaen_Hydra> oh, nasty
[17:11:39] <Lerneaen_Hydra> ;)
[17:13:40] <jepler> in the last 24 hours, the strongest sustained winds were 24MPH and the gusts were 32MPH (10m/s and 14m/s)
[17:14:26] <skunkworks> sounds like here - the news had to educate people about 'wind chill' again
[17:16:17] <cradek> the sun is out so it's up to 9F now :-/
[17:18:24] <skunkworks> 11 here
[17:19:43] <skunkworks> http://news.yahoo.com/s/nm/20061207/wl_nm/britain_tornado_dc
[17:22:24] <SWPadnos> oh great - I'm about to go to England, and they start having tornadoes
[17:22:26] <SWPadnos> that figures
[17:22:57] <skunkworks> I thouht that was only a midwest thing :)
[17:23:56] <SWPadnos> I thought it was only trailer parks
[17:23:57] <SWPadnos> (rimshot)
[19:23:08] <cradek> I wonder where mac went...
[19:23:27] <cradek> I'm anxious to see the first edm part
[19:27:49] <cradek> hi mac
[19:27:52] <cradek> how's the leaf?
[19:27:52] <mdynac> hi
[19:28:16] <mdynac> got about 3 inches per hour cut rate....not too bad for the first try..
[19:28:33] <mdynac> i should get about 10 iph
[19:28:43] <cradek> haha iph
[19:28:44] <mdynac> with aluminum
[19:28:53] <mdynac> yeah i know.....
[19:28:56] <cradek> what kinds of things do you adjust to improve it?
[19:28:58] <mdynac> we be slowpokes
[19:29:44] <mdynac> well i can increase the servo speed, jack up the flushing and increase the cutting current and on time....
[19:30:10] <Lerneaen_Hydra_> 10 IPH! :|
[19:30:10] <anonimasu> hm, what does commercial machines do?
[19:30:11] <mdynac> i was just easing it thru today
[19:30:23] <cradek> did your part finish?
[19:30:31] <Lerneaen_Hydra_> 0.1mm/s
[19:30:34] <anonimasu> how's the finish on your parts?
[19:30:57] <jepler> * jepler would love to see a picture
[19:30:59] <mdynac> i have one other problem to overcome before i move on.....i need to stop motion when the wire breaks....where in the code would be a good place to do such a thing?
[19:31:17] <anonimasu> hm, build a wire feeder?
[19:31:18] <mdynac> pictures are coming in about an hour or so when i get home......
[19:31:29] <anonimasu> nice
[19:31:35] <anonimasu> how did the finish end up?
[19:31:50] <mdynac> finish = excellent....however it is aluminum.....
[19:31:54] <cradek> mdynac: you could do estop, abort, or feedhold
[19:32:26] <cradek> kind of depends what you want to happen after you fix it
[19:32:45] <mdynac> i take it any of those three would stop the machine immeadiately....
[19:32:51] <cradek> yes
[19:33:01] <anonimasu> * anonimasu nods
[19:33:15] <anonimasu> * anonimasu is really interested in EDM:ing
[19:33:15] <SWPadnos> but feedhold is the only one that'e easily recoverable later
[19:33:19] <mdynac> okay, wire breakage is normal....i just re-thread the machine and resume the burn....
[19:33:22] <anonimasu> especially if you can machine gears in tool steel..
[19:33:34] <anonimasu> though might be overkill for a home setup
[19:33:49] <mdynac> dude this aint no home setup....
[19:33:52] <SWPadnos> you should be able to machine gears in titanium
[19:33:51] <Lerneaen_Hydra_> how's milling in tool steel?
[19:33:52] <cradek> mdynac: start at exactly the same spot?
[19:33:56] <anonimasu> if a gear takes 5 days it'd still be ok :)
[19:34:01] <mdynac> yes
[19:34:02] <anonimasu> mdynac: what is it then?
[19:34:22] <mdynac> a charmilles EF330 wire edm now with emc2 control
[19:34:31] <cradek> mdynac: maybe you just want pause or feedhold then
[19:34:51] <mdynac> i take it i gotta whip up some hal pins.....
[19:35:01] <cradek> using halui pause would just let you hit resume in the gui after you fix it
[19:35:19] <cradek> sure, it's always more hal pins :-)
[19:35:20] <mdynac> kewl that is exactly what i need...
[19:35:37] <SWPadnos> but it's not RT (which shouldn't be an issue at IPH speeds ;) )
[19:35:39] <cradek> I assume you've got a "wire broke" digital input?
[19:35:41] <mdynac> i'm starting to see a hal pattern here....
