#emc | Logs for 2005-10-24

[00:38:07] <alex_joni> * alex_joni needs some beauty sleep
[00:38:11] <alex_joni> nigth guys
[00:38:23] <fenn> g'nigth
[00:39:57] <Jymmm> alex needs a *LOT* of beauty sleep!
[00:41:12] <Jacky^> hehe :)
[00:41:50] <Jacky^> Jymmm: how the steppers are going ?
[00:41:59] <Jacky^> solved ?
[00:42:03] <Jymmm> idle at the moment
[00:42:19] <Jymmm> think it's a thermal issue.
[00:42:33] <Jacky^> maybe
[01:24:19] <fenn> hell and damnation
[01:24:35] <fenn> there's got to be some way to convert .svg files to .hpl
[01:25:55] <rayh> rayh is now known as rayh-away
[01:42:08] <fenn> heh jymmm it was your post on cnczone that led me to freesteel
[01:42:15] <fenn> how did you hear about it?
[01:43:14] <fenn> whoops nevermind, it was someone else that mentioned it
[02:23:49] <Jacky^> g nite all
[02:23:57] <Jacky^> Jacky^ is now known as Jacky^afk
[02:47:50] <fenn> god the people on cnczone are so ignorant
[02:48:42] <fenn> "You can't change software. It is what it is. You're saying you want to "change" Mach2 or TurboCNC to output to the usb port?"
[03:14:36] <djb_rh_> lol
[03:14:36] <djb_rh_> djb_rh_ is now known as djb_rh
[04:33:36] <Jymmm> fenn what is 'freesteel'?
[04:33:59] <Jymmm> fenn: I cna convert SVG to HPGL
[04:35:55] <Jymmm> fenn: Hell, I can convert SVG to whatever for that matter =)
[04:36:28] <Jymmm> hi les
[06:14:15] <CIA-6> 03paul_c * 10emc2-auto/wiki/ (21 files in 16 dirs): "Auto update wiki from a cron job. Mon Oct 24 05:30:01 BST 2005 "
[09:52:21] <Jymmm> alex_joni hi and bye!
[09:52:34] <alex_joni> hey Jymmm
[09:52:39] <alex_joni> go to sleep :P
[09:52:46] <Jymmm> G'Night =)
[11:19:13] <Jacky^afk> Jacky^afk is now known as Jacky^
[12:57:13] <anonimasu> hello
[12:58:55] <alex_joni> hey
[13:01:40] <anonimasu> what's up?
[13:03:54] <Jacky^> hello :)
[13:04:24] <Jacky^> alex_joni: are you familiar with html ?
[13:06:18] <Jacky^> i want to pass a term to this website from a python module http://www.demauroparavia.it/cerca
[13:10:32] <alex_joni> Jacky^: a bit
[13:10:40] <alex_joni> bu tnot with python
[13:12:36] <anonimasu> http://www.sigmatek.at/english/e-produkte/e-pdias.html
[13:12:54] <alex_joni> sigmatek sounds familiar
[13:13:12] <alex_joni> the ones with your PLC?
[13:13:28] <cradek> Jacky^: http://docs.python.org/lib/httplib-examples.html
[13:13:57] <alex_joni> hi chris
[13:14:07] <alex_joni> wanted to bug you.. but didn't catch you :D
[13:14:14] <alex_joni> finally I did it myself :D
[13:14:16] <cradek> hi, just here for a second
[13:14:22] <alex_joni> np
[13:14:36] <cradek> too early on a monday
[13:14:39] <alex_joni> fixed that stupid tkemc bug (jogging multiple axes)
[13:15:51] <cradek> yay!
[13:16:10] <cradek> can you jog in multiple axes simultaneously now?
[13:16:15] <alex_joni> yup
[13:16:32] <cradek> slick
[13:16:38] <alex_joni> up to 4, as there are no more key-bindings :)
[13:16:47] <alex_joni> if there were it would probably work with even more
[13:17:00] <alex_joni> can you try smthg on axis?
[13:17:03] <cradek> that bug has offended me since day one, but I did axis instead of fixing it
[13:17:07] <alex_joni> not now ofcourse
[13:17:14] <cradek> sure, later, what?
[13:17:17] <alex_joni> I wonder how it works:
[13:17:24] <alex_joni> define a machine with XZA instead of XYZ
[13:17:27] <alex_joni> and let me know :)
[13:17:52] <cradek> it won't work right
[13:17:57] <alex_joni> * alex_joni wonders about jogging, position, etc
[13:18:07] <alex_joni> it SHOULD
[13:18:11] <cradek> well, it won't *display* right
[13:18:15] <alex_joni> it SHOULD
[13:18:16] <cradek> the axes are hardcoded xyzabc
[13:18:39] <anonimasu> alex_joni: yeah
[13:18:45] <anonimasu> alex_joni: product news ;)
[13:19:01] <cradek> I hear some talented guy named alex can fix things like that
[13:19:14] <cradek> gotta run, bbl
[13:19:46] <alex_joni> later :)
[13:19:49] <Jacky^> hy guys..
[13:20:04] <Jacky^> http://www.demauroparavia.it/cerca?stringa=blahblah seems to work
[13:20:07] <Jacky^> thanks
[13:20:27] <Jacky^> geckodrives received right now
[13:20:37] <alex_joni> Jacky^: don't break them :D
[13:20:38] <Jacky^> E. 103 more..
[13:20:42] <Jacky^> ghgh :/
[13:20:46] <alex_joni> shipping?
[13:20:49] <Jacky^> nope
[13:20:50] <alex_joni> or taxes?
[13:20:57] <Jacky^> shopping $ 67
[13:21:03] <Jacky^> taxes :(�
[13:21:08] <Jacky^> spipping*
[13:21:14] <Jacky^> ouch ..
[13:21:23] <Jacky^> *_*
[13:21:30] <alex_joni> never mind
[13:21:35] <alex_joni> you'll tell us when you're sober
[13:21:37] <alex_joni> ROFL
[13:21:47] <Jacky^> i payed:
[13:22:01] <alex_joni> you drank :D
[13:22:04] <Jacky^> E. 420 + 103 for 3 geckos :/
[13:22:04] <alex_joni> grappa?
[13:22:07] <Jacky^> damn..
[13:22:15] <alex_joni> LOL
[13:22:19] <Jacky^> * Jacky^ take a bottle of wineeee
[13:22:32] <Jacky^> * Jacky^ want forget..
[13:22:36] <alex_joni> wine is too soft for you :D
[13:22:42] <Jacky^> right
[13:22:52] <Jacky^> let e try ballantines
[13:22:56] <Jacky^> :/
[13:23:17] <Jacky^> E. 103 .. are a lot
[13:23:20] <Jacky^> :(
[13:23:56] <anonimasu> comes in january ;)
[13:24:01] <Jacky^> i was expected a big box ..
[13:24:11] <Jacky^> these are 8x8 cm
[13:24:22] <Jacky^> are these the drivers ? O_o
[13:25:01] <alex_joni> Jacky^: yes :D
[13:25:08] <Jacky^> ugh
[13:25:09] <alex_joni> 1.5 kW .. imagine that :D
[13:25:18] <Jacky^> it fit on a hand
[13:25:28] <alex_joni> I mounted it on a P4 cooler (copper)
[13:25:32] <alex_joni> and it works great
[13:25:38] <alex_joni> Jacky^: amazing how small :D
[13:25:40] <Jacky^> good
[13:25:50] <alex_joni> open it up, and you'll know why it's worth the money :))
[13:26:24] <Jacky^> all smd components i suppose..
[13:27:04] <alex_joni> lots of them
[13:27:07] <anonimasu> Jacky^: as I said, you cant build better drives then them..
[13:27:07] <anonimasu> ;)
[13:27:23] <Jacky^> anonimasu: I hope .. :/
[13:31:06] <anonimasu> Jacky^: pick them apart ;)
[13:37:53] <anonimasu> * anonimasu will be ordering a new collet holder soon
[13:40:04] <anonimasu> about 500eur for taper & holder..
[13:43:03] <anonimasu> din2080
[13:47:47] <alex_joni> * alex_joni has no clue what that is :D
[13:48:34] <anonimasu> a toolholder
[13:48:57] <alex_joni> what-ever :D
[13:49:21] <anonimasu> http://www.ss-maskin.com/PDF/i-si.jpg
[14:03:35] <anonimasu> * anonimasu aywns
[14:04:21] <alex_joni> looks like plastic
[14:07:59] <alex_joni> * alex_joni aywns too
[14:10:43] <alex_joni> what do you guys think of positioning speeds of 700 m/min ?
