Back
[00:00:00] <Thetawaves> i've got a document specifying setup instructions for mach3, can somebody help me derive some good guess values for step/dir setup and hold times?
[00:00:08] <Thetawaves> http://www.amonstar.com/template/file/3020/desktop-CNC user manual.doc
[00:00:33] <Thetawaves> i wish there were not spaces in that link
[00:00:51] <Thetawaves> http://www.amonstar.com/template/file/3020/desktop-CNC%20user%20manual.doc
[00:24:18] <seb_kuzminsky> Thetawaves: neat-looking machine
[00:24:23] <seb_kuzminsky> the cable guides are classy
[00:24:50] <Thetawaves> thanks, i am very excited to get it running
[00:25:28] <seb_kuzminsky> you're running mach3 on it now?
[00:25:38] <Jymmm> Never thought of this
http://i44.tinypic.com/346uwxk.jpg
[00:26:13] <Thetawaves> no i haven't set it up yet
[00:26:39] <seb_kuzminsky> Thetawaves: nice-looking control box too
[00:27:19] <seb_kuzminsky> i didnt see any stepper timing info in the manual you linked
[00:27:22] <seb_kuzminsky> did i just miss it?
[00:28:12] <jdhnc> Jymm: real men wiggle cables
[00:28:23] <Thetawaves> seb_kuzminsky, that's the problem, i don't see any info
[00:28:37] <Jymmm> jdhnc: I am woman, hear me roar!
[00:28:47] <jdhnc> jpegs?
[00:28:56] <Jymmm> perv!
[00:29:16] <seb_kuzminsky> Thetawaves: it's almost visible in the screenshot on page 14
[00:29:34] <seb_kuzminsky> i bet we can read it all off there
[00:30:35] <Thetawaves> page 13 right?
[00:31:04] <seb_kuzminsky> the page that's labeled "15" at the bottom
[00:32:22] <Thetawaves> hmm! page 15 has two pictures (front and back of controller) and some text
[00:33:39] <seb_kuzminsky> this pic:
http://highlab.com/~seb/mach3.png
[00:34:31] <seb_kuzminsky> in my .doc viewer (open office 3.2, i think), that's on the page labeled 15
[00:34:35] <seb_kuzminsky> anyway
[00:34:38] <seb_kuzminsky> in that pic
[00:35:03] <seb_kuzminsky> there's a field labeled 'Step Pulse 1-5 us', and its value appears to be 10 for all three steppers
[00:35:10] <Thetawaves> ahhh yeah 13 for me (weird)
[00:35:14] <seb_kuzminsky> so that's steplen=10000 for linuxcnc
[00:36:09] <Thetawaves> alright, what about setup time?
[00:36:30] <seb_kuzminsky> then Dir Pulse is 3, that's probably what we call 'dirsetup'
[00:36:44] <mrsun> how fun to get sick right at the freakin weekend :P
[00:37:07] <seb_kuzminsky> there's a field unhelpfully called 'Steps per', i wonder if that's inch or metric?
[00:37:18] <seb_kuzminsky> prolly metric on a chinese machine?
[00:37:32] <seb_kuzminsky> 400 steps/mm seems like a lot
[00:38:57] <Thetawaves> yes, i believe all units are mm
[00:39:10] <seb_kuzminsky> i guess it's reasonable, it's 0.0001" per step, in the units i'm used to...
[00:39:40] <seb_kuzminsky> if so, and if the max vel is given in mm/min, then we can compute the stepspace from that
[00:40:03] <seb_kuzminsky> 200 mm/min * 400 steps/mm = 80,000 steps/min
[00:40:37] <Thetawaves> 2000mm/sec actually
[00:40:48] <seb_kuzminsky> 80,000 steps/min * 1/60 min/second = 1333 steps/second
[00:40:54] <seb_kuzminsky> oh you're right
[00:41:07] <seb_kuzminsky> so 13,333 steps/second
[00:42:02] <seb_kuzminsky> that's 0.000075 seconds/step, 75 us/step
[00:42:24] <seb_kuzminsky> we know steplen is 10us, so stepspace must be the other 65
[00:42:41] <Thetawaves> how what?
[00:42:50] <Thetawaves> i'm calculating 800,000 steps a second
[00:43:06] <Thetawaves> 2000mm/sec * 400steps/mm = 800,000mm/sec
[00:43:08] <Thetawaves> errrr
[00:43:14] <Thetawaves> 800,000steps/sec
[00:43:24] <seb_kuzminsky> it's 2000 mm/min according to the label on the gui
[00:43:32] <Thetawaves> ooooohhhh
[00:44:09] <Thetawaves> ok i understand now thanks
[00:44:15] <seb_kuzminsky> great! :-)
[00:44:47] <seb_kuzminsky> so i guess there's some newfangled tool called 'stepconf' where you can poke these numbers in, and it'll write a linuxcnc config for you
[00:45:00] <Thetawaves> yeah i've got that open
[00:45:10] <seb_kuzminsky> or you can spend an evening with an irc window, a manpage window, and a text editor ;-)
[00:45:25] <seb_kuzminsky> is the machine under power?
[00:45:28] <Thetawaves> no
[00:45:33] <seb_kuzminsky> ah, too bad
[00:45:39] <seb_kuzminsky> i think stepconf has a "test axis" button
[00:45:44] <mrsun> mm
[00:45:46] <Thetawaves> i can plug it in :)
[00:45:52] <seb_kuzminsky> gulp!
[00:47:21] <Thetawaves> so dir setup = 3000
[00:47:28] <Thetawaves> what about direction hold?
[00:47:41] <seb_kuzminsky> i dont think i see it anywhere on that screen :-(
[00:47:52] <Thetawaves> should i just leave it at the 20us
[00:48:06] <seb_kuzminsky> that should be safe
[00:51:34] <Thetawaves> max step rate 6896hz
[00:52:11] <seb_kuzminsky> should be twice that
[00:52:15] <seb_kuzminsky> something's not adding up
[00:53:03] <Thetawaves> active low == not inverted
[00:55:06] <Thetawaves> so what is wrong?
[00:55:47] <Thetawaves> steptime = 10,000 step space = 65,000 direction hold = 20,000 dir setup = 3,000
[00:56:58] <seb_kuzminsky> what's your base period?
[00:57:24] <Thetawaves> reported as 6364, but i entered 7064
[00:58:17] <Thetawaves> min base period 72065ns
[00:59:11] <seb_kuzminsky> wait, which of those numbers is the base period you're using? the 75 us one or 7 us?
[00:59:48] <Thetawaves> sorry, 7064 is the jitter number, the base period which is a calculated field says 72064
[00:59:55] <seb_kuzminsky> ah ok
[01:00:10] <seb_kuzminsky> so that explains the problem
[01:00:23] <seb_kuzminsky> you get to change the state of the step and/or dir output pin once per base period
[01:00:41] <seb_kuzminsky> so the step freq is limited to base period / 2
[01:01:05] <seb_kuzminsky> there's some special way to do two writes to the pins per period
[01:03:03] <Thetawaves> i sorta want to hook my driver up to an oscilloscope and measure the cutoff freq
[01:04:05] <Thetawaves> where does the min base period number come from?
[01:04:07] <seb_kuzminsky> can you ask the manufacturer/reseller for timing specs for the stepper amps? or can you find the docs online somewhere?
[01:04:32] <Thetawaves> yeah i asked, still waiting
[01:04:41] <seb_kuzminsky> the base period is determined by measuring the scheduling latency of your computer
[01:04:46] <seb_kuzminsky> it's highly hardware dependent
[01:05:38] <Thetawaves> so it asks the latency test for that number?
[01:05:49] <seb_kuzminsky> right
[01:06:12] <seb_kuzminsky> ah, i was looking for the parport.<portnum>.reset function:
http://linuxcnc.org/docs/2.4/html/hal_parallel_port.html#r1_1_4
[01:06:39] <seb_kuzminsky> i think that would let you do one full step per base period, instead of one half step
[01:06:49] <seb_kuzminsky> i wonder if stepconf uses that
[01:06:52] <seb_kuzminsky> it doesnt seem to
[01:07:11] <Thetawaves> so 7khz step speed is pretty lousy?
[01:07:31] <seb_kuzminsky> it might be useful to accept the slower step rate for now, and get the machine moving, and then optimize and tune later
[01:08:03] <Thetawaves> 17.5mm/sec
[01:08:06] <seb_kuzminsky> i'm not sure if 7 khz is good or bad... you're using software step generation, out the parallel port, and i dont have any machines that are set up like that
[01:08:36] <Thetawaves> seems really fast
[01:08:39] <seb_kuzminsky> all the machines i use have hardware step generation, and hundreds of khz is normal
[01:09:01] <Thetawaves> heh
[01:09:37] <seb_kuzminsky> 17.5 mm/sec is 41 inches/minute
[01:09:42] <seb_kuzminsky> seems like a fine place to start
[01:09:49] <Thetawaves> 1050mm/min hmmm
[01:10:04] <seb_kuzminsky> one minute per meter
[01:10:34] <seb_kuzminsky> see if it'll even turn the motors, then if it does, see if you can enable the parport reset and double your steprate
[01:10:44] <seb_kuzminsky> are you on linuxcnc 2.4 or 2.5?
[01:11:07] <Thetawaves> the one on ubuntu 10.4 linuxcnc edition
[01:11:31] <seb_kuzminsky> we provide both 2.4 and 2.5 on ubuntu 10.04
[01:11:35] <Thetawaves> v2.4
[01:11:38] <seb_kuzminsky> ah ok
[01:11:54] <Thetawaves> that's what the user manual claims, but i have updated it
[01:12:13] <seb_kuzminsky> ok that's good
[01:12:27] <seb_kuzminsky> it'll say when you start the axis gui later
[01:12:38] <Thetawaves> ok
[01:12:46] <seb_kuzminsky> or you can ask the packaging system by running "dpkg --status emc2" or "dpkg --status linuxcnc", depending on version
[01:13:08] <seb_kuzminsky> but i guess it's not important now
[01:13:26] <Thetawaves> 2.4.7
[01:14:10] <Thetawaves> active low = invert right?
[01:14:56] <seb_kuzminsky> let's see...
[01:15:02] <seb_kuzminsky> i'm looking here:
http://linuxcnc.org/docs/2.4/html/config_stepconf.html#sub:Parallel-Port-Setup
[01:15:15] <seb_kuzminsky> it says "Turn on the "invert" check box if the signal is inverted (0V for true/active, 5V for false/inactive)."
[01:15:26] <seb_kuzminsky> i'm not sure what your stepper amps want
[01:15:41] <seb_kuzminsky> it might be in that doc you linked earlier, somewhere
[01:16:03] <seb_kuzminsky> oh yes it is, it's the picture before the one i linked back to you
[01:16:22] <seb_kuzminsky> looks like only estop is active low
[01:16:31] <seb_kuzminsky> aka inverted
[01:18:30] <Thetawaves> same with each step axis
[01:18:30] <mrsun> http://www.youtube.com/watch?v=FuwFlW3n_3E&feature=g-vrec&context=G208964eRVAAAAAAAAAQ how cute, lets cnc it? :)
[01:18:45] <Thetawaves> turns out that only 7 of the parallel port pins are in use
[01:18:51] <Thetawaves> plenty of expansion for my own circuits :D
[01:19:14] <seb_kuzminsky> no limit or home switches?