[19:35:57] <mdynac> yes
[19:36:00] <mdynac> on the old system
[19:36:05] <cradek> mdynac: yeah, it's how you wire between the brains and the hardware... pretty flexible isn't it
[19:36:15] <mdynac> quite
[19:36:38] <mdynac> i suppose i will get to be well versed in it very soon
[19:36:59] <cradek> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Halui
[19:37:04] <cradek> here's a little information
[19:37:12] <anonimasu> mdynac: well it's not as amazing that you are getting good finishes with a real power supply :)
[19:37:19] <jepler> http://linuxcnc.org/docs/html/gui/halui/index.html
[19:37:24] <cradek> instead of using halvcp (virtual control panel) you'll use your real digital inputs
[19:37:29] <anonimasu> * anonimasu saw some homebrew thing, but it look to amazing
[19:37:45] <mdynac> eventually what needs to happen when i start a program is that the wire turns on, then the flushing , then the cutting voltage, and then the machine starts to move.....
[19:38:11] <cradek> mdynac: you can coordinate all that with classicladder's timers etc
[19:38:23] <Lerneaen_Hydra_> what type of finish are we talking about? Ra 0.4µm?
[19:38:45] <mdynac> if the wire breaks the machine and all mentioned above goes off, then i rethread and start the process over wireon, water on, cutiingon, move
[19:39:13] <anonimasu> Lerneaen_Hydra_: the ones I saw were rough
[19:39:30] <mdynac> yes classic ladder seems to be what i should use.....
[19:39:40] <SWPadnos> mdynac, pause doesn't turn off those things by default (I don't think so anyway)
[19:39:40] <mdynac> never even looked at it yet tho.....
[19:40:05] <SWPadnos> you'll need some interesting ladder logic to get all the pieces right for pause/resume
[19:40:08] <mdynac> but if i ladder it up i can do it correct?
[19:40:18] <SWPadnos> should be able to
[19:40:38] <cradek> now that we have feedhold you can do about anything
[19:40:42] <SWPadnos> it's probably easier if you're not using M codes to control fluid / water as well
[19:41:06] <mdynac> actually most of the control of these items are external of the old pdp8 and i could just emulate the three signals in ladder
[19:41:23] <mdynac> too much typing to explain here....
[19:41:36] <mdynac> i hate typing...
[19:41:57] <skunkworks> how hard would it be to have a fastforward/_rewind_ sort of a feed over ride that goes through 0 into negative feed rates
[19:42:02] <skunkworks> :)
[19:42:16] <SWPadnos> cradek, in halui, does pause/resume result in an NML message or a HAL pin changing
[19:42:20] <SWPadnos> skunkworks, hard ;)
[19:42:21] <skunkworks> I could see edm people wanting that
[19:43:26] <SWPadnos> we discussed it at fest
[19:43:40] <mdynac> basically the emc2 M52 code works as well as an old Japax machine which did not "backup".....congrats guys.....
[19:44:04] <Lerneaen_Hydra_> anonimasu: oh, how rough?
[19:44:13] <anonimasu> Lerneaen_Hydra_: .... rough
[19:44:19] <Lerneaen_Hydra_> ok ;)
[19:44:32] <anonimasu> I think it was in the emc dropbox..
[19:49:06] <skunkworks> look at the email I just got http://www.electronicsam.com/emcemail.JPG
[19:50:09] <skunkworks> and no - I don't delete any of the emails I get.
[19:50:31] <mdynac> finish is about 1.1uM Ra
[19:51:23] <mdynac> howeer this is a wire edm.....
[19:52:11] <Lerneaen_Hydra_> skunkworks: wtf? does that product really exist?
[19:52:36] <Lerneaen_Hydra_> skunkworks: you have 77000 unread messages, better get cracking ;)
[19:52:39] <cradek> SWPadnos: all of halui just sends NML message
[19:52:42] <cradek> s
[19:53:11] <skunkworks> http://www.emcinsignia.com/promos/retrospect1?utm_id=130
[19:53:26] <mdynac> be back in about an hour, gotta close out here and go home to celebrate the working wire machine!
[19:53:34] <SWPadnos> ok - so pause/resume doesn't affect HAL per se, it just causes the TP to stop issuing moves
[19:54:13] <cradek> SWPadnos: yes, it's just like the rest of the guis
[19:54:28] <SWPadnos> ok - that's a problem for using CL to control water / pump / cooling ...