[14:10:48] <alex_joni> * alex_joni grins widely
[14:13:12] <anonimasu> 2haha
[14:13:19] <anonimasu> makes me kind of jealos ;)
[14:13:22] <anonimasu> is that a new robot?
[14:13:24] <alex_joni> it's for a laser
[14:13:25] <alex_joni> :)
[14:13:29] <alex_joni> welding laser
[14:13:33] <anonimasu> argh
[14:13:36] <alex_joni> but that's spot speed
[14:13:46] <alex_joni> and the laser gets there through mirrors
[14:13:54] <alex_joni> so it's easy to do 700 m/min..I guess :D
[14:14:04] <anonimasu> :D
[14:14:09] <anonimasu> _easy_
[14:14:13] <alex_joni> right :D
[14:14:22] <anonimasu> * anonimasu wishes emc could handle that kind of speed
[14:14:53] <alex_joni> http://www.photonics.com/spectra/applications/read.asp?aoaid=345
[14:15:01] <alex_joni> anonimasu: will in time :)
[14:15:14] <alex_joni> but I don't want to go anywhere near a machine doing 100m/min :D
[14:15:27] <anonimasu> if emc will push a machine along at > 100m/min
[14:15:33] <anonimasu> I'll build a machine that'll do it..
[14:15:34] <anonimasu> :D
[14:15:41] <alex_joni> ok.. do we have a deal?
[14:15:46] <anonimasu> * anonimasu nods
[14:15:51] <alex_joni> heh :D
[14:16:09] <alex_joni> some scales for that machine: http://www.sonysms.co.jp/e/products/magnetopticalscale/
[14:16:28] <alex_joni> - Accuracy: 3 + 3 ML / 1000p-p
[14:16:28] <alex_joni> - Offers high-speed response of 150 m/min and supports linear motor control
[14:16:33] <alex_joni> should be ok :))
[14:16:59] <alex_joni> this is hilarious:
[14:17:02] <anonimasu> hehe
[14:17:03] <anonimasu> yeah
[14:17:04] <alex_joni> The resolution of the AB quadrature signals can be switched between 0.05, 0.1, 0.5, and 1um
[14:17:10] <alex_joni> at 150 m/min
[14:17:12] <alex_joni> ROFLMAO
[14:17:15] <anonimasu> 1 um at 150m/min
[14:17:17] <alex_joni> who can count that?
[14:17:33] <anonimasu> insanity..
[14:17:47] <anonimasu> servo spindle at 40k rpm.. at a lathe..
[14:18:17] <alex_joni> on 0.05 um and 150 m/min, if my calcs are ok it gives 50.000.000 counts / sec
[14:18:19] <alex_joni> ROFL
[14:18:23] <anonimasu> and constant speed contouring.. ;)
[14:18:35] <anonimasu> that's a project for emc & hal
[14:19:06] <alex_joni> place one of these on a lathe: - High resolution: 1 deg/10000 rotary. :))
[14:19:31] <anonimasu> you can probably buy a house for the price of a scale ;)
[14:19:50] <anonimasu> so violent stuff
[14:21:25] <anonimasu> * anonimasu imagines cutting alu at 100m/min
[14:45:56] <Jacky^> anyone know how much ampere are 500va for 220V-75V transformer ?
[14:50:09] <Jacky^> 6.6 ampere ?
[14:50:25] <Jacky^> 500/75 ..?
[14:50:49] <fenn> depends if its 75 AC or DC
[14:50:55] <Jacky^> DC
[14:51:01] <fenn> then, yes
[14:51:19] <Jacky^> thanks
[14:58:29] <fenn> mornin
[14:58:35] <etla> hi
[14:59:24] <etla> whats up ?
[14:59:54] <fenn> gonna learn python today
[15:00:05] <fenn> then i'm gonna write a .svg to g-code script
[15:00:09] <etla> to hack axis ?
[15:00:19] <etla> .svg ?
[15:00:21] <fenn> figure i should learn it anyway
[15:00:31] <fenn> .svg is a vector graphics format, kinda like eps
[15:00:39] <fenn> bezier curves and such
[15:01:03] <anonimasu> re-write the planner to machine .svg parts ;)
[15:01:16] <fenn> that's a good idea :)
[15:01:33] <anonimasu> but that's probably too much work :)
[15:01:38] <fenn> just need to write a 3d .svg editor :P
[15:03:10] <anonimasu> hehe
[15:03:39] <fenn> ayam already handles 3d bezier curves, so it's not totally unfeasable
[15:04:03] <anonimasu> ayam?
[15:04:36] <fenn> www.ayam3d.org
[15:04:52] <fenn> 3d modeler with a somewhat reasonable UI
[15:05:00] <fenn> compared to blender at least
[15:05:18] <jepler> it's not hard to turn bezier curves into sequences of segments.
[15:05:32] <jepler> cradek's ttt does that for true-type fonts. I forget which kind of curve it is.
[15:11:55] <jepler> huh, I thought I contributed an adaptive subdivision algorithm to ttt, but it just uses a hardcoded 10 segments per bezier curve
[15:19:33] <alex_joni> btw, cradek posted an stl to dxf convertor
[15:19:41] <alex_joni> so that's not very far from what you want
[15:19:49] <fenn> uh, it's not?
[15:20:06] <fenn> stl != svg
[15:20:07] <alex_joni> dunno :D
[15:21:09] <alex_joni> like they say
[15:21:13] <alex_joni> don't know don't care :D
[15:23:30] <etla> anyone thinking about drives for the surpluscenter/sanyo denkis ?
[15:29:50] <alex_joni> etla: nope :)
[15:30:01] <alex_joni> anyone heard of a CAD file ending in .src ?
[15:30:03] <rayh-away> rayh-away is now known as rayh
[15:31:26] <alex_joni> hello rayh
[15:33:09] <fenn> etla: why cant you get rid of the pin for brake
[15:33:17] <fenn> then you'd have 11 pins
[15:34:41] <fenn> when you turn the power supply on, it energizes the brake coils
[15:34:52] <fenn> any problems with that?
[15:41:50] <alex_joni> cradek: chris, are you around?
[15:45:08] <rayh> Hi alex.
[15:45:30] <alex_joni> how's it going ray? still at the mazak?
[15:45:59] <rayh> Yes. And I feel like an old dog and someone's trying to teach me a new trick.
[15:46:22] <alex_joni> why is that?
[15:46:41] <rayh> Hal + CL
[15:46:55] <alex_joni> heh.. not the most intuitive stuff I reckon
[15:46:58] <rayh> Some sort of mental block
[15:48:23] <etla> supid Q: what is classicladder ?
[15:48:41] <alex_joni> etla: stupid answer, google it :D
[15:49:01] <fenn> etla: a PLC programming language
[15:49:43] <alex_joni> classicladder
[15:49:45] <alex_joni> darn
[15:49:50] <alex_joni> http://membres.lycos.fr/mavati/classicladder/
[15:51:56] <rayh> rayh is now known as rayh-away
[15:55:46] <etla> so what kind of chips are you programming ?
[15:57:42] <alex_joni> chio chips :D
[15:58:48] <fenn> etla: it's a software PLC emulator
[15:59:03] <fenn> PLC= programmable logic controller
[15:59:27] <alex_joni> etla: a PLC is a device that does some logic stuff
[15:59:38] <alex_joni> if one input goes high, then it might wait for a while
[15:59:42] <alex_joni> turn some outputs on, etc
[15:59:54] <alex_joni> and you usually program that using some graphic language
[16:00:03] <anonimasu> ^_^
[16:00:30] <alex_joni> a
[16:00:44] <alex_joni> at least that's the birdview of it :)
[16:01:25] <etla> and you want to motion control with the plc ?
[16:04:26] <anonimasu> lots of stuff
[16:05:50] <etla> ok...
[16:06:07] <etla> how do they compare with microcontrollers and DSPs ?
[16:06:22] <alex_joni> higher abstraction level
[16:06:34] <alex_joni> they usually have a microcontroller or PC inside
[16:06:44] <alex_joni> but you don't need to take that into account
[16:07:45] <anonimasu> yep
[16:07:50] <etla> hmmm I think I will stick to PICs ...
[16:08:00] <anonimasu> heh
[16:08:05] <anonimasu> you cant really compare them...
[16:09:57] <anonimasu> plc's are very very neat in induatrial stuff
[16:11:09] <anonimasu> like toolchangers. as a example
[16:11:34] <etla> ok...
[16:11:49] <etla> do you know much about the I2C bus/protocol ?
