#emc | Logs for 2008-10-06

[00:30:00] <steves_logging> steves_logging is now known as steve_stallings
[02:07:18] <steve_stallings> steve_stallings is now known as steves_logging
[02:13:36] <skunkworks_> have one side of the circuit board clear milled.
[02:14:17] <skunkworks_> Only had a bit of time. Working well - I used the latest gcode.ulp for eagle - cradek: works awsome.
[10:09:42] <anonimasu> hmm..
[10:09:55] <anonimasu> anyone have a nice circuit to convert a 24v encoder signal to 5v?
[10:14:36] <archivist_ub> if open collector, trivial
[10:15:16] <archivist_ub> else diodes and resistors
[10:16:44] <KimK_IA_AFK> KimK_IA_AFK is now known as KimK_IA
[10:17:44] <archivist_ub> cheap diode is 1n914 or 1n4148
[10:18:33] <KimK_IA> a comparator (LM339?) could add hysteresis
[10:19:15] <KimK_IA> how fancy do you need to be?
[10:21:24] <archivist_ub> that depends on speed and any noise in system and if inputs already have schmitt
[10:23:31] <anonimasu> hmm.. grckos..
[10:23:34] <anonimasu> geckos..
[10:24:03] <anonimasu> mabe I should build a differential opamp stage..
[10:24:25] <anonimasu> and compare agaionst 21v
[10:24:38] <anonimasu> that err 19..
[10:25:06] <anonimasu> 24-19 = 5v out
[10:25:17] <anonimasu> archivist_ub: low speed and 500ppr..
[10:25:22] <archivist_ub> diodes and R to limit range and LM339 to clean
[10:25:37] <anonimasu> I'd like to test one of my servos with a encoder I have before I order new encoders
[10:26:03] <anonimasu> I were thinking lm321
[10:26:11] <anonimasu> as I have thoose in my drawer
[10:26:25] <anonimasu> they handle 32v conviniently enough :)
[10:27:49] <archivist_ub> dunno that part number, but at some part the range needs limiting
[10:28:15] <anonimasu> well, if I use it as a comparator
[10:28:40] <anonimasu> hm, I have zeners too. :p
[10:28:49] <anonimasu> mm.. zener + resistor might be a ides..
[10:28:50] <anonimasu> idea..
[10:28:56] <anonimasu> r-Z
[10:29:05] <anonimasu> err resistor>zener
[10:29:24] <archivist_ub> or diode using 5v rail
[10:30:28] <anonimasu> I dont know how that would look(i suck at this)
[11:01:31] <anonimasu> toastydeath: hey
[11:02:54] <KimK_IA> anonimasu: archivist_ub is right, do your inputs already have a schmitt trigger? If so, maybe you can modify or extend them to cover the 24V encoders: see http://en.wikipedia.org/wiki/Schmitt_trigger
[11:11:03] <coden4life> lost as ever..... step conf holding me back
[11:13:24] <coden4life> if I set pin 4 to amplifier enable and pin 7 to e stop out ..... when I test axis x and y they jog forwards and back ..... anyone tell me how to set parallel pins 4, 7, and 14 to always be high.... jmkasunich gave me a few clues but everytime I edit my hal file emc won't load.....
[11:14:12] <coden4life> the reason I set those pins to amplifier enable and e stop out was to get a logic high on pin 4 and 7
[11:16:07] <KimK_IA> KimK_IA is now known as KimK_IA_AFK
[11:17:04] <archivist_ub> coden4life, are you doing all edits with stepconf or hand editing afterwards
[11:18:14] <archivist_ub> note stepconf overwrites hand edits to the inifiles
[11:20:40] <coden4life> hand afterwords
[11:21:14] <archivist_ub> setting estop out to multiple pins is easy enough in stepconf as far as I can see
[11:22:07] <coden4life> all I am really trying to do is set pins 4, 7 and 14 to always be a 1 or logic high
[11:23:01] <coden4life> I was going to see if running a c program I wrote using parapin to just set em high but I am not sure if emc clears all the pins when it runs
[11:23:59] <archivist_ub> or just use the "amplifier enable" signal option in stepconf
[11:24:35] <coden4life> when I do that for all the pins the only axis that jogs when testing the steppers is the x axis
[11:25:04] <coden4life> that is why I set pin 7 to estop so I had a logic high there too
[11:25:35] <coden4life> all the pins meaning 4, 7 and 14
[11:25:41] <archivist_ub> go to parallel port setup screen and select as needed signal for each pin and make sure invert is set as needed
[11:41:55] <SWPadnos> coden4life, you can set those pins always high by adding a couple of lines in the "custom.hal" file that stepgen makes for you
[11:42:07] <SWPadnos> that file doesn't get overwritten
[11:42:39] <SWPadnos> if those are amplifier enables though, you should think about whether you want them always on when EMC is running
[11:43:21] <SWPadnos> they would turn on when you start EMC and remain on until you exit, regardless of estop, machine on/off, etc.
[11:43:35] <SWPadnos> the lines you need are basically what jmkasunich emailed:
[11:44:14] <SWPadnos> net one parport.0.pin-04-out parport.0.pin-07-out parport.0.pin-14-out
[11:44:17] <SWPadnos> sets one 1
[11:44:34] <SWPadnos> bye
[11:46:17] <BigJohnT> morning guys
[11:46:36] <SWPadnos> hiya
[11:46:43] <micges> hi
[11:58:28] <BigJohnT> * BigJohnT heads out to work
[11:58:33] <BigJohnT> see you guys later
[12:10:52] <SWPadnos> hello again
[12:11:22] <coden4life> wife needed to reboot into windows to print tests out for her students today
[12:11:23] <coden4life> lol
[12:11:43] <SWPadnos> just access the windows partition with Linux and use openoffice :)
[12:12:05] <coden4life> couldn't the format is from a program called examvies
[12:12:10] <coden4life> examview
[12:12:21] <SWPadnos> oh. then you'd need to use wine
[12:12:26] <coden4life> going to set her up with wine when she gets
[12:12:33] <SWPadnos> and install and all that, so it's a lot easier to reboot ;)
[12:12:36] <coden4life> the next textbook with the install disk
[12:14:33] <coden4life> so I found something strange... when I setup with step wiz I had to have pin 4,7, and 14 set to amplifier enable... but to test the motors I had to check invert on pins 7 and 14 to test the jog...... when I ran emc after everything worked the y and z axis would not jog manuallly
[12:15:09] <SWPadnos> I'd bet that stepconf only turns on the enable for the axis it's testing
[12:15:14] <coden4life> went back and set pins 7 and 14 to not be inverted and then when I ran emc they jogged
[12:15:29] <SWPadnos> it's uncommon to need all the enables on to run one motor
[12:16:26] <coden4life> I tried a real quick mdi and all axis worked ... motors getting warm though
[12:16:57] <SWPadnos> are those 3 pins the enables for the 3 axis motor drives?
[12:17:45] <coden4life> ya the manufacturer config for k-cam has those pins set for axis enable x,y,z
[12:18:18] <coden4life> I could not get K-Cam to work in wine so I tried on a windows box and no luck there
[12:18:19] <SWPadnos> huh. but you need all of them high to allow any motor to move??
[12:18:25] <coden4life> yup
[12:18:38] <coden4life> the manual says that those need to be logic high
[12:19:25] <SWPadnos> well, with EMC running, the machine-on output and all motor enable outputs will go high when you go out of estop and into machine on state (F1/F2)
[12:19:50] <SWPadnos> so you should be able to connect the 3 pins to the 3 amplifier enable signals and have it work
[12:20:31] <SWPadnos> but it may not work in stepconf testing
[12:21:57] <coden4life> http://cgi.ebay.com/AIO-500-STEPPER-MOTOR-SYSTEM-WITH-3-MOTORS-CNC_W0QQitemZ320307133111QQcmdZViewItem is the deal I bought
[12:23:05] <coden4life> now how does touching off the tools work or do I have to manually set the tool.tbl file after I get my readings from the position screen (view)
[12:25:38] <coden4life> being a programmer (using c and gtk mostly) adding to the tkemc interface to allow for a calbrate or measure of the tool would be nice.
[12:27:45] <SWPadnos> patches gratefully accepted ;)
[12:27:55] <SWPadnos> you do have to edit the tool table file manually
[12:28:22] <SWPadnos> there should be a menu item in AXIS and tkemc to do that. I think mini has a table editor built in (a simple text editor)
[12:29:20] <coden4life> So is it possible to look at the code on the Brothers at work to borrow thier emc setup / gui
[12:29:22] <coden4life> lol
[12:29:34] <SWPadnos> uh
[12:29:35] <coden4life> maybe dd the entire drive
[12:29:37] <SWPadnos> what? :)
[12:30:17] <coden4life> ya the Brother Mills at work when they boot you can see all the verbose linux boot shit .... and EMC stuff loading
[12:30:27] <SWPadnos> huh. interesting
[12:30:48] <SWPadnos> well, if they're using the AIO-500 stepper kit, you could probably copy the config ;)
[12:31:06] <coden4life> just wanna see the gui code
[12:31:23] <SWPadnos> how old are the machines?
[12:31:27] <SWPadnos> (or the retrofits)
[12:31:29] <coden4life> new
[12:31:37] <coden4life> like the latest models
[12:31:57] <SWPadnos> hmmm. interesting. can you tell if it's EMC2 or a modified EMC?
[12:32:25] <SWPadnos> the original EMC was public domain, but EMC2 is GPL, so they'd have to give you the source if it's EMC2
[12:33:57] <coden4life> Brother TC - 32B
[12:34:06] <SWPadnos> out of curiosity, what's so great about their GUI?
[12:35:24] <coden4life> nothing really it is pretty plain but they have lots of screens that do lots of well things
[12:35:43] <SWPadnos> hmmm. interesting. the screenshots look like old AutoCAD
[12:35:52] <SWPadnos> but they're small so it's hard to tell
[12:36:23] <coden4life> I don't run em I am stuck on a dual spindle / turret okuma lathe and some hitachi sieki's and couple tsungami's
[12:36:33] <SWPadnos> heh
[12:36:45] <SWPadnos> I'm stuck on a Bridgeport that I ahven't gotten around to retorfitting yet ;)
[12:36:49] <SWPadnos> (it's only been 3 or 4 years)
[12:36:59] <coden4life> lol
[12:37:12] <archivist_ub> get on with it SWPadnos !
[12:37:21] <SWPadnos> yeah yeah - I'm getting to it :)
[12:37:22] <archivist_ub> :)
[12:38:04] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Simple_EMC_G-Code_Generators
[12:38:10] <archivist_ub> I admit using, gets in the way of upgrading
[12:38:11] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?G-Wiz
[12:38:20] <SWPadnos> oh, I don't use it much either :(
[12:38:47] <SWPadnos> I really should figure out how to prevent rust a little better - being in the garage in New England doesn't help much
[12:38:52] <SWPadnos> the unheated garage
[12:39:35] <archivist_ub> yes that is a problem, there are some nice preventative oils/greases out there
[12:39:51] <coden4life> I am retrofitting like a WW2 drill press lol
[12:40:24] <Guest954> logger_emc: bookmark
[12:40:24] <Guest954> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-10-06.txt
[12:42:39] <coden4life> been looking at old equiptment on ebay.... but I don't know if I could do an upgrade on an older machine
[12:42:58] <coden4life> like getting a mazak to thread
[12:44:06] <SWPadnos> that was fun
[12:44:17] <coden4life> the old machines are nice but how big of a pain is it to get a nice interface hooked up instead of LEDs and dials
[12:44:53] <SWPadnos> it's relatively easy. you rip out everything from the motor drives back to the UI and replace it with a PC and EMC2
[12:45:12] <SWPadnos> insert additional hardware as needed (like a Mesa/Motenc card)
[12:45:19] <archivist_ub> ex cnc will already have ballscrews and servos but may be worn
[12:46:07] <archivist_ub> one with readout may have old acme screws and loads of backlash
[12:47:25] <archivist_ub> note school type educational cnc is likely to have acme screws and steppers
[12:48:01] <archivist_ub> Boxford and Denford can be/are in that class
[12:48:44] <archivist_ub> stick the intarweb back together!
[13:14:02] <christel> [Global Notice] Hi all, Apologies for taking you on a detour via Splitsville this morning, the issues were a result of late-scheduled maintenance at NERO. It should all be done now and we'll strive to ensure that we can give you advance warning next time! Thank you for using freenode and apologies for the inconvenience caused.
[14:01:26] <toastatwork> i need to find a good tattoo artist
[14:02:11] <archivist_ub> cnc tattoo?
[14:02:59] <alex_joni> ouch
[14:03:24] <archivist_ub> * archivist_ub does not want holes not provided in manufacture of self
[14:04:18] <toastatwork> hah
[14:04:27] <toastatwork> wasn't someone actually working on that
[14:04:59] <toastatwork> i recall hearing an article that they successfully did a tattoo on pig skin (dead)
[14:11:50] <archivist_ub> we have made a thousand cores for tattoo machines here (local coil winders did the coil)
[14:13:11] <toastatw1rk> cool!
[14:31:10] <renesis> cnc tattoo wouldnt be hard with feedback
[14:40:11] <Dallur> just painfull
[14:40:20] <toastatw1rk> renesis: i'll leave you to test the prototype =)
[14:40:51] <archivist_ub> if scream change tool larger needle
[14:41:45] <toastatw1rk> :o
[14:43:03] <renesis> it would prob be less painful when you perfected it
[14:43:26] <renesis> and you could test on leather
[14:44:34] <renesis> you basically need a built in inspectioin tool that can find range immediately ahead of the needle
[14:45:06] <renesis> hell fuck leather test that thing on corrugated cardboard
[14:45:39] <archivist_ub> cnc mill piercings :)
[14:45:46] <renesis> but you could control ink rates, feed rates, could be continuous
[14:46:01] <renesis> omg no
[14:46:06] <toastatw1rk> they use dead pigs and watermelon to train apprentices
[14:46:21] <toastatw1rk> so i'm assuming you'd want to use the same
[14:46:39] <renesis> not really
[14:46:58] <renesis> i need to test for bumps
[14:47:47] <renesis> because yeh that feedback based height offset thing is pretty nuts!