[19:54:36] <Lerneaen_Hydra_> at least the products are so different that there's no risk of them suing for trademark infringement (IANAL)
[19:54:46] <SWPadnos> there's no change in HAL to use as an input to CL
[19:55:12] <SWPadnos> and no way for CL to detect that the resume (or pause) has actually happened
[19:55:14] <cradek> not true
[19:55:20] <jepler> (BIT) halui.program.is-paused
[19:55:25] <SWPadnos> ok - good :)
[19:55:25] <cradek> you have the class of pins like halui.lube.is-on
[19:55:44] <SWPadnos> this isn't lube - it shouldn't be controlled with M<whatever>
[19:55:46] <cradek> just like you can also "see" the state with the other guis
[19:56:00] <cradek> oh it shouldn't?
[19:56:06] <cradek> I think I don't understand the problem
[19:56:44] <SWPadnos> as I understand it, it's like a motor enable - whenever you want to move, you have to turn things on in some sequence
[19:56:51] <cradek> oh you need to do some sequence at run and resume?
[19:56:55] <SWPadnos> when you stop, you want to turn those things off
[19:56:57] <SWPadnos> yes
[19:57:02] <cradek> ok
[19:57:14] <cradek> you do have that state information, and feedhold
[19:57:34] <cradek> so I think you could do whatever, you just have to be clever with your feedhold
[19:57:35] <SWPadnos> yes - feedhold may be the way to do it instead of pause/resume
[19:57:39] <SWPadnos> yes
[19:58:00] <cradek> not sure I'm that clever
[19:58:43] <SWPadnos> use a mux driven by a bit from CL, which selects between the "actual" feed value or a 0
[19:58:56] <jepler> so CL would turn on feedhold and juggle these other outputs in response to the error signal. you'd get some "reset" button in halvcp, which would do the "reverse" operations, finally clearing feedhold when it was OK to do so...
[19:59:01] <SWPadnos> after doing all its shenanigans with the machine
[19:59:09] <SWPadnos> yes
[19:59:34] <SWPadnos> oh - is feedhold a bit input, or is it done by setting feed_override to 0?
[19:59:42] <cradek> it's a bit
[19:59:47] <SWPadnos> cool then - even easier ;)
[19:59:58] <cradek> of course you could also set adaptive feed to 0 (if AF is enabled)
[20:00:09] <SWPadnos> which I think it is for mac
[20:00:11] <cradek> I think feedhold always works, there is no feedhold override
[20:04:55] <SWPadnos> yay - UK pounds have gone down by 1 cent
[20:15:17] <jepler> SWPadnos: planning travel? or just like to keep an eye on currency markets?
[20:16:58] <SWPadnos> planning travel
[20:17:08] <SWPadnos> the UK is frighteningly expensive right now
[20:18:35] <jepler> the exchange rate for euros is depressing too
[20:18:57] <cradek> if only we had bought some of those $1.10 euros
[20:19:01] <SWPadnos> I've seen that at $1.40+, so it's not too bad
[20:19:22] <SWPadnos> the UKP hasn't been this high since the Thatcher days though
[20:19:49] <jepler> http://www.codejedi.com/cgi-bin/blog.cgi/2006/12/02#20061202mssucks
[20:20:00] <jepler> not allowed to make "obscene images" in microsoft's drawing program?
[20:21:35] <cradek> Scarier is Microsoft Money, in whose EULA notes that the application makes unannounced connections to the net to in order to send details of transactions and bank account names to Microsoft.
[20:21:57] <SWPadnos> it should say that MS isn't allowed to make obscene uses of your personal data
[20:22:03] <SWPadnos> but that'll never happen
[20:29:33] <Lerneaen_Hydra_> remember the clause in frontpage that prohibited creation of webpages critical of microsoft?
[20:29:59] <jepler> no but I'm not surprised
[20:30:09] <SWPadnos> Borland had a similar clause for C++ Builder at one point - "you may not use this software to make a competing product" ;)
[20:30:18] <SWPadnos> they cahnged it once some programmers noticed
[20:35:37] <skunkworks> http://dotnetmasters.com/HistoryOfBasic.htm
[20:37:34] <jepler> Visual Basic developers invent a new way of developing applications that completely bypasses requirements gathering, based on the principle that "if you don't care where you're going, you don't need a map". They listen to a user for an hour, then do a bunch of drag and drop screens, then see if it's what the user wants, and then do the same cycle over and over again until either (1) the user finally says "It's not what I want, but I'm sick of working wit
[20:37:41] <jepler> haha
[20:37:43] <Lerneaen_Hydra_> haha!