[01:19:21] <Thetawaves> nope
[01:19:26] <seb_kuzminsky> 7 is step&dir for 3 axes, then estop, right?
[01:19:28] <Thetawaves> i will have to build them
[01:19:34] <seb_kuzminsky> nice :-)
[01:19:35] <Thetawaves> yeah
[01:20:14] <seb_kuzminsky> i'd be ready to cut power to the steppers manually, before clicking on the Test Axis button
[01:20:25] <seb_kuzminsky> i wouldn't trust the estop to work just yet
[01:20:34] <Thetawaves> ok
[01:21:20] <Thetawaves> how many pulses per revolution are my stepper motors?
[01:21:29] <seb_kuzminsky> i don't know
[01:23:20] <seb_kuzminsky> what's your leadscrew pitch?
[01:23:30] <Thetawaves> my best guess is these screws are 2mm/rev
[01:23:47] <seb_kuzminsky> ok
[01:24:34] <seb_kuzminsky> and it's 400 steps/mm according to that one picture
[01:24:57] <Thetawaves> 800 steps / rev? wtf?
[01:25:26] <seb_kuzminsky> that's what the manual says
[01:25:28] * seb_kuzminsky shrugs
[01:25:44] <Thetawaves> ?8mm 1204 trapezoidal screw , pitch of 1.25 (double nut + spring consumer backlash)
[01:25:47] <Thetawaves> hmm
[01:25:52] <seb_kuzminsky> ah
[01:25:56] <seb_kuzminsky> 1.25 mm/rev pitch
[01:28:06] <Thetawaves> i can't seem to verify that number somewhere else
[01:28:48] <seb_kuzminsky> if it's in one place in the docs, that's probably enough
[01:29:22] <seb_kuzminsky> once we're done configuring we'll measure the actual movement and sanity-check it
[01:29:45] <seb_kuzminsky> if it's wrong we'll debug it, but until then i'd trust the manual
[01:31:09] <Thetawaves> another site says Max Speed 4000mm/min
[01:31:10] <Thetawaves> Max Engraving Speed 2000mm/min
[01:31:47] <Thetawaves> i didn't find that number in the manual, on another site found searching for 1204 trapezoidal screw
[01:31:48] <seb_kuzminsky> i'd guess "Max Speed" means rapids, without the tool cutting the work, and "Max Engraving Speed" means max speed when the tool is cutting
[01:32:46] <seb_kuzminsky> stepper motors lose torque at higher speeds, and moving the machine requires more torque when cutting than when just traversing
[01:38:13] <Thetawaves> seb_kuzminsky, thank you for your help
[01:38:31] <Thetawaves> i have to give some drunks a ride home :)
[01:38:56] <Thetawaves> i'll be back
[01:39:02] <seb_kuzminsky> heh
[01:39:04] <seb_kuzminsky> ok, see you later
[01:39:10] <seb_kuzminsky> good luck with your machine :-)
[01:39:19] <Thetawaves> thanks :D
[02:00:22] <DJ9DJ> moin
[02:07:47] <joe9> any suggestions on how I can make the small notch on the right with a drill press, please?
http://i.imgur.com/XghCZ.png is the notch that I am talking about. The little one on the bottom right.
[02:08:30] <joe9> the notch size is approximately 4mm in width and 6mm in height.
[02:08:53] <joe9> I am thinking of looking for a square drill bit, but am not sure if there is a better way of doing something like this?
[02:09:18] <jdhnc> does it really need to be square?
[02:09:44] <joe9> the piece is square, but the notch is not.
[02:09:50] <joe9> jdhnc: does that make sense?
[02:10:06] <jdhnc> do the corners of the notch really need to be square?
[02:10:07] <joe9> the size in the image is the actual size/dimension.
[02:10:21] <mrsun> a file? :P
[02:10:28] <joe9> no, the notch is not square. it is 6mm x 4mm.
[02:10:47] <jdhnc> do the corners of the notch really need to be right angles.
[02:10:55] <joe9> jdhnc: oh, yes, please.
[02:11:42] <jdhnc> you could use a broach in a drill press
[02:12:03] <joe9> let me google and read up on "broach". Thanks for the advice.
[02:12:14] <jdhnc> but, if it isn't precision customer oriented whatever, perhaps a drill and a file would be sufficient
[02:13:13] <joe9> I have a dremel. would you recommend using that over a file?
[02:13:26] <mrsun> no imo
[02:13:34] <mrsun> alot harder to control power tools =)
[02:13:36] <jdhnc> a flat file can make the edges square, a dremel will never do that.
[02:13:49] <joe9> mrsun: with just the file, it will be hard to go to such depth. I am a noob.
[02:14:05] <mrsun> joe9, a good file and 4mm deep
[02:14:09] <joe9> mrsun: I meant "will it be "
[02:14:09] <mrsun> no problem
[02:15:00] <mrsun> how thick is the material ?
[02:15:00] <joe9> 1/32" inches
[02:15:00] <jdhnc> drill lots of holes next to each other, then file down the rough edges
[02:15:00] <mrsun> gah
[02:15:00] <jdhnc> oh, I'd start with the file
[02:15:00] <joe9> any files that you recommend?
[02:15:00] <mrsun> inches, whats that in mm now :P
[02:15:00] <joe9> probably, 1 mm
[02:15:07] <mrsun> 0.8 about
[02:15:08] <joe9> or just a little bit more, I think.
[02:15:17] <mrsun> you will file that away with your teeth if you want :P
[02:15:22] <mrsun> wont take more then a couple of minutes
[02:15:29] <joe9> oh, really.
[02:15:37] <mrsun> files can cut material
[02:15:49] <mrsun> and fast, specialy thin material =)
[02:16:36] <joe9> mrsun, any file that you would recommend?
[02:17:36] <mrsun> a flat one? :P
[02:17:47] <mrsun> thats not wider then 6mm on the thin side? :P
[02:17:55] <joe9> btw, is the square drill bit a bad idea, if I have to do this more than a few times?
[02:18:25] <joe9> s/if I have/even if i have/
[02:18:32] <mrsun> or a square file
[02:18:37] <mrsun> thats not wider then 6mm =)
[02:18:39] <joe9> mrsun, thanks.
[02:19:01] <mrsun> joe9, i would say it depends on how many more times
[02:19:09] <joe9> 10 times.
[02:19:18] <jdhnc> what's a square drill bit?
[02:19:25] <mrsun> joe9, do a test, see how long it takes to file that slot in a piece =)
[02:19:43] <mrsun> joe9, and if you are into metal working you realy need to learn to use a file and see how effective a tool it can be =)
[02:20:00] <joe9> mrsun, ok, thanks. will try a file then.
[02:20:10] <mrsun> if it would be alot of times i would rather use some kind of a punch press to do the notching =)
[02:20:31] <jdhnc> what is the material?
[02:20:37] <joe9> FR-4
[02:21:04] <jdhnc> file should be fine
[02:21:19] <jdhnc> or a cut-off wheel in a dremel maybe
[02:21:22] <mrsun> oh its a pcb ?
[02:21:25] <joe9> mrsun, for filing, do I need something special while clamping the piece to protect it?
[02:21:29] <joe9> mrsun: yes.
[02:21:58] <mrsun> joe9, it will cut like butter, so just put some protective sheeting in the vice before clamping it, like a piece of paper or something =)
[02:22:47] <joe9> files seem to be available in these varieties "Rough, Bastard, Second Cut and Smooth" . which one would you recommend?
[02:23:33] <mrsun> ough =)
[02:23:41] <jdhnc> smooth is slow, rough is fast
[02:23:42] <mrsun> i like the ones that is cross cut =)
[02:23:59] <joe9> mrsun, not to question your advice, with a file, will it be precise enough? with the dremel, I was under/over cutting by a little bit. I agree it was hard to get it exact with the dremel.
[02:24:14] <jdhnc> a plain flat wood file \
[02:24:37] <mrsun> joe9, depeneds on you, with a dremel its alot easier to overcut as its a very effective cutting tool
[02:24:45] <mrsun> a file cuts as far as you apply preasure =)
[02:24:57] <mrsun> for very fine cuts, no preasure, for rough cuts, alot of preasure
[02:25:01] <mrsun> alot more control =)
[02:25:09] <joe9> mrsun, that is what I found. hence, am looking for something more precise and faster.
[02:25:32] <joe9> mrsun, jdhnc thanks a lot.
[02:25:36] <mrsun> joe9, filing is also an art, but if you have lines to go after you should have no problems getting it quite precise ==)
[02:25:48] <mrsun> as long as you take it slow to start with until you get more adapt with the tool =)
[02:26:40] <joe9> jdhnc: that is what I had used earlier. a dremel cut-off tool.
[02:27:35] <jdhnc> you need a dado blade for it
[02:28:10] <joe9> are punch presses cheap? they seem to be pretty expensive.
[02:28:23] <mrsun> dunno
[02:28:25] <joe9> and, for a punch press, I probably need a die-set oslt. correct?
[02:28:32] <jdhnc> yes
[02:28:52] <jdhnc> punch tooling is also pricey. you could use an arbor press though.
[02:29:23] <mrsun> for cutting pcbs i was more thinking one of these notching tongs =)
[02:29:42] <jdhnc> never cut pcb with a notcher, but it might work
[02:29:54] <mrsun> or whatever its called =)
[02:31:19] <mrsun> tho cant seem to find one that cuts squares :P
[02:32:02] <jdhnc> nibbler
[02:32:39] <jdhnc> maybe a wood mortising bit
[02:35:17] <joe9> i am reading up on it. what do you think of mortising bits?
[02:35:59] <archivist> mortising bit wont like being used on pcb material
[02:38:23] <jdhnc> 10 times, might last.
[02:38:30] <joe9> http://www.youtube.com/watch?v=Iz3pDMP7IsE
[02:39:24] <joe9> but, I do not know if I can get those bits at homedepot/lowes or mcmastercarr.
[02:39:55] <jdhnc> afaik, they are not really square
[02:40:18] <joe9> jdhnc: i think you are right, the edges seem a little curved
[02:40:21] <mrsun> http://www.youtube.com/watch?feature=endscreen&NR=1&v=Wt46yoQkBDA
[02:40:22] <mrsun> :P
[02:42:52] <jdhnc> I'd start with scribe & cut/file
[02:48:58] <joe9> mrsun, good video. thanks.
[02:49:03] <joe9> jdhnc: what is a scribe?
[02:50:31] <jdhnc> a pointy tool that scratches
[02:50:40] <joe9> ok, thanks.
[02:50:57] <jdhnc> scribe the area to be removed, then use a file or dremel or whatever up to the scribe mark
[02:51:41] <jdhnc> diamond scribe, xacto knife, razor blade, electric engraver
[02:54:08] <joe9> got it.
