#emc | Logs for 2007-10-06

[00:00:00] <Skullworks-PGAB> that is - unless you were on the receiving end - like me.
[00:08:17] <Skullworks-PGAB> logger_emc bookmark
[00:08:51] <Skullworks-PGAB> logger bookmark
[00:09:01] <Skullworks-PGAB> guess not
[00:09:08] <toast> bookmark
[00:09:18] <toast> logger_emc: die in a fire
[00:09:18] <toast> I'm logging. I don't understand 'die in a fire', toast. Try /msg logger_emc help
[00:09:18] <cradek> logger_emc: bookmark
[00:09:18] <cradek> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-10-06.txt
[00:09:29] <cradek> * cradek kicks logger_emc
[00:10:59] <Skullworks-PGAB> was snoozing on the job - nothing before I joined
[00:44:39] <tomp> dovetail exit along the half-angle?
[00:46:44] <cradek> yeah that was the general idea, 60 deg dovetail, exit at 30 deg
[00:47:01] <tomp> right, that wont scar :)
[00:47:16] <cradek> on a bridgeport boss it sure will :-)
[00:47:41] <cradek> (if you use G00)
[00:48:32] <tomp> oh, dont do that, you should be able to move pretty fast G01 XYZ far enuf before moving in rapud
[00:48:41] <tomp> rapid
[00:49:02] <tomp> got a labjack today :) 8 adc's and 1 dac , usb
[00:49:47] <cradek> a while back, someone proposed that emc should do what the bp does with G0, because it's "safer" - that's the only reason I brought it up
[00:50:14] <tomp> the dogleg thing?
[00:50:28] <cradek> no it makes it into two moves, straight up, then straight over
[00:50:56] <toast> that's slow
[00:50:56] <tomp> oh, tool axis the g1x plane maybe
[00:51:17] <Skullworks-PGAB> several machines had that option - to split up any XYZ rapid
[00:51:31] <tomp> /then g1x
[00:51:37] <Ziegler> you could do it in the code couldnt you
[00:51:37] <toast> how do you decide which moves first
[00:51:50] <Skullworks-PGAB> some you could change a parameter
[00:51:50] <Ziegler> g0 z[up]
[00:51:59] <cradek> toast: it goes up then over, or over then down
[00:51:58] <Ziegler> g0 x[x]y[y]
[00:52:07] <toast> how does the machine know that's a clear move
[00:52:21] <toast> could be using a slotter
[00:52:36] <cradek> toast: it is always safe with an end mill, but yeah other cutters are bad news
[00:52:46] <cradek> that's why it's clearly (in my opinion) a bad idea
[00:52:53] <toast> i agree
[00:52:56] <cradek> should do what the program says
[00:53:02] <toast> if you want to move one axis at a time, program it that way
[00:53:27] <toast> that is my theory.
[00:54:03] <Skullworks-PGAB> I like all axis moving at same time - extra scary at 1500ipm
[00:54:05] <Ziegler> fwiw (if anything) programing it the way you want it to move makes sence to me
[00:55:33] <cradek> it's no fun if nobody wants to argue the other side
[00:55:43] <cradek> (I don't remember who that even was)
[00:55:50] <toast> some other guy who isn't here
[00:56:02] <Ziegler> heh
[00:56:08] <tomp> T: Bus=01 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 10 Spd=12 MxCh= 0
[00:56:08] <tomp> D: Ver= 1.10 Cls=ff(vend.) Sub=ff Prot=ff MxPS=16 #Cfgs= 1
[00:56:08] <tomp> P: Vendor=0cd5 ProdID=0003 Rev= 0.00
[00:56:08] <tomp> S: Manufacturer=LabJack
[00:56:08] <tomp> S: Product=LabJack U3
[00:56:09] <tomp> C:* #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr= 0mA
[00:56:11] <tomp> I: If#= 0 Alt= 0 #EPs= 4 Cls=ff(vend.) Sub=ff Prot=00 Driver=(none)
[00:56:11] <cradek> oh well
[00:56:11] <tomp> E: Ad=01(O) Atr=02(Bulk) MxPS= 64 Ivl=0ms
[00:56:13] <tomp> E: Ad=82(I) Atr=02(Bulk) MxPS= 64 Ivl=0ms
[00:56:14] <toast> uh oh
[00:56:17] <tomp> E: Ad=03(O) Atr=02(Bulk) MxPS= 64 Ivl=0ms
[00:56:17] <tomp> E: Ad=83(I) Atr=02(Bulk) MxPS= 64 Ivl=0ms
[00:56:19] <tomp> labjack is alive, usb analog i/o, now for some fun stuff, get hal to notice it
[00:56:59] <toast> i don't like it when people ask for machines to be smarter than they are
[00:57:05] <toast> that, to me, implies trouble
[00:59:26] <Ziegler> I dont want my machine to climb into my room at watch me sleep at night
[01:00:05] <Ziegler> 0_o
[01:00:23] <Skullworks-PGAB> but machine vision does have its place...