[20:37:45] <Lerneaen_Hydra_> nice page :D
[20:37:57] <Lerneaen_Hydra_> (too bad basic is all that I can code ;) )
[20:38:01] <Lerneaen_Hydra_> atm at least
[20:38:10] <Lerneaen_Hydra_> Lerneaen_Hydra_ is now known as Lerneaen_Hydra
[20:38:44] <jepler> "A spelling error results in the command to launch the interpreter becoming "BASICA"."
[20:38:50] <jepler> I'd like to see another source for this claim
[20:39:09] <cradek> yeah, boguuuus
[20:39:23] <skunkworks> I think it is supposed to be funny
[20:39:58] <cradek> pc dos had "basic" and "basica" and I don't remember what the difference was
[20:40:20] <SWPadnos> and gwbasic, eventually
[20:40:44] <jepler> Microsoft BASICA (short for "Advanced BASIC") is a simple disk-based BASIC interpreter written by Microsoft for PC-DOS. BASICA allows use of the ROM-resident BASIC included with early models of IBM's PC while DOS is loaded (the ROM BASIC itself runs when nothing is loaded when booting) and adds functionality such as file access and storage of programs on disk. It does not run on non-IBM PCs and even on later IBM models, since those lack the requisite ROM
[20:40:51] <jepler> </wikipedia>
[20:41:15] <cradek> ok, basica invoked the rom basic
[20:41:27] <jepler> * jepler has vague memories of trying to learn GWBASIC, since after all he'd been pretty good at BASIC on his Commodore.
[20:41:49] <SWPadnos> well, everyone should remember gorilla.bas
[20:42:05] <skunkworks> I have a ton of programs I wrote in gwbasic. one is a gcode previewer
[20:42:11] <skunkworks> that was qbasic
[20:42:18] <skunkworks> gorilla.bas
[20:42:22] <SWPadnos> was it?
[20:42:26] <skunkworks> pretty sure
[20:42:32] <cradek> I think gwbasic and qbasic were different
[20:42:37] <SWPadnos> oh well - I haven't used basic since - well since VB came out ::)
[20:42:44] <SWPadnos> probably earlier
[20:43:30] <skunkworks> the first verion of the k&t tape emulator was in gwbasic :)
[20:44:24] <cradek> are you going to be sad now that you won't need that huge controller and tape emulator?
[20:44:41] <skunkworks> not one bit.
[20:44:47] <skunkworks> that thing was scary
[20:45:12] <cradek> I look at something like that and I can't believe it ever worked
[20:45:27] <skunkworks> and for as long as it did.
[20:45:53] <skunkworks> we have some of the maintenece records from the company that had it before. It seemed like it never worked ;)
[20:46:14] <cradek> ha
[20:46:43] <skunkworks> they seemed to be always trying to swap boards from one controller to the other (we actually have 2 of these) to try to troubleshoot the problemss.
[20:47:14] <cradek> how long since you've been able to use it?
[20:47:49] <skunkworks> the last thing I did was bore out some turbo-charger housings - so about a year I suppose.
[20:49:02] <skunkworks> some people have boats.
[20:50:33] <skunkworks> mdynac: picture time?
[20:50:48] <mdynac> i gotta email them to someone.....
[20:51:12] <cradek> you can send them to me: chris@timeguy.com
[20:51:16] <mdynac> i have nowhere to post them........
[20:51:32] <mdynac> k
[20:51:36] <cradek> or you can put them on the wiki yourself unless they're huge
[20:52:35] <skunkworks> or you can just keep them on your computer and cradek can hack in and get them. ;)
[20:52:58] <mdynac> i'd like to see him try....
[20:53:11] <mdynac> cradek i just sent a tarball to you....
[20:53:16] <cradek> ok
[20:53:56] <mdynac> not the greatest pics, some of the shop and the machinery available to me, and some of the emcedm, and some of the cut piece....
[20:54:05] <mdynac> and one of my helpers..
[20:55:14] <skunkworks> cradek: I would not call going to emc2 and upgrade. As owhite had mentioned going form emc1 to emc2 - a its a whole different universe
[20:58:05] <skunkworks> things like - all g1,2,3 moves where all incramental - and you had to give the finish posision in absolute going back to g0
[20:58:49] <jepler> skunkworks: huh?
[20:58:56] <cradek> yeah I don't follow either
[20:58:59] <jepler> skunkworks: the behavior of g1 is about the same between emc1 and emc2
[20:59:11] <cradek> no he means his old machine
[20:59:15] <jepler> oh oh
[20:59:21] <skunkworks> sorry - I am taking about the k&t to emc2
[20:59:23] <jepler> I thought he was talking about emc1 to emc2
[20:59:27] <cradek> (but I still don't follow)
[21:00:33] <cradek> mdynac: nothing yet...