[14:48:43] <renesis> do it alien style, a z-axis inside a z-axis
[14:49:46] <renesis> have one axis/range sensor do course movement, like raising up for a knee or arm or whatever
[14:50:50] <renesis> you should be able to lock that one, and do it manually
[14:51:49] <renesis> then a super tiny thing that does little depressions in skin and like nipples and moles or whatever
[14:52:46] <renesis> hmm, but then you have multiple setup issues
[14:53:27] <renesis> now you have some sort of optical thinger and some software that can realign the machine based on the existing tattoo
[14:53:47] <renesis> cnc tatoo machine just got messy =(
[14:56:02] <renesis> oh but yeh id think youd want the head with the pokey part super light, no torque, and with much travel
[14:56:30] <renesis> so if they move they dont like stab themselves
[14:57:59] <renesis> heh, low torque stepper needle
[15:04:47] <toastatw1rk> lol
[15:04:54] <toastatw1rk> toastatw1rk is now known as toastatwork
[15:06:43] <archivist_ub> current method is balistic entry at mains frequency
[15:07:20] <toastatwork> the guns have a little vfd.
[15:08:04] <archivist_ub> variable hurts
[15:08:26] <toastatwork> ?
[15:09:25] <toastatwork> that is the other thing i am a little worried about
[15:09:43] <toastatwork> i have had 6x 12 gauge surface piercings before, but i have never had a tattoo
[15:10:19] <toastatwork> but i know some real wimps who have tattoos
[15:20:43] <Guest41041> HUZZAAAH
[15:20:58] <Guest41041> Guest41041 is now known as fragalot
[15:21:11] <fragalot> school donated me a PC.
[15:21:12] <fragalot> :D
[15:23:31] <archivist_ub> have you run the latency test :)
[15:23:48] <fragalot> I need to figure out how to get it home on my bike first.
[15:23:49] <fragalot> :p
[15:27:15] <fragalot> I'm hoping it 'l boot, :p
[15:32:47] <toastatwork> is there a trick to upping the txpower on a wireless card
[15:33:08] <SWPadnos> yeah. usually the trick is to redesign the card
[15:33:24] <fragalot> ^
[15:33:28] <toastatwork> lol
[15:33:29] <jepler_> toastatwork: yeah, almost all systems with wireless cards have a port called "RJ-45" which can be used to boost signal strength in a point-to-point topology.
[15:33:42] <toastatwork> noted
[15:33:57] <SWPadnos> if you want to connect to a single AP, the best bet is to use an antenna with high directional gain
[15:34:16] <SWPadnos> remember, you need to increase receiver sensitivity also if you increase transmit power
[15:34:28] <SWPadnos> a directional antenna effectively does both
[15:37:22] <jepler_> http://forum.openwrt.org/viewtopic.php?id=9612
[15:38:00] <toastatwork> i'm not really that pressed over it, i was just wondering why iwconfig eth0 txpower wasn't changing the power
[15:38:15] <toastatwork> the room i'm in is really crappy on wireless
[15:44:42] <fenn> brother CNC control has a lot of neat features: http://www.brother.com/europe/machine/info/tc32b/tc32b_spe.html
[15:45:12] <fenn> er, http://www.brother.com/europe/machine/info/tc32b/tc32b_spe.html#1
[15:45:49] <fenn> and some weird limitations that make me think it's not emc-derived, like 9999.999?
[15:45:53] <cradek> huh, I wonder what you use conical interpolation for
[15:46:17] <fenn> i think that means parabolas and hyperbolas?
[15:46:29] <cradek> oh maybe
[15:46:35] <cradek> I assumed "cones" not conic sections
[15:46:44] <fenn> same thing isnt it?
[15:47:00] <SWPadnos> I'd like an ice cream conic section please
[15:47:01] <fenn> or maybe they mean spirals with axial movement
[15:47:05] <cradek> well, I was thinking of a move like a helix, but with radius increasing/decreasing linearly
[15:47:11] <cradek> yes
[15:47:48] <SWPadnos> they mention that circular moves only simultaneously control 2 axes, so I'd lean toward the simpler version
[15:48:07] <SWPadnos> but they also say that helical/conic are optional
[15:48:12] <SWPadnos> (and control 3 axes)
[15:48:51] <fenn> maybe i'm out of touch but it seems easier to just write the most general function you can (traverse along a cone surface)
[15:49:10] <toastatwork> hyperbolas and parabolas are conics
[15:49:22] <cradek> toastatwork: we know :-)
[15:49:26] <toastatwork> GOOD
[15:49:29] <toastatwork> #;2~%$$%
[15:49:33] <SWPadnos> that may make sense at a low level, but getting the paarmeters from the users brain to the low level is an exercise in futility for more complex moves
[15:49:54] <SWPadnos> it's hard enough with circular arcs
[15:50:25] <fenn> that's why we should just let cam programs do all the work
[15:50:51] <SWPadnos> I'm awaiting your useful Linux-based CAM solution
[15:51:17] <toastatwork> *free
[15:51:18] <fenn> i have to go schlep someone else's trash around today
[15:51:38] <toastatwork> Pro/E has full linux support iirc
[15:51:47] <fenn> s/free/Free/
[15:52:31] <SWPadnos> free or low cost would be nice
[15:52:33] <SWPadnos> :)
[15:53:00] <SWPadnos> though I admit I haven't tried SheetCAM
[15:56:03] <jepler_> arcs are nice because they have a dead simple arc length parameterization and the acceleration inherent in the curve is always perpendicular to the direction of movement. lines and arcs (circles) have these properties, I don't know of any other curve families that do. Ellipses, parabolas, splines: none of them have these two properties. To support these other curves, you have to have a nice arc-length parameterization and a nice way to figure out how
[15:56:32] <SWPadnos> how ...
[15:56:51] <jepler_> To support these other curves, you have to have a nice arc-length parameterization and a nice way to figure out how much of the acceleration is "taken away" from creating the velocity profile for whatever shape is specified by the new curved movement primitive.
[15:58:07] <anonimasu> sheetcam is useful..
[15:58:11] <anonimasu> I run it with the oxyfuel table
[15:58:15] <tomp> any confirmation of emc on Brother?
[15:58:36] <SWPadnos> no, someone would have to look at one or ask
[15:59:08] <SWPadnos> it's also possible that it's a Delta Tau controller - they have a Linux -based version
[15:59:24] <tomp> goin to Taipei on 15th, will look if i can get away. DT has linux based????!!