[16:12:30] <anonimasu> I havent used it but it seems simple
[16:13:27] <etla> would it be fast enough for axis command / encoder count transfers between a servo amp and a controller
[16:13:37] <anonimasu> no idea
[16:14:05] <fenn> awww crap FreeCAD is written for windows
[16:15:30] <etla> what do you need to draw ?
[16:16:16] <fenn> everything
[16:16:41] <fenn> freecad is supposed to be a 3d parametric cad based on opencascade
[16:17:40] <etla> I havent seen any free 3D cads that are any good
[16:17:49] <fenn> that's because they havent been written yet
[16:18:16] <etla> I like the login in MultiSurf (www.aerohydro.com)
[16:18:19] <etla> logic
[16:20:41] <fenn> "relational geometry?"
[16:20:59] <etla> once you have drawn everything
[16:21:14] <etla> you can just pick up one point and move it and everything moves with it
[16:21:25] <etla> you start with points
[16:21:27] <fenn> isn't that the same thing as parametric modeling?
[16:21:35] <etla> curves are defined by points
[16:21:42] <etla> surfaces are defined by curves
[16:21:46] <etla> yes parametric
[16:44:03] <etla> approx: rumor has it that you are designing a driver for the surpluscenter servos ??
[16:50:47] <approx> yes
[16:51:05] <etla> how far along are you
[16:51:07] <approx> I will get on it as soon as I recieve the motors
[16:51:24] <etla> ok, lots of other guys doing the same thing
[16:51:25] <approx> i have done some tests with another 900W servo
[16:51:46] <approx> it already tracks motion by step/dir
[16:52:06] <etla> ok, I don't think I want to do step/dir control
[16:52:13] <etla> does it have sinewave commutation ?
[16:52:33] <approx> yes, it has sinewave
[16:52:54] <etla> is the sinewave based on encoder counts or interpolation btw hall signals ?
[16:53:04] <approx> btw, it can do step multiplation so pc won't be overloaded
[16:53:13] <approx> encoder based
[16:53:24] <etla> yes but with step multiplication resolution will suffer
[16:53:31] <etla> are you coding in asm or c30 ?
[16:53:41] <approx> not badly if those sanyos have 8000 CPR encoders
[16:54:06] <approx> now in C but will be optimized with asm
[16:54:19] <etla> but it will allways be a tradeoff between rapid speed and resolution
[16:54:31] <etla> OK, would you cate to share some code and/or schematics ?
[16:54:38] <etla> care
[16:54:46] <approx> there is not much to share at this point :b
[16:54:51] <etla> ok...
[16:56:35] <etla> are you using an ICD2 programmer ?
[16:58:53] <approx> what's that?
[16:58:56] <approx> :)
[16:59:15] <etla> found anything cheaper ?
[16:59:23] <approx> I have my own icd for avr
[16:59:38] <approx> ahem, isp programmer
[16:59:46] <etla> so it is avr based ?
[16:59:51] <approx> yep
[17:00:03] <lerman_> lerman_ is now known as lerman
[17:00:08] <etla> does it have hardware encoder counters and pwm generators
[17:00:25] <approx> PWMs yes, but no encoder interface
[17:00:41] <approx> but encoder can be read quite fastly with interrupts
[17:00:51] <etla> what clock speed is it running
[17:00:57] <alex_joni> approx: use an LS7166 for encoder
[17:01:11] <alex_joni> etla: I'd suggest one of these for programming http://www.olimex.com/dev/avr-jtag.html
[17:01:18] <approx> i try to keep part number low :p
[17:01:45] <approx> avr can run 20MHz
[17:02:01] <alex_joni> approx: I found it WAY better than any other solution
[17:02:03] <alex_joni> even FPGA
[17:03:16] <approx> how that is so good?
[17:03:41] <alex_joni> does take care of filtering garbage out
[17:03:45] <alex_joni> easy to access
[17:03:53] <alex_joni> does take care of index-pulses
[17:03:54] <alex_joni> etc.
[17:03:56] <alex_joni> google it up
[17:04:00] <approx> hmm
[17:04:15] <alex_joni> not very cheap... but surely worth the money
[17:04:57] <anonimasu> approx: if you poll your counter at 20 mhz, ;) what's left for calculating
[17:05:18] <alex_joni> you can't poll at 20 MHz
[17:05:28] <alex_joni> you still need about 4-5 instructions per poll
[17:05:31] <anonimasu> I know..
[17:05:33] <approx> but one can use INT0 to count encoder
[17:05:46] <alex_joni> if it's properly decoded
[17:05:50] <approx> no wasted time
[17:05:54] <alex_joni> direction, etc
[17:06:00] <alex_joni> only need one?
[17:06:20] <approx> another channel can be on ordinary i/o
[17:06:45] <approx> I already tested a 4000crp encoder with dc motor
[17:06:55] <alex_joni> how fast?
[17:06:59] <alex_joni> 2-3k RPM?
[17:07:02] <approx> it coudn't generate high enough frequency to lose counts
[17:07:12] <anonimasu> approx: how fast did you run it?
[17:07:16] <alex_joni> how do you know it doesn't lose counts?
[17:07:18] <approx> ~3000rpm
[17:07:39] <approx> well I have monitored the CPU usage.
[17:07:47] <approx> cpu was over 50% idling at full speed
[17:07:57] <alex_joni> CPU?
[17:08:07] <anonimasu> how do you measure that at a avr?
[17:08:33] <approx> build a non-realtime loop that counts how fast it runs idle process
[17:08:44] <anonimasu> oh
[17:08:54] <fenn> gee isn't that 48mhz?
[17:08:55] <anonimasu> are you using a operating system on the avr?
[17:08:56] <approx> not wery easy to explain :p
[17:09:07] <approx> no, just my code
[17:09:09] <fenn> oops n/m
[17:09:10] <anonimasu> since you seem to be doing multitasking?
[17:09:19] <approx> no, just interrupts
[17:09:37] <approx> "non realtime" task is all non-interrupt stuff
[17:09:38] <alex_joni> so where's the non-realtime loop ?
[17:09:41] <alex_joni> ahh I start to understand
[17:09:50] <alex_joni> ok.. try not to mix terms,
[17:09:58] <approx> :p
[17:09:58] <alex_joni> so you had code taking care of int's
[17:10:06] <approx> yep
[17:10:09] <alex_joni> and code doing nothing (setting a pin high or smthg)
[17:10:13] <alex_joni> and you emasured that
[17:10:22] <jepler> approx: are you talking about using a single interrupt pin to do interrupt-driven quadrature decoding? I don't quite follow how you would do that.
[17:10:29] <approx> yes I have leds that will lit if load is over 50%
[17:11:06] <approx> jepler: need only one signal to get edges. other signal gives direction
[17:11:35] <alex_joni> jepler: run the quadrature through a J/K
[17:12:04] <alex_joni> actually D
[17:12:19] <alex_joni> it will give you step and direction
[17:12:23] <alex_joni> from the quadrature
[17:12:26] <jepler> approx: so at each interrupt you look at the state of the other quadrature output to decide the direction of the movement, then change the interrupt to be triggered by the opposite-going edge?
[17:12:54] <approx> yes but no need to change interrupt. avr can interrupt on both edges
[17:13:26] <alex_joni> that's utterly stupid.. I reckon
[17:13:50] <etla> and you monitorr the Halls also ?
[17:13:57] <alex_joni> you could run it through a bistable, and get only counts
[17:14:03] <alex_joni> 1 count up, 1 count down
[17:14:23] <approx> halls are not much needed when using encoder
[17:14:48] <etla> but you need to get an index signal before you know where you are
[17:14:54] <approx> true
[17:14:59] <alex_joni> what happens if your counts get screwed?
[17:15:05] <alex_joni> the motor stops running?
[17:15:15] <jepler> approx: which AVR can do that? I'm most familiar with at90s2313 (an older part) and its datasheet gives "low level", "falling edge" and "rising edge" as the only option. Not "both edges".
[17:15:35] <approx> atmega-series have "both" option
[17:15:59] <anonimasu> but well, a encoder chip might spare you lots of pain..
[17:16:09] <approx> maybe
[17:16:28] <approx> but isn't that serial communicated?
[17:17:00] <anonimasu> that's just a few polls compared to hogging the cpu with interupts..
[17:17:36] <approx> how much those LS7166s cost? are they available?
[17:18:15] <anonimasu> 11$
[17:18:20] <anonimasu> approximately...
[17:18:24] <approx> pretty expensive
[17:18:55] <alex_joni> Standard Pricing:
[17:18:55] <alex_joni> $11.65 / 1
[17:18:55] <alex_joni> $9.35 / 25
[17:18:55] <alex_joni> $7.45 / 100
[17:18:55] <alex_joni> $6.35 / 500
[17:18:56] <alex_joni> $5.40 / 1K
[17:19:06] <alex_joni> http://www.usdigital.com/products/ls7166/
[17:19:18] <approx> it is cheaper to put a dedicated atmega on counting :)
[17:19:32] <anonimasu> what do you pay for a atmega?