[21:00:53] <mdynac> hmmmmm
[21:01:40] <jepler> cradek: and it's not your greylisting?
[21:02:10] <mdynac> the mailto: was in there....
[21:02:29] <mdynac> my bad...
[21:02:30] <cradek> aha, got it
[21:02:52] <skunkworks> say you your starting at g0x1y1 and want to draw a 45deg line to x2y2 you would need g0x1y1 ; g1x1y1 (incramental) ; g0x2y2
[21:04:11] <skunkworks> to get out of the g1,2,3 command you needed to put your acutall absolute position
[21:06:49] <skunkworks> plus the thing that was the most annoying. There was no position readout. no clue where you where.
[21:07:44] <Lerneaen_Hydra> that sounds exactly like the hardware controller in the lathe I retrofitted
[21:08:00] <anonimasu> hm, isnt that the way emc does it also?
[21:08:16] <anonimasu> ah.. not toggled incremental ;)
[21:08:22] <Lerneaen_Hydra> and add that all input of programs was done through a numpad and a 4 digit LED display, and 4 other LEDs. programs could be saved to tape
[21:08:46] <Jymmmm> saved to tape?!
[21:08:48] <skunkworks> yeck.
[21:09:34] <Lerneaen_Hydra> fun fun
[21:10:05] <skunkworks> plus you can't surf the web on it while machining
[21:10:05] <Lerneaen_Hydra> http://lerneaenhydra.shacknet.nu/images/files_projects/compact5/img_5238.jpg
[21:10:09] <Lerneaen_Hydra> that too
[21:10:54] <skunkworks> hey - one of those lathes was at the cnc workshop
[21:10:58] <mdynac> http://www.edmnetwork.com/images/saleItemImages/EF20-330%20Beige-1985.jpg
[21:10:59] <skunkworks> -controller
[21:11:37] <mdynac> the above link is the machine i put the emc2 on
[21:11:46] <cradek> http://timeguy.com/cradek/emc/edm
[21:12:08] <cradek> mdynac: I don't see the part or emc!
[21:12:19] <skunkworks> mdynac: http://www.electronicsam.com/images/KandT/oldkandt.JPG
[21:12:31] <mdynac> gee i sent the wrong pics....however that machine is next for emc.....
[21:13:26] <mdynac> that is one of my customers machine, he cuts a keyway in that little ol cam....
[21:13:39] <mdynac> 6000 lbs of cam.....
[21:13:58] <Lerneaen_Hydra> that's a cam?!
[21:14:06] <Lerneaen_Hydra> now that's a nice size :D
[21:14:09] <mdynac> yes it is....
[21:14:27] <cradek> wow
[21:14:33] <mdynac> sorry cradek, i will send you the correct files....
[21:14:34] <Lerneaen_Hydra> cool
[21:14:49] <cradek> mdynac: no problem
[21:14:51] <Lerneaen_Hydra> what does the cam do?
[21:15:04] <Lerneaen_Hydra> don't tell me it's for a motor (valves)
[21:15:07] <mdynac> i think it's the kotex cam....
[21:15:31] <mdynac> one rotation and you get one packaged kotex...
[21:16:18] <anonimasu> :)
[21:17:35] <mdynac> i'm not kidding....
[21:17:48] <mdynac> they also make the tampon cam....
[21:18:15] <mdynac> hey somebodys gotta do it....you think they just fall out of the sky??
[21:18:32] <cradek> thank goodness they don't
[21:19:14] <mdynac> that machine has 30" z clearance....
[21:19:37] <Lerneaen_Hydra> wtf?! tampon cam?!
[21:19:51] <mdynac> see the aluminum riser we put in behind the cam???
[21:20:32] <Lerneaen_Hydra> when I read tampon I do not think metalworking stuff
[21:21:04] <mdynac> the cam is the programming for the machine.....
[21:21:08] <skunkworks> Lerneaen_Hydra: do you think they are raised on a farm?
[21:21:13] <mdynac> all mechanical...
[21:21:17] <Lerneaen_Hydra> ?
[21:21:19] <Lerneaen_Hydra> huh?
[21:21:41] <mdynac> the cam is the timing....
[21:21:42] <Lerneaen_Hydra> I'm not following any of you now
[21:21:46] <Lerneaen_Hydra> yeah, that I know
[21:22:07] <mdynac> change the cam = different output....