[01:00:46] <Skullworks-PGAB> we use it to be sure the robots have not mis loaded
[01:01:55] <Skullworks-PGAB> also the check that the part has complete the punch process
[01:09:02] <toast> that's a fine, predictable process
[02:25:27] <tomp> is there a python version of hal_joystick.c ?
[02:26:02] <tomp> re: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-06-22.txt where 'hal_joystic' was cited as a model for a user space driver
[02:31:01] <fenn> i think the general idea is to use hal_input.py
[02:33:05] <tomp> tanks
[02:34:13] <fenn> joystick would be an absolute type input i think
[02:34:25] <fenn> then again it might not work with hal_input at all..
[02:35:34] <Skullworks-PGAB> from the Gecko yahoo group
[02:36:05] <Skullworks-PGAB> Posted by: "Mariss Freimanis" Tue Oct 2, 2007 10:44 am (PST) Tom,
[02:36:03] <Skullworks-PGAB> We are planning a reorganization of our tech support in January. We
[02:36:03] <Skullworks-PGAB> will be setting up a phone support staff composed entirely of
[02:36:03] <Skullworks-PGAB> non-English speaking Latvians unreachable via our new endlessly
[02:36:03] <Skullworks-PGAB> recursive automated 900-number phone menu system at $3 a minute. It's
[02:36:04] <Skullworks-PGAB> about time we enter the 21st century.:-)
[02:36:08] <Skullworks-PGAB> Mariss
[02:36:35] <SWPadnos> I, for one, welcome out Latvian-speaking gecko-support overlords
[02:36:38] <fenn> hey whatever works
[02:36:39] <SWPadnos> our
[02:37:12] <SWPadnos> hal_input should work with any joystick supported by Linux
[02:37:22] <Skullworks-PGAB> I keep looking at those little geckos, they just seem too small
[02:37:45] <SWPadnos> it supports bit inputs and output (buttons and lights, respectively), and absolute and relative "analog" inputs
[02:38:07] <SWPadnos> well, they're nice and compact, but it's true they don't have adequate cooling for their rated power
[02:38:07] <fenn> what's 'EV_SYN'?
[02:38:19] <SWPadnos> event something
[02:38:27] <tomp> thanks, i have a labjack, a usb device with 8 adc's and one dac ( besides other stuff ). it has non rtai src for drivers with kernel 2.6 http://www.labjack.com/labjack_u3.php?prodId=25
[02:38:34] <fenn> yeah its the something i was wondering about.. :)
[02:39:04] <SWPadnos> http://www.gelato.unsw.edu.au/lxr/ident?i=EV_SYN
[02:39:26] <SWPadnos> tomp, if it acts like an 8-axis joystick, then it's no problem
[02:39:40] <SWPadnos> if not, then you're in for a world of fun
[02:40:20] <fenn> "Synchronization events" oh well
[02:40:37] <tomp> i'm looking at hal_joystick ( seems to depend on the system recognizing it as /dev/input/jsN , this device is listed in /proc/usb/ but not in /dev
[02:41:19] <SWPadnos> I think anything on the USB bus ends up in /proc/usb (not positive though)
[02:41:29] <fenn> is it in /sys/bus/usb/devices?
[02:41:48] <fenn> cat /sys/bus/usb/devices/*/id*
[02:42:00] <fenn> er, you probably dont know the id in hex :)
[02:43:31] <fenn> ok, look on lsusb for the hex code
[02:46:04] <tomp> lsusb sees it ... i do it again
[02:46:18] <fenn> lsusb doesnt necessarily mean there's a driver
[02:46:27] <tomp> Bus 001 Device 017: ID 0cd5:0003
[02:46:38] <Ziegler> im using geckos on my new machine... heck of an upgrade from HobbyCNC
[02:46:39] <tomp> thats it from the mfctr id & dev id
[02:47:23] <tomp> right, no driver, need to write one, thought there was a away to do it in python per that irc conversation
[02:48:05] <fenn> ziegler can you really tell the difference?
[02:48:29] <fenn> i guess the motors would run a lot cooler
[02:48:36] <Ziegler> more stable and smoother
[02:48:50] <fenn> can you get higher accels out of it now?
[02:49:05] <Ziegler> well yes.. voltage allowance difference too
[03:04:15] <tomp> [526128.569239] /home/tomp/tmp/hdwr/labjack/src-mfctr--Linux_C_NativeUSB_U3UE9/Driver/U3/labjacku3.c: USB LabJack device now attached to labjacku3-208
[03:04:17] <tomp> [526128.569250] usbcore: registered new driver labjacku3
[03:04:17] <tomp> [526128.569254] /home/tomp/tmp/hdwr/labjack/src-mfctr--Linux_C_NativeUSB_U3UE9/Driver/U3/labjacku3.c: USB LabJack U3 Driver v0.32
[03:04:17] <tomp> woohoo
[03:04:59] <jmkasunich__> cleveland 2, new york 1, in 11 innings.... ;-)
[03:05:05] <Skullworks-PGAB> tomp - what you toying with?
[03:05:08] <fenn> oh i thought you were trying to get a joystick working.. silly me
[03:05:29] <tomp> http://www.labjack.com/labjack_u3.php?prodId=25
[03:06:27] <Skullworks-PGAB> so I gathered - but your goal?
[03:06:58] <tomp> read analog of a process, control speed and direction of a motor ( EDM )
[03:08:26] <Skullworks-PGAB> ah
[03:08:28] <Skullworks-PGAB> sinker?
[03:10:57] <Ziegler> woah.. what is the labjack (laymans terms)
[03:11:46] <fenn> reads a voltage and turns it into a number
[03:13:51] <Ziegler> ah
[03:14:07] <tomp> analog and digital i/o, timer and counter on a 2x3 usb box uhoh, it sez its 203 degrees kelvin in here !
[03:14:44] <tomp> Iteration 18
[03:14:44] <tomp> AIN0(SE) : 0.385 volts
[03:14:44] <tomp> AIN1(SE) : 0.333 volts
[03:14:44] <tomp> AIN0(Neg. chan. 1) : 0.038 volts
[03:14:44] <tomp> AIN1(Neg. chan. 0): 0.039 volts
[03:14:45] <tomp> AIN0(Neg. chan. Vref) : -0.452 volts
[03:14:49] <tomp> AIN1(Neg. chan. Vref): -0.270 volts
[03:14:51] <tomp> FIO3 state : 1
[03:14:53] <tomp> Counter1(FIO6) : 0
[03:14:55] <tomp> Temperature : 306.130 K
[03:15:01] <fenn> you better hope it's not 203K
[03:15:10] <tomp> nothings hooked up yet ;) it's reading air
[03:15:13] <SWPadnos> 306 is much nicer
[03:15:50] <SWPadnos> I wonder what microcontroller they're using
[03:16:16] <tomp> and i didnt know theres another db15 with more i/o on it nto bad 99$ toy
[03:16:31] <fenn> what's the sample rate?
[03:16:34] <tomp> hope i can hack their code into a hal driver
[03:16:44] <tomp> dunno, time to rtfm
[03:17:16] <fenn> 1.2-ksamples/s
[03:17:28] <SWPadnos> if you can read from a device or proc file, then making a HAL (userspace) driver should be pretty easy
[03:17:57] <SWPadnos> making a HAL driver that also talks over USB directly is another story
[03:18:09] <Ziegler> heh
[03:18:23] <fenn> or 8192 samples/s in burst mode?
[03:26:47] <tomp> SWPadnos: so if 'it' is /dev/usb/labjack0 and it's driver has init/read/write funcs.... then i can make a user space driver?
[03:32:58] <fenn> tom have you tried loadusr hal_input -A /dev/usb/labjack0
[03:35:56] <tomp> will try
[03:37:31] <fenn> from the man page i get the impression i should be able to load my mouse into hal, but i get permission denied
[03:39:02] <Ziegler> woah that sounds interesting
[03:40:11] <SWPadnos> the device can't be in use by another app, and I think X is set up by default to use any pointing devices that get plugged in
[03:40:52] <fenn> poo
[03:41:06] <fenn> i can cat the file; think i could set up a pipe?
[03:41:11] <Ziegler> Did I hear correctly that someone managed to use a joystick to jog?
[03:41:17] <SWPadnos> fenn, dunno
[03:41:22] <SWPadnos> Ziegler, sure, that's been done
[03:41:25] <fenn> my knowledge of named pipes is a bit flaky to say the least :)
[03:41:39] <SWPadnos> don't remember who did it though (xemet, I think)
[03:41:40] <Ziegler> how much more of a stretch would it be to use a mouse?
[03:41:55] <SWPadnos> it depends on whether the mouse is in use now, doesn't it? ;)
[03:42:51] <Ziegler> keyboard is technically in use isnt it?
[03:43:19] <SWPadnos> yes, but you're not using the hal_input driver for it, you're using "keyboard events"
[03:43:29] <tomp> fenn: bin/halcmd loadusr hal_input -A /dev/usb/labjacku3_0 yields LookupError: No input device matching '/dev/usb/labjacku3_0' was found
[03:43:29] <tomp> yet ls /dev/usb gives tomp@cncbox:~/emc2.1.6$ ls /dev/usb yields one entry labjacku3_0
[03:43:40] <Ziegler> I see SWPadnos
[03:44:31] <tomp> Ziegler: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Using_A_Joypad_To_Move_Your_CNC_Machine
[03:44:54] <SWPadnos> tomp, that may be a poorly worded error for "the device you told me to use is not a HID, so I can't tell what the hell to do with it" :)
[03:44:55] <Ziegler> thanks tomp... I was googling it but had not found it yet
[03:45:50] <fenn> grr why cant hal_input just read a device file
[03:46:03] <SWPadnos> because there's no standard for whata "device file" tells you
[03:46:12] <SWPadnos> there is a standard for how a HID communicates
[03:46:46] <tomp> hal_input is only for HID devices?
[03:47:01] <SWPadnos> correct
[03:47:03] <fenn> i made a pipe 'event9' to try to fake it out
[03:47:08] <fenn> but i get No input device matching '9' was found
[03:47:32] <SWPadnos> dunno, but I think you'll have a heck of a time typing in binary HID packets anyway
[03:47:59] <fenn> i'm not typing them, its cat'ing from event1
[03:48:17] <SWPadnos> ok - whichever
[03:48:20] <fenn> cat event1 | sudo tee -a event9
[03:48:29] <fenn> then cat event9 gives me a stream of gunk matching the mouse movement
[03:48:46] <fenn> but i dont understand how hal_input is searching for names
[03:48:49] <Ziegler> hehe cool
[03:49:34] <Ziegler> could you trick hal_joystick to think your mouse is a joystick?
[03:49:46] <fenn> no
[03:49:57] <fenn> well, maybe, but it wouldnt work the way you want
[03:51:04] <fenn> mouse is a relative position sensor, joystic is an absolute position sensor
[03:56:09] <tomp> does HID have a driver? its not listed by lsmod and the docs on labjack say it is classified as an HID device but doesnt work with the HID driver because "the hid driver does not support the "Interrupt" out endpoint used by the Labjack and thus cannot communicate with the Labjack" reference: http://search.cpan.org/src/CDRAKE/Device-LabJack-0.02/linux-labjack/INSTALL
[03:56:37] <tomp> /labjack is classified as an HID....
[04:00:53] <SWPadnos> tomp, did you notice the LabJackPython.zip file on their software downloads page?
[04:01:13] <tomp> no :-[
[04:01:19] <SWPadnos> bonehead :)
[04:08:30] <toast> toast is now known as toastydeath
[05:19:59] <user_> user_ is now known as Skinnypuppy
[05:54:37] <toast> toast is now known as toastydeath
[09:52:48] <_epineh> hey all
[09:54:21] <fenn> noone here but us chickens
[10:00:32] <_epineh> ha
[10:01:47] <_epineh> the ol time zone difference is always a prob...
[11:14:30] <wh173y> hello there
[12:27:52] <jlmjvm> good morning
[13:38:36] <jmkasunich_> jmkasunich_ is now known as jmkasunich
[13:56:17] <Ziegler> fenn: any luck getting the mouse to work?
[14:44:18] <jlmjvm> jepler:my keyboard isnt working with axis for some reason,the r,p,and s keys arent working
[14:44:52] <jlmjvm> and now they are?
[14:50:00] <jlmjvm> never mind
[15:59:50] <ADucci> hey.
[16:03:19] <ADucci> Trying to calibrate my steppers, i how do i set the "Steps Per" setting?
[16:18:13] <SWPadnos> ADucci, change INPUT_SCALE in the ini file
[16:18:20] <SWPadnos> if you're asking the question I hink you're asking
[16:18:23] <SWPadnos> think
[16:21:36] <ADucci> ok, thanks, just read that in the help file... the only thing, i dont really know how may steps per, on my motor or gear ratio...