[17:19:33] <alex_joni> cheaper yes, better.. no
[17:19:39] <alex_joni> 3-5$
[17:19:55] <anonimasu> what mega?
[17:20:11] <approx> any model with 8k flash is enough
[17:20:15] <alex_joni> 64?
[17:20:18] <alex_joni> I think
[17:20:39] <anonimasu> hm
[17:20:41] <approx> currently i have atmega8
[17:20:53] <anonimasu> I pay about $12 for atmega168's..
[17:21:06] <alex_joni> 128 I hope
[17:21:11] <anonimasu> 8$
[17:21:44] <anonimasu> ah well you wont do the job of that chip as good with a micro..
[17:22:08] <alex_joni> let him try
[17:22:16] <alex_joni> if he succeeds he'll prove us wrong
[17:22:19] <fenn> wonder why it is so expensive...
[17:22:36] <alex_joni> fenn: suggest looking at the pdf
[17:22:38] <anonimasu> yep
[17:26:33] <approx> how about really simple PLDs for encoder counting?
[17:26:40] <approx> would be fast
[17:26:58] <alex_joni> approx: what kind of drive are you designing?
[17:26:59] <anonimasu> have you had a look at the datasheet
[17:27:18] <fenn> approx: how bout a flipflop and a binary up/dn counter
[17:27:32] <alex_joni> fenn: I did that on a project...
[17:27:37] <alex_joni> it's a lot more complicated :D
[17:27:42] <fenn> really?
[17:27:44] <approx> too much analog components
[17:27:45] <alex_joni> I could get only 4-bit counters
[17:27:54] <alex_joni> approx: what analog components?
[17:28:12] <approx> i have seen some schematics of such circuit..
[17:28:24] <approx> it has delays made with capacitors etc
[17:28:54] <alex_joni> what circuit?
[17:29:07] <approx> some circuit at cnczone forums :p
[17:30:12] <anonimasu> lol
[17:31:09] <fenn> i doubt it matters if you miss a count every once in a while at 8000 ppr with index
[17:32:50] <approx> no pulses should be lost
[17:33:14] <approx> eventually motor will lose torque and stall when commutation phase drifts enough
[17:33:34] <fenn> but you have index to reset on and compare count values to
[17:33:53] <approx> then it might be ok
[17:34:29] <fenn> if you are off enough that the motor is out of sync then you have a problem, agreed
[17:34:51] <fenn> but 1/8000th of a revolution = .000024" with 5mm ballscrews
[17:35:47] <approx> but if one does lots of repetive short moves then index pulse will not be encountered :p
[17:35:59] <fenn> good point
[17:37:02] <etla> use the halls ?
[17:37:23] <etla> you get the count reset at least six time per rev or is it more ?
[17:37:32] <etla> 4 pole motor so is it 4*6 times ??
[17:37:34] <approx> but halls are inaccurate
[17:37:49] <etla> why
[17:38:12] <approx> aren't those just analog magnetic switches?
[17:38:33] <etla> should be ttl outputs ?
[17:38:38] <fenn> it's a semiconductor
[17:38:46] <etla> depends on motor I guess
[17:38:52] <approx> yes but the trip point may wary too much
[17:40:39] <etla> hmmm need to test that somehow...
[17:42:25] <fenn> compare it to the encoder output when you get the motors
[17:42:36] <fenn> it should trip on the same number each time
[17:42:50] <approx> i'm not counting on that. i have zero counting error tolerance on my controller :)
[17:43:02] <approx> i will not be satisfied with nothing less
[17:43:25] <etla> yes, hook signals up to daq board and somehow spin the motor... then analyze waveforms for hours and hours :)
[17:43:38] <approx> exactly
[17:43:40] <fenn> well you gotta do it some time
[17:44:03] <etla> the counting error for the sine drive can be reset every index or hall pulse
[17:44:15] <approx> and testing should be done with higher rpm than actual usage
[17:44:20] <etla> but if your closing the position loop also with your drive then counting errors are a problem
[17:44:49] <etla> are you planning on using the irams10 for the power stage ?
[17:45:07] <approx> no decisions for that yet
[17:45:17] <approx> maybe discrete mosfets
[17:45:31] <approx> cheper to fix if it blows :)
[17:46:38] <etla> does anyone know how torque command would be done with these motors ?
[17:47:18] <fenn> what do you mean?
[17:47:33] <fenn> through emc?
[17:47:49] <etla> yes, is that the normal way to command servos
[17:47:58] <etla> pc sends out torque commands and receives encoder count
[17:48:08] <fenn> its pretty flexible with hal
[17:48:30] <etla> ok so what is the 'best' way ?
[17:48:36] <fenn> i think most servo drive cards send out torque or velocity
[17:48:53] <fenn> it depends where you want the PID loop to be
[17:49:09] <etla> I think I want it in the pc
[17:49:14] <etla> more flexibility that way
[17:49:34] <fenn> if you send velocity commands you can't control torque damping from the pc
[17:50:21] <etla> hmm.. and torque corresponds to current in the motor ?
[17:50:39] <fenn> yeah
[17:51:20] <etla> so the drive needs to monitor the current and control it
[17:52:13] <fenn> the drive puts out xxx amps and makes sure the controller hasn't gone crazy
[17:52:27] <fenn> i mean current limiting is still in the drive
[17:52:35] <etla> yep
[17:52:37] <fenn> but basically its just a big amp
[17:52:59] <etla> so need to monitor current at max of each of the sine-phases and adjust pwm according to that
[17:53:46] <fenn> * fenn kick-starts his brain
[17:55:00] <fenn> don't you calculate pwm based on the commands you get from the controller?
[17:55:27] <etla> but the pwm controls the voltage.... not the current
[17:55:52] <fenn> how do you limit current when the motor is reversing?
[17:56:17] <etla> what do you mean by reversing ?
[17:56:21] <fenn> changing direction
[17:57:01] <etla> you switch to the commutation scheme for reverse
[17:57:10] <etla> and then you start applying the commanded current
[17:57:12] <etla> I guess
[17:57:28] <fenn> that's the part i dont understand
[17:57:42] <fenn> it all boils down to the fact that i dont understand v=ir
[17:57:54] <fenn> er, v=iz i guess
[18:00:06] <fenn> how do you monitor current? (my ignorance is showing)
[18:00:31] <etla> there is a sense resistor on the negative bus
[18:00:56] <etla> you take that voltage and ad convert it in the microcontroller
[18:01:26] <fenn> ok
[18:01:54] <fenn> i keep forgetting microcontrollers have all sorts of nifty doohickies on them
[18:02:00] <etla> but I think it needs filtering cause it is pcm...
[18:02:10] <etla> and it would be nice to have that also isolated
[18:05:11] <etla> I'd like to use the EPP interface for communicating with a pc
[18:05:17] <etla> like the pico system does
[18:05:20] <etla> cheap.
[18:08:15] <cncuser> good evening folks
[18:08:23] <etla> hi
[18:08:30] <cncuser> hi etla
[18:08:45] <cncuser> well, what emc2 version should i take ?
[18:08:51] <etla> hi, just learning to use eagle....
[18:08:55] <etla> no idea about emc2.
[18:09:04] <cncuser> for the next llivecd ?
[18:09:32] <fenn> um, the latest?
[18:09:36] <etla> I'm using emc1 for various hobby projects, haven't touched emc2
[18:09:39] <cncuser> etla: cool, eagle is nice. im going to lern it too, somedays, when i have time. i want to do pcb milling :)
[18:09:53] <fenn> cncuser: does puppy-emc have a web browser?
[18:09:55] <cncuser> fenn: ok, so cvs
[18:10:03] <cncuser> fenn: it could.
[18:10:10] <fenn> dillo is only 357k compressed
[18:10:17] <cncuser> fenn: you can install a dotpup package : dillo, mozilla
[18:10:19] <cncuser> i know
[18:10:24] <cncuser> maybe i put it in
[18:10:35] <fenn> you should put in dillo-i18n-misc
[18:11:01] <etla> cncuser: we have Pads at work so it's all familiar but internet collaborations etc. requires the free eagle...
[18:11:32] <cncuser> fenn: ok, ill do :)
[18:12:36] <fenn> cncuser: thanks
[18:12:50] <cncuser> mah, thos sf cvs is lame
[18:13:19] <fenn> alex has a daily .tgz on the web somewhere
[18:13:39] <fenn> i doubt it has changed much
[18:13:47] <fenn> since yesterday
[18:14:01] <cncuser> *waiting till 3 minutes for CVS login*
[18:14:07] <cncuser> ha, its done
[18:16:35] <cncuser> hmm
[18:17:07] <cncuser> ahhh src/
[18:17:16] <cncuser> :) gets me everytime
[18:17:56] <rayh-away> cncuser: We compiled emc2 yesterday and are using it here.
[18:18:03] <rayh-away> some additional module stuff there.
[18:18:55] <cncuser> rayh-away: i c, it updated much :)
[18:19:09] <cncuser> *compiling*
[18:20:22] <cncuser> clean compile :)
[18:24:28] <rayh-away> There were some additions to config and a couple hal modules.
[18:25:44] <cncuser> hmm, whats the --prefix for if tehres no install target
[18:27:45] <cncuser> someone want to help me separete only the necessary files for emc from the sourcetree ?
[18:29:27] <cncuser> do i only need scripts, configs lib and bin ?
[18:34:08] <fenn> you also need rtlib and either tcl or axis
[18:34:37] <Jymmm> Mornin Folks!
[18:34:51] <cncuser> fenn: i know, axis will not be in it right now. having trouble getting python smaller
[18:34:56] <cncuser> hi jymmm
[18:44:20] <fenn> whatever happened with crapahalic?
[18:46:10] <anonimasu> hm
[18:46:28] <anonimasu> surpluscenter seems to have declined my purchase..
[18:46:57] <fenn> "go away, you are too much of a bother"
[18:47:26] <anonimasu> * anonimasu sighs
[18:47:31] <anonimasu> no clue why it happened..
[18:47:39] <anonimasu> I should have more then enough cash on the card..
[18:49:16] <fenn> maybe i will send you mine, then I wont have to think about motor driver design
[18:49:54] <anonimasu> ^-^
[18:50:03] <anonimasu> I wonder how much a logosol driver cost
[18:51:18] <Jymmm> Yeah, you know how those romainian's are.
[18:51:19] <fenn> bah 91v DC max
[18:51:44] <anonimasu> ?
[18:51:54] <Jymmm> and those Sweeeds too
[18:53:12] <Jymmm> anonimasu didn't someone offer to get them for you ?
[18:53:22] <anonimasu> yeah cradek
[18:53:23] <anonimasu> I think
[18:53:28] <anonimasu> :)
[18:53:37] <Jymmm> fenn dont you live aroudn the corner from them?
[18:53:56] <anonimasu> http://cgi.ebay.com/SVG-Track-Systems-Logosol-Controller-4-Axis-99-57301-01_W0QQitemZ7556392565QQcategoryZ97184QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[18:54:19] <Jymmm> fenn were you asking about a way fto convert SVG to HPGL ?
[18:54:51] <fenn> yeah
[18:55:03] <Jymmm> fenn I can do that easily.
[18:55:08] <fenn> poo
[18:55:09] <anonimasu> *yawns*
[18:55:12] <fenn> i want to do it
[18:55:14] <anonimasu> I'll take care of it tomorrow night
[18:55:26] <Jymmm> fenn : Fine, get yourself Corel then.
[18:56:23] <fenn> i'm going to write a svg to g-code script
[18:56:32] <Jymmm> already been done.
[18:56:40] <fenn> yeah but it sucks
[18:56:49] <fenn> what's the point of using .svg if you cant do smooth curves?
[18:57:17] <anonimasu> yep
[18:57:29] <Jymmm> See, your blaming the code when you didn't bother to look at the source application tha generated the SVG's.
[18:57:40] <fenn> um
[18:57:46] <cncuser> hmm got emc2 down to 6220 kb
[18:57:53] <anonimasu> cncuser: what are you doing?
[18:58:04] <cncuser> anonimasu: i shrink
[18:58:05] <Jymmm> SodiPie and InkSpace have FP issues
[18:58:06] <cncuser> :)
[18:58:18] <anonimasu> cncuser: I hope you arent commiting what you shrink ;)
[18:58:20] <Jymmm> have had for years.
[18:58:28] <fenn> what do you mean FP issues?
[18:58:40] <Jymmm> Decimal, floating point.
[18:59:04] <cncuser> anonimasu: youll see the iso on wednesday. im heading for <20mb
[18:59:49] <fenn> you mean like rounding errors?
[19:00:07] <anonimasu> cncuser: bdi?
[19:01:18] <cncuser> anonimasu: puppy sit! (aka cooltoolix)
[19:01:24] <Jymmm> fenn you'll have to look at the bug list, I just know it's not accurate enough to be able to say 5.000 is actually 5.000 instead of 4.975682227
[19:01:24] <anonimasu> hm
[19:01:26] <anonimasu> ok
[19:01:52] <fenn> oh well
[19:01:56] <cncuser> i wonder i upx has impact on emc2
[19:02:18] <anonimasu> I think you should shrink something else then emc
[19:02:19] <anonimasu> :)
[19:02:30] <cncuser> anonimasu: i shrink everything
[19:02:36] <cncuser> anonimasu: no excuses
[19:02:38] <cncuser> :)
[19:02:41] <fenn> kill!
[19:02:57] <cncuser> anonimaus: rtai also a big peace
[19:03:25] <fenn> anonimaus, i like that
[19:03:43] <fenn> like mighty maus, but not
[19:04:11] <anonimasu> we are chasing weight at work ;)
[19:05:07] <fenn> for high-performance forestry equipment?
[19:05:58] <anonimasu> nope..
[19:06:10] <anonimasu> I am involved in racing right now ;)..
[19:07:46] <Jacky^> what the equivalent power of 250W in NM for DC motor ?
[19:08:27] <fenn> you have to specify what rpm
[19:08:40] <Jacky^> 1.500, 2.000, 3.000, 4.000 min-1
[19:08:55] <Jacky^> i'm looking here: http://www.bonfiglioli.com/bcspres_it.html
[19:09:47] <Jacky^> better here: http://www.bonfiglioli.com/bcspres_uk.html
[19:09:51] <Jacky^> is english
[19:10:02] <Jacky^> now, i know i want 250w dc motor ..
[19:10:26] <Jacky^> i just see type and torque in NM
[19:10:34] <fenn> well, torque changes with what rpm you are at
[19:10:52] <fenn> still you can calculate it to get an estimate
[19:11:43] <Jacky^> ok, i'm a bit confused ..
[19:12:00] <Jacky^> i will ask to les when hes around
[19:12:20] <fenn> units is throwing a fit at me
[19:14:02] <fenn> units 250W N*m/min
[19:14:12] <fenn> 15000
[19:15:05] <Jacky^> what are 15000 ?
[19:15:11] <fenn> so 3.75 Nm at 4000 rpm
[19:15:18] <fenn> etc
[19:15:33] <icee> assuming 100% efficiency
[19:15:35] <Jacky^> oh ok..
[19:15:39] <icee> which you don't get
[19:15:41] <fenn> right, that's only an estimate
[19:18:32] <Jacky^> so, i think bcs70 model , 4,5 Nm shoud be enough
[19:19:03] <fenn> yes
[19:19:42] <Jacky^> instead the bc60 has 3,0 Nm and could be few..
[19:20:12] <fenn> that is continuous torque rating, peak rating is higher (i dont know if you need 250W continuous or peak)
[19:20:13] <Jacky^> assuming 250w as reference
[19:20:39] <fenn> peak torque rating is usually the one that counts for CNC applications
[19:21:09] <Jacky^> uhm.. les sayd 250W so, i think he meant peak ..
[19:21:36] <fenn> you should probably ask him
[19:21:48] <Jacky^> yeah, thanks
[19:22:13] <Jacky^> btw i will request quotes for these 2 models
[19:23:50] <fenn> good luck
[19:24:03] <fenn> maybe you should request a free sample :)
[19:24:10] <Jacky^> :)
[19:24:20] <Jacky^> whats a free sample ??
[19:24:38] <Jacky^> free servomotor ? :P
[19:24:49] <anonimasu> somehow I doubt it
[19:25:18] <Jacky^> anonimasu: yeah
[19:25:32] <Jacky^> very hard to get it
[19:31:14] <Jacky^> speed 1.500, 2.000, 3.000, 4.000 min-1
[19:31:36] <Jacky^> supply 24/ 180 Vdc
[19:31:54] <anonimasu> voltage gives rpm..
[19:32:00] <anonimasu> amperage privides torque :)
[19:32:13] <Jacky^> this mean it can run 1t 4k rpm with 180 V
[19:32:16] <Jacky^> :/
[19:32:20] <anonimasu> yes
[19:32:31] <Jacky^> too high voltage..
[19:32:42] <anonimasu> not really
[19:32:53] <anonimasu> that's the kind of voltages you need to have..
[19:32:56] <Jacky^> gkevo run at 80 right ?
[19:33:00] <Jacky^> gecko*
[19:33:09] <anonimasu> 80v 8a I think..
[19:33:37] <Jacky^> this mean i will never run the motor at max speed ..
[19:33:44] <anonimasu> yep
[19:33:51] <anonimasu> it's still probably a good motor
[19:33:56] <anonimasu> look the the torque curve
[19:33:59] <anonimasu> torque vs rpm
[19:34:20] <Jacky^> :/
[19:34:24] <Jacky^> i will try..
[19:34:40] <anonimasu> it's not sure that you will have usable torque there either
[19:35:41] <anonimasu> bbl
[19:35:45] <Jacky^> right
[19:35:52] <Jacky^> understood
[19:44:26] <fenn> http://cgi.ebay.com/Large-BLDC-servo-motor-robot-generator-phase-converter_W0QQitemZ7555889474QQcategoryZ25290QQrdZ1QQcmdZViewItem
[19:45:34] <Jacky^> lo alex_joni
[19:45:40] <alex_joni> 'lo
[19:58:37] <fenn> alex_joni: any particular reason why you wouldn't be able to use 4-bit binary counters in a homebrew quadrature counter?
[19:58:49] <alex_joni> I was able
[19:58:54] <alex_joni> but I needed to use lots of them
[19:58:59] <alex_joni> 3 channels 24-bit
[19:59:01] <alex_joni> you do the math
[19:59:06] <fenn> why not just use 1 and let the PC remember the rest of the bits
[19:59:20] <fenn> er, if you 're using a pc that is
[19:59:22] <alex_joni> it counted faster than that
[19:59:24] <alex_joni> no PC
[19:59:28] <alex_joni> microcontroller
[19:59:32] <alex_joni> not very much flash
[19:59:52] <fenn> 24*3 is not very much flash
[20:00:05] <fenn> er, you need ram not flash
[20:00:22] <alex_joni> ram
[20:03:26] <Jymmm> * Jymmm flashes fenn
[20:03:42] <alex_joni> * alex_joni flashes fenn's ram
[20:03:55] <fenn> eek!
[20:04:33] <Jacky^> * Jacky^ erase all
[20:06:08] <Jacky^> .seen les
[20:06:15] <Jacky^> mmmhhh
[20:09:10] <Jacky^> .tell les Jacky^ need heeelp ! :P
[20:11:04] <alex_joni> .ok
[20:11:12] <Jacky^> .lol
[20:11:38] <Jymmm> .- . - -- . .-.-.-
[20:11:59] <alex_joni> Jymmm: not nice to talk dirty in here
[20:11:59] <alex_joni> :D
[20:12:07] <Jymmm> alex_joni since when?
[20:12:18] <alex_joni> dunno
[20:12:20] <alex_joni> just now
[20:12:25] <Jymmm> bullshit
[20:13:04] <Jymmm> DonkeyDooDoo
[20:13:29] <alex_joni> Jymmm: remember yucko the clown?
[20:14:49] <fenn> * fenn .o0( aet me? )
[20:15:28] <Jymmm> yep
[20:15:34] <alex_joni> http://www.yourfilehost.com/media.php?cat=video&file=yukko_again.wmv
[20:16:32] <Jacky^> alex_joni: dou you think a mini-helicopter can be drived as a cnc machine ?
[20:16:43] <alex_joni> hmm.. harder
[20:16:48] <alex_joni> but it might.. :D
[20:16:56] <fenn> how would you know where it is?
[20:17:01] <Jacky^> found an interesting article here: http://pdv.cs.tu-berlin.de/~musial/bristol.pdf
[20:17:13] <Jacky^> it run linux, gps, autopilot ..
[20:17:18] <Jacky^> very interesting
[20:17:33] <Jacky^> but not enough info :/
[20:18:22] <icee> I built an autonomous helicopter a couple years ago
[20:18:30] <icee> figuring out attitude is very difficult.
[20:18:42] <Jacky^> yeah ..
[20:18:43] <icee> Lots of noisy, derivative signals
[20:18:49] <Jacky^> a frien of mine failed too ..
[20:19:04] <icee> No, mine flew and was successful, mostly.
[20:19:06] <Jacky^> http://stor.altervista.org/4c/4c.php
[20:19:23] <alex_joni> * alex_joni suggests differential GPS
[20:19:26] <icee> in order to autoland it required this very cumbersome radio differential GPS thing
[20:19:27] <Jacky^> he tried to 'copy' the draganflyer
[20:19:45] <alex_joni> icee:right :D
[20:19:51] <icee> I've been meaning to revise it, and to build a laser altimeter.. but
[20:19:54] <alex_joni> best I've seen was over DECT
[20:20:11] <icee> alex: the only thing i used the radio dgps for was precise estimates of altitude
[20:20:16] <Jacky^> laser altimeter ? never heard.. interesting
[20:20:33] <alex_joni> icee: yes but you could get accurate positioning (about 1-2cm)
[20:20:45] <icee> j: yah, you fm modulate a laser, and mix what you get back with your current modulation
[20:20:54] <Jacky^> cool
[20:21:05] <icee> yah.. accurate positioning would translate to better attitude estimates from the gps, too
[20:21:39] <icee> that was my approach; using a 3 accelerometer + 3 gyro IMU i built to update a kalman filter with attitude estimates, and cleaning it up based on the macro-scale accelerations measured over GPS
[20:22:06] <icee> when asked to hover the random walk was within a 15' diameter circle
[20:22:13] <icee> which really isn't bad
[20:22:19] <Jacky^> :)
[20:24:20] <Jacky^> icee: already know this ? http://robots.net/proj/Autopilot/
[20:24:26] <icee> yah.
[20:24:38] <icee> I originally bought one of rotomotion's boards to try and do this with
[20:24:40] <icee> total piece of shit.
[20:25:05] <Jacky^> how was working ?
[20:26:07] <icee> it had major noise and stability issues
[20:26:07] <Jacky^> im registered to the website since few days but not received passd yet..
[20:26:15] <icee> the hardware, that is.
[20:26:26] <icee> their software supposedly works well
[20:26:51] <Jacky^> understood ..
[20:34:42] <alex_joni> holy sh*t
[20:34:46] <alex_joni> try this out: http://y.20q.net/anon?K-k9gBgjZfZ7v-DWSIoIyQZhfvxK9lRTUYZUfIyFp2mIwQ,,d
[20:36:01] <Jymmm> alex_joni that was pretty good, short though =(
[20:36:08] <alex_joni> Jymmm: how short?
[20:36:26] <Jymmm> alex_joni only 5 minutes
[20:36:33] <alex_joni> heh
[20:38:27] <Jymmm> Can someone (basically) tell me the diff with 1, 2, or 3 flute cutters?
[20:38:57] <icee> the more flutes, the less removed per flute
[20:39:44] <icee> fewer flutes are generally better at plunge milling, more flutes are better at surfacing, etc.
[20:40:07] <Jymmm> ok, what about cutout...
[20:40:22] <Jymmm> like if I wanted to cout out a 6" circle
[20:40:29] <icee> out of what?
[20:40:38] <Jymmm> MDF or plywood
[20:41:12] <icee> oh, for routing. hm. Out of my area of expertise
[20:43:29] <Jymmm> bah humbug!
[20:45:29] <icee> sorry ;)
[20:45:46] <Jacky^> alex_joni: i maded the test but its wrong :/
[20:46:05] <Jacky^> I can do better,, i can do a miracle !
[20:46:12] <Jacky^> http://utenti.quipo.it/base5/scienze/f01gesu.gif
[20:46:28] <Jacky^> look the center of this image for about 30 sec
[20:46:40] <Jacky^> after look at the wall :)
[20:47:21] <anonimasu> hm I've got the new spindle bearings in place
[20:47:26] <Jacky^> look ath the 4 small points
[20:48:08] <Jacky^> an open and close the eyes
[20:48:23] <Jacky^> it rechargiable
[20:48:26] <Jacky^> heh
[20:48:36] <alex_joni> right.. maybe next tiem
[20:48:37] <alex_joni> time
[20:49:35] <anonimasu> Jymmm: the difference between flutes is what feedrate you have to keep to get the desired chipload
[20:50:15] <Jymmm> anonimasu what if I want to cut as fast as I can?
[20:50:25] <anonimasu> and also the stability of the cutter, 3 flutes are the most rigid ones, and 2 the least..