[15:59:26] <jepler_> since that code hasn't magically appeared in CVS, it seems easier to use algorithms that fit arcs and lines to whatever output you already have. You can determine that a bunch of consecutive points are nearly colinear or cocircular and change the lot to a single linear or circular move. Or if you start with a spline or other parametric description of the tool path, fit directly with arcs using a biarc algorithm.
[16:00:28] <SWPadnos> do you need perpendicular accel?
[16:00:37] <SWPadnos> it doesn't seem like it's necessary
[16:01:04] <jepler_> SWPadnos: It complicates the computation of how long the acceleration and deceleration phase of the move last
[16:01:14] <SWPadnos> hmmm
[16:01:45] <fenn> i think parabolic/hyperbolic interpolation is only really useful for optics anyway
[16:01:47] <jepler_> SWPadnos: when it's perpendicular you get to ignore it when calculating the acceleration time
[16:02:01] <SWPadnos> linear moves don't have that simplification
[16:02:05] <anonimasu> toastatwork: no diamond turning lathe -_-
[16:02:45] <SWPadnos> fenn, depends. I'd bet shipbuilders, or people designing computer mice for that matter, would be able to use "higher order" curves
[16:03:10] <fenn> yes but piecewise simplification works just as well for them, nobody will notice the 'bumps'
[16:03:17] <SWPadnos> sure
[16:05:26] <archivist_emc> I would like to interpolate gear shapes and I dont want bumps
[16:05:58] <fenn> hobknobble
[16:06:50] <archivist_emc> for some shapes 1 off hob would not be justified but slow interpolation would be ok
[16:07:18] <archivist_emc> like the escape wheel I did the other day
[16:07:43] <jepler_> biarcs have C1 continuity (consecutive arcs are always tangent), and as long as they are not "too short" emc will also keep up the requested velocity blending between them. I really like the biarc algorithm for approximating arbitrary planar curves in gcode. assuming the plane you're talking about is the X, Y, or Z plane and not some oddly oriented plane :-P
[16:08:20] <archivist_emc> erm like mine where Im rotating A
[16:08:30] <archivist_emc> and feeting x or y
[16:08:32] <anonimasu> cant you blend arcs if P1 == P2 ie, start and end are coolinear
[16:08:41] <anonimasu> start of the second arc
[16:08:48] <anonimasu> (I'm not sure if it's already done..
[16:10:12] <jepler_> I'm not sure what a YA arc would mean
[16:10:34] <anonimasu> A would be rotary right?
[16:10:38] <archivist_emc> yes
[16:11:02] <anonimasu> jepler_: same ting as cutting something on a lathe with live tooling..
[16:11:10] <anonimasu> (I think)
[16:11:16] <archivist_emc> jepler_: did you see my youtube vids
[16:12:46] <jepler_> if the moves you want are better described as curves in a plane (YZ plane?) then maybe what you want is not a machine configured as XZA or XYA (trivial kinematics), but a machine configured as XYZ (nontrivial kinematics)
[16:13:00] <archivist_ub> http://nl.youtube.com/watch?v=adT8Dr5JZ4c
[16:13:39] <anonimasu> I'll ask again, does anyone know of a easy way to convert a 24v encoder signal to 5v?
[16:13:52] <archivist_ub> and this from the other direction http://nl.youtube.com/watch?v=Ge9_Qbd09zU&feature=related
[16:17:04] <archivist_ub> but the "next" if I can get my head around the curves will be epycyclic gear teeth using a slitting saw, saves a lot in milling cutter costs
[16:17:35] <anonimasu> it sounds like you need non trivial kins :)
[16:18:14] <archivist_ub> I hates nontrivial maffs
[16:18:22] <pjm_> anonimasu 24v to 5v, i'd just use either a potential divider with resistors, and perhaps a diode to +5v just as a safety measure
[16:18:44] <pjm_> or a 4.7v zener
[16:18:49] <anonimasu> I have thoose..
[16:18:59] <archivist_ub> or two diodes and one extra R to guarantee
[16:19:11] <pjm_> the potential divider with a couple of R's is prolly the easiest
[16:19:11] <anonimasu> I doubt the encoder will throw spikes..
[16:20:25] <anonimasu> * anonimasu nods
[16:20:37] <pjm_> so a 3.9K and 1K R will do the job
[16:21:00] <pjm_> 1K down to deck, other end to 5V in and one side of 3.9K
[16:21:06] <pjm_> other side of 3.9K to 24V
[16:21:11] <pjm_> that will give u 5V
[16:21:14] <pjm_> out from 24V in
[16:21:34] <anonimasu> yep
[16:22:05] <anonimasu> what size of resistors(since you already did the calc and im lazy)
[16:22:06] <anonimasu> :p
[16:22:13] <pjm_> 1/8th watt
[16:22:15] <pjm_> or smt
[16:22:20] <pjm_> anything will work
[16:22:21] <anonimasu> thanks :)
[16:22:38] <pjm_> or just find a 10K preset and use that
[16:22:54] <pjm_> add 24V across the track and use a DVM to set 5V on the wiper
[16:23:02] <anonimasu> yeah, I can do that..
[16:23:09] <anonimasu> I've got some smt resistors at work..
[16:23:14] <anonimasu> and smt pot's
[16:23:15] <anonimasu> :)
[16:23:25] <anonimasu> think 100K is too much?
[16:23:33] <anonimasu> the pot is 250 and the resistor is 100k..
[16:23:42] <pjm_> depends on what the current sink is on the 5v input pin
[16:23:45] <anonimasu> ttl
[16:23:56] <anonimasu> it's a geckodrive..
[16:24:01] <pjm_> so prolly wants 5-10Ma
[16:24:06] <pjm_> mA i mean
[16:24:19] <jepler_> http://emergent.unpy.net/files/sandbox/voltage-divider.png
[16:24:22] <anonimasu> it's still easier for me to have a voltage follower after it :p
[16:24:38] <anonimasu> * anonimasu nods
[16:24:55] <anonimasu> thanks :)
[16:25:17] <anonimasu> I'd like to try spinning a motor so I see if it'll work nicely enough before spending cash on ordering new encoders from the us
[16:27:14] <anonimasu> pjm_: yeah, but what kind of resistors does that end up as?
[16:28:00] <pjm_> a 100K R at 24V will give 0.2mA
[16:28:04] <pjm_> which probably isnt enough
[16:28:28] <pjm_> the 1 / 3.9K will give u about 5mA
[16:28:34] <pjm_> which 'should' be enough
[16:28:41] <anonimasu> ok..
[16:28:46] <anonimasu> I'll build with that
[16:28:50] <pjm_> dunno also how much I the output of the encoder will sink
[16:29:08] <anonimasu> I think quite a bit..
[16:29:18] <pjm_> if it will only supply 1mA then i'd measure the input R of the 5V port
[16:29:22] <pjm_> and just use a single R
[16:29:43] <anonimasu> * anonimasu nods
[16:30:17] <pjm_> i usually stick a 390ohm R in the parport pin I'm testing and so far i've not blown anything up feeding 12V up the R
[16:34:13] <jepler_> fwiw, the G320 datasheet shows that there is a 1K pull up to +5V integrated into the drive
[16:35:25] <jepler_> you should also note this item in the geckodrive faq:
[16:35:27] <jepler_> Q.) Can I run a 0 – 24V encoder with my G320 or G340?