[21:22:14] <Lerneaen_Hydra> translation of some object depending on the rotational position of the axle the cam is connected to
[21:22:36] <Lerneaen_Hydra> for example in a four-stroke combustion engine to control the intake/exhaust valves
[21:22:41] <mdynac> those cams were drawn up in the 1940's
[21:22:56] <mdynac> same machines used today
[21:23:03] <Lerneaen_Hydra> but when I see tampon I do apparently not think of the correct term
[21:23:05] <cradek> new images up...
[21:23:41] <mdynac> kewl
[21:23:44] <EldonB46> Question: Where can I find good C-code Programming manual/examples that describes "subroutines", "computed offsets coordinates systems from current (which may already be offset) tool positions", and "Parametric Programming". I have read all of the EMC2_User_Manual pages, but there are only a few examples of what I am looking for.
[21:24:30] <cradek> mdynac: great pictures!
[21:25:34] <Lerneaen_Hydra> whee KDE! :D
[21:25:38] <Lerneaen_Hydra> kde ftw!
[21:26:07] <cradek> EldonB46: do you mean Gcode or C?
[21:26:07] <skunkworks> Lerneaen_Hydra: a cam can have many shapes. :)
[21:26:24] <EldonB46> Sorry, G-code
[21:26:42] <mdynac> thx
[21:26:47] <Lerneaen_Hydra> skunkworks: I don't know about you but for me tampon = http://en.wikipedia.org/wiki/Tampon
[21:27:17] <mdynac> i am going to put a little presentation together....as you can see the machine is a rats nest right now...
[21:27:36] <skunkworks> uh - yah. but the cam is in the machine that makes the tempons.
[21:27:45] <cradek> EldonB46: I only understand the first part of your question (subroutines) and the answer is here: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Oword (examples) and here: http://linuxcnc.org/docs/html/gcode/main/#SECTION00500000000000000000 (language description)
[21:28:23] <cradek> mdynac: was that PDP8A the old control?
[21:28:34] <mdynac> yep
[21:28:57] <cradek> http://timeguy.com/cradek-files/emc/edm/MVC-028S.JPG
[21:28:59] <cradek> wooo!!
[21:29:11] <Lerneaen_Hydra> skunkworks: oh.
[21:29:13] <Lerneaen_Hydra> duh
[21:29:13] <Lerneaen_Hydra> >.<
[21:29:19] <Lerneaen_Hydra> then say so ;)
[21:29:31] <skunkworks> mdynac: cool
[21:29:32] <alex_joni> cradek: wooo
[21:29:37] <alex_joni> mdynac: congrats
[21:29:41] <alex_joni> good work
[21:30:00] <mdynac> thx, but you guys really did the tough work....
[21:30:01] <anonimasu> hm I wonder if this coil wire will work.
[21:30:37] <mdynac> i just applied it....
[21:31:04] <anonimasu> got a closeup of the parts?
[21:31:17] <anonimasu> nice :)
[21:31:22] <mdynac> there is a couple of part pics....
[21:33:12] <mdynac> rayh and dave eng need to get the news...
[21:33:32] <alex_joni> mdynac: how about pasting to the users list?
[21:33:36] <alex_joni> this is great stuff
[21:33:56] <mdynac> okay
[21:34:38] <EldonB46> cradek: Thanks, I have also looked at the two pages, but still have question. Like while using one coordinate system, and another be specified relative the current - like a five hole bolt pattern around a know location? Maybe someone here will know. Thanks.
[21:38:57] <alex_joni> mdynac: what feed did you have for that part?
[21:39:39] <mdynac> well it took 2 hours 32 minutes to go about 8 inches..
[21:40:12] <mdynac> emc is cool cause it tells you how long it took to burn the line of g code....
[21:40:32] <mdynac> if you start it in a terminal...
[21:40:45] <alex_joni> if you have debug turned on
[21:40:54] <mdynac> i was not pushing it hard at all
[21:41:37] <alex_joni> nice
[21:41:56] <mdynac> i was mainly concerned about the cutting generator screwing up the computer.....rf noise....
[21:42:13] <Lerneaen_Hydra> 'nigt all
[21:42:15] <Lerneaen_Hydra> err
[21:42:16] <Lerneaen_Hydra> 'night all
[21:42:25] <Lerneaen_Hydra> see how much I need sleep? ;)
[21:42:34] <alex_joni> night LH
[21:42:40] <anonimasu> night lh
[21:42:45] <mdynac> which it did not do....when these machines run they wipe out the entire am broadcast band.....
[21:42:49] <cradek> EldonB46: yes there are 9? relative coordinate systems
[21:42:56] <anonimasu> hehe
[21:42:58] <alex_joni> mdynac: nice
[21:43:10] <cradek> EldonB46: if you put one at the center of your pattern, straightforward trig would give you the pattern
[21:43:14] <mdynac> so my grounding scheme is okay
[21:43:14] <alex_joni> I know this from welding (also a nice electrical discharge)
[21:43:22] <mdynac> ya
[21:43:45] <mdynac> they don't get the FCC seal of approval....
[21:44:23] <alex_joni> we care more about CE than FCC
[21:44:34] <mdynac> true
[21:45:00] <alex_joni> but they are mostly equivalent
[21:45:09] <alex_joni> there's a new joke around these parts..
[21:45:16] <alex_joni> do you know what CE stands for?
[21:45:20] <cradek> EldonB46: coordinate systems: http://linuxcnc.org/docs/html/gcode/main/#SECTION00500000000000000000
[21:45:57] <mdynac> i just got a call from a customer in alabama, he is getting rid of an Andrew EF330.......anyone need a cheap wire machine?
[21:46:28] <mdynac> custom engineered?
[21:46:36] <alex_joni> mdynac: china export ;-)
[21:46:45] <mdynac> hah!
[21:47:29] <EldonB46> cradek: I understand, but just wondering, if I am using one coordinate systems (G54) and would like to jump to a new location (i,e, G00X+4Y-2.5) is
[21:47:41] <EldonB46> sorry, more to come
[21:47:49] <jepler> ... To define CE marking you first need to understand what it represents. By affixing the CE mark you are declaring that your product complies with all the relevant European Union standards and it is illegal to sell a product in the EU unless it has passed the relevant compliance tests to achieve a CE mark. ...
[21:48:28] <anonimasu> yes.
[21:48:31] <anonimasu> it's hell.
[21:49:41] <EldonB46> Is there an internal register that knows the current ABS location so "G10 L2 P4 X#<somenumber> Y#<somenumber>" would set the next system?
[21:50:31] <jepler> no, there is no such register
[21:50:50] <cradek> to offset from a current location, you could use G92 instead
[21:51:11] <cradek> but why not set up your coordinate systems first instead of making them relative to some intermediate position?
[21:51:28] <EldonB46> But, G92 does not like tool offsets.
[21:51:31] <jepler> you can accept the reference location as a parameter to your O- sub and add it to all moves within the sub
[21:51:53] <mdynac> later folks, tis nap time.....
[21:51:55] <cradek> I'm pretty sure G92 does work with tool offsets
[21:51:58] <eholmgren> tampon what?
[21:52:08] <cradek> mdynac: later, thanks again for the photos, nice to see a new application
[21:52:14] <jepler> so instead of: "G0 X4 Y4 / O100 call" you would "O100 call [4] [4]"
[21:54:31] <jepler> then to move X=.5 Y=.8 compared to reference location you'd write: "G1 X[#1+.5] Y[#2+.8]"
[21:55:05] <EldonB46> cradek: but why not set up your coordinate systems first - A family of parts that have objects (holes) relative to a center (or edge) could be created without manually doing all of the trig for each part, especially when I am not the author of the part and do not want to understand all of the structures in the part G-code. Maybe I am just lazy?
[21:56:53] <eholmgren> trig is fairly simple compared to alot of other things ...
[21:57:18] <cradek> if you want to work with offsets from the current position it seems to me you have two tools at your disposal: g91 (incremental distance) and g92 (set the origin relative to here)
[21:57:55] <cradek> (but it's possible I still don't quite understand your question, sorry if so)
[22:13:54] <EldonB46> cradek: Thanks for the help, Here is my current project, http://www.mr-fab.com/Shop/CNC/ - just looking for way to make programming easier. Do you think the book stores have G-code books - I may look at Amazon.
[22:14:30] <cradek> EldonB46: the problem with gcode is every manufacturer is different...
[22:15:03] <cradek> ah, interesting part
[22:15:29] <cradek> every time I see a screenshot of AXIS, I try to navigate in it
[22:15:34] <eholmgren> Eldon: how big are the western scenes?
[22:15:46] <EldonB46> I understand, I am working with EMC2 and wondering about the code flexibilities
[22:15:53] <cradek> what are you trying to add to this program?
[22:18:46] <EldonB46> This part is a programming exercise that I am playing with, and just looking for way to program for future parts.
[22:19:00] <cradek> ah, I see
[22:19:03] <EldonB46> eholmgren: ?
[22:19:50] <jepler> EldonB46: you should probably put a G20 in the setup codes if the coordinates are in inches, or G21 if they are in millimeters.
[22:19:53] <EldonB46> cradek: every time I see - I also, the part is there as code if you would like
[22:20:11] <eholmgren> http://www.mr-fab.com/online/Preview/FAB/gallery/FunProjects/ArtWorks?tn
[22:20:23] <jepler> EldonB46: in your "o104 sub", the first two parameters are the center of the hole, according to the comment
[22:20:29] <EldonB46> Good Idea, I thought I had all of that stuff done.
[22:20:39] <jepler> EldonB46: so instead of doing G54 L10 P2 ... and then ... call [0] [0] ...
[22:20:49] <jepler> you could: call [4] [4] ...
[22:21:15] <cradek> EldonB46: did you cut these with emc2?
[22:22:32] <jepler> I like this one the best: http://www.mr-fab.com/online/Preview/FAB/gallery/FunProjects/ArtWorks/DCP_0421.JPG?1/2 -- those are neat
[22:22:47] <jepler> bbl
[22:23:02] <EldonB46> cradek: No not yet, my CNC controller is still not complete, photos of the case (controller) is on http://www.mr-fab.com/Shop/
[22:23:31] <cradek> I like the depth in the mounting of 421
[22:24:07] <EldonB46> OH, sorry I did not know what you were looking at, Yes/No - the art things were cut on "PlasmaCAM"
[22:25:23] <EldonB46> I am in the process of creating, controllers for my Mill and creating an XYZ router bed.
[22:26:44] <cradek> wow, what cnc controller needs six hard disks?
[22:28:31] <EldonB46> cradek: No, those are the "web" server disk.
[22:28:55] <cradek> ah
[22:29:15] <cradek> oops, have to run, later everyone
[22:30:17] <ve7it> ve7it is now known as LawrenceG
[22:30:19] <EldonB46> Cya
[23:47:42] <robin__sz> moo
[23:48:05] <Jymmmm> Gribbit
[23:48:31] <robin__sz> ah ha! now I know your day job!
[23:48:41] <Jymmmm> ?
[23:48:51] <robin__sz> you are the Budweiser frog!
[23:48:58] <Jymmmm> lol
[23:49:13] <tomp> hello
[23:49:15] <robin__sz> going well?
[23:49:22] <Jymmmm> It that was the case, wouldn't it be..... Bud Weis Errrrrrrrr
[23:49:41] <robin__sz> ah yes ... sorry, i dont speak frog as well as you ;0
[23:50:53] <Jymmmm> robin__sz: Not good at all (home finding), but I'd rather not go there, been trying to keep the anxiety of it all down for the last 9 days.
[23:51:07] <robin__sz> sigh
[23:51:48] <robin__sz> well, as they say : smile, be happy, things could be worse ...
[23:51:50] <robin__sz> so ..
[23:51:57] <robin__sz> I smiled and was happy
[23:52:03] <robin__sz> and sure enough, things got worse :)
[23:52:35] <Jymmmm> *IF* things continue as there already are, there will be a couple of lawsuits. But that takes forever and still won't help with having a place to live now (before Jan 1)
[23:52:44] <robin__sz> well yeah
[23:52:50] <robin__sz> parents?
[23:52:55] <robin__sz> in laws?
[23:53:15] <Jymmmm> None - closest option looks like motel + storage
[23:53:28] <robin__sz> youch
[23:53:37] <robin__sz> this is NY right?
[23:53:46] <Jymmmm> San Jose California
[23:54:26] <robin__sz> rental is that hard out there?
[23:54:44] <Jymmmm> 90% have a NO PETS policy.
[23:54:51] <robin__sz> bummer
[23:55:15] <robin__sz> surely the same applies to motels?
[23:55:45] <Jymmmm> Well, mayn people are traveling with their animals, then stop off to sleep/shower etc
[23:55:59] <robin__sz> uh huh
[23:56:22] <robin__sz> trailer park? or is that not an option you wanna consider?
[23:57:23] <Jymmmm> Months ago we looked into a Mobile Home. Even if you buy the MH the mortgage can be $500-$700/mo PLUS space renat FOR LIFE.
[23:57:37] <Jymmmm> at another $750/mo
[23:58:07] <robin__sz> sounds expensive
[23:58:18] <Jymmmm> Not a good invested long or short term when you look at it (unless you own the land or lot)
[23:58:28] <robin__sz> well yeah
[23:59:00] <robin__sz> land is always a good investment
[23:59:03] <Jymmmm> And there are no places within 150 miles that have where you buy the lot and pay $50/mo association fees. (which is reasonable imo)
[23:59:05] <robin__sz> they stopped making it years ago