[02:54:09] <joe9> thanks.
[04:03:42] <|n0b0dy|> http://www.sherline.com/tip20.htm
[04:21:14] <Thetawaves> a lot of the square hole uses could possibly be done with a round bore and keyway
[04:21:34] <Thetawaves> anybody agree?
[04:34:34] <Loetmichel> *google... where is Vilnius? Bossjust ordered me to show some computers a t a military fair... 1500km drive, 2 days demonstration of Computers, 1500km drive back ... THANKS colleague woh has to go to the doctor in thios time :-(
[04:42:54] <psha[work]> Loetmichel: with your roads 1.5k km is not so much )
[04:59:32] <archivist> Thetawaves, do you mean a broach
[05:00:30] <Thetawaves> yeah
[05:01:26] <archivist> I would hate to use my HSS broaches on pcb material
[05:03:47] <Thetawaves> because it will break the pcb?
[05:03:56] <Thetawaves> or because it will gum up your broaches?
[05:04:16] <archivist> no, wear the broach
[05:04:59] <Thetawaves> more than metal?
[05:05:52] <archivist> glass fibre is abrasive
[05:07:21] <archivist> the pcb manufacturers use air driven spindles at 60k rpm and carbide tooling
[05:12:31] <Thetawaves> ahh cool
[05:12:50] <Thetawaves> make some
[05:13:23] <Thetawaves> hell you can probably get a press setup to cut it all in one pass
[05:13:40] <Thetawaves> instead of the many teeth of a broach
[05:14:25] <archivist> the machines mill the outline
[05:24:23] <archivist> if it is ready to run it is also not faulty
[05:48:57] <jdhnc> if it is not faulty, is it ready to run?
[05:55:41] <mrsun> hmm, i wonder using a better opamp if i could use the +10.5V analog output to power an opamp for the analog input ref
[05:55:49] <mrsun> analog frequency ref
[05:56:43] <mrsun> using 324N i would get 0-9V max ... :/
[07:56:29] <DJ9DJ> rehi
[08:41:57] <Jymmm> erlo
[09:07:07] <mazafaka> What 'erlo', oh? 'Waterloo' ??
[09:11:20] <mazafaka> Why remind of this battle?
[10:07:51] <Jymmm> http://images.nationalgeographic.com/wpf/media-live/photos/000/426/cache/truck-moss-woods_42607_990x742.jpg?01AD=3RI9KG1Y5wGX3VroyqohykdYy7accIPTuV39KMJB8y69UqVUKsOItFw&01RI=CFF855C1ABCE61F&01NA=
[10:13:15] <Loetmichel> jymm: sorry sight.... could have been a gread muscle car somerhere other tan standing more or less rotten in the woiods.
[10:13:15] <Loetmichel> woods
[10:13:37] <Jymmm> http://images.piccsy.com/cache/images/ice-ice-192756-500-667.jpg
[10:16:04] <Loetmichel> Jymmm: looks like a normal winter not far from here (north ) ;-)
[10:22:11] <pingufan> Hello. Yesterday I installed (manually) the hal-entry for the tooltip sensor. Is this all I have to do, or are there some other things I have to do to make it work? I would like to test the small program I got, but I am afraid to damage something.
[10:23:21] <archivist> afraid.... not much we can do for you
[10:24:37] <pingufan> Sure. Is something to be configured in addition, or should it work right now?
[10:24:52] <awallin_> start halmeter and look at probe-input
[10:24:55] <awallin_> does it work?
[10:25:15] <archivist> he got that far already
[10:30:47] <awallin_> so then you run some safe probe move and enable the switch by hand. does that work?
[10:31:28] <joe9> I am using this attachment (
http://www.amazon.com/Dremel-196-High-Speed-Cutter/dp/B00008ZA01) with a drill press to drill 5mm diameter (not through holes) 0.4 mm depth holes.
[10:31:36] <joe9> my drill press is a craftsma 38142.
[10:31:44] <joe9> Is it a bad idea?
[10:31:56] <joe9> s/craftsma/craftsman/
[10:32:19] <joe9> any thoughts, please?
[10:32:46] <archivist> does it work, or does it fail, does the tool wear..., all things we cannot see from here
[10:33:20] <joe9> I think it works. I was not sure if it is a bad idea to use a dremel drill bit with a proper drill press.
[10:33:33] <joe9> I have not tried it yet with a drill press.
[10:33:44] <pingufan> awallin_: I had to boot up everything first. Therefore the delay. Yes, by default it us "FALSE", and when I press it (simulate a touch with the tool), it becomes true.
[10:34:05] <joe9> archivist: I am checking to make sure that it is not "off-the-charts" bad to use dremel bits with a drill press
[10:34:42] <awallin_> joe9: with a 5mm bit your drill-press probably has enough rpm. running a 0.5mm cutter at low rpm would not be that great
[10:35:29] <joe9> awallin_: ok, thanks. as long as it works, it is not such a bad idea. then.
[10:36:55] <pingufan> So a "G38.2 z0 f20" should do a touch off on the sensor now? Move downwards until the button/sensor is triggered?
[10:37:56] <pingufan> Only to be sure, this is my very first steps.
[10:38:32] <awallin_> try it without a tool in the spindle, with lots of z-clearance, and flipping the probe-switch by hand.
[10:38:55] <pingufan> Good idea. But it should behave as I described?
[10:39:47] <awallin_> yes, you have to start above z=0, maybe z=5 or something. and the probe needs to be FALSE when you enter g38.2.
[10:46:06] <pingufan> Excellent! :D Works!
[10:46:57] <jthornton> Program G38.x axes to perform a straight probe operation. The axis words are optional, except that at least one of them must be used. The axis words together define the destination point that the probe will move towards, starting from the current location. If the probe is not tripped before the destination is reached G38.2 and G38.4 will signal an error.
[10:47:01] <pingufan> The z coordinate is in machine coordinates or address space of workpiece ?
[10:47:26] <jthornton> http://linuxcnc.org/docview/2.5/html/gcode/gcode.html#_g38_x_straight_probe_a_id_sec_g38_probe_a
[10:47:31] <awallin_> whatever is the current coordinate system
[10:48:36] <joe9> i have this vise (
http://www.harborfreight.com/5-inch-drill-press-milling-vise-94276.html). I am not planning on using it for milling, but, just as a drill press vise to hold a piece for a few straight lines of holes. I find that the scale is very inaccurate and am thinking of replacing the scale with something more accuarte, such a calipers or something like that.
[10:48:55] <joe9> wondering if anyone has come across something like that.
[10:49:12] <pingufan> jthornton: What do i understand with 38.4 ? How does this work? Does it move down until it gets contact and then retract?
[10:50:01] <awallin_> pingufan: no. with 38.4 you start with the probe in TRUE state, and move until you see FALSE
[10:50:42] <joe9> the scale on the vise is not held down by a screw but some small screw like attachment. If i remove it with a chisel, I am not sure if I can fit something else in that hole.
[10:50:46] <pingufan> So this is actually not really practicable in my case.
[10:51:21] <Jymmm> Loetmichel:
http://laughingsquid.com/paradise-parking-beautiful-photos-of-abandoned-cars-decaying-in-nature-by-peter-lippmann/
[10:51:49] <pingufan> I just noticed that the microswitch I used closes before it clicks. is this often the case?
[10:54:02] <pingufan> Can I fix the position of G30 in the system (permanently store it), or must this be part of my milling program? Also: Does G30 work with machine coordinates or user coordinates?
[10:56:19] <awallin_> the parameter values are saved in a file, they don't need to be in every program (unless you want to change them)
[10:56:41] <awallin_> docs seem to say G30 parameters are in absolute coords
[10:59:52] <pingufan> Where can I read that?
[11:01:00] <awallin_> http://linuxcnc.org/docs/2.4/html/gcode_main.html#sec:G30,-G30.1
[11:02:46] <Jymmm> Got Mouse?
http://25.media.tumblr.com/tumblr_lzbpnzUwUQ1qbklpto1_500.jpg
[11:03:39] <pingufan> It is sad that I cannot store a special move to be done for moving to the sensor. I.e. first up, then back (max y) then right (where the sensor is).
[11:04:27] <pingufan> This picture remembers me to my Janich&Class Industrial PC. MY first computer. Running CPM/3
[11:04:58] <awallin_> I'm sure there are "free" parameter numbers where you can store your probe-position. You could then make a custom M-code for your probe-sequence
[11:09:23] <Loetmichel> re @ home
[11:10:26] <pingufan> Hmm. Ok. This I will try later, when I am familiar with this probing at all. Any idea which file contains the G30 position?
[11:11:06] <awallin_> VAR file
http://linuxcnc.org/docs/2.4/html/config_ini_config.html
[11:12:48] <pingufan> Whow. Thank you. :)
[11:13:46] <pingufan> Ah! I do not have to fiddle around at all! G30.1 should do that for me?
[11:15:10] <awallin_> yes, use g30.1 to store the reference position parameters in the VAR file. In general editing the VAR file by hand is a bad idea..
[11:15:36] <pingufan> I thought that, too.
[11:16:31] <pingufan> Only for understanding. The tool table contains tool lengths. They are ignored when using the G38.2 method ?
[11:19:51] <awallin_> the probe-move will not set tool length. it will only store the probed position in the VAR file.
[11:20:19] <awallin_> you have to figure out the tool-length-compensation from that yourself, and then change the tool-length for the current tool with another g-code
[11:33:27] <sH`> anyone have a good source of carbide bits other than homedepot or flea-bay. ;q
[11:37:47] <pingufan> awallin_: Ok. So as the tools are almost of same length and have no stopper ring, I can set all tool lengths to 0 and only work with sensed corrections. right?
[11:39:53] <awallin_> pingufan: yes that would probably work..
[11:41:49] <awallin_> sH`:
http://www.maritool.com/ ? depends one what you are looking for I guess
[11:42:26] <pingufan> Nice! Is there a well working converting program (linux, if possible) for converting i.e. a dxf -> G-Codes? I currently test cut2d, but that is windows and not free.
[11:44:49] <Loetmichel> sH`: which country?
[11:44:56] <Loetmichel> and which size?
[11:45:00] <sH`> US.
[11:45:22] <sH`> just looking for 10x piece sets.
[11:45:51] <sH`> 3/32" 2.4mm
[11:45:53] <Loetmichel> for germany and the small 1/8" bits i would recommend
http://www.mmetoolshop.de
[11:46:02] <Loetmichel> sorry, no idea
[11:46:46] <sH`> or even 1.5mm
[11:46:47] <sH`> k.
[11:46:56] <sH`> no biggie, i'll dig around the interwebs.
[11:47:52] <Loetmichel> but i think the PCB manufacturers should know where to get the small milling/drilling bits in TC, 'cause they are usimg them like corn in a popcorn machine ;-)
[11:59:36] <Loetmichel> bad manners on criticvs and machines which can only mill swiss cheese, on parmesan it would bend ;-)
[12:00:51] <pingufan> Well, it's fine for me. Cheap tools. I had a look on his machines, they really don't appear strong for their size. My little Gravograph VX is definitely stronger (but also much smaller).