[16:35:27] <jepler_>
[16:35:27] <jepler_> A.) You must use a +5V encoder with the gecko servo drives. Anything other than that, regardless of how it is powered, is not recommended with the G320 or G340 and will void the warranty.
[16:35:41] <jepler_> (from http://www.geckodrive.com/faq.aspx?n=342824_
[16:35:44] <jepler_> s/_/)
[16:36:24] <anonimasu> jepler_: I didnt get a new drive even though I had warranty and jymm has talked to mariss on the phone about it before..
[16:36:29] <anonimasu> had..
[16:41:28] <archivist_ub> if driving an LED on an opto, then they do not like any high reverse voltage
[16:42:25] <anonimasu> it was a new drive that was doa..
[16:42:37] <anonimasu> :)
[16:52:43] <Paragon> Hello All
[16:53:08] <tomp> http://www.observatorysciences.co.uk/client/epics.asp epics linux delta tau
[16:53:26] <Paragon> What do you think of the spec of this servo drive and pros / cons with using it with EMC? http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=110277724698
[17:04:25] <SWPadnos> http://www.aceshowbiz.com/news/view/00018830.html
[17:07:26] <tomp> 'member when palin meant michael?
[17:08:13] <archivist_ub> still does over here
[17:08:27] <buckie555> Hello - I'm setting up a 4th axis. Can anyone tell me what the hot keys are for jogging the a axis from the axis gui?
[17:08:40] <archivist_ub> []
[17:08:43] <Paragon> lol
[17:08:48] <buckie555> I couldn't find anything in the docs
[17:09:00] <archivist_ub> buckie555, [ ]
[17:09:21] <buckie555> excellent - thanks very much.
[17:09:26] <archivist_ub> they are keys I use
[17:09:39] <SWPadnos> they should be in the quick reference on the help menu
[17:09:44] <archivist_ub> and they are documented but cannot remember where
[17:10:30] <buckie555> I'll recheck - the docs I found mentioned lots of others including the hotkeys for x,y and z but nothing beyond that
[17:10:59] <archivist_ub> see help as sw[ just said
[17:11:07] <buckie555> great
[17:11:42] <Paragon> Not sure if you guys saw my previous question regarding servo controller. What benifits are ther of having a servo driver use step / dir over EMC pwm etc?
[17:12:14] <SWPadnos> if you have a drive that only accepts step/dir, then it's better
[17:12:45] <SWPadnos> if you have a drive that can take some "analog" input, such as voltage or PWM, then that should be better in a couple of ways:
[17:12:55] <SWPadnos> 1) you should get better part finish
[17:12:55] <Paragon> But as far as emc is concerned is it better to get a pwm drive or a step / dir drive?
[17:13:11] <SWPadnos> 2) you can tune the PID in software instead of hardware
[17:13:22] <SWPadnos> EMC doesn't care, depending on the hardware you use
[17:13:49] <buckie555> I've set up the axis as angular (it's a standard 3 axis machine with a 4th rotary axis setup on the bed of the machine). I understand that I can index to a position with a g1 a0, etc, but is there also a way for me to tell the axis to rotate at a given speed until told otherwise?
[17:13:55] <SWPadnos> if you'd use PWM on the parallel port, then it's more or less equivalent to step/dir, with the exception that step/dir doesn't need feedback to the computer (saving some port pins)
[17:14:03] <Paragon> What about the over head of step pulses compared to pwm and counter pulses?
[17:14:08] <SWPadnos> buckie555, no, not really
[17:14:19] <SWPadnos> buckie555, you'd have to set something up in HAL for that
[17:14:31] <SWPadnos> overhead is minuscule compared to port access
[17:15:00] <buckie555> can i command a value much greater than 360 degrees to get it to keep revolving
[17:15:04] <SWPadnos> ie, reading and writing data from/to the parport takes orders of magnitude more time than operating on that data
[17:15:23] <SWPadnos> buckie555, yes, but that move will have to complete before the next will start
[17:15:24] <Paragon> I see....
[17:15:37] <Paragon> Does this look like a good deal? http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=110277724698
[17:15:46] <SWPadnos> what kind of machine?
[17:16:34] <Paragon> It's for a rotary table conversion for my denford starmill
[17:17:44] <fragalot> *g* after the CD finishes burning i'll be able to like,.. run EMC o.0
[17:17:53] <SWPadnos> I've heard good (enough) things about MCP, but I don't know that drive specifically
[17:17:56] <tomp> buckie555: so you want to use a rotary axis to postion to an angle (sometimes) and use it as a speed controlled spindle (sometimes) ?
[17:18:03] <SWPadnos> it's in the same price range as a Gecko, with just about the same specs
[17:18:08] <buckie555> exactly
[17:18:34] <SWPadnos> buckie555, you can do that with a HAL setup to switch between the two types of command
[17:19:03] <SWPadnos> it's not quite the same (and much more complex), but the Mazak demo config has this type of setup for tool changes
[17:19:16] <buckie555> yes, ii suppose i could have hal monitor a switch on the panel to switch from a position pid loop to a velocity pid loop
[17:19:24] <SWPadnos> the spindle is usually controlled by the S word, but for spindle orient it has to be rotated to a specific angle
[17:19:37] <buckie555> right
[17:19:41] <SWPadnos> you can do it in G-code as well - change an output with an M code (I think) to change modes
[17:20:27] <buckie555> I didn't think hal could monitor any changes from a g code
[17:20:38] <SWPadnos> an M code can change a HAL pin though
[17:21:16] <SWPadnos> either M1xx or one of the M6x codes I think (I know there are input codes, I don't know if the output codes have also been implemented)
[17:21:31] <buckie555> right - I'll take a look at those.
[17:21:34] <tomp> is the top rpm of such an device limited by precision of encoder ( high count + high rpm = too fast too count ) or, does the spindle mode use a single pulse per rpm?
[17:21:55] <SWPadnos> you may end up having to switch the feedback path also
[17:22:49] <SWPadnos> you could do reasonable speed control with only the index pulse (as a 1PPR encoder), but you'd still have to be sure the width is enough to always be detected by the software encoder module
[17:23:08] <Paragon> Sorry Chaps the wife just beconned me away. Thanks for the info.
[17:23:11] <buckie555> well the motor has 4096 counts per rev and I'm using the mesa 5i20 which I think can read 4Meg transitions per second so I should be OK
[17:23:25] <SWPadnos> oh, in that case you should be fine :)
[17:23:43] <buckie555> Yeah I'm only looking for 2000rpm
[17:24:39] <buckie555> Mainly It will be used for indexed machining but I can envisage situations where it would be nice to do some rudimentary turning on it
[17:25:46] <Guest954> rigid tapping..
[17:25:55] <Guest954> huh
[17:26:07] <buckie555> Has anyone got any suggestions for tuning that will ensure a very tight lock
[17:26:21] <Guest954> Guest954 is now known as skunkworks__
[17:27:09] <buckie555> I'm not currently using a brake and the motor is a 1kW AC servo with a continuous torque rating of 6nm so I figure it should be able to hold for the light routing that I'm planning to use it for.
[17:27:09] <fragalot> *burp*
[17:27:46] <SWPadnos> stick the drive in velocity mode and use no tuning for the spindle mode
[17:28:14] <SWPadnos> then tune a PID for the positioning mode, which should be relatively easy with a good AC drive helping
[17:28:27] <fragalot> this "national strike" thing sucks balls
[17:28:28] <fragalot> >.>
[17:29:10] <anonimasu> * anonimasu yawns
[17:29:30] <fragalot> all stores closed, no trains, no bus, no tram, some schools closed too, no government buildings open..
[17:29:40] <anonimasu> where are you canada?
[17:29:50] <fragalot> belgium
[17:29:55] <fragalot> for the 4523th time
[17:29:58] <anonimasu> ah
[17:30:02] <fragalot> :p
[17:30:10] <fragalot> * fragalot waits for the EMC cd to finish burning
[17:30:18] <fragalot> I really hope this machine still runs
[17:30:18] <anonimasu> well, it's your own fault for having a masked ip;)
[17:30:23] <fragalot> anonimasu: hehe
[17:30:33] <fragalot> even if it wasn't masked you still wouldn't know ;)
[17:30:47] <anonimasu> haha, I would..
[17:30:53] <anonimasu> I'd assume you were from belgium..
[17:30:57] <fragalot> lol.
[17:31:04] <anonimasu> if your host ended in .be ;)
[17:31:10] <fragalot> obviously, it doesn't ;)
[17:31:16] <fragalot> atleast not on the part that freenode sees
[17:31:16] <anonimasu> and if you dont have one whois works ;)
[17:31:32] <fragalot> you'd end up in the UK, or in the US,.. depending on my mood
[17:31:36] <fragalot> :p
[17:31:52] <anonimasu> close enough
[17:31:55] <fragalot> SSH tunnels ftw.
[17:32:27] <anonimasu> * anonimasu sighs
[17:32:47] <anonimasu> damn car, requiring mainentance... :p
[17:32:50] <anonimasu> keeping me from machining stuff
[17:33:57] <anonimasu> hm, im going to cut something that looks like a squid.
[17:34:08] <anonimasu> to hold my bottle holder in the bathroom..
[17:34:21] <fragalot> xd
[17:34:36] <fragalot> is the ubuntu livecd dog slow to boot up for everyone?
[17:34:41] <anonimasu> yes
[17:34:42] <fragalot> 'cos seriously, this is the 6th machine, and.... urgh
[17:34:55] <anonimasu> I have thid bottle holder and it looks like a blue silicone squid :p
[17:34:55] <fragalot> i'd have installed Gentoo by now :p
[17:36:10] <fragalot> * fragalot kicks slowbuntu
[17:38:05] <SWPadnos> stage 1 or stage 2?
[17:47:22] <fragalot> fiddlepoop!
[17:47:27] <fragalot> I don't have a mouse,.. lol
[17:49:56] <tomp> be waring of working off center ( i have customers mounting long arm fixtures on rotary axis, and complain 'it wont hole position' ) (give me a place to stand...)
[17:50:18] <anonimasu> what?
[17:50:21] <anonimasu> hold position?
[17:50:40] <anonimasu> tomp: -_- disturbing
[17:50:52] <tomp> buckie555 wanted to hold position w/o a brake
[17:51:12] <anonimasu> tomp: oh, is that the guy without the brake and the Z axis only supported by motor?
[17:51:31] <anonimasu> (or am I confusing someone)
[17:51:31] <anonimasu> :p
[17:51:38] <buckie555> it's a rotary axis on the bed of the table
[17:51:39] <tomp> didnt know about Z ( best get a brake and relase it on power up)
[17:51:51] <anonimasu> tomp: that's someone else then
[17:51:54] <fragalot> <3 Z crashing on powerfaillures
[17:51:59] <anonimasu> haha :D
[17:52:08] <anonimasu> if it does your design suckd :)
[17:52:12] <fragalot> *trip* SHBLAM
[17:52:46] <tomp> bendix 24V ac brakes are cheap (electro-clutch) releases when powered, grabs when power lost
[17:52:55] <fragalot> Aye
[17:53:37] <buckie555> well the motor is rated at 6nm continuous and 13nm peak. The greatest distance we'll be machining from the axis centreline is 60mm and we're only light routing ash
[17:54:08] <tomp> sounds cool
[17:54:48] <buckie555> I've done the calcs and we should have plenty of headroom re required torque - I'm more concerned about getting the pid loop tight enough to eliminate any slop in the axis
[17:55:18] <tomp> practical test it with a bar mounted on face , some known distance, and a force tester pulling on it ( fish scale ;)
[17:55:36] <buckie555> There's room on the back of the headstock spindle shaft to fit a brake but I was hoping I wouldn't have to at this stage - rush job!
[17:55:48] <buckie555> tomp - yes that's a good idea
[17:55:58] <tomp> you need to see it react to 1 unit of error ( hard to do )
[17:57:10] <buckie555> In practical terms only experience will show how much movement we can get away with, the encoder is 4096 counts and I would like to think I could get it to hold within 5 counts
[17:58:27] <buckie555> taking into account the drive reduction that gives a potential movement of 0.3 degrees
[17:59:39] <cradek> is there no worm?
[17:59:44] <buckie555> well I'll be tuning tonight and machining tomorrow so I'll report back whether it was OK or whether we need to add a brake
[17:59:46] <tomp> jack screw from lathe bed to face plate/chuck ( just during the time when you want it still) ?
[18:00:43] <buckie555> unfortunately I'm using it to do a radius in lots of small increments so that would be rather cumbersome
[18:01:32] <fragalot> /YEAH/
[18:01:39] <fragalot> I made it to step 2 of the installation >.> timezone.
[18:01:50] <fragalot> lol this is so slow >.>
[18:01:54] <cradek> fragalot: it can be very slow if you're short on ram
[18:02:10] <fragalot> cradek: I think this box has like 128 or so, lol.
[18:02:17] <cradek> yeah that's just not enough
[18:02:24] <tomp> leather belt around edge, pull it to hold it, slack it to index
[18:02:24] <fragalot> nope
[18:02:29] <cradek> it may not even install
[18:02:33] <buckie555> I noticed in the archives peter wallace recommended upping the servo rate beyond 1ms, has anyone had any experience at higher rates - good or bad?
[18:02:38] <fragalot> cradek: We'll see.
[18:02:49] <buckie555> tomp - interesting
[18:03:25] <SWPadnos> buckie555, it certainly works better for some things (like rigid tapping), so if your CPU can handle it, it's probably a good idea
[18:03:42] <SWPadnos> the main downside is that it's more of a CPU load, AFAIK
[18:04:02] <skunkworks__> skunkworks__ is now known as skunkworks
[18:04:11] <fragalot> (if your controller can handle it)
[18:04:22] <buckie555> well being a 5i20 setup I'm doing very little in the base thread so I'm sure I should have plenty of overhead in the servo thread
[18:04:27] <SWPadnos> he's gota Mesa card, so that's not a problem
[18:04:38] <SWPadnos> you don't even need a base thread with the Mesa
[18:04:44] <fragalot> * fragalot shuts up, and waits for the box to load the partitioner
[18:05:06] <buckie555> I'm using it for a pulse generator to drive the vfd
[18:07:27] <jepler_> SWPadnos: about those pancake servos I bought back at workshop -- do you have a datasheet for them that dimensions the mounting hole locations?
[18:07:36] <SWPadnos> no, sorry
[18:07:40] <jepler_> ok
[18:07:45] <jepler_> no big deal
[18:12:04] <jepler_> piasdom: welcome
[18:13:25] <fragalot> woo
[18:13:28] <fragalot> it started the install.
[18:14:53] <Paragon> I know this seems dumb but ... I having trouble in choosing a horizontal / vertical rotary table due to size restriction of my mill (Max Travel x=170, y=90, z=115 mm) Iwas looking at a Vertex HV4 100mm but I think it mabe to big. Does any one have any sugestions?
[18:15:15] <cradek> what do you need to do with it?
[18:16:31] <Paragon> This kind of thing and gear cutting. http://www.youtube.com/watch?v=pQ20hmdfN7Y
[18:17:02] <cradek> ok, so it will always be vertical
[18:17:04] <anonimasu> I have a verted it's got a bit of slop.. in the gearing
[18:17:13] <cradek> (there's virtually no use for a horizontal rotary on a cnc)
[18:17:45] <toastatwork> ^^^ truf
[18:18:10] <Paragon> I agree cradek... here is a picture of the same mill I own but with a ATC
[18:18:33] <Paragon> oopps ;-) http://www.denfordata.com/bb/viewtopic.php?t=382&highlight=starmill
[18:19:22] <cradek> ok that's pretty small - Z travel is your big issue
[18:19:32] <anonimasu> vertex has smaller sizes too
[18:19:35] <cradek> you will want to use short tools
[18:19:38] <Paragon> Yes it is.
[18:19:52] <cradek> if you load up a typical tool and move all the way up, how far from the table is the tip?
[18:20:15] <Paragon> I am currently using r8 > e25 and that take up space below spindle
[18:20:36] <cradek> I don't know what that means
[18:22:09] <cradek> er25?
[18:22:17] <Paragon> Mmmm good question... I am not sure right now with out measuring. but at a guess about 90mm maybe less
[18:22:18] <cradek> surely not r8 like bridgeport size collets?
[18:22:36] <Paragon> Yep!
[18:22:45] <cradek> bizarre
[18:22:54] <anonimasu> hm.. 10arcsec seems to be what people run with vertex tables..
[18:22:59] <anonimasu> of repeatability
[18:23:33] <cradek> looks like the lower one is CAT40
[18:23:52] <cradek> I bet you want no bigger than 4"/100mm table
[18:24:20] <Paragon> Could be. I also have R8 to EZ Change ( quick change ) system
[18:24:56] <anonimasu> hmm that's ~0,016mm
[18:25:13] <anonimasu> not bad.
[18:25:55] <Paragon> This is the table I was looking at http://cgi.ebay.co.uk/HV-4-HORIZONTAL-VERTICAL-ROTARY-TABLE-BY-VERTEX_W0QQitemZ320305072452QQihZ011QQcategoryZ12584QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[18:27:39] <cradek> the center looks like it might be about 80mm above the table
[18:27:45] <cradek> that seems too tall
[18:28:25] <Paragon> Yeah that was my worry too.
[18:29:52] <cradek> http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=350106136252
[18:29:55] <Paragon> This is the quick change system that came with the mill. I only have three holders though and not all the collets
[18:30:08] <cradek> you would have to mount it to a little angle plate
[18:30:34] <cradek> but the round part goes almost to the edges, so the center might only be 55mm high
[18:31:02] <cradek> oh it even says 55 in the text, ha
[18:31:33] <Paragon> cradek you gem! :-) that would be perfect. Hope I can convert it to CNC though.
[18:32:13] <cradek> yeah I wonder what kind of thrust bearings it has - sometimes the knob is one bearing surface
[18:32:21] <cradek> I bet you can do it with some experimentation.
[18:33:24] <fragalot> Everything can be CNC'
[18:33:24] <fragalot> d
[18:33:48] <anonimasu> * anonimasu nodse:)
[18:33:51] <anonimasu> * anonimasu nods
[18:34:05] <Paragon> I own a Granville Senior Lathe too, Which is almost identical to a Myford ML7 but heavier ... I have the ML7 milling slide adapter so two bird with one stone :-) http://www.lathes.co.uk/granville/index.html
[18:35:01] <Paragon> nodding ...
[18:35:42] <anonimasu> http://www.emeraldinsight.com/fig/0330240402006.png
[18:35:46] <anonimasu> cute.
[18:36:22] <anonimasu> matsuura of some kind
[18:37:34] <cradek> Paragon: a CDROM is 120mm diameter - you could put one on the mill and load some tools and see how it looks
[18:38:30] <Paragon> Cradek: Good idea, I will do that.... :-)
[18:38:53] <cradek> stick it up on edge with some chewing gum :-)
[18:38:56] <fragalot> my god i just realized my old school had a machine just WAITING to get retrofitted
[18:39:14] <fragalot> it was a conventional machine, but with pushbuttons instead of handwheels
[18:39:15] <fragalot> :p
[18:40:13] <Paragon> BTW my irc usage is pretty lame how do you send a message as you did to me cradek? Do you type the nik of the person first followed by a colon?
[18:40:20] <cradek> yes
[18:40:33] <cradek> often there is tab completion to help you: type cr[TAB]
[18:40:46] <Paragon> Thought so just needed clarification. thanks
[18:41:20] <Paragon> tab completion ... cool :-)
[18:41:48] <Paragon> Saves on my fingers.
[18:44:20] <toastatwork> anonimasu: that looks like a jig grinder
[18:44:32] <toastatwork> is it really a matsuura?
[18:44:34] <toastatwork> (mill, obv)
[18:46:01] <anonimasu> yes
[18:46:04] <anonimasu> :)
[18:46:59] <anonimasu> http://vids.myspace.com/index.cfm?fuseaction=vids.individual&VideoID=43960967
[18:48:08] <toastatwork> connection not stable enough for myspace
[18:48:17] <anonimasu> :/
[18:48:20] <anonimasu> it's a really cool video
[18:48:29] <anonimasu> airfoils
[18:48:30] <anonimasu> :)
[18:49:56] <tomp> gnu opensource pmac c http://www.phys.unsw.edu.au/~mcba/pmac.c ( not the same as the NIST one )
[18:50:07] <anonimasu> pmac?