[20:50:29] <anonimasu> what rpm are you going to use?
[20:50:30] <anonimasu> err
[20:50:32] <anonimasu> do you have
[20:50:35] <anonimasu> and how fast can you feed?
[20:50:41] <anonimasu> and what chipload do you want?
[20:50:42] <Jymmm> I've been doing stepdown of .25 but want .76
[20:50:53] <Jymmm> like cutting out stencils
[20:51:00] <anonimasu> that's just a horsepower issue..
[20:51:11] <anonimasu> depth of cut isnt as relevant if you have the right chipload and enough hp..
[20:51:13] <anonimasu> hp/kw
[20:51:30] <Jymmm> not really... even at slow feeds I've snapped bits.
[20:51:39] <Jymmm> (not intentional)
[20:51:45] <anonimasu> how fast are you turning your but=?
[20:51:47] <anonimasu> bit?
[20:51:52] <Jymmm> 18K RPM
[20:51:59] <anonimasu> and how fast are you feeding?
[20:52:04] <Jymmm> 60IPM
[20:52:19] <Jymmm> this is just 3/4" MDF too
[20:52:32] <Jymmm> but I'd like get get away from the three passes
[20:52:57] <anonimasu> what size of bit
[20:53:05] <Jymmm> from 1/8 to 1/4
[20:53:17] <Jymmm> "
[20:53:36] <anonimasu> hm :/
[20:53:54] <Jymmm> I mean there has to be some significant reason for having various number of flutes that I'm not aware of.
[20:54:03] <Jymmm> That's what I'm really asking about.
[20:54:15] <anonimasu> surface finish and material removal..
[20:54:18] <Jacky^> Jymmm: are you roughing first ?
[20:55:06] <anonimasu> Jymmm: so you have 0.13 per tooth of feed?
[20:55:10] <anonimasu> 0.13mm..
[20:55:14] <anonimasu> with 1/4
[20:55:14] <Jymmm> I need to order some more bits, and they have 1,2, 3 flutes and I'd like to know the diff/purpose
[20:56:25] <Jymmm> This is in general, not specific chipload.
[20:56:33] <anonimasu> Jymmm: that's all I can say, there's issues with more then 3 flutes in alu..
[20:57:13] <Jymmm> ok, then when/why would you use a 3 flute bit? Under what cercumstances is what I'm asking.
[20:57:29] <anonimasu> for plunging..
[20:57:41] <Jymmm> speed, accuracy, finish, material, etc
[20:57:49] <anonimasu> and if I need a higher material removal rate..
[20:57:59] <anonimasu> but when you mill al it's a matter of heat..
[20:58:02] <anonimasu> as I dont have any cooling
[20:58:16] <icee> more flutes == better finish in general.
[20:58:18] <anonimasu> also, 3 flutes are better for killing resonance..
[20:58:20] <Jymmm> this is all plastic/wood
[20:58:28] <icee> less flutes == faster cutting
[20:58:33] <anonimasu> no that's wrong..
[20:58:48] <anonimasu> more flutes = more teeth
[20:58:54] <anonimasu> more cutting edges..
[20:59:03] <anonimasu> that means higher feed to obtain the desired chipload..
[21:00:12] <alex_joni> icee: less flutes means you need higher RPM, not feed
[21:00:31] <anonimasu> * anonimasu nods
[21:00:35] <icee> I'm just saying what i just read on the freud site :P
[21:00:45] <Jacky^> hehe
[21:01:30] <anonimasu> brb kiosk
[21:01:36] <anonimasu> it's -10c outside now
[21:02:09] <Jacky^> is 1/8 " = 3.17 mm ?
[21:02:19] <Jacky^> it seem to me too small
[21:02:38] <Jacky^> i'd use 10 mm for the first pass :P
[21:02:40] <Jymmm> Jacky^: Sometimes not small enough.
[21:02:58] <Jacky^> maybe
[21:03:06] <Jacky^> i prefer biggest bit
[21:03:30] <icee> http://www.techsavvy.com/industry/file/national/03txp/nap09.html?id=118358&comp_id=03TXP&base_region=*
[21:03:34] <icee> another reference saying what i said:
[21:03:40] <icee> Micro-Grain/Solid Carbide Bits
[21:03:40] <icee> The single flute design facilitates high feed and material removal rates
[21:03:40] <icee> The double flute is designed for smoother finishes
[21:05:06] <icee> http://www.amanatool.com/faqs_st.htm#question8
[21:05:25] <icee> When should I use a single flute, two-flute and three-flute bit?
[21:05:26] <icee> Single flute bits are primarily used for cutting to size when speed is more important than finish. Two and three flute bits are used when finish is more important then speed. The more flutes you use the slower the feed rate
[21:07:27] <Jacky^> I agreed
[21:07:38] <Jymmm> Jacky^ oh you do not!
[21:07:44] <Jacky^> :)
[21:10:09] <alex_joni> hello bob
[21:10:11] <alex_joni> long time no see
[21:12:59] <Jacky^> Jymmm: anyway, it seem to me hard to find best performace changing jobs often
[21:13:35] <Jacky^> if the model is only one, after a bit of 'hacking' is not difficult to fine the best way..
[21:14:08] <Jymmm> Jacky^: It's more getting familure with various cutters and their capabilities
[21:14:10] <Jacky^> there are a lots of toolpath ways :)
[21:14:26] <Jacky^> toolpath is important too ..
[21:14:50] <Jacky^> maybe more than bit used
[21:15:04] <Jacky^> i mean strategy, not sure if term is correct
[21:21:51] <anonimasu> iab
[21:22:17] <Jacky^> its hard enough to made manually a router job, with the cnc and toolpaths we can learn for years :(
[21:22:40] <anonimasu> cnc is easier then manual machining in many cases
[21:22:48] <anonimasu> although skill comes with years of practice :)
[21:22:56] <Jacky^> anonimasu: yeah
[21:23:16] <Jacky^> but who have a 'carpentier' experience have no difficult in many cases
[21:23:48] <Jacky^> at least, if hes familiar with the software
[21:26:23] <Jacky^> anonimasu: how much the price of aluminium there ?
[21:26:28] <anonimasu> I dont know
[21:26:30] <anonimasu> expensive
[21:26:43] <Jacky^> here too
[21:28:01] <anonimasu> or well
[21:28:09] <anonimasu> I can check it later
[21:28:27] <Jacky^> ok..
[21:28:33] <Jacky^> let me know
[21:28:48] <Jacky^> just curious :)
[21:33:32] <anonimasu> Jymmm: http://en.wikipedia.org/wiki/Milling_cutter
[21:39:58] <CIA-6> 03alex_joni * 10emc2/src/emc/iotask/ioControl.cc: changed iocontrol.0.spindle-speed-out to go from -speed to +speed
[21:43:56] <Jymmm> anonimasu ty
[21:54:37] <alex_joni> hey Martin
[21:55:07] <Imperator_> Hi Alex
[21:55:17] <Imperator_> my card is running again :-D
[21:55:37] <Imperator_> i have added a request_mem_region
[21:55:57] <Imperator_> now it is in /proc/iomem also
[21:56:30] <Imperator_> hm, Paul is not here to say thanks
[21:56:57] <alex_joni> cool
[22:02:15] <K`zan> Yea! Driver is "done" :-)!!!
[22:03:28] <Jacky^> K`zan: what driver ?
[22:08:23] <K`zan> AVR/ULN2803 to interface with my (cheapo) steppers and EMC.
[22:08:36] <Jacky^> :)
[22:08:39] <Jacky^> nice
[22:09:14] <Jacky^> right today i received the geckos too :P
[22:09:37] <Jacky^> tomorrow i will tri to run them with a small motor ..
[22:13:46] <K`zan> http://wrlabs.shacknet.nu/~vw/CNC/adriver-M16-1.pdf
[22:14:09] <K`zan> Nice, but WELL out of my budget :-).
[22:14:46] <K`zan> Decided to get the HobbyCNC 3 axis driver, but my steppers don't draw enough.
[22:15:04] <lerman> The way these things are supposed to work is that you look at one signal to change state (say A), and when it does, you look at the state of the other signal. Then you look for the other signal to change state (B in this case) and sample the first signal. By alternating back and forth this way, you guarantee that the number of interrupts (or edge transitions) that you care about is limited...
[22:15:06] <lerman> I just saw the discussion of quadrature counters. It is NOT as easy as it sometimes seems. You should NOT interrupt on both edges of one of the signals. Suppose the encoder hovers around one edge (say there is some vibration). Interrupt, interrupt,...
[22:15:07] <lerman> ...by the rate of rotation (for rotary encoders) or of motion (for linear encoders). Doing this has the downside that it introduces hysteresis of one edge (or one count) into the system. But it is the only way that is guaranteed to work.
[22:15:23] <K`zan> Sooo, until I get "real" steppers (draw over 500mA) I decided to roll my own :-).
[22:15:28] <K`zan> Educational.
[22:16:20] <icee> ler: or you can use hardware that samples B on the edge of A
[22:16:52] <K`zan> Now to see if EMC works with it :-).
[22:16:54] <icee> ler: like the quadrature decoder in dsPIC*
[22:17:44] <alex_joni> lerman: nice to see someone grasping the idea finally :)
[22:19:37] <lerman> alex_joni: I grasped the idea about thirty years ago :-). But this is the first time I've found someone to agree with me. ;-) Just goes to show that great minds think alike.
[22:20:09] <anonimasu> yep
[22:20:53] <alex_joni> * alex_joni mind is telling me not to fool around and buy some decent drives
[22:21:07] <alex_joni> and in case I do something stupid like building my own, to use proper chips
[22:21:13] <alex_joni> like the LS7166 I was talking about
[22:22:12] <lerman> Well, do you want to build a tool or do you want to use a tool? (I suppose you could always build your own chips.)
[22:22:49] <lerman> That's chips as in integrate circuit, not swarf.
[22:25:24] <alex_joni> yes
[22:25:25] <alex_joni> IC
[22:48:56] <K`zan> I'm burning the BDI-4.30.iso I hope that was a good choice :).
[22:49:10] <alex_joni> K`zan: what were the others?
[22:49:34] <K`zan> I don't remember, live and some lesser versions were out there IIRC.
[22:49:46] <alex_joni> what machine?
[22:49:49] <alex_joni> PC I mean
[22:49:53] <K`zan> Yes
[22:50:03] <alex_joni> not Yes..
[22:50:04] <K`zan> Thunderbird 1100
[22:50:08] <alex_joni> ahh.. ok then
[22:50:08] <K`zan> AMD
[22:50:17] <alex_joni> * alex_joni remembers what a Tbird is
[22:50:51] <alex_joni> it's from the old days.. when I could still follow processor talks :D
[22:50:54] <K`zan> :-), scrap pile here, prolly better than the K6/500 :) for this, or anything except possibly a server :-).
[22:50:59] <K`zan> LOL
[22:51:02] <alex_joni> nowaday it's pretty weird :D
[22:51:08] <alex_joni> nowadays
[22:51:27] <alex_joni> about twice as fast as the K6/500
[22:51:32] <K`zan> Still doing sneakernet, but now with DVDs :-)
[22:51:36] <alex_joni> heh
[22:51:39] <K`zan> mount /cd1
[22:51:41] <K`zan> ww
[22:51:57] <alex_joni> * alex_joni just remembered his old K7-550 on a ASUS K7M
[22:52:02] <alex_joni> the first one out there :
[22:52:03] <alex_joni> :D
[22:53:10] <alex_joni> anyways.. going to bed now
[22:53:24] <alex_joni> catch you later
[22:54:11] <Jymmm> G'Night alex_joni
[22:54:31] <alex_joni> Jymmm: bye
[22:54:32] <alex_joni> :D
[22:56:23] <K`zan> basic install ok or is there something in custom I might need ?
[22:56:45] <K`zan> ok, basic it is.
[23:16:12] <dmess> hi
[23:17:36] <Jymmm> bye
[23:40:02] <dmess> high all..
[23:40:13] <Jymmm> back again
[23:40:25] <Jymmm> [14:34:05] <dmess> hi
[23:40:25] <Jymmm> [14:34:27] * dmess (n=dmess@ has left #emc
[23:40:25] <Jymmm> [14:35:29] <Jymmm> bye
[23:40:25] <Jymmm> [14:57:32] * dmess (n=dmess@ has joined #emc
[23:40:25] <Jymmm> [14:57:55] <dmess> high all..
[23:40:25] <Jymmm> [14:58:06] <Jymmm> back again
[23:40:50] <dmess> hmm..
[23:40:58] <Jacky^> hey
[23:41:00] <Jacky^> :)
[23:41:02] <LawrenceG> hey icee got his surplus 1kw ac servo motors from ups
[23:41:06] <dmess> time warp... again...
[23:41:15] <icee> they're awesome. :)
[23:41:51] <Jacky^> LawrenceG: taxes ?
[23:41:53] <LawrenceG> I am jealous
[23:42:06] <icee> well, they're still just large paperweights right now
[23:42:09] <icee> until we figure out the drive ;)
[23:44:02] <Jacky^> LawrenceG: i received my drivers too today ..
[23:44:17] <Jacky^> but E. 103 tax, damn..
[23:44:36] <Jacky^> how it is calculated ?
[23:44:48] <Jacky^> weight ?
[23:44:57] <icee> j: that's up to your customs department
[23:45:04] <Jymmm> usually value
[23:45:06] <LawrenceG> hey its your government... probably a % of value
[23:45:26] <Jacky^> Jymmm: on 420 euro I payed 103 tax..
[23:45:40] <Jacky^> so, should be proportional ?
[23:45:44] <dmess> 23 %
[23:46:01] <Jacky^> and the weight ?
[23:46:04] <Jacky^> nope ..
[23:46:19] <dmess> 26%
[23:46:19] <Jacky^> it just influence the shipping cost right ?
[23:46:22] <LawrenceG> here, we have "free trade" with USA, but we get dinged for 7% provincial tax + 7% gst (goods and services tax)
[23:46:32] <icee> 20-30% is in the typical european VAT ballpark, right?
[23:46:46] <dmess> lawrenceG.. where are you??
[23:46:56] <Jacky^> i choosed ups, 67 $ few days from CA to IT
[23:46:57] <LawrenceG> Canada, Vancouver Island
[23:47:24] <dmess> ok im in oshawa .near toronto
[23:47:57] <LawrenceG> cool... did you buy any of the infamous motors
[23:48:00] <Jacky^> dmess: nice ive an haunt nera there :)
[23:48:10] <Jacky^> near*
[23:48:17] <dmess> im wondering... WHAT to buy
[23:48:24] <dmess> where??
[23:48:30] <Jacky^> dmess: wht you need ?
[23:48:50] <dmess> i haunt the province of ontario
[23:48:56] <Jacky^> dmess: the city is Hamilton
[23:49:07] <dmess> you can have it
[23:49:13] <LawrenceG> http://www.surpluscenter.com/sort.asp?UID=2005102417040240&catname=&byKeyword=yes&search=sanyo
[23:50:25] <LawrenceG> the 300 watt w/o brakes, the 400w and the 1kw versions are sold out
[23:51:22] <Jacky^> dmess: aunt :)
[23:51:33] <Jacky^> aunt elena hehe
[23:52:09] <Jacky^> anyone already tried that motor ?
[23:52:11] <icee> in the 2 feet ballpark
[23:52:12] <icee> er
[23:52:50] <icee> i have the motor here, and i've energized the brake so far.. will look at the encoder later tonight or tomorrow
[23:53:12] <Jacky^> good torque ?
[23:53:18] <icee> haven't fired up the motor yet
[23:53:20] <icee> have no drive.
[23:53:36] <Jacky^> thats the prob
[23:53:54] <Jacky^> but the motor should be great
[23:53:56] <Jacky^> :)
[23:54:23] <Jacky^> i'm looking around for dc motors..
[23:54:25] <icee> well, we're making a drive
[23:54:32] <icee> should be fairly cheap.
[23:54:33] <dmess> just remember to put the BRAKE OFF into the GO command... or its NOT nice electronically.
[23:54:43] <icee> supposedly the motor is good for 12 N-m instantaneously (the 1kW one)
[23:54:50] <Jacky^> icee: homebuilt driver ? nice
[23:55:05] <icee> dm: well, we have current limiting etc
[23:55:32] <icee> and it should become a following error and stop very quickly in emc
[23:55:33] <dmess> still cook SOMETHING... even a relay..
[23:55:55] <Jacky^> chicken ?
[23:56:04] <Jacky^> ugh
[23:56:43] <dmess> i gotta build the pIII-2.8 for this box... that surplus page just come up
[23:58:10] <dmess> you have these drives jacky??
[23:58:28] <Jacky^> ac drivers ?
[23:58:32] <Jacky^> nah ..
[23:58:36] <Jacky^> never seen
[23:58:47] <Jacky^> i bought 3 G340
[23:59:04] <Jacky^> and rousted a lot of L298
[23:59:08] <Jacky^> lol :)
[23:59:58] <Jacky^> but never tried with chicken ..