[12:01:08] <pingufan> ... nd has a weight of 50kg.
[12:01:43] <syyl_ws> gravograph is a well known manufacturer :)
[12:02:25] <Loetmichel> the boss of "heiz" has made a criminal complaint against ALL cometitors he was aware of for not having the german "safety features" like emergency sto switch and some manuals
[12:03:15] <pingufan> I got my mill there for 500 Euro. It had a totally burned out electronics. But the steppers were ok. Then I bought three gecko drives (thanks to cradek, this recommendation was golden), and now I have a perfect machine.
[12:03:17] <Loetmichel> and has reactet on critics on his construction in the forum with MORE complaints rather than answer the questions
[12:04:42] <pingufan> My mill also has no emergency stop. But it has a main switch. and this switch is on the machine, easy to reach. (But it's my machine and I don't sell it.)
[12:04:51] <pingufan> That's a bit different.
[12:04:55] <Loetmichel> to be precise: he has filed the complaints at the police
[12:05:35] <pingufan> Haha - Crazy. Well, I only buy my tools there, he is cheap.
[12:06:13] <pingufan> And I save costs for shipment, it is a drive of 15 minutes with my car.
[12:06:33] <Loetmichel> ... so he hasnt the best standing at cncecke.de ;-)
[12:07:06] <pingufan> Ok, back to my Gcodes. So after entering a G30.1 the current point is PERMANENTLY stored in vars?
[12:07:24] <Loetmichel> dont misiinterpret me,. i dont talk about your .at dealer, i talk about the manuifacturer of the "heiz" machines!
[12:07:34] <Loetmichel> high-Z
[12:07:56] <pingufan> Don't worry.
[12:09:02] <pingufan> Well, supper is ready. M;ust leave now. Have a nice weekend!
[12:15:47] <IchGuckLive> hi all around the world
[13:06:28] <IchGuckLive> skunkworks: are you online ?
[13:35:58] <mrsun> any circuits for PWM to -10 to +10V ? :)
[13:42:56] <pcw_home> There are a number of way to do it UP/DOWN, PWM/DIR, 50% = 0
[13:44:35] <mrsun> so UP and 0-10V, and DOWN and 0-10V?
[13:44:55] <mrsun> in that case foward/reverse run/stop two wire sequenc is about the same thing? :)
[13:45:16] <mrsun> ahh
[13:45:25] <pcw_home> up = 0 to +10V, down = 0 to -10V
[13:45:28] <mrsun> up down of the internal frequency ref ?
[13:45:42] <mrsun> Using the Up/Down function allows the frequency reference to be set by two push buttons. One digital input must be
[13:45:43] <mrsun> programmed as the Up input (H1-oo= 10) to increase the frequency reference, the other one must be programmed as the
[13:45:43] <mrsun> Down input (H1-oo= 11) to decrease the frequency reference
[13:45:59] <pcw_home> that is PWM feed to two pins depending on polarity
[13:46:41] <mrsun> but that would require using a closed loop i guess to use that kind of command that i pasted there
[13:46:50] <pcw_home> 50% = 0 is the simplest but really needs a good enable circuit otherwise it will come up full -10
[13:47:47] <mrsun> Example: Terminal A1 is set to supply the frequency reference, and the bias (H3-04) is set to -100%. The frequency reference
[13:47:47] <mrsun> can be set from 0 to 100% with an analog input of 5 to 10 V. With an input of 0 to 5 V, the frequency reference can be set
[13:47:47] <mrsun> from -100% to 0%. The drive reverses the motor rotation with negative input
[13:47:50] <Loetmichel> mrsun: i would go the easy way and make a RC-filter behind the PWM, then a operational amplifier to amplify the signal from 0-5V to 20V and shift it 10V "down"
[13:48:07] <Loetmichel> implies a negative and positive 12V to 15V supply
[13:48:11] <mrsun> now theres something, i guess that would be so ic an use 0-10V signal and still ahve forward/rev on the it by 0-5V (rev) 5-10V (forward)
[13:48:20] <pcw_home> Right thats the 50% = 0 option
[13:48:32] <mrsun> pcw_home, ahh =)
[13:49:04] <mrsun> or ill just use the start forward/start rev and a pwm 0-10V to set the speed
[13:50:11] <pcw_home> Ya, If you have a for/rev digital input I would use that and a unipolar 0-10V its simpler
[13:50:12] <mrsun> got the basic forward/rev, e-fault and VFD fault return board built up
[13:50:50] <mrsun> just would have to add a terminal to connect a low pass filter for the 0-10V, tho the 0-10V requires me to have a +12V supply somewhere in there also :P
[13:51:35] <pcw_home> PWM from parallel port?
[13:52:03] <mrsun> pcw_home, i guess, from emc
[13:52:13] <mrsun> got no mesa card :P
[13:53:02] <pcw_home> When we do this we first standardize the PWM with a CMOS gate running from a reference supply (3.3 or 5V)
[13:53:40] <mrsun> pcw_home, huh ? :)
[13:54:21] <Loetmichel> pcw_home: taht would be a good idea, also ;-)
[13:54:25] <pcw_home> (the parallel port signal is not stable voltage wise)
[13:54:45] <mrsun> pcw_home, got a breakout board that gives me 0 - +5V
[13:55:30] <pcw_home> OK then just a op amp X2
[13:55:49] <mrsun> low pass + opamp X2 then ? :)
[13:56:26] <pcw_home> breakout give you 0 to 5V analog or just 5V signal level?
[13:56:33] <pcw_home> gives
[13:56:57] <mrsun> gah things gets a bit strange to me with all this, the bob wants a constant +5V voltage applied to the things im driving then it pulls stuff low by shorting to gnd ... feels all backwards =)
[13:57:06] <mrsun> pcw_home, digital levels
[13:57:27] <mrsun> so the pwm will be at the output but 0-5V guaranteed =)
[13:58:21] <pcw_home> OK so low pass filter, maybe multiple pole then op amp (matched resistors, input to +)
[13:59:05] <mrsun> ofc, as i need a second supply i could make the voltage 0-12V using the opto isolator on the board for the VFD
[13:59:25] <mrsun> gah i suck at electrical design =)
[14:00:59] <pcw_home> Yep that how some of our output potentiometer things like the spindle output on our 7I47S work
[14:01:01] <pcw_home> OPTO --> CMOS BUFFER --> RAIL-RAIL OP AMP
[14:01:40] <pcw_home> (CMOS buffer because OPTO doesn't swing rail to rail)
[14:04:09] <mrsun> ahh might be true, depending on the voltage drop over the opto on the breakout board i guess i might not have a full +5V
[14:04:24] <mrsun> see, here my electrical engineering stupidity comes in again :P
[14:04:48] <pcw_home> also depends on how good you need it to be
[14:06:32] <mrsun> i guess as i have opto couplers already from the breakout board i could just do with a buffer for the vfd signals? :)
[14:07:37] <pcw_home> Does the VFD have a 10V (potentiometer top lead) output?
[14:07:45] <mrsun> pcw_home, yes
[14:07:50] <mrsun> 10.5V
[14:07:55] <mrsun> for the frequency ref stuff
[14:08:54] <pcw_home> so maybe the easiest thing would be to use the OPTO output, low pass filter it and run it from this 10.5V
[14:09:55] <pcw_home> bbl
[14:10:21] <mrsun> pcw_home, ahh you mean couple the 10.5V to the opto coupler, then low pass filter the output fromt he opto coupler for the analog input ?
[14:11:38] <jthornton> all that is left now is to clean out the boil pot and wait for the beer to make itself
[14:11:40] <pcw_home> yes 10.5 to pullup res to OPTO
[14:12:08] <pcw_home> slave driving those poor little yeasties
[14:12:23] <mrsun> pcw_home, pull up res ? to where? :)
[14:13:49] <mrsun> i wonder if the frequency ref common is the same common as everything else on the vfd
[14:14:10] <pcw_home> Well depending on polarity OPTO+ connects to 10.5V and OPTO- pulls up a resistor to ground
[14:15:25] <pcw_home> _or_ 10.5V connects to resistor, other resistor side connects to OPTO+, and OPTO- connect to ground
[14:16:05] <pcw_home> so you now have ~10.5V PWM from the OPTO output
[14:16:15] <mrsun> as ive got it now +10.5V connects to opto, output of opto connects to resistor and then a cap to GND and back to the vfd
[14:18:10] <pcw_home> Sounds right but the resistor needs to connect to ground and you probaby need a series resistor as well since you will have very unsymmetrical drive resistance
[14:18:39] <jthornton> you know the little buggers love it till they get drunk
[14:18:56] <pcw_home> (this is why we follow the OPTO with a 4000 series CMOS gate in our stuff)
[14:19:43] <pcw_home> then a series resistor after the gate for the first RC pole
[14:20:06] <pcw_home> really BBL
[14:20:17] <mrsun> http://i44.tinypic.com/2hf2b8p.png thats how ive done it =)
[14:20:22] <mrsun> pcw_home, just go =)
[14:20:26] <mrsun> im just happy i get help :P
[14:50:22] <PCW> mrsun: a 1K or so resistor from pin 9 of the PC847 to ground would probably make that work a lot better
[14:51:17] <mrsun> PCW, before or after the resistor to the cap ? :)
[14:54:11] <PCW> before
[14:54:23] <mrsun> oki =)
[14:55:26] <PCW> makes the output more of a variable voltage source instead of variable current source
[14:56:39] <mrsun> ahh ok =)
[14:59:04] <PCW> looks like about 20 mA of LED drive on the OPTOs and 50% minimum CTR so 1K should just work with the lousiest OPTO
[15:02:19] <mrsun> PCW, the LED drive current is calculated to 20mA yes, 1.2V 20mA
[15:02:44] <mrsun> tho they can take 50mA
[15:03:22] <mrsun> PCW, CTR?
[15:03:42] <PCW> Current Transfer Ratio
[15:04:45] <PCW> so 50% CTR and 20 mA in means ~10 mA out on the worst possible optocoupler (when new)
[15:07:39] <mrsun> ahh ok =)
[15:07:56] <PCW> If this was not a one-off I would use a 1.5K K pulldown so the worst possible OPTO could tolerate a 33% loss of CTR over time and still work
[15:07:58] <PCW> but its extremely likely that the CTR in the OPTO you have is better than 200%
[15:08:35] <PCW> (since its somewhere between 50% and 600%)
[15:11:53] <PCW> This is much more fun than figuring out why my FPGA processor code for DMA packet reception from Ethernet
[15:11:55] <PCW> does not work but alas I must return to banging my head against the wall...
[15:11:56] <PCW> bbl
[15:12:13] <mrsun> hehe =)
[15:12:44] <mrsun> but the A1 terminal should be high impedance right ?
[15:12:53] <mrsun> analog input terminal
[15:13:00] <PCW> I swear I make 2 mistakes in eveyy line of code
[15:13:01] <mrsun> so it shouldnt draw like any power ? :)
[15:13:13] <mrsun> PCW, haha days like that can just go away imo =)
[15:13:53] <PCW> Yes it normally should be high impedance (so potentiometers are reasonably linear)
[15:19:32] <mrsun> so now those parts are done, putting vfd into source mode, connect the ouputs to the optos outputside inputs and let them go to vfd gnd, let VFD 10.5V go to opto coupler output side and low pass filter that and pass it back, now its just to get the relay to talk to me =)
[15:19:55] <mrsun> i guess i could use the NC side of the relay as a "READY TO RUN" signal and the NO side as FAULT ...
[15:20:10] <mrsun> or ill just go for getting the fault side
[15:20:19] <mrsun> dont want to hog all the inputs i have just for the vfd =)
[15:24:16] <mrsun> hopefully i wont burn stuff up :P
[15:24:32] <mrsun> quite expensive stuff to play with when you dont know what the hell you are doing :P
[15:35:01] <Evertick> Howdy@
[15:35:04] <Evertick> !
[15:36:50] <Evertick> I'm having some trouble figuring out ngcgui. Can anyone help?
[15:38:50] <mrsun> PCW, grounding of signal cable, should it be done in one end or two ends? :)
[15:43:19] <PCW> the analog? Both ends but just to the local GND of your PWM-potentiometer simulator
[15:43:31] <PCW> at the breakout end
[15:44:18] <mrsun> PCW, this board will be at the vfd unit and i will ahve a signal cable from the pc to it
[15:48:24] <mrsun> PCW, the 1k was that supposed to go to the analog input gnd terminal ?
[15:48:30] <mrsun> =)
[15:49:13] <Jymmm> cradek:
http://i41.tinypic.com/a9web8.jpg
[15:50:52] <cradek> meh, 1 should have been overbar(.9)
[15:51:31] <DJ9DJ> hmm, looks interesting
[15:51:59] <cradek> cute but not the greatest implementation
[15:52:20] <cradek> in particular the mishmash of different kinds of minus signs - poor typesetting
[15:53:05] <Jymmm> cradek: I thought you might enjoy it is all =)
[15:53:26] <PCW> mrsun yes
[15:53:27] <PCW> when you are done and before you connect to VFD you should have a isolated three terminal device like a potentiometer (V+ wiper GND)
[15:53:29] <PCW> and you should be able to test the voltage output with any handy 5- 12V power supply
[15:53:45] <cradek> and the 5 is just wrong!
[15:53:57] <mrsun> PCW, ahh ok =)
[15:54:04] <Thetawaves> Mach3 configuration guide for this 3 axis machine specifies 800 steps per (what i believe to be mm). These screws are 2 start 4mm pitch lead screws. SO 800 = x * (1/4) ... x = 3200 steps per rev
[15:54:16] <mrsun> PCW, yeah as it is now (ive already changed it) ive got 3 from eachother isolated terminals for the functions
[15:54:19] <cradek> Jymmm: I did, but my inner geek thinks it wasn't done by a math person :-)
[15:54:30] <Thetawaves> does my math add up?
[15:54:49] <mrsun> one for the analog (A+, Wiper, GND) one for digital inputs (S1, S2, S3 and common) and one for the relay output back to the pc
[15:54:55] <Jymmm> cradek: I'd love to see how you've improved upon it.
[15:54:58] <Thetawaves> 3200 steps / rev seems high...
[15:57:00] <mrsun> and i think ive managed to turn every single screw terminal upside down
[15:57:06] <mrsun> or back to front
[16:03:41] <mrsun> oh well, fixed =)
[16:06:00] <Jymmm> http://i39.tinypic.com/142fmgh.jpg
[16:11:59] <raynerd> Hi guys, not stricktly linuxcnc but general cnc. I`m converting one of these TEP machines as someone else is on here. I need to add a z axis. The machine is only tiny and I was just wondering what peoples opinions were on using something like a vertical slide and adapting it for the z axis:
http://www.ebay.co.uk/itm/RDGTOOLS-SMALL-VERTICAL-SLIDE-MINI-LATHES-TAIG-CONQUEST-CLARKE-/29065604275
[16:11:59] <raynerd> 6?pt=UK_Home_Garden_PowerTools_SM&hash=item43ac733f04
[16:13:58] <mrsun> got to love clients that cuts lines into half :P
[16:15:35] <raynerd> or no use?
[16:16:16] <raynerd> If my machine was bigger I`d make one, but I don`t know how to get it all so compact and just thought a prebuilt slide would work OK.
[16:23:21] <mrsun> gonna make this card tomorrow and see what happends i think =)
[16:23:26] <mrsun> thanks alot for the help PCW =)
[16:23:54] <PCW> Thank me if it works :-)
[16:28:41] <mrsun> PCW, hehe =)
[16:28:45] <mrsun> well you cant have made it much worse =)
[16:29:10] <mrsun> hopefully i wont burn out the breakoutboard and the parport :P
[16:42:50] <Jymmm> http://forgifs.com/gallery/d/198097-1/Rims-they-see-me-rollin.gif
[16:52:04] <PCW> and... I have packet data!
[16:56:41] <raynerd> Anyone about?
[16:57:26] <raynerd> I`m just about to order some parts to make my z axis. My travel on the x and y axis is only 13" in both directions... what would you be looking for in z axis on such a small machine?
[16:57:42] <raynerd> I thought about 4" ?
[16:58:53] <raynerd> But to get this, it means I have to have the y axis suspended a miniumum of 4" above the bed... on such a small machine I`m just wondering if this is all then getting a bit high
[17:01:42] <DJ9DJ> gn8
[17:01:48] <raynerd> gn8
[17:32:32] <Jymmm> Say Awwwwww (and even the cat with cabbage on it's head is cool)
http://www.cutestpaw.com/articles/best-of-the-week-cute-animal-pictures-1/
[17:54:08] <Jymmm> While the pipe are starburst candy, da shit inside looks legit! lol
http://i42.tinypic.com/17e3xd.jpg
[18:01:44] <alex4nder> woot
[18:01:56] <alex4nder> my taig did its first "professional" milling job
[18:02:31] <cstop> subdued clapping from near and far
[18:04:04] <Thetawaves> i told the machine to move 5mm but instead it moves 2.5mm
[18:04:44] <pfred1> your axis scale correct?
[18:05:47] <Thetawaves> axis scale?
[18:06:03] <pfred1> it is on your machine's conf file
[18:06:10] <pfred1> in not on
[18:06:26] <Jymmm> "I don't want to hide anymore." - Waldo
[18:06:37] <Thetawaves> i was using stepconf...
[18:06:58] <pfred1> yes stepconf makes up axis scales
[18:07:04] <Thetawaves> where should i look for this file?
[18:07:22] <pfred1> in your EMC directory?
[18:07:33] <pfred1> you can change it in stepconf too
[18:08:26] <Thetawaves> i see a textual description of the scale setting but no actual place to set it
[18:08:27] <pfred1> it is a sum of your ratio your lead steps per revolution and your step mode
[18:08:53] <Thetawaves> ahhhhh
[18:09:26] <pfred1> basically it is how many step pulses your machine thnks it needs to move a unit measure
[18:09:35] <pfred1> yours sounds about half off to me
[18:14:53] * pfred1 just loaded up drill.py here
[18:15:08] <pfred1> I need to learn Python
[18:15:41] <Tom_itx> i need to learn it a bit too
[18:16:25] <pfred1> looking at the source it is a rather introspective language
[18:16:40] <pfred1> seems to keep on mentioning "self" an awful lot
[18:17:13] <pfred1> http://wiki.linuxcnc.org/uploads/drill.py
[18:17:26] <Thetawaves> self is similar to this in C++
[18:18:07] <pfred1> http://www.douglas-self.com/
[18:19:01] <pfred1> any algebra wizards here?
[18:20:08] <pfred1> I have this equation π X Dia. X RPM / 12 = SFPM
[18:20:43] <pfred1> but I usually know what SFPM I want and what Dia. I am running too so I usually need to solve it for RPM
[18:20:46] <Thetawaves> something is weird about that equation
[18:21:39] <pfred1> so I made up this equation SFPM / (Ï€ X (Dia. / 12)) = RPM
[18:21:51] <pfred1> it works but it is a pain to calculate
[18:21:55] <Thetawaves> what is x?
[18:22:03] <pfred1> times
[18:22:38] <Thetawaves> SFPM / (* (Dia. / 12)) = RPM?
[18:22:45] <Thetawaves> seems i am missing a symbol
[18:22:58] <pfred1> oh you can't see π ?
[18:23:27] <Thetawaves> you should use utf8 as your character set
[18:23:39] <alex4nder> looks good here
[18:23:41] <pfred1> I am
[18:24:07] <pfred1> alex4nder you can see the PI symbol?
[18:24:43] <alex4nder> yah
[18:24:43] <Thetawaves> pfred1, here is a simpler equation
[18:24:50] <Thetawaves> there*
[18:24:54] <pfred1> Thetawaves that is what I was thinking
[18:25:31] <pfred1> PI X Dia. X RPM / 12 = SFPM
[18:25:38] <Thetawaves> cutting speed * 4 / D
[18:25:47] <pfred1> no that is inaccurate
[18:25:59] <Thetawaves> ohhh
[18:26:12] <pfred1> my mill is pretty picky about its surface speeds
[18:26:37] <Tom_itx> is it tough to scan a file for a string in python?
[18:27:12] <pfred1> seems more a perlish thing to do
[18:27:20] <Thetawaves> load the file into a string and do a string search?
[18:28:47] <pfred1> didn't Nebraska try to legislate PI = 4 once?
[18:29:03] <pfred1> one of them backwards western states
[18:29:14] <Thetawaves> rpm = (sfpm * 12) / (dia * pi) ?
[18:29:37] <pfred1> http://en.wikipedia.org/wiki/Indiana_Pi_Bill
[18:32:04] <Thetawaves> or really rpm = sfpm * (12/ (dia * pi))
[18:32:41] <pfred1> Thetawaves that would be a lot easier to solve
[18:33:12] <Thetawaves> or actually rpm = sfpm * ( 3.819 / dia)
[18:33:26] <Thetawaves> calculate the 3.819 number to your desired descimal places via 12/pi
[18:34:41] <Thetawaves> run your numbers and let me know if that actually works
[18:34:49] * Thetawaves *thinks* he can do algebra
[18:35:03] <pfred1> yeah that was what I had to do for the formula I made
[18:35:46] <pfred1> I know PI X DIA. X RPM / 12 = SFPM works
[18:36:08] <Thetawaves> any of those three i supplied should also work
[18:36:24] <pfred1> so when my other formula joved with its results i knew it worked but it seems it could be simplified to me
[18:37:07] <pfred1> I'll try rpm = sfpm * ( 3.819 / dia)
[18:37:57] <AitalMAC> It's possible to set a spindle with PWM or PDM with acceleration and decelleration?
[18:39:39] <Thetawaves> pfred1, success?
[18:40:55] <pfred1> Thetawaves donno haven't tried it yet will soon
[18:42:12] <pfred1> Thetawaves it is pretty close
[18:42:37] <pfred1> just the difference in how many decimal places 3.819 is
[18:42:43] <Thetawaves> add some more digits to your 3.819
[18:42:55] <pfred1> Thetawaves then it gets harder to use
[18:43:06] <Thetawaves> it would be handy to keep that in one of the memory slots on your calculator
[18:43:50] <Thetawaves> just remember to save 12/pi before you start your session :)
[18:44:18] <pfred1> I got 1739.6726
[18:44:35] <pfred1> when the actual value was 1740
[18:46:40] <Thetawaves> nahhhh
[18:46:45] <pfred1> now I get 1739.9999 if I do 12/PI first
[18:46:47] <Thetawaves> the 'actual value' is some rounding
[18:47:06] <Thetawaves> 1739.6726???? pfffff 1740 good enough
[18:47:10] <pfred1> pft I never get the exact RPM out of my mill anyways
[18:47:12] <Thetawaves> :)
[18:47:58] <pfred1> I used to hunt wiht the different speeds to figure out which one I should run was a pain
[18:48:55] <pfred1> then one day I had the bright idea to write the speeds into my endmill box
[18:49:02] <pfred1> that saved me a lot of time
[18:49:17] <Thetawaves> if you only cut one material?
[18:49:26] <pfred1> but not every tool I own is in a box and yes that is the other issue too
[18:49:40] <pfred1> mostly it matters when I mill steel
[18:49:58] <pfred1> other stuff I can be off on the speed and it isn't the end of the world or the tool I'm using
[18:50:04] <Thetawaves> so you only have the calcs for steel?
[18:50:24] <Thetawaves> thanks for the tip
[18:50:45] <pfred1> well if you have the steel speeds yo ucan sort of figure for softer materials
[18:51:00] <pfred1> figure double or triple it depending
[18:51:41] <pfred1> softer than steel I think the general rule is if there ain't fire gouting out you're good to go
[18:52:11] <Jymmm> This is kinda creepy
http://i42.tinypic.com/2rwvzeq.jpg
[18:52:12] <Jymmm> (if it's true that is)
[18:52:31] <Thetawaves> i bet my microstepping setting is messing up my axis scale
[18:52:39] <Thetawaves> test axis functions perfectly
[18:52:47] <pfred1> Thetawaves it is one of the factors for your axis scale figure
[18:53:05] <pfred1> if you double it you should be spot on
[18:53:19] <pfred1> your 2.5 will turn into a 5
[18:53:24] <Thetawaves> i have it set to 1
[18:53:26] <Thetawaves> should set to 0
[18:53:33] <pfred1> so you're half stepping then?
[18:53:47] <pfred1> 1 is full stepping
[18:54:04] <pfred1> I don't think there is a 0
[18:54:33] <Thetawaves> i don't really know what i should set there, i don't think my driver does any microstepping
[18:54:34] <pfred1> 2 is half stepping 4 quad etc
[18:54:51] <pfred1> full stepping is prety horrible
[18:55:09] <Thetawaves> so set it to 2?
[18:55:22] <pfred1> I've run in full step mode it is how can I put it? limiting
[18:55:24] <pfred1> yes
[18:55:40] <pfred1> half stepping is about the minimum i can stomach
[18:56:52] <pfred1> Thetawaves what kind of driver do you have?
[18:58:45] <Thetawaves> noname chinese
[18:58:56] <pfred1> what IC is it based on?
[18:59:05] <Thetawaves> i'll have to open it up tonight
[18:59:25] <pfred1> it doesn't even say what it is?
[18:59:37] <pfred1> is it a TB6560?
[18:59:37] <Thetawaves> i set it up to do half stepping, now the test area definitely moves 8mm in total when i told it 4mm
[18:59:50] <Thetawaves> i have a feeling emc2 will move 5mm when i tell it to now however
[19:00:24] <Thetawaves> there are no numbers on the outside of the box
[19:00:40] <pfred1> got a link with a pic of the board?
[19:01:06] <pfred1> thoug hthe old SLAs and the TB do kind of look alike from a distance
[19:01:22] <pfred1> and the old toshibas look like the new one too
[19:02:40] <pfred1> if it is a cheapie multi axis Chinese board odds are it is a TB based driver
[19:03:16] <pfred1> lets hope they didn't saddle you with an old TA board
[19:03:32] <pfred1> they're still floating around though
[19:05:10] <Thetawaves> http://www.amonstar.com/template/file/3020/desktop-CNC%20user%20manual.doc
[19:05:18] <Thetawaves> thats got some detail pictures
[19:05:30] <Thetawaves> a ton of the connectors are different though
[19:05:54] <Thetawaves> i'll open it up now
[19:06:23] <pfred1> your machine came with Mach3?
[19:07:13] <pfred1> hmm yeah they're not being entirely forthcoming about what exact drivers you have
[19:07:21] <pfred1> they might be old TAs
[19:07:56] <pfred1> which with the right freewhelling diodes I guess there is nothing wrong with them
[19:08:52] <Thetawaves> i took the cover off and the main ICs are on the bottom of the board
[19:08:59] <Thetawaves> only the optos on top
[19:09:34] <pfred1> I wonder what this one chinese character means?
[19:09:56] <pfred1> it probably means i run to casino with roud eye's money or something
[19:10:43] <pfred1> Thetawaves the mach3 your machine came with is the limited trial version?
[19:10:53] <Thetawaves> no idea
[19:10:58] <Thetawaves> i assume it is a hacked version
[19:11:00] <pfred1> you never ran it?
[19:11:06] <Thetawaves> i don't own windows
[19:11:07] <pfred1> that was just what I was thinking
[19:11:32] <pfred1> if it came from China they have some kind of a moral obligation to crack western commercial software or something
[19:12:13] <Thetawaves> it would make the machine pretty much useless to a huge market
[19:12:18] <Thetawaves> without mach3
[19:12:23] <pfred1> everyone is all hot to trot on mach land because mach doesn't support win7 or some such nonsense
[19:13:14] <pfred1> closed source makes about as much sense as dinosaurs staring at comets
[19:13:24] <Thetawaves> pfred1, you want me to send you the exe?
[19:13:30] <pfred1> in the long run it is about as survivable too
[19:13:30] <Thetawaves> beware, there be sharks
[19:13:46] <pfred1> I can't DCC firewalled
[19:14:18] <Thetawaves> you probably don't want to run shady chinese exes anyway
[19:14:26] <pfred1> ya gotta say the chinese language sure is pretty in print
[19:14:50] <pfred1> I seen pics of their small keyboards though
[19:15:03] <pfred1> you know the ones with only about 2,500 keys?
[19:15:21] <pfred1> gotta be murder to touch type on
[19:15:50] <Thetawaves> no
[19:16:08] <Thetawaves> each character is like 3 keys pressed together
[19:16:13] <Thetawaves> or something
[19:16:21] <Thetawaves> there is an encoding scheme
[19:16:32] <pfred1> we got nothing to worry about from over there!
[19:16:59] <pfred1> least ways not until they give up that chicken scratch they pass off as text
[19:18:07] <pfred1> their Internet is freaky too you ever check it out?
[19:18:50] <Thetawaves> no
[19:18:51] <pfred1> not to say that our Internet is wrapped too tightly but theirs is really bizarre
[19:20:04] <pfred1> lets just say that as a culture they fixate an awful lot
[19:21:54] <pfred1> if one Chinese person is into something you can bank on a half a billion of them being into the same thing
[19:22:49] <pfred1> and when they go in they go all in
[19:23:41] * pfred1 wonders what the Chinese translation for rugged individualist is ....
[19:24:22] <pfred1> probably political prisoner or something to that effect
[19:28:17] <pfred1> Chinglish is a riot!
[19:28:25] <pfred1> Notice:
[19:28:32] <pfred1> 2. You’d better use the PC to control the machine, especially for this machine and do not install other software if conditions permit.
[19:28:54] <pfred1> Valen you hear that? You'd better!
[19:30:02] <pfred1> what scares me is when I read too much of it and it starts making sense
[19:34:28] <pfred1> Thetawaves EMC2 is listed as acceptable software for your machine
[19:37:17] <pfred1> Thetawaves all your motor settings are on page 14
[19:43:15] <Valen> if its a chinese controller they are probably wrong
[19:43:32] <pfred1> Valen come on they got it to work
[19:43:40] <pfred1> http://www.newegg.com/Product/Product.aspx?Item=N82E16856119041
[19:43:44] <pfred1> hawt!
[19:43:45] <Valen> there are like 3 different similar controllers that most of the chinese "mills" use
[19:44:05] <Valen> and they all copy the instructions from some other different controller that has a different pinout
[19:44:12] <pfred1> heh
[19:44:19] <pfred1> this was a software setup page
[19:44:24] <Valen> not that i'm generalising
[19:44:39] <Valen> yeah they have info to put into stepconf, thats just wrong info
[19:44:44] <Valen> but good luck
[19:44:47] <pfred1> Foxconn SFF R50-D4 Intel Atom D525 (1.8GHz, Dual-core) Intel NM10 Intel GMA 3150 Barebone $109.99
[19:44:48] <Valen> you might get lucky
[19:45:08] <Valen> i want quad core atoms!
[19:45:28] <pfred1> how come?
[19:45:41] <pfred1> one core for OS one for RTAI
[19:45:55] <Valen> more cores for OS
[19:45:58] <Valen> gets kinda laggy
[19:45:59] <pfred1> why?
[19:46:09] <pfred1> shouldn't
[19:46:22] <pfred1> what are you doing ripping movies while you're cutting?
[19:46:23] <Valen> webcam often uses 100% cpu
[19:46:40] <pfred1> why are you running a webcam with your CONC controller?
[19:46:45] <pfred1> CNC controller even
[19:47:18] <Valen> alignment camera
[19:48:14] <Valen> http://www.cnczone.com/forums/attachment.php?attachmentid=143600&d=1318136270
[19:48:25] <Valen> from
http://www.cnczone.com/forums/1004601-post29.html
[19:48:33] <Valen> the camera is off to the left of that image
[19:48:52] <pfred1> too advanced for me I just use stop blocks
[19:48:56] <Valen> and here it is integrated into emc
http://www.cnczone.com/forums/attachment.php?attachmentid=143603&d=1318136270
[19:49:00] <Valen> tis easy
[19:49:55] <Valen> but yeah, more cores would be handy when your doing stuff other than cutting
[19:50:09] <pfred1> that is what I have other PCs for
[19:50:38] <pfred1> my CNC box is dedicated to task
[19:50:55] <Valen> i use it for everything in the shop except the radio ;->
[19:51:02] * pfred1 pulled it out of that trash heap special!
[19:51:20] <pfred1> nope i have a separate PC in my garage for other nonsense
[19:51:52] <Valen> i often use stuff i find on the interweb for cutting with ;->
[19:51:57] <Valen> looking up feeds n speeds
[19:52:53] <pfred1> I just got a Starret catalog in a thrift store has spme feeds and speed tables in it
[19:53:09] <pfred1> cost me a quarter
[19:54:05] <pfred1> I just bought a tool last weekend I always wanted
[19:54:14] <pfred1> never thought I'd ever get one either
[19:55:32] <Valen> hrm?
[19:55:48] <pfred1> I got a left handed circular saw
[19:56:07] <pfred1> they're pretty rare you know?
[19:56:22] <pfred1> I never could justify what they cost either
[19:56:29] <pfred1> even though I'm left handed
[19:56:52] <pfred1> I got to admit is is kind of weird to run a circular saw wiht my left hand now
[19:57:06] <pfred1> I been doing it right handed for so long
[19:57:24] <Valen> lol
[19:57:25] * pfred1 can't mouse with his left hand either ....
[19:58:34] <pfred1> years ago I was like I'm going to learn right handed because there is no way I'm going to be moving mice for the rest of my life
[19:59:57] <pfred1> I think it is because of that that I can't draw on computers though
[20:00:36] <pfred1> and trying to mouse with my left hand feels positively perverse to me today
[20:00:55] * pfred1 has a Wacom though ...
[20:01:22] <pfred1> but it's kind of strange too
[20:02:09] <pfred1> I almost got it all setup once in Linux with the pressure sensitivity and everything
[20:03:21] <pfred1> I don't think I ever quite got the eraser to work right though some part of it I couldn't manage to get 100% right
[20:16:01] <blockbuster> what's a good cnc milling software?
[20:16:37] <pfred1> blockbuster LinuxCNC
[20:16:54] <blockbuster> i don't have a linux.. is there a windows base software?
[20:17:02] <blockbuster> what's a good windows based software?
[20:17:04] <pfred1> blockbuster Linux is free
[20:17:18] <blockbuster> is there a link you can provide?
[20:17:31] <pfred1> http://linuxcnc.org/
[20:17:49] <blockbuster> can it do 4th axis rotary?
[20:17:50] <pfred1> when you download the CD image it comes with the OS
[20:19:43] <pfred1> blockbuster you mean like this?
http://www.youtube.com/watch?v=9Ut4fG6lO80
[20:19:58] <blockbuster> yes
[20:21:31] <blockbuster> can it do 4th axis ??
[20:22:51] <pfred1> blockbuster you just watched a video what do you think?
[20:23:07] <blockbuster> i like it.. i wan to do something like that. can the software do that?
[20:23:35] <blockbuster> i need a software that converts my stl files to gcode.
[20:24:02] <pfred1> blockbuster why stop at just 4 axises?
http://www.youtube.com/watch?v=7WCrqqoZkPg
[20:24:17] <blockbuster> i only have a 4 axis cnc mill
[20:40:14] <joe9> I love this idea:
http://www.instructables.com/id/Accurized-Craftsman-Drill-Press/ , but my drill press quill does not have the metal bracket around it.
[20:40:33] <joe9> Any suggestions on how I can add a vernier calipers to a drill press.
[20:41:11] <joe9> http://www.harborfreight.com/heavy-duty-16-speed-bench-drill-press-38142.html is my drill press.
[20:41:35] <joe9> I am thinking I could attach a round bracket or something like that to the quill.
[20:41:44] <pfred1> this is my drill press
http://img535.imageshack.us/img535/8026/millpic.jpg
[20:42:19] <joe9> pfred1: it does seem to have something there.
[20:42:27] <joe9> did you put a calipers too?
[20:42:35] <pfred1> no
[20:42:35] <joe9> or, is it just the depth guage?
[20:43:00] <pfred1> yes if i need to bore a hole to a depth I just take the piece out and measure it
[20:43:12] <pfred1> it doesn't really come up that often
[20:44:44] <pfred1> although if i could get glass scales for the X and Y table I'd do that
[20:45:03] <joe9> pfred1: quick question, Is it possible to get a O ring shaped bracket that I can attach to the quill?
[20:45:16] <joe9> at Home depot ?
[20:45:42] <pfred1> maybe I don't know
[20:45:42] <joe9> I am told that the machinist could help with it, but, I want to see if I could do it with just the home depot stuff.
[20:46:11] <pfred1> lowes seems to have more stuff than home depot
[20:48:44] <pfred1> joe9 this is my secret for how I can unload a piece
http://www.instructables.com/id/Vise-Stop/
[20:49:07] <pfred1> pop it out measure pop it right back in
[20:50:07] <Thetawaves> what should i do as a first cut?
[20:50:17] <pfred1> Thetawaves air
[20:50:26] <Thetawaves> of what?
[20:50:34] <pfred1> lots do the logo
[20:50:35] <joe9> pfred1: how does it work? I could not understand it much from the article. let me google "vise stop"
[20:51:05] <pfred1> joe9 you load the piece into the vise bring the stop up to it and lock it down
[20:51:20] <pfred1> then when you open the vise you can remove the piece and put it right back where it was
[20:51:49] <pfred1> as in you'll be drilling the same hole if you are drilling a hole
[20:53:27] <pfred1> Thetawaves then when you're done cutting air the next most popular thing is to load a pencil as a tool and let it draw on a piece of paper
[20:53:41] <pfred1> pencils are cheap enough
[20:54:44] <joe9> pfred1: why not use parallels?
[20:55:00] <pfred1> joe9 I do
[20:55:07] <pfred1> they go under the work though
[20:55:25] <pfred1> they don't really align it side to side
[20:55:56] <joe9> oh, ok. vise stop is for side-to-side alignment.
[20:55:59] <joe9> got it.
[20:56:05] <Thetawaves> pfred1, so i've got my z axis setup -40 to 0
[20:56:29] <Thetawaves> emc logo claims to go outside the bounds of my z axis
[20:56:32] <pfred1> why not -2 to +2 ?
[20:56:42] <pfred1> yeah that'll happen
[20:57:02] <pfred1> that took me a bit to figure out
[20:57:07] <Thetawaves> so all my thing should be setup with +-
[20:57:14] <Thetawaves> s/thing/axis/
[20:57:21] <pfred1> well the Z axis for the logo
[20:57:52] <Thetawaves> ok
[20:58:16] <pfred1> in the comments it gives some hints to get it to run
[20:58:23] <pfred1> I forget what it is now
[20:59:41] <pfred1> joe9 lots of people make vise stops that screw right to their vises
[20:59:54] <pfred1> I didn't feel like drilling a hole in mine though
[21:00:04] <joe9> i am searching to see if I can buy it from HF or HomeDepot.
[21:00:10] <pfred1> not because it is a great vise but I just didn't feel like getting another vise out to hold it in
[21:00:43] <joe9> i am thinking of something that I can clamp onto the vise.
[21:00:59] <pfred1> yup people og that route too sort of like a custom C clamp
[21:01:04] <joe9> I do not think I can get a good hole in my vise.
[21:01:24] <pfred1> there is on on instructables
[21:01:36] <pfred1> http://www.instructables.com/id/Machinist-Vise-Stop-Clamps-on-jaw-/
[21:01:39] <pfred1> that idiot
[21:02:16] <pfred1> with his precision vise stop woo woo
[21:02:33] <pfred1> you cna jamb a rock into a tee slot and it'd work
[21:03:51] <Tom_itx> use a dowel pin in a tool holder
[21:03:56] <pfred1> but yeah work stops are really handy now that i have one I use it a lot
[21:04:08] <Tom_itx> use allthread off the end of the vise
[21:04:15] <Tom_itx> there are dozens of ways to make a stop
[21:04:24] <pfred1> and a million uses too
[21:04:43] <pfred1> it really ups the repeatibility game
[21:05:09] <Connor> http://www.cnczone.com/forums/1043298-post104.html
[21:05:17] <Connor> My End stop I made using 80/20.
[21:05:39] <pfred1> ah the third word for cnczone
[21:05:47] <pfred1> after mach3 and Gecko
[21:06:06] <Connor> Eh?
[21:06:26] <pfred1> you can basicilly sum that entire site up with those 3 words
[21:06:37] <Connor> I don't think so.
[21:06:43] <Connor> Lots of good info on that site.
[21:08:18] <pfred1> cnczone emc2 About 34,400 results cnczone mach3 About 204,000 results
[21:09:21] <pfred1> I'm surprised it is so even
[21:10:37] <pfred1> it doesn't seem like mach3 is much longer for this world anyways
[21:11:02] <pfred1> XP licenses are getting harder to come by :)
[21:12:35] <pfred1> hey i thought this site got taken offline?
http://www.rapidsharedata.com/tag/cnc+zone+mach3+crack/
[21:13:43] <pfred1> check it out someone has posted the EMC2 crack!
http://www.rapidsharedata.com/tag/EMC2+crack/
[21:14:03] <pfred1> mad hacking skillz
[21:20:09] <joe9> any suggestions on where I can find a round clamp that I can attach to the quill of a drill press?
[21:20:20] <Tom_itx> make one?
[21:20:34] <joe9> I am planning to use that clamp to either attach the jaw of a calipers or a dial indicator
[21:21:39] <joe9> Tom_itx: I am a noob. I doubt I can make one that is flawless.
[21:21:41] <pfred1> joe9 I made this to hold a little router
http://i.imgur.com/7u0zS.jpg
[21:21:49] <pfred1> drilling that hole wasn't easy!
[21:22:28] <pfred1> I used an adjustable spade bit
[21:23:30] <pfred1> I made it so I could fit that tool into the column
http://i.imgur.com/zWNB9.jpg
[21:23:43] <pfred1> almost like it was made for it huh?
[21:24:39] <tehDarkAura> hardware?
[21:24:54] <pfred1> no, software
[21:25:20] <tehDarkAura> can someone point me to some prefered hardware setups
[21:25:23] <tehDarkAura> ;)
[21:25:38] <pfred1> for LinuxCNC?
[21:26:02] <tehDarkAura> yeah to drive the steppers etc
[21:26:43] <pfred1> the really hardcore here seem to lean towards servos
[21:27:13] <tehDarkAura> ahh wow
[21:27:26] <pfred1> well I run open ended steppers
[21:27:49] * pfred1 dreams of someday running servos but who knows ...
[21:28:31] <pfred1> my drivers just have step and direction inputs and it was easy to get it to work with LinuxCNC
[21:28:41] <pfred1> I mean even I could do it!
[21:29:28] <tehDarkAura> hehehe
[21:29:36] <pfred1> tehDarkAura
http://www.youtube.com/watch?v=cgbeyNNBZ68
[21:30:38] <pfred1> you can get drivers similar to what I have for about $22 USD for 3 axis boards
[21:30:40] <tehDarkAura> very cool :)
[21:31:27] <pfred1> but if you want higher performance you have to go for nicer more expensive hardware
[21:31:38] <tehDarkAura> ahhh I see
[21:31:49] <tehDarkAura> can I bug you for a link or name?
[21:31:54] <pfred1> that'd set you back about $150 or more
[21:32:23] <tehDarkAura> even that isnt too bad -- rather start small and move if i identify a need though
[21:32:43] <pfred1> well the more expensive drivers hold their value better
[21:32:54] <tehDarkAura> ahh okay
[21:33:09] <pfred1> the cheap ones are hardly worth their weight to ship around
[21:33:40] <tehDarkAura> oh i just realised the title of the vid is a driver ;)
[21:34:07] <pfred1> well the main driver IC
[21:34:11] * tehDarkAura brews more coffee
[21:34:19] <pfred1> that is a TB6560AHQ based motor driver
[21:34:36] <pfred1> they're about as cheap as it gets well you can go al itle cheaper
[21:34:40] <pfred1> but not much
[21:37:33] <pfred1> dang G540s are that expensive?
[21:48:34] <tehDarkAura> one on ebay with 4 motors for 650 :P
[21:48:50] <tehDarkAura> looks like nice motors though
[21:49:38] <pfred1> I bought surplus motors and they were $13 a piece 200 oz/in bipolar
[21:49:51] <tehDarkAura> ohh nice
[21:50:12] <pfred1> ah most would consider them underpowered for a homebrew CNC router
[21:50:15] <tehDarkAura> what V?
[21:50:20] <pfred1> so I figured I'd use 2 an axis
[21:50:29] <tehDarkAura> makes sense
[21:50:40] <tehDarkAura> nema23?
[21:50:43] <pfred1> cheaper than one big powerful motor
[21:50:52] <pfred1> yeah twin stack nema 23s
[21:51:00] <tehDarkAura> sweet
[21:51:45] <tehDarkAura> you use acme threaded rod or V cut?
[21:51:52] <pfred1> tehDarkAura watch this
http://www.youtube.com/watch?v=fHPKaHLzXes
[21:51:58] <tehDarkAura> kk
[21:53:04] <tehDarkAura> ahhh
[21:53:18] <tehDarkAura> looks like it goes pretty fast
[21:53:22] <pfred1> I raised my microstep mode and got it to run faster never made a video though
[21:53:44] <pfred1> I think i made it to 1.5 or 1.7 IPS
[21:53:50] <tehDarkAura> nice!
[21:54:17] <pfred1> yeah not too shabby for 24VDC power in I guess
[21:54:29] <pfred1> remember you have to power this stuff
[21:54:35] <pfred1> more power usually costs more money
[21:54:40] <tehDarkAura> yeah im in no hurry -- if i can get my pcbs made here instead of shipping to china ill be happy :)
[21:55:04] <pfred1> high perf steppers run 48VDC and up
[21:55:11] <tehDarkAura> ahhh wow
[21:55:25] <pfred1> so I'm firmly entrenched in the middling performance category presently
[21:55:53] <pfred1> the way i look at it is I have lots of room for improvement
[21:56:06] <tehDarkAura> :) no doubt
[21:56:28] <tehDarkAura> better to get rolling and make your own pcbs for the drivers ;)
[21:56:58] <tehDarkAura> im intrigued what the servo stuff looks like
[21:57:06] * tehDarkAura goes to youtube to search
[21:57:10] <pfred1> these are the drivers I made
http://www.instructables.com/id/TB6560-Microstepping-Bipolar-Chopper-Stepper-Motor/
[21:57:58] <pfred1> yeah for servos folks here seem to prefer mesa boards
[21:58:27] <pfred1> that is all peak performance stuff steppers usually can't keep pace with any of that
[21:58:47] <pfred1> but it seems pretty complicated to me to get it all working
[21:59:52] <Tom_itx> i used a mesa board for my steppers
[22:00:03] <pfred1> sure they work for that too
[22:00:18] <Tom_itx> don't have new driver boards yet, still using the original ones i had
[22:00:30] <pfred1> Tom_itx I'm trying to keep my entire build under $100 :)
[22:00:52] <pfred1> we all have to have goals you know?
[22:00:55] <tehDarkAura> right on!
[22:01:25] <tehDarkAura> boards look nice -- will definately be back to look at your schem again
[22:01:34] <pfred1> 100 IPM for under 4100
[22:01:39] <pfred1> 100 IPM for under $100
[22:01:51] <tehDarkAura> thats awsome
[22:02:05] <Tom_itx> what are you driving with them?
[22:02:12] <pfred1> router
[22:02:24] <Tom_itx> less load than a mill then?
[22:02:32] <pfred1> I picked up at a flea market for $15
[22:02:42] <pfred1> oh yes much less
[22:02:55] <pfred1> my mill cost $800 so that blows the budget right there
[22:03:22] <pfred1> I mean the whole machine under $100
[22:04:26] <pfred1> if I was to CNC my mill, and I'm not, I'd need to replace its lead screws
[22:05:08] <pfred1> they'd likely run me more than $100 an axis to do
[22:05:20] <pfred1> and render the machine useless ot me for manual use
[22:05:49] <pfred1> so that whole plan is scotched
[22:06:28] <pfred1> plus I've no work i am interested in doing CNC on a mill
[22:10:16] <Tom_itx> i do
[22:10:23] <Tom_itx> but my mill is really small
[22:10:33] <pfred1> mine is no giant
[22:10:48] <Tom_itx> sherline
[22:10:53] <Tom_itx> i've had for years
[22:10:55] <pfred1> Ok yours would fit on mine
[22:11:00] <Tom_itx> :)
[22:11:13] <pfred1> but I've run a mill 4 of mine would fit on too
[22:11:14] <Tom_itx> i used to program/run the big ones
[22:11:19] <jdhnc> joe9:
http://www.ebay.com/itm/180586633070
[22:11:45] <pfred1> that thing was awesome
[22:11:55] <pfred1> it's run a 3 inch hogger like nothing
[22:12:13] <pfred1> just go blasting right through M3 steel
[22:12:40] <Tom_itx> these were mostly tree, fadal and okuma
[22:12:44] <Tom_itx> vmc
[22:12:58] <Tom_itx> various sizes
[22:13:03] <pfred1> this thing said Milwaukee #3 on it
[22:13:27] <Tom_itx> we ran mostly aircraft stuff
[22:13:41] <pfred1> maybe you used tools i made then?
[22:13:59] <pfred1> we made clamps for the aircraft industry
[22:14:11] <pfred1> I think Boeing
[22:14:15] <Tom_itx> we made most of our fixtures
[22:14:25] <pfred1> we just made the clamps
[22:14:34] <Tom_itx> we were certified for boeing but they never sent much our way
[22:14:35] <pfred1> they'd attach them to their own fixtures
[22:15:07] <AitalMAC> it's possible to connect stepgen of a fourth axis to the spindle controll?
[22:15:35] <Tom_itx> seems almost anything is possible with linuxcnc
[22:16:22] <pfred1> it is all ones and zeros
[22:16:46] <pfred1> the matrix!
[22:16:54] <AitalMAC> So if i get the Hal right i can connect it?
[22:17:11] <pfred1> connect it to what?
[22:17:32] <AitalMAC> Praticly i need acceleration and decelaration for my spindle
[22:17:41] <AitalMAC> but PWMgen does not provide that
[22:17:49] <pfred1> ah ha
[22:17:51] <AitalMAC> so i was sondering if i could use stepgen
[22:18:01] <AitalMAC> wonderings
[22:18:15] <AitalMAC> just wild guess
[22:18:17] <pfred1> so PWM just puts out one frequency?
[22:18:23] <AitalMAC> Yes
[22:18:36] <pfred1> not much modulaiton going on there
[22:18:44] <pfred1> modulation even
[22:19:03] <AitalMAC> Yes
[22:19:12] <pfred1> that doesn't sound right to me
[22:19:21] <AitalMAC> But i like to be able to use my spindle with the M3 command from Gcode
[22:19:27] <pfred1> the big deal with PWM is that it M's
[22:19:29] <AitalMAC> and not as an actual fourth axis
[22:20:30] <pfred1> EMC can't do spindle speed control?
[22:20:46] <Tom_itx> why not?
[22:20:56] <Tom_itx> just need feedback
[22:21:20] <pfred1> OK open loop spindle speed control then
[22:21:56] <pfred1> I find even that hard to believe
[22:23:26] <pfred1> AitalMAC
http://linuxcnc.org/docs/2.5/html/examples/spindle.html#_spindle_soft_start
[22:23:27] <AitalMAC> I don't think i can get to 8000 RPM with feedback
[22:23:55] <pfred1> AitalMAC is that what you need?
[22:24:12] <AitalMAC> thank you
[22:24:19] <pfred1> sure np
[22:24:20] <AitalMAC> i'll look into it i'll let you know
[22:27:36] <AitalMAC> I see how it works for acceleration, will it do the same for decellaration?
[22:27:58] <AitalMAC> if i send an M5 command, will it decellerate?
[22:29:55] <pfred1> it seems like if you want all the spindle control you do maybe a spindle encoder isn't asking too much?
[22:38:48] <tehDarkAura> could make a spot for a mag and use a hall effect sensor
[22:39:15] <pfred1> tehDarkAura most do it with make break light plates
[22:39:42] <tehDarkAura> ahhh
[22:39:58] <pfred1> then with a couple sensors you can extraploate
[22:40:16] <pfred1> extrapolate even
[22:40:20] <tehDarkAura> ;)
[22:40:36] <pfred1> getting late here been trying to get away for a while now
[22:40:37] <tehDarkAura> im no spelling nazi -- no spell check in my client
[22:40:48] <tehDarkAura> later
[22:41:01] <tehDarkAura> thanks for the links & info :)
[22:41:09] <pfred1> that is just my sausage fingers getting tied up over the keyboard
[22:43:01] <pfred1> tehDarkAura if you are building your own machine the best advise I can think to offer is take your time planning and that should save you from some of the more lethal pitfalls that catch folks in more of a hurry to get up and running
[22:44:28] <pfred1> well that and have some realistic targets you wish to achieve
[22:44:40] <pfred1> you know some concrete goals
[22:44:58] <pfred1> as far as speed size accuracy etc go
[22:45:11] <tehDarkAura> okay cool -- appreciate the advise :)
[22:45:18] <tehDarkAura> *c
[22:45:30] <pfred1> once you've outlined all of that then your decisions get easier to make
[22:45:48] <pfred1> when you can come up with some numbers
[22:47:56] <tehDarkAura> totally makes sense -- I've been checking a lot of other people attempts / successes -- has helped
[22:48:43] <pfred1> successes are easy enough to find examples of failures can be more telling though
[22:50:03] <tehDarkAura> the blogs that end mid project ;)
[22:50:13] <pfred1> exactly
[22:50:20] <pfred1> yeah last update in 2008
[22:50:30] <tehDarkAura> lol have seen a few
[22:56:20] <morfic> pcw_home: you wouldn't happen to have some little script that creates a test setup out of all used pins the hardware+firmware have available? if not i'll figure out how to copy/paste some sort of *vcp "test panel" together.
[22:58:01] <pfred1> tehDarkAura see if you can find a copy of this file
http://www.scribd.com/doc/50421588/CNCMechanics-v1-2
[22:58:44] <tehDarkAura> aha! very cool thank you ! :)
[22:59:49] <pfred1> in their viewer that file looks kind of scrambled to me
[23:00:01] <pfred1> could be my browser I don't know
[23:03:02] <tehDarkAura> not to bad in mine -- first page got cut short i think
[23:03:27] <pfred1> tehDarkAura the original filename is CNCMechanics v1.2.pdf