[16:22:07] <SWPadnos> can you hand-crank the table?
[16:22:59] <ADucci> mmm, good idea, i think i can.. ps thanks for the "s" command, got the plasma unit cutting.
[16:23:10] <SWPadnos> cool!
[16:23:15] <SWPadnos> err hot! :)
[16:23:56] <ADucci> so, turn one revolution, and measure the distance moved!
[16:24:03] <SWPadnos> steppers are almost all 200 steps per rev. multiply that by any half/quarter/microstepping the drive does
[16:24:09] <SWPadnos> yep
[16:24:17] <SWPadnos> or 10 - decreases error
[16:38:29] <tomp> labjackpython works, while emc2 runs, now to get hal communications to work (pass data to/from hal comps to labjack)
[16:40:40] <SWPadnos> import hal or some such :)
[16:46:36] <tomp> thx
[17:07:42] <tomp> whats version of python will be used by the upcoming emc2 release? labJackPython uses ctypes (with RH & MIT licensing) on python <2.6, while ctypes is builtin to python2.6 and (i hope) gpl'd
[17:09:20] <SWPadnos> apparently, it's supposed to be included in python 2.5 as well
[17:09:32] <SWPadnos> according to this: http://python.net/crew/theller/ctypes/
[17:12:08] <tomp> great must've been <2.5
[18:49:56] <jlmjvm> is classic ladder in emc,or does it need to be downloaded?
[19:59:17] <alex_joni> jlmjvm: it is in emc2
[20:27:56] <alex_joni> good night all
[20:37:01] <jepler_> jepler_ is now known as jepler
[20:48:40] <JymmmEMC> I swear i never can figure out why rewiring an outlet is always so involved.
[20:57:40] <tomp2> it's called 'previous creativity' :)
[21:12:00] <Skullworks-PGAB> Jym try doing that at a gas station....
[21:24:18] <kollabierer> hi
[21:27:41] <JymmmEMC> Skullworks-PGAB: lol, no thanks =)
[21:33:26] <SWPadnos> JymmmEMC, how big is your router?
[21:33:41] <JymmmEMC> SWPadnos: 24 x 24 x 5.5
[21:33:46] <JymmmEMC> travel
[21:33:55] <SWPadnos> hmmm. what does it weigh?
[21:33:59] <Skullworks-PGAB> (they pour epoxy into the box to seal the conduit as a path for explosive vapors)
[21:34:26] <JymmmEMC> SWPadnos: I COULD pick it up, but it's aweful bulky to do so
[21:34:55] <SWPadnos> ok, so less than 200 pounds for sure, movable by 2 people
[21:35:01] <Skullworks-PGAB> block & tackle
[21:35:31] <JymmmEMC> SWPadnos movable by one if needed
[21:36:01] <SWPadnos> ok. any chance I could borrow it (and you, if you like) fora demo at ESC next year?
[21:36:10] <JymmmEMC> ESC ?
[21:36:22] <SWPadnos> the House of Prime Rib conference I go to every year :)
[21:36:30] <SWPadnos> err - Embedded Systems Conference
[21:36:57] <JymmmEMC> ah, sure (Disclaimer, you break it you bought it of course =)
[21:37:07] <SWPadnos> no problem. it's what, $200? ;)
[21:37:25] <JymmmEMC> SWPadnos: try more like $6000
[21:37:38] <SWPadnos> yeah, right. for something one person can move? I think not!
[21:37:40] <SWPadnos> :)
[21:37:57] <JymmmEMC> The spindle alone is $250
[21:38:11] <SWPadnos> no problem. I'm pretty careful
[21:38:34] <SWPadnos> I'm not sure I'd cut anything anyway, just demo realtime control with something people can see
[21:39:02] <SWPadnos> the "ping-pong ball supported by a fan" demo gets old pretty quickly
[21:39:10] <JymmmEMC> SWPadnos: For the record, I really do have about $6000 in it.
[21:39:27] <SWPadnos> I don't doubt it. I have close to $15k into my shop
[21:40:24] <JymmmEMC> SWPadnos: it needs to be on a roughyl 4'x4' table
[21:40:48] <JymmmEMC> I have it sitting on a gorilla rack now
[21:41:11] <SWPadnos> ok. I'll need to see if that's an OK size - I may need to get a sherline or something small
[21:41:39] <JymmmEMC> let me know
[21:41:58] <SWPadnos> will do. the show is in April, so there's a fairly long time to decide the specifics
[21:42:02] <SWPadnos> thanks, just in case :)