[18:50:08] <fenn> what's pmac?
[18:50:39] <SWPadnos> aservo controller card
[18:50:48] <SWPadnos> they also have external controller boxes
[18:50:48] <anonimasu> nice :)
[18:51:10] <fenn> ISA, does anyone even have computers with ISA bus in them?
[18:51:58] <fragalot> OMG IT BOOTED
[18:52:02] <fragalot> lol, sound eve nworks
[18:52:04] <fragalot> even,..
[18:52:22] <fenn> would a timing belt on the lathe headstock be able to give me enough torque to do the same sort of things as back gear?
[18:52:58] <toastatwork> fenn: imho yes
[18:53:07] <toastatwork> our big damn lathes all use timing belt type deals
[18:53:07] <anonimasu> yes
[18:53:31] <fragalot> Indeed. :p
[18:53:33] <cradek> mine uses dual V belts - that might be easier
[18:53:36] <fenn> ok good :) timing belt pulleys are much easier to come by/make
[18:53:46] <fragalot> cradek: V belts tend to slip
[18:54:12] <cradek> slip or strip, pick your poison
[18:54:20] <anonimasu> hehe
[18:54:24] <toastatwork> dunno about the strip part
[18:54:35] <toastatwork> one 30 hp machine uses 3x timing belts and we stalled it in low gear
[18:54:37] <cradek> it's a 2HP machine, I'd rather it slip when something goes wrong
[18:54:46] <cradek> eek
[18:54:50] <toastatwork> made quite the noise but everything was fine
[18:55:00] <tomp> some early versions of emc talked to pmacs, then the pmac was removed and emc became a software only control.
[18:55:42] <SWPadnos> the original EMC was done that way - the PC only sent position commands to the PMAC, which used an onboard DSP for PID
[18:56:06] <SWPadnos> it also ran on VXWorks I think (I'd have to look at my paper to find out ;) )
[18:56:18] <SWPadnos> oh. I should stick that on the web somewhere
[18:56:21] <fenn> pmac at omac
[18:56:56] <anonimasu> ^
[18:57:02] <anonimasu> * anonimasu goes to rest
[18:58:36] <fenn> i wonder how lathes.co.uk got so many good pictures of machines
[18:59:02] <fenn> they are always clean and new on white background
[19:03:00] <fragalot> fenn: probably because they are
[19:04:45] <fenn> http://www.dugnorth.com/blog/uploaded_images/drill_lathe.jpg
[19:05:02] <fenn> reminds me of my pitiful beginnings
[19:05:44] <tomp> i visited lots of european shops, hospital clean. i been in a zillion us shops, and really clean are not the norm. so that may account for some of it.
[19:07:28] <anonimasu> I love hospital clean shops ^_^
[19:07:59] <tomp> definitely affects the way you work
[19:08:06] <anonimasu> but new buildings/refurbrished and stuff and no welding..
[19:08:19] <anonimasu> and no non work surfaces to collect crap.
[19:08:37] <tomp> important point
[19:09:08] <anonimasu> welding/grinding stuff creates airborne particles that stick to everything and makes it black -_-
[19:15:08] <tomp> what do you call the center like thing that bites into the wood on a wood lathe? ( crossed v's drive the stock )
[19:15:40] <tomp> look like that on the grozzly drill lathe
[19:15:47] <tomp> grizzly
[19:15:59] <cradek> 'spur center'
[19:16:25] <tomp> spur right thx
[19:18:08] <fragalot> (sigh) I keep forgetting to pick up the rest of my controller circuit componements..
[19:18:18] <fragalot> then i need to resolder all the capacitors because i misread the label on them >.>
[19:19:06] <fragalot> .. I might just try to let them run with the wrong capacitors in,.. might not be filtered as good, but the filter level is insane anyhow :p
[19:20:29] <fragalot> CRUD
[19:38:32] <fenn> http://www.pacificsun.ca/~robert/NAMES2003/cincinnati30v.jpg
[19:39:18] <fenn> now _that_'s a micro mill
[19:42:35] <Paragon> Wonderfull :-)
[19:43:30] <Paragon> cradek: Will this vice be ok for my mill also? Sorry aboutthis just trying to tool up :-)
[19:43:32] <Paragon> http://cgi.ebay.co.uk/5-MILLING-VICE-Swivel-Base-for-Milling-Machine_W0QQitemZ360082646645QQihZ023QQcategoryZ12584QQrdZ1QQssPageNameZWD1VQQcmdZViewItemQQ_trksidZp1638Q2em118Q2el1247
[19:44:23] <Paragon> It appears to be quite a low profile
[19:45:14] <fenn> indeed, ebay is a vice
[19:45:45] <Paragon> certainly is ...
[20:25:58] <tomp> fenn thats a nice site http://www.pacificsun.ca/~robert/
[20:33:35] <cradek> Paragon: IMO, a pivoting vise is a waste of height and money if your mill is CNC
[20:33:48] <cradek> height, money, and stiffness
[20:34:28] <skunkworks> but emc doesn't have a rotation matrix.. ;)
[21:26:32] <Paragon> cradek: the swivel base is removable on that item i think. Eventualy I will probably purchase a manual mill which I could also use it on.
[21:46:24] <toastydeath> pivoting vises are also a waste of money if you don't own a cnc mill
[21:47:06] <Paragon> :-)
[21:47:52] <anonimasu> uh..
[21:48:00] <anonimasu> toastydeath: not really
[21:48:11] <anonimasu> toastydeath: What if you want to machine /\ stuff..
[21:48:20] <toastydeath> then you use something that's not a pivoting vise
[21:48:39] <anonimasu> I thought the use for sine vises was just htat..
[21:48:40] <anonimasu> that..
[21:48:52] <toastydeath> we're not talking about sine vises
[21:49:19] <toastydeath> the pivot block for a 6" milling vise
[21:49:32] <SWPadnos> it is very tempting to buy a few 1TB hard drives at $129.99 each
[21:49:46] <SWPadnos> they're even low power
[21:51:01] <anonimasu> toastydeath: I dont find thoose useful at all
[21:51:03] <anonimasu> :P
[21:51:24] <Paragon> I picked up a Western Digital 1TB World Book NAS the other day. Found a hack so have now got ssh connectivity 'It runs Linux' :-)
[21:51:44] <SWPadnos> http://www.newegg.com/Product/Product.aspx?Item=N82E16822136151
[21:51:54] <SWPadnos> heh - cool
[21:52:23] <SWPadnos> the WD is only 5.4W when active O_O
[21:53:14] <Paragon> One thing I will say though is that it is SLOW 5mbs max due to the ARM processor.
[21:54:02] <Paragon> http://www.lambda-tek.com/componentshop/index.pl?origin=gbase6.4&prodID=B71191
[22:35:10] <skunkworks_> skunkworks_ is now known as skunkworks