#emc | Logs for 2010-01-15

Back
[00:00:19] <danimal_garage> but not 8====D--- ()
[00:00:27] <cradek> I just checked and iocontrol.0.tool-change DOES stay true until you acknowledge the tool change is done by setting iocontrol.0.tool-changed to TRUE
[00:00:37] <cradek> so maybe you have tool-changeD stuck at TRUE
[00:01:15] <cradek> do you realize that the tool-prepare and tool-change are totally separate? you have no prepare on your turret so you should just leave that looped through so it finishes right away.
[00:02:16] <danimal_garage> well i basically linked them in case i accidentally did an m6 without a t#, so it just doesnt spin forever
[00:02:34] <cradek> them = what?
[00:02:45] <danimal_garage> tool prepare and tool change
[00:02:49] <cradek> if you do m6 without a t first you'll get an error
[00:02:55] <danimal_garage> oh ok
[00:03:35] <danimal_garage> ok the coil i have after toolchange and tool prepare is supposed to turn off the tool prepared and tool changed coils
[00:04:10] <cradek> are you saying you have -|change|---|prep|-----(something)- ?
[00:04:28] <danimal_garage> yes, the something is an internal coil
[00:04:56] <cradek> -|change|---|prep|- kind of makes no sense
[00:05:05] <cradek> I bet prep goes low as soon as you ask for the change
[00:05:20] <danimal_garage> yes
[00:05:30] <danimal_garage> ok i'll get rid of it and try, brb
[00:05:36] <cradek> just loop back prepare=>prepared and ignore it
[00:13:29] <danimal_garage> ok thats better, now it's doing somethnig
[00:13:46] <danimal_garage> however i think i have my weighted_sum wrong
[00:14:04] <cradek> woo
[00:14:14] <cradek> spins forever?
[00:14:45] <danimal_garage> no, triggers up/spin for like a fraction of a second and stops
[00:14:52] <cradek> ah
[00:15:03] <danimal_garage> like it always thinks its in position
[00:15:22] <danimal_garage> when i watch the ladder editor, it always stays pink
[00:15:38] <cradek> isn't it cool how the ladder editor shows you its state?
[00:15:50] <danimal_garage> yes, VERY helpful
[00:16:34] <danimal_garage> crap i gotta run and pick up some aluminum before they close. I'll be back in like 30 min
[00:17:18] <kevinh_> hello
[00:18:25] <cradek> hi
[00:19:52] <kevinh_> i'm looking in to an insane project, porting at least some portion of emc to the mac os.
[00:20:24] <cradek> can you get a realtime kernel?
[00:20:43] <kevinh_> we are... investigating... what we can do.
[00:20:48] <cradek> (if you just want to run simulator mode and not control hardware, it'll probably just work as-is)
[00:21:15] <kevinh_> xnu is soft real-time out of the box
[00:21:19] <cradek> if you want to control hardware, you need rtai or some other realtime system
[00:21:20] <kevinh_> but that's not good enough
[00:21:38] <kevinh_> is there an external hardware system we can program with emc?
[00:21:51] <cradek> can you be more specific?
[00:22:02] <Jymmm> These strippers are made in Germany and OEM'ed to Ideal, which does NOT carry the smaller size. Does anyone know/recognise who the mfg might be? I have no clue on German tool brands http://www.idealindustries.com/prodDetail.do?prodId=45-227&div=3&l1=wire_strippers&l2=grip-n-strip
[00:22:16] <kevinh_> i'm still trying to get my head around the whole system to determine if this is even possible
[00:22:27] <kevinh_> so i'm not really suer what all the options are
[00:22:29] <kevinh_> *sure
[00:22:39] <cradek> no matter what hardware you use or not use, emc's architecture depends on realtime
[00:23:04] <cradek> we don't do the kind of motion control where you queue up motion ahead of time in some external thingy
[00:23:10] <cradek> so if that's what you're asking, no
[00:23:43] <cradek> there are lots of devices that work with EMC to count encoders, generate pwm, generate analog, etc., faster and better -- but they are all run in real time
[00:23:59] <kevinh_> is the original emc core real-time dependent, or is it the extensions unique to emc that are?
[00:24:22] <cradek> emc has been a realtime system since day 1
[00:25:14] <kevinh_> there is nothing that's adaptable to a non-real time system?
[00:26:11] <archivist_emc> you need deterministic time for loop control
[00:26:31] <kevinh_> hmm.
[00:26:32] <cradek> if you wanted to build a motion control that uses a thing outside the PC to queue up motion, I'm sure there are some parts of EMC you could reuse, but that's not at all the current architecture
[00:27:37] <kevinh_> would there be any benefit/possibility of porting that type of code back to the linux version?
[00:27:39] <cradek> our current architecture is better for several reasons. the tradeoff is (of course) it requires realtime in the PC.
[00:28:14] <skunkworks> but it allows 'real cool' tm flexabillity
[00:28:28] <cradek> I'm not sure. that's definitely something the emc board would discuss if there was a proposal/offer.
[00:28:58] <cradek> yes you'd lose most of what makes EMC better than other systems if you lost the realtime-in-the-PC
[00:29:35] <cradek> for that tradeoff there would have to be a huge gain, IMO, and running on a non-free OS on proprietary hardware wouldn't be it
[00:30:34] <kevinh_> hmm...
[00:30:39] <kevinh_> * kevinh_ ponders for a moment
[00:31:32] <archivist_emc> what advantage does macos give you
[00:31:33] <kevinh_> cradek: I would hope that the mac os x code is separate from the non-real time code.
[00:31:55] <kevinh_> contributing mac os x code would eliminate the emc princiuple of eliminating other archs
[00:32:14] <kevinh_> archivist_emc: a client wants it, so i'm investigating it.
[00:32:16] <cradek> it is not a goal of the emc project to eliminate other OSes
[00:32:37] <kevinh_> that's not quite what i meant
[00:33:02] <kevinh_> somewhere on the site there was a goal of removing non-linux os code from emc, to streamline the software
[00:33:05] <kevinh_> maybe i misunderstood
[00:33:22] <cradek> oh, when we started emc2, we got rid of support for some platforms including Sparc
[00:33:32] <cradek> they weren't used and we dropped the code
[00:34:04] <kevinh_> gotcha ;-)
[00:34:32] <cradek> the simulator will build on about anything. it will even run on ppc processor (running linux) currently.
[00:34:53] <cradek> it would surely run on other unixlike systems with minor tweaks
[00:34:55] <kevinh_> archivist_emc: i think the biggest motivation behind the port is that there is NOTHING for cnc on the mac
[00:35:04] <kevinh_> absolutely nothing.
[00:35:35] <cradek> is that the only reason?
[00:35:37] <kevinh_> and the mac os is gaining quite the following among tinkerers of various kinds. there is demand for a mac app that can do some of this stuff
[00:36:01] <cradek> what interface hardware is available? (I don't know anything about macs)
[00:36:03] <kevinh_> and there is the goal of making it easy to use, to lower the barrier to entry
[00:36:15] <kevinh_> interface hardware you mean the ports etc...?
[00:36:26] <cradek> yeah hookups
[00:36:28] <kevinh_> nothing that's supported
[00:36:30] <kevinh_> :-(
[00:36:38] <cradek> do they have pci slots?
[00:36:38] <kevinh_> there is usb....
[00:36:45] <kevinh_> depends on the mac, but not a lot of em
[00:36:49] <kevinh_> only the mac pros
[00:36:59] <kevinh_> and other towers (g5's etc...)
[00:37:05] <cradek> as far as we know, usb just can't be used for realtime communication
[00:37:14] <kevinh_> too much latency
[00:37:20] <kevinh_> plus it queues and buffers...
[00:37:21] <cradek> you are already stuffed if that's all you have
[00:37:21] <cradek> yep
[00:38:00] <cradek> yep
[00:38:09] <kevinh_> that's the genesis of the 'external hardware' idea
[00:38:58] <archivist_emc> external hardware is then most of a pc... just use a pc
[00:39:48] <cradek> like I said, you could use parts of emc to make one of these queueing-based systems. we'd probably want you to call it something other than emc though, since a lot of functionality would be lost
[00:40:35] <kevinh_> is there any sort of support for a remote front end? a daemon of some sort? (not remote x11)
[00:40:46] <cradek> finally! I found the wiki page I was wanting to show you: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Emc2HardwareDesign
[00:40:57] <kevinh_> already got that one open ;-)
[00:41:01] <kevinh_> read it.
[00:41:03] <cradek> ah
[00:41:14] <cradek> don't know why I couldn't find it
[00:41:27] <kevinh_> which led me here to ask questions ;-)
[00:45:02] <kevinh_> what about a remote front-end (besides remote x)?
[00:46:51] <kevinh_> (I gotta do my due diligence before i say it's not possible)
[00:48:42] <cradek> that part is pretty well handled. the guis are pretty separate and communicate over a channel.
[00:48:44] <archivist_emc> there has been work on remote gui's
[00:49:06] <cradek> the only thing they need other than that channel is file access to the gcode programs (if they make a preview based on it or something)
[00:49:56] <kevinh_> hmm... this might seem like a doable route. let linux do what it does best, let the mac do what it does best
[00:50:05] <archivist_emc> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?EMC_Components
[00:50:20] <cradek> heh, them's fighting words :-)
[00:51:09] <kevinh_> by that, i mean make it easy for newbies to use.
[00:51:10] <cradek> people have tried 2-computer solutions in other ways, often involving windows+linux but lately beagle+linux
[00:53:04] <archivist_emc> is there any cam software on mac
[00:53:10] <kevinh_> no cam software
[00:53:22] <kevinh_> plenty of converters and such
[00:53:27] <kevinh_> but no actual cam software
[00:53:54] <kevinh_> (that i've found)
[00:54:21] <cradek> as much as it pains me to say it, a port to windows would probably be more useful
[00:54:22] <archivist_emc> I know an attempt to compile HeeksCAD and cnc on mac but its not running yet
[00:54:52] <cradek> you'd have all the same problems, including a closed kernel and no realtime
[00:55:11] <cradek> but at least you'd have cam software and pci slots
[00:55:31] <cradek> (of course I have zero interest in doing that...)
[00:55:47] <kevinh_> well.. you can at least get the xnu source
[00:55:57] <kevinh_> the core of mac os x is oss, under the apple license
[00:56:16] <cradek> maybe you have a chance at a realtime kernel then
[00:56:27] <cradek> but again, if there's no realtime-capable hardware you're still screwed
[01:06:01] <kevinh_> it seems like this just isn't feasible, at least with emc2. only choice might be a remote-gui, but that wouldn't offer an all-in-one solution.
[01:08:10] <PCW> Probably best realtime hardware possibility on mac is Ethernet, but you would have to dedicate the Ethernet port to the realtime net
[01:08:56] <kevinh_> that's a lot easier to accomplish. wireless for internet, wired for real time
[01:09:46] <cradek> kevinh_: would you be making the hardware part too?
[01:11:56] <PCW> Is there a RTnet or Ethercat driver for EMC?
[01:11:58] <PCW> I thought someone on the mailing list was poking at RTNet
[01:12:12] <cradek> iirc, ethercat has a weird license
[01:12:17] <cradek> not sure about rtnet
[01:14:13] <danimal_garage> http://img.photobucket.com/albums/v20/danielwilcox/Screenshot-2.png
[01:14:31] <danimal_garage> any chance you can take a look at that and see if there's anything blatantly obvious?
[01:14:42] <PCW> Theres an open source master for Ethercat
[01:14:43] <PCW> (slaves are proprietary bu if you use their chips the proprietary stuff is all in the chip)
[01:14:51] <danimal_garage> i'll give you a list of the inputs
[01:15:18] <danimal_garage> I22= iomotion.toolchange
[01:15:24] <alex_joni> PCW: there are some Ethercat steps out there
[01:15:41] <danimal_garage> i23=iomotion.toolprepare
[01:15:59] <danimal_garage> Q9=turret up/spin
[01:16:01] <alex_joni> there is at least a guy in germany using ethercat to control some parts of his machine (not sure if all or not)
[01:16:10] <PCW> Steps?
[01:16:16] <danimal_garage> Q11=turret stop
[01:16:39] <cradek> danimal_garage: if you open your Symbols window you can name them with words - it's a lot easier to follow then
[01:16:45] <danimal_garage> q7= tool changed
[01:16:51] <danimal_garage> oh ok
[01:17:17] <alex_joni> PCW: starting points
[01:17:20] <danimal_garage> i'll do that now
[01:17:25] <alex_joni> wouldn't call it a finished thing ..
[01:18:29] <cradek> danimal_garage: brb, sorry
[01:19:19] <PCW> Seems like you would want some alternate read/write packets or finer scheduling for packet based I/O
[01:19:21] <PCW> (this is also an issue for SPI if you want read-compute-write-wait- a-longtime
[01:22:38] <PCW> alex_joni Were going to play a bit with Ethercat, we have a new FPGA board that has a Ethercat option in layout now
[01:23:40] <alex_joni> cool
[01:23:58] <alex_joni> if you want I can look up what I had on ethercat&emc2
[01:24:16] <alex_joni> * alex_joni still thinks it's the best thing for rt-eth
[01:24:25] <alex_joni> except the high prices for the slaves
[01:24:42] <alex_joni> or the FPGA slave code
[01:25:47] <kevinh_> back. sry
[01:26:52] <PCW> I didnt want to use Ethercat initially but its the best latency of the RT Etthernet families and using the slave ASIC
[01:26:54] <PCW> we can bootstrap the FPGA like all our AIO cards so other than complexity it looks pretty nice
[01:27:24] <kevinh_> cradek: if the need arises.
[01:28:06] <kevinh_> if i can't do it myself, i'll find someone who can
[01:28:39] <alex_joni> PCW: found something, hang on for a link
[01:29:04] <alex_joni> PCW: http://juve.ro/~juve/ethercat/
[01:30:02] <danimal_garage> http://img.photobucket.com/albums/v20/danielwilcox/Screenshot.png
[01:30:02] <PCW> Thanks Alex. got it!
[01:30:11] <alex_joni> there are 2 different comps there
[01:30:26] <danimal_garage> there's my updated ladder
[01:31:14] <kevinh_> comp = compiled driver?
[01:31:18] <alex_joni> the rar one is connecting to an encoder
[01:31:30] <alex_joni> kevinh_: no, something that gets preprocessed into a full .c file
[01:31:43] <alex_joni> it's a tool to keep writing HAL drivers simple
[01:32:28] <alex_joni> kevinh_: here's an example: http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=src/hal/components/and2.comp;h=345eeb0841ee92e9c5993675d2b825eefca3e27c;hb=HEAD
[01:32:47] <kevinh_> i opened it up as soon as i realized it was ascii
[01:33:01] <alex_joni> the link's a bit confusing ;)
[01:33:30] <alex_joni> * alex_joni is off to bed
[01:33:34] <alex_joni> 3:30 am is a bit late
[01:33:39] <kevinh_> looks a little like vhdl
[01:34:13] <kevinh_> by that i mean verilog hdl
[01:36:41] <danimal_garage> it's changing one position every t#m6 no matter what tool i enter
[01:50:12] <danimal_garage> ok now i got it so it just spins. i'm pretty sure i have the weighted sum wrong
[01:50:27] <danimal_garage> which would be a hal issue
[01:51:01] <danimal_garage> seems like the ladder might be right now
[01:52:07] <danimal_garage> the output of the compare block never triggers the coil it's supposed to when it gets to the right tool position
[02:14:20] <danimal_garage> ok it works, but the tool numbers are messed up
[02:15:37] <danimal_garage> t8=t8, t7=t9, t6=t10, t5=t11, t4=t12, t3=t13, t2=t14, and t1=?
[02:15:48] <kevinh_> what are you making?
[02:15:58] <Jymmm> a sx4
[02:15:59] <danimal_garage> trying to get my turret working on my lathe
[02:16:02] <Jymmm> a 2x4
[02:16:33] <danimal_garage> but the good news is that it works
[02:23:35] <danimal_garage> which is alot farther than i was this morning
[02:23:42] <kevinh_> congrats.
[02:24:03] <kevinh_> i wish i had gotten something so interesting working
[02:33:36] <kevinh_> how accurate is the timer in emc?
[02:33:48] <kevinh_> like how precisely does the system operate
[02:34:18] <danimal_garage> i would assume it's extremely accurate
[02:34:58] <danimal_garage> figure if my machine can hold .0002", it's gotta be an accurate program
[02:36:14] <kevinh_> how often does it fire? to the nanosec? to the milli?
[02:36:49] <danimal_garage> you're asking the wrong person
[02:36:57] <danimal_garage> i'm a n00b
[02:36:58] <kevinh_> heh. you're the only person here
[02:37:26] <danimal_garage> i figure it's as fast as your computer's latency
[02:37:29] <GonMD> lol, im in the hospital
[02:37:38] <kevinh_> but that could be different depending
[02:37:53] <kevinh_> it has to be reliably discrete
[02:38:07] <GonMD> danimal_garage: get your turret up and running?
[02:38:26] <danimal_garage> GonMD: Gonorrhea again??
[02:38:36] <GonMD> har har. appendicitis
[02:38:47] <danimal_garage> ouch, you ok?
[02:38:54] <danimal_garage> considering
[02:39:12] <GonMD> get in at 6am, in surgery after lunch, they letting me go toworrow morning
[02:39:39] <danimal_garage> turret is working, however my weighted_sum is off, since t12=t4 and so on
[02:39:52] <danimal_garage> thats not too bad
[02:39:59] <danimal_garage> enjoy the meds
[02:40:36] <GonMD> dont have any meds atm
[02:42:10] <kevinh_> what time is it where you are?
[02:42:20] <i-pink> hii alll
[02:42:24] <danimal_garage> 6:41pm
[02:42:28] <GonMD> hola
[02:42:42] <danimal_garage> hi i-pink
[02:42:54] <i-pink> someone can help me with spelling
[02:43:13] <i-pink> i want to make T shert
[02:44:22] <i-pink> on the back i print cell phone barcode to my blog
[02:45:00] <kevinh_> | I ||||| II II IIIIII II II ||| || II| |III |||
[02:45:10] <i-pink> ????
[02:45:16] <GonMD> lol
[02:45:45] <i-pink> this is the pic to the back
[02:45:55] <i-pink> http://www.speedyshare.com/files/20343220/qrcode20.png
[02:47:06] <GonMD> lol. clever advirtising
[02:48:21] <LawrenceG> nice
[02:48:28] <i-pink> and on the front i want to put a crazy something
[02:48:38] <i-pink> thanks
[02:49:27] <i-pink> on the front in the tits area i think to print this
[02:49:42] <i-pink> Don't look, read!
[02:51:30] <i-pink> and under that to put a barcode are send this "I know i am look HOT, but i have a boyfriend"
[02:51:41] <i-pink> what do you think?
[02:53:22] <LawrenceG> what instrument can read the bar codes?
[02:53:44] <i-pink> all cell phone with camera
[02:54:21] <LawrenceG> ah....
[02:55:30] <i-pink> yes
[02:55:45] <i-pink> what to you think?
[02:56:47] <i-pink> this is me and my first T shert
[02:56:48] <i-pink> http://inbarhovav.blogspot.com/2009/12/blog-post_22.html
[02:57:25] <LawrenceG> I like to look..... boy Friend should share the beauty
[02:58:21] <i-pink> what do you mean?
[03:02:37] <i-pink> haa
[03:02:58] <i-pink> "I know i am look beauty, but i have a boyfriend"
[03:03:02] <i-pink> LawrenceG ?
[03:04:14] <jepler> /lastlog -clear
[03:04:23] <cradek> ditto
[03:04:47] <i-pink> ?
[03:04:50] <LawrenceG> sorry
[03:05:02] <i-pink> ok
[03:05:11] <i-pink> is ok.. :-)
[03:06:22] <jepler> kevinh_: the "dwell" timer (e.g., for G4P#) is not particularly precise -- the design is generally to 'wait no less than' the amount specified
[03:08:28] <danimal_garage> is there really nothing on weighted sum in the wiki?
[03:09:55] <jfigie> cradek: yesterday I asked about initializing a signal in hal to a value. setp does not work because the signal already has a writer. Is there any other way?
[03:11:12] <cradek> if you have a writer hooked to the signal, how to initialize it to a value is a meaningless question because the writer is writing
[03:11:50] <cradek> in hal, the writer writes every period (except in unusual cases like bidirectional pins)
[03:11:57] <jfigie> well I have it hooked up in a feedback loop
[03:12:12] <danimal_garage> cradek: i got the ladder working great, except the weighted sum is off
[03:12:13] <cradek> can you back up and say generally what you're trying to do?
[03:12:17] <cradek> danimal_garage: slick!
[03:12:30] <danimal_garage> the tool # i enter doesnt equal the tool number it changes to
[03:12:43] <cradek> danimal_garage: mine are all off by 1...
[03:12:49] <danimal_garage> but it does change tools nicely
[03:12:55] <cradek> very cool.
[03:13:03] <cradek> that didn't take you long.
[03:13:05] <danimal_garage> had to put a timer for the turret stop
[03:13:14] <danimal_garage> it was latching before it settled
[03:13:24] <danimal_garage> but now it works great
[03:13:26] <cradek> yep it has to settle against the stop before you let it drop down, I have a timer there too
[03:13:36] <danimal_garage> cool
[03:13:57] <danimal_garage> i havent found a number for t1 yet
[03:14:02] <cradek> haha
[03:14:25] <cradek> does it spin forever if you say T20?
[03:14:45] <cradek> if so, that's a bug you'll want to fix eventually :-)
[03:14:45] <danimal_garage> but t14=t2, t13=t3, t11=t5, and so on
[03:14:47] <danimal_garage> yes, it does
[03:14:59] <cradek> if you draw out the bit patterns, maybe you'll see which wires you have wrong
[03:15:47] <danimal_garage> yea, i'm not sure on where to enter the bit patterns... i coppied that section of the hal
[03:15:56] <jfigie> cradek: I am using two unused keyboard keys to increase or decrease a signal value. I have basically have a mux and an adder and a limiter hooked up in a feedback loop around the adder. This seems to work pretty well so far. but it would be nicer if it could start at some other value than the lower limit.
[03:16:05] <danimal_garage> i see a pattern, but it has # in front of each line
[03:16:21] <danimal_garage> hold on, i'll pastebin it
[03:16:56] <cradek> jfigie: what kind of adder is it?
[03:17:08] <jfigie> sum2
[03:17:52] <cradek> are you using classicladder for anything else or are you pure hal blocks so far?
[03:18:09] <Danimal> just hal
[03:18:23] <Danimal> but so far looks easy
[03:18:35] <jfigie> pure hal blocks
[03:18:50] <Danimal> took me a while to understand what i was doing, but it looks easy
[03:18:56] <Danimal> oh i thought you were talking to me
[03:19:36] <Danimal> http://pastebin.ca/1751952
[03:19:56] <cradek> jfigie: I'm having trouble understanding your setup. pastebin the relevant hal?
[03:20:47] <Danimal> lines 277-285.... is that where i put in the bit patterns, or is that just for reference?
[03:21:03] <cradek> danimal_garage: just invert 'em all
[03:21:19] <jepler> Danimal: "#" means it's a comment, hal won't pay any attention to it..
[03:21:22] <cradek> er maybe not
[03:21:52] <Danimal> jepler: yea, i know, just making sure i wasnt missing something
[03:22:05] <jepler> documentation of weighted_sum: http://linuxcnc.org/docs/html/man/man9/weighted_sum.9.html
[03:22:13] <jepler> (also 'man weighted_sum' in the terminal)
[03:22:36] <cradek> you aren't setting your wsum weights
[03:22:48] <jepler> wsum.N.bit.M.weight signed rw
[03:22:48] <jepler> The weight of the m'th input of weighted summer n. The default
[03:22:48] <jepler> value is 2^m.
[03:22:54] <cradek> oh ok
[03:22:57] <jepler> they're supposed to have sensible-ish defaults
[03:23:20] <Danimal> what is 2^m?
[03:23:24] <cradek> brb
[03:23:56] <jepler> ^ denotes "to the power of"
[03:24:16] <Danimal> oh ok
[03:24:17] <jepler> it means the default is 1, 2, 4, 8, 16, ...
[03:24:24] <jfigie> cradek: I will add some comments to the file to make it easier to read. I don't want to waist too much of your time. Maybe in the next few days I will have something to show.
[03:26:10] <Danimal> man i still dont get it
[03:26:50] <Danimal> what do they mean by weight
[03:27:10] <Danimal> "Its output is the offset plus the sum of the weight of each TRUE input bit."
[03:27:49] <jepler> Danimal: think of a balance scale. You have a bunch of weights. The label one one of them is 1, the label on another is 2, and so forth. If you put some combination of them on the scale, the weight on the scale is tha many ounces.
[03:28:25] <jepler> that is the idea that makes them be "weights"
[03:28:32] <Danimal> isnt each input bit only equal 1?
[03:29:31] <jepler> when bit.M.in that is like putting the M'th weight on the scale
[03:29:47] <jepler> er, when bit.M.in is TRUE that is like putting the M'th weight on the scale
[03:29:54] <jepler> and when bit.M.in is FALSE it's like taking that weight off
[03:30:09] <Danimal> each bit equals 1 though, right?
[03:30:30] <Danimal> or does that depend on a parameter for each bit?
[03:30:47] <jepler> the bit.M.weight says how weighty the Mth weight is
[03:31:06] <jepler> so if the bits are 1, 0, 1 and the weights are 1, 2, 4 then the sum is 1+4 = 5
[03:31:12] <jepler> (and if offset is 0)
[03:31:34] <Danimal> ah ok
[03:33:13] <Danimal> i think i got it
[03:33:16] <Danimal> thank you
[03:33:40] <Danimal> sometimes i need it dumbed down to understand it :)\
[03:33:43] <jepler> Danimal: I hope so. I'm not sure I'm doing the best job at explaining it ..
[03:33:59] <Danimal> yea, it seems to make sense now
[03:34:37] <Danimal> i'll go give it a shot
[03:35:41] <Danimal> after this, my lathe is usable :)\
[03:35:51] <Danimal> -\
[03:40:12] <jepler> jfigie: I am pretty sure there's not a way to "preset" limit2 or limit3's output. One possibility without changing how limit2/limit3 work would be to use an additional sum2 to give the 'initial offset' you want.
[03:46:24] <mozmck> I think gcc 4.4 broke my emc2 compiling...
[03:48:54] <jepler> 'night all
[03:50:00] <danimal_garage> night jepler
[03:50:01] <jepler> mozmck: hm, ubuntu 9.10 is using gcc 4.4.1 and builds emc ok for me .. if you pastebin the errors I'll take a look tomorrow though
[03:50:28] <mozmck> there's a million of them! all in the kernel includes
[03:50:43] <mozmck> 'night
[04:38:08] <Danimal> this thing is driving me nuts
[04:46:41] <i-pink> you can help me with the geek comix site
[04:59:12] <tom3p> kevinh consider firewire, thats used in hard realtime, google aerotech cnc. there are no existing drivers but thats sounds like what you're capable of
[05:02:10] <tom3p> when is the new LTS coming out? April? I spent a long day trying to get some hardware up that is supported in kernel 2.6.28, might as well wait for LucidLynx and an emc2 to go with it
[05:13:29] <kevinh_> member:tom3p: now there's an idea that i hadn't thought of.
[05:13:44] <kevinh_> gah stupid auto links
[05:18:49] <tom3p> and i have an aeroteech 3200 for sale (hint hint)
[05:23:38] <kevinh_> it looks like i could use isochronous mode over firewire
[05:25:12] <kevinh_> looks like usb supports same
[05:25:25] <kevinh_> at least in theory
[05:25:39] <kevinh_> but i know the mac os has isochronous fw
[05:27:20] <kevinh_> i'll be around in about 9 hours. i need to sleep or i'm gonna pop.
[05:27:43] <tom3p> i dont understand it like you, but i just found maybe 6 cnc manufactures ( controls manufacturers ) touting firewire.
[05:27:45] <tom3p> gnite
[05:29:29] <kevinh_> kevinh_ is now known as kevinh
[06:31:52] <Danimal> well cradek, i got it to where you have it (1 tool off)
[06:32:26] <Danimal> best i can do for now
[06:35:25] <Danimal> at least it's working
[06:36:20] <Danimal> once i get the covers back on i'll take a video
[06:42:00] <toastydeath> most take videos with the covers off
[06:42:11] <toastydeath> you don't get quite as much publicity with them on
[06:43:31] <Danimal> not much to see with them off
[06:43:50] <toastydeath> i was making a sex tape joke
[06:43:56] <Danimal> ha
[06:44:06] <Danimal> sorry, long day
[06:44:09] <toastydeath> haha s'ok
[06:57:58] <GonMD> evening ladies
[08:12:29] <GonMD> you guys ever hear of a Yurt?
[08:15:02] <toastydeath> yes
[08:15:10] <toastydeath> girlfriend's friend in the sca camps in one
[08:25:05] <ChanServ> [#emc] "This is the #emc channel - talk related to the Enhanced Machine Controller and general machining. Website: http://www.linuxcnc.org/, wiki at http://wiki.linuxcnc.org/"
[09:04:54] <lilalinux> lilalinux is now known as lilalinux_away
[09:07:13] <christel> [Global Notice] Hi all, it appears one of our volunteers set a network ban which were way too wide last night in an attempt to deal with the current attacks; as a result a truckload of genuine users have been banned from the network, with a misleading and unhelpful message. For more information please enable +w in your client. Apologies for the inconvenience, we'll try tidy up after them now.
[09:09:34] <lilalinux_away> lilalinux_away is now known as lilalinux
[10:08:12] <krushia_> krushia_ is now known as krushia
[10:09:45] <alex_joni> logger_1: bookmark
[10:09:45] <alex_joni> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2010-01-15.txt
[10:58:25] <MrSunshine_> logger_1, bookmar
[10:58:25] <MrSunshine_> I'm logging. I don't understand 'bookmar', MrSunshine_. Try /msg logger_1 help
[10:58:28] <MrSunshine_> logger_1, bookmark
[10:58:28] <MrSunshine_> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2010-01-15.txt
[10:58:30] <MrSunshine_> hah
[11:23:31] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[11:43:45] <_Poincare> _Poincare is now known as Poincare
[12:29:50] <MarkusBec> MarkusBec is now known as MarkusBec_away
[12:34:12] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[12:41:07] <MarkusBec> MarkusBec is now known as MarkusBec_away
[12:41:55] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[13:25:25] <jepler> http://failblog.files.wordpress.com/2010/01/epic-fail-sign-fail1.jpg
[13:26:36] <jthornton> lol
[13:46:57] <ries> jepler: http://failblog.files.wordpress.com/2009/12/epic-fail-paternity-fail.jpg
[13:48:12] <JustinXJS> hehe
[14:32:30] <JT-Work> if you have a choice between a jacobs 2 and a jacobs 33 for a small drill chuck is there any advantage to either one?
[14:33:24] <ries> * ries managed to create a proper g-code file for axis using ProEngineer and Pro/NC hurraaay!
[14:39:42] <JT-Work> looks like the od of the 33 is bigger...
[14:43:19] <JT-Work> * JT-Work looks for the MSC discount code dan1mal_garage posted the other day
[14:50:59] <cradek> 11:25 < danimal_g> up to 40% off on metalworking stuff at msc today only. promo code MWMANIA40W
[14:51:14] <cradek> JT-Work: try typing /lastlog msc
[14:51:37] <cradek> (I don't know if that's universal in new irc clients but old good ones have it)
[14:56:02] <JT-Work> cradek: thanks
[14:56:39] <JT-Work> won't work on chatzilla :(
[14:59:39] <cradek> it means show me the lines from the scrollback matching 'msc'
[15:00:01] <cradek> I use /lastlog http all the time when I remember someone gave a link to something
[15:30:29] <isssy> hi all
[16:26:09] <tom3p> anyone know how kernel 2.6.8 has 'fixed' usb? i have a 'labjack' adio and the drivers require a hardware ugrade (buy a new one $$) or a new kernel (2.6.8+)
[16:27:38] <Danimal> cradek: it's working now, but it's 1 off like yours
[16:28:16] <Danimal> no biggie, i can live with that
[16:29:05] <cradek> Danimal: neat
[16:29:20] <cradek> Danimal: just draw an arrow with a sharpie that points to the right number :-)
[16:29:38] <Danimal> yea, i'll get used to it
[16:34:00] <dan1mal_garage> cradek: what if i just move the little magnet over one position?
[16:34:14] <dan1mal_garage> on the switch for the tool position
[16:35:08] <cradek> would it work for 8 and 1?
[16:35:24] <cradek> if you can fix it by moving the magnet, you can fix it in hal
[16:36:00] <dan1mal_garage> it didnt work with offset
[16:36:14] <lilalinux> lilalinux is now known as lilalinux_away
[16:36:16] <dan1mal_garage> i tried both 1 and -1
[16:36:32] <dan1mal_garage> but physically moving the magnet should work
[16:36:55] <dan1mal_garage> because it only goes from 1-8, where the wsum can be infinite
[16:37:00] <dan1mal_garage> right?
[16:37:14] <cradek> maybe you need %IW0 % 8 or similar
[16:39:19] <dan1mal_garage> oh
[16:39:29] <dan1mal_garage> well i just moved the bagnet and it works great
[16:39:39] <cradek> brute
[16:39:40] <dan1mal_garage> magnet*
[16:39:50] <lilalinux_away> lilalinux_away is now known as lilalinux
[16:40:27] <dan1mal_garage> i do have one issue (even before i moved the magnet)... it wont move one position at a time
[16:40:43] <dan1mal_garage> i have to move several and come back to it
[16:40:55] <cradek> huh
[16:41:07] <dan1mal_garage> and it doesnt seem to be a physical issue, it wont even try
[16:41:31] <dan1mal_garage> so if i'm at 2, it wont go to 3 unless i go to 5 and come back to 3
[16:41:32] <cradek> maybe your magnet is not quite in the right place
[16:41:43] <dan1mal_garage> this is before i touched the magnet
[16:41:52] <cradek> I mean, maybe once it settles at 2, it reads 3 instead
[16:41:53] <dan1mal_garage> but you still could be right i guess
[16:42:07] <cradek> you should debug :-)
[16:42:18] <cradek> sounds like you are really close.
[16:42:38] <dan1mal_garage> how do you have yours connected to the mesa boards?
[16:43:08] <cradek> all my IO is through the two 7i37s
[16:48:43] <dan1mal_garage> it doesnt seem to be the magnet
[16:50:17] <JT-Work> robh did mention something about having fun getting the timing correct on his super slant turrets
[16:52:49] <dan1mal_garage> i might have the wiring wrong
[16:53:23] <dan1mal_garage> i have the wires from the switches going to the negative terminals on the 7i37
[17:01:29] <cradek> hi seb!
[17:02:13] <lilalinux> lilalinux is now known as lilalinux_away
[17:10:43] <jimbo> Does anyone know where to source Inch dimensioned angular contact bearings? IE .625 ID 1.375 OD
[17:15:14] <kevinh_> kevinh_ is now known as kevinh
[17:18:52] <seb_kuzminsky> hey there :-)
[17:28:03] <alex_joni> hi seb
[17:37:17] <dan1mal_garage> cradek: i figured out the problem... the turret doesnt drop untill 1 position after it gets the signal for that position
[17:37:49] <dan1mal_garage> so when it gets the signal for tool 1, it actually drops in position 2 because it's ahead
[17:38:09] <dan1mal_garage> so thats why it wont go forward one position, because it's actually already in that position
[17:38:24] <cradek> aha
[17:38:44] <cradek> so your magnet is off?
[17:38:47] <dan1mal_garage> no
[17:39:17] <dan1mal_garage> it doesnt stop till one full position after it gets the signal
[17:39:31] <dan1mal_garage> it takes that long to physically stop
[17:40:17] <cradek> hmm
[17:40:39] <dan1mal_garage> no matter where the magnet is, it will always stop one position farther than it triggered the sensor
[17:41:28] <cradek> I don't see why that would be - you just have to get it oriented so when the magnet says it's in position, there's still time to get the stop triggered in time to stop it
[17:42:07] <cradek> all of this is in realtime - it will be the same every time
[17:42:44] <dan1mal_garage> that wont work, because it will never be in the position that it was in when it triggered. it takes a full position or more to slow down
[17:43:04] <dan1mal_garage> i tried it, when you get the the very edge to where it kinda works, it wont repeat
[17:43:06] <cradek> but it doesn't slow down - it smashes into a hard stop at full speed
[17:43:23] <cradek> maybe mine is different?
[17:43:32] <dan1mal_garage> hmmm
[17:43:40] <dan1mal_garage> mine is older, possible i guess
[17:44:06] <cradek> I assure you it can index one position at a time...
[17:44:20] <cradek> the turret would have been an unacceptable design if not
[17:44:28] <dan1mal_garage> hmm
[17:44:35] <dan1mal_garage> did you have to adjust your magnet?
[17:45:02] <cradek> I think I had it off to figure out wiring, so I can't say
[17:45:10] <dan1mal_garage> oh ok
[17:45:16] <cradek> I did fiddle with it to make it work right
[17:45:22] <dan1mal_garage> i didnt touch mine till this morning
[17:45:29] <cradek> but whether it ended up in the original position or not, I have no idea
[17:45:36] <dan1mal_garage> ah ok
[17:45:44] <dan1mal_garage> i'll play with it some more
[17:47:39] <dan1mal_garage> the only other thing i was thinking is to put a block in there that adds one to the tool number every time it finishes a tool change
[17:49:04] <dan1mal_garage> so basically -[%IW0=%IW1]-----------------(%B1)-
[17:49:50] <dan1mal_garage> -[%B1]---------------------[%IW1+1]-
[17:50:10] <dan1mal_garage> or something
[18:04:55] <dan1mal_garage> hmm guess i cant do that
[18:21:11] <lilalinux_away> lilalinux_away is now known as lilalinux
[18:23:31] <dan1mal_garage> ok i got it so it's still 1 off, but it starts at t0 and goes to t7
[18:24:57] <dan1mal_garage> and i fixed that with offset
[18:25:06] <dan1mal_garage> w00t! it works perfect!
[18:27:51] <dan1mal_garage> cradek: i put -[turret position= (tool prep number+1)]------------ in my ladder
[18:27:59] <dan1mal_garage> then i added 1 to my wsum offset
[18:28:45] <dan1mal_garage> and the magnet is right above the correct sensor, and t1=t1 ----t8=t8
[18:29:09] <dan1mal_garage> and it moves one tool at a time
[18:31:49] <dan1mal_garage> i'm friggin excited
[18:31:55] <cradek> yay!
[18:32:25] <dan1mal_garage> it does in fact stop one position after it triggers the sensor
[18:33:14] <dan1mal_garage> so maybe ours are different
[18:33:34] <cradek> yeah, who knows
[18:33:46] <cradek> or our ladders and/or magnets are different
[18:34:18] <dan1mal_garage> yea the size of the magnet will effect that
[18:34:29] <dan1mal_garage> bigger the magnet=more overlap
[18:35:00] <dan1mal_garage> all that matters is it works well
[18:35:17] <lilalinux> lilalinux is now known as lilalinux_away
[18:38:36] <cradek> yep, just has to reliably get the right tool
[18:38:52] <cradek> glad you got it - that's really cool.
[18:40:49] <dan1mal_garage> yea, thanks for the help once again
[18:42:12] <dan1mal_garage> so all that's left is the lube, coolant, and collet closer
[18:42:21] <dan1mal_garage> and it's 100%
[18:42:25] <mozmck> I compiled emc2 here on 9.10 with my rtai patched 2.32.2 kernel a couple of days ago and it worked fine.
[18:42:30] <dan1mal_garage> all that stuff is easy
[18:42:48] <mozmck> Now it gives me these errors http://pastebin.com/m5e47f545
[18:42:51] <cradek> no problem - you're clueful and easy to work with. :-)
[18:43:24] <mozmck> any idea what might be wrong?
[18:43:26] <cradek> #
[18:43:28] <cradek> /usr/src/linux-headers-2.6.32.2-rtai/arch/x86/include/asm/bitops.h:12:2: error: #error only <linux/bitops.h> can be included directly
[18:43:33] <dan1mal_garage> i dunno if i'm clueful lol i get stumped by alot of easy stuff
[18:44:03] <cradek> mozmck: what changed? it wasn't in emc...
[18:44:15] <mozmck> cradek: I saw that and changed it but it only removed one or two errors
[18:44:37] <mozmck> I don't know. I think I may have done an update that had an update to gcc...
[18:45:30] <cradek> is this a 64 bit install?
[18:45:40] <mozmck> no
[18:46:54] <cradek> sorry, I have no idea what it all means, except it looks like your kernel headers are screwy somehow
[18:47:54] <mozmck> looks like. I think they have changed some locations in the latest kernels, but I don't know why it worked once.
[18:49:33] <celeron55_> i think the problem is in the files including bitops.h
[18:49:51] <celeron55_> they supposedly include asm/bitops.h while they should include linux/bitops.h
[18:50:14] <celeron55_> probably changed to the newest kernel to be an error instead of a warning or something?
[18:50:17] <cradek> but he said he fixed that...?
[18:50:43] <celeron55_> ah
[18:50:54] <mozmck> I did in one file, maybe I should check more thoroughly
[18:51:26] <mozmck> it's weird anyhow.
[18:51:36] <celeron55_> do a grep -R 'asm/bitops.h' in /home/moses/Projects/emc2/src/
[18:51:49] <celeron55_> grep -R 'asm/bitops.h *
[18:51:53] <celeron55_> ...
[18:51:55] <celeron55_> grep -R 'asm/bitops.h' *
[18:52:21] <mozmck> will do. thanks. don't know why it worked once :(
[18:52:42] <celeron55_> i suppose you updated your kernel?
[18:53:09] <celeron55_> if not, then it's a weird error
[19:14:12] <jepler> do you have an include/linux/bitops.h inside your emc source directory? Some of those messages refer to include/linux/bitops.h with a relative path.
[19:14:33] <jepler> there shouldn't be one, but maybe you created it at some point with all the best intentions at heart
[19:15:23] <celeron55_> i see it's included from a file which is in /usr/src/linux-headers-2.6.32.2-rtai/arch/x86/include/asm/
[19:15:31] <celeron55_> so i guess it's not a problem
[19:16:06] <frallzor> http://pici.se/p/qsFtKedkt/ does this montage look ok?
[19:16:38] <jepler> anyway, if the kernel says to include linux/bitops.h instead of asm/bitops.h you can try changing that in src/rtapi/rtapi_bitops.h
[20:09:24] <lilalinux_away> lilalinux_away is now known as lilalinux
[20:34:47] <gene_> Hi guys, anybody home?
[20:35:22] <Jymmm> Hi gene_
[20:35:36] <gene_> Helllooooooo??
[20:35:51] <gene_> hi Jymmm
[20:36:42] <gene_> which figure in the latency-test is the critical one aqnd what can it be maximum for a stepper system?
[20:40:50] <cradek> base period jitter, about 25000 depending on what step rate you want
[20:42:22] <SWPLinux> gene_: there are labels at the top of the test results window. the one labeled "max jitter" is the one to look at
[20:42:28] <SWPLinux> maybe that should be renamed
[20:43:15] <GonMD> afternoon all
[20:43:27] <gene_> base period is a hair over 11,000 max after messing with it for a while, max interval so far is 35,000
[20:43:39] <SWPLinux> wow. latency is really really bad in sim mode on a machine with continuous USB errors
[20:43:48] <gene_> base period jitter is a hair over 11,000 max after messing with it for a while, max interval so far is 35,000
[20:44:28] <SWPLinux> max interval will be roughly equal to max jitter plus the programmed interval
[20:44:50] <SWPLinux> ie, a 25 uS thread with 10 uS max jitter will have roughly 35 uS max interval
[20:44:56] <gene_> I think the old TNT card was better, but not hugely so. programmed is 25 u-secs
[20:46:19] <gene_> Different video card, I finally found a machine this X-1650Pro agp ati card will run in. Video sure looks a lot better too!
[20:47:21] <SWPLinux> I can imagine
[20:47:26] <gene_> Would this be a candidate to install the catilyst drivers?
[20:47:32] <SWPLinux> in the TNT days, the only company that had usable output was Matrox
[20:47:43] <SWPLinux> Catalyst aren't realtime compatible
[20:47:53] <SWPLinux> and they aren't open source either, so you can't build your own
[20:48:09] <gene_> And I see acc the wiki, that they weren't all that fast. ;-)
[20:48:27] <gene_> Ok, thanks Steve
[20:48:30] <SWPLinux> the latest should be OK, and certainly faster than any TNTN
[20:48:32] <SWPLinux> -N
[20:48:40] <gene_> yup
[20:49:52] <tlab2> tlab2 is now known as tlab
[20:59:43] <gene_> stopped the test, running emc, with a 60 us base period. This card isn't a bit better than the TNT in showing the backtrace
[21:00:27] <SWPLinux> if you're using Mesa GL, then that's more or less to be expected
[21:00:46] <SWPLinux> you need to use something slightly, but not completely, accelerated to get better performance :)
[21:01:14] <gene_> if I don't slow the feed override down to about 15%, it cuts nice big raduis curves in the emc logo in place of nice square corners, just like every other card I've had in this thing.
[21:01:45] <skunkworks_> um.. isn't that your accelleation settings?
[21:02:06] <andypugh> G61.2?
[21:02:31] <SWPLinux> or cycle time
[21:02:44] <SWPLinux> display cycle time
[21:02:56] <andypugh> G61.1 I mean.
[21:03:00] <gene_> I can see it slowing for the corners, and where there are short bridges like in the corners of some chars, that follows nicely.
[21:03:09] <Jymmm> * Jymmm found some cool Lime Green LED's yesterday
[21:03:21] <gene_> G61.1 from the mdi screen keeps when I run the logo?
[21:03:26] <andypugh> http://linuxcnc.org/docs/html/gcode_main.html#sub:G61,-G61.1,-G64:
[21:07:05] <gene_> Doesn't make any diff. This is 2.3.4, and I cannot issue an mdi command, then return to manual, f, and rerun, it throws a can't do that in manual error. All I can do is restart 2.3.4 to recover.
[21:07:35] <andypugh> The line-segment blending is deliberate, they call it "Naïve CAM detection" which I thought was a naïve way to machine cams, but apparently means that is is designed to work with unsohisticated CAM G-code creation software.
[21:08:03] <Jymmm> SWPLinux: What size ExpressSlot do you have?
[21:08:09] <andypugh> Can't you add a G61 to the G-code?
[21:08:27] <SWPLinux> Jymmm: the wide one, 54?
[21:08:31] <Jymmm> k
[21:08:46] <SWPLinux> no idea if the USB part works though
[21:08:48] <andypugh> Though it sounds like your accell rates are a bad match to your speeds.
[21:09:04] <SWPLinux> so a plug-in card would have to be specifically PCIe to USB
[21:09:48] <andypugh> Hmm, is this rounding happening at machining speed? That demo file runs pretty fast.
[21:10:07] <Jymmm> SWPLinux: Yep, that's what I'm looking for =)
[21:10:48] <gene_> Andy, I've found I can go faster overall, if the accels are lower. Yes, I have seen it whiloe machining, but the corners of the parts are nice & sharp while the display isn't. Weird.
[21:10:49] <andypugh> Jymmm: No extra USB headers on the motherboard?
[21:11:07] <SWPLinux> andypugh: not on a laptop
[21:12:05] <SWPLinux> http://www.newegg.com/Product/Product.aspx?Item=N82E16839229007
[21:12:47] <gene_> and now, if I hiut an F5, issue the G61.1 command, then the error says it has to be in mdi mode to do that. And yes, I'm sober!
[21:13:13] <tom3p> SWPLinux: re: latency and usb errs... a few days ago i reported how a munged fat on a usb stick caused 'unexpected realtime delays' and repairing the fat removed the errors.
[21:13:30] <SWPLinux> oh, I'm sure
[21:13:42] <andypugh> gene_: Yeah, sure, we believe you.
[21:13:58] <SWPLinux> this laptop now has a fried USB subsystem, so there is crap happening all the time with the system trying to initialize those ports/devices
[21:14:15] <SWPLinux> and I was running sim, so I didn't really expect to have good numbers :)
[21:15:20] <gene_> Ideas, or is this new card making emc perverse? I'm not running sim, but the parport is empty since I'm in a relatively warm basement, setup on my pool table.
[21:16:33] <SWPLinux> gene_: set the feed override very very low, and see if the corners still get rounded
[21:16:45] <Jymmm> SWPLinux: http://www.frys.com/product/5209127?site=sr:SEARCH:MAIN_RSLT_PG
[21:16:56] <SWPLinux> if not, then you have either too low an acceleration setting (preventing sharp corners)
[21:17:12] <Jymmm> SWPLinux: http://www.frys.com/product/5534050?site=sr:SEARCH:MAIN_RSLT_PG
[21:17:13] <SWPLinux> or you have a display cycle time that's long enough to miss some positions
[21:17:33] <SWPLinux> scenario 1 is that the corner is actually getting rounded and the display is accurate
[21:17:39] <andypugh> I can issue G61, G61.1 and G64 P0.001 happily enough in MDI here.
[21:17:56] <SWPLinux> scenario 2 is that the machine runs correctly, but the display update is too slow to show it right
[21:18:11] <SWPLinux> Jymmm: I wonder where the nearest Fry's is :)
[21:18:23] <SWPLinux> I think it may be in Hollywood or Santa Monica
[21:18:25] <Jymmm> SWPLinux: Where you at?
[21:18:39] <SWPLinux> Culver City (apparently)
[21:18:44] <Jymmm> Burbank
[21:18:51] <Jymmm> Oxnard =)
[21:18:52] <SWPLinux> near the intersection of 405 and Jefferson
[21:19:04] <SWPLinux> yeah, Burbank is annoying, and Oxnard is a ways away
[21:19:04] <andypugh> Hollywood, Burbank, Santa Monica.. It all sounds so exotic.
[21:19:11] <Jymmm> http://www.frys.com/isp/manhattanbeach.html
[21:19:25] <Jymmm> SWPLinux: http://www.frys.com/ac/storelocator/index.jsp
[21:19:29] <SWPLinux> watch the Golden Globes this weekend (or at least the red carpet show), you may catch me in a tux
[21:19:38] <SWPLinux> (probably not, but it's possible :) )
[21:19:40] <alex_joni> wow
[21:19:43] <gene_> after restarting emc the third time, The G61.1 on an F5 screen, followed by an r on the F3 screen, and its fixed. You guys have warm weather I think, you are making me jealous.
[21:19:44] <Jymmm> SWPLinux: =)
[21:20:01] <alex_joni> SWPLinux: :P
[21:20:04] <Jymmm> SWPLinux: That t-shirt tux of yours?
[21:20:05] <andypugh> Though it is concievable (though unlikely) that London (down the road from here) sounds exotic to you.
[21:20:05] <SWPLinux> it's 70 here, but 38 at home :)
[21:20:17] <SWPLinux> not a penguin tux
[21:20:28] <gene_> Steve, it seems to me that scenario 2 is the diagnosis here.
[21:20:31] <SWPLinux> though that would be funny, for me to show up in a "tux" shirt :)
[21:20:34] <Jymmm> SWPLinux: Why not, that be funny
[21:20:40] <andypugh> It's above freezing today for the first time this year.
[21:20:48] <SWPLinux> I didn't bring one thouhg
[21:20:50] <SWPLinux> though
[21:21:06] <gene_> Might be 40 here, we're having a heat wave, so maybe some of the snow will leave, hint hint...
[21:21:21] <SWPLinux> heh
[21:21:25] <SWPLinux> that's the same as Vermont
[21:21:38] <SWPLinux> I wonder how much ice there'll be when I get back
[21:21:43] <gene_> IzZat where you normally are?
[21:21:48] <SWPLinux> yep
[21:22:08] <SWPLinux> and there's supposed to be between 4 and 20 inches of rain here (!), starting Sunday evening or so
[21:22:23] <SWPLinux> hopefully after we've shot the Globes, and after we
[21:22:29] <SWPLinux> we've struck the equipment
[21:22:32] <gene_> didja bring a mud sled and a life jacket?
[21:22:33] <SWPLinux> as we're going to be outside
[21:22:38] <SWPLinux> no, not this time
[21:22:49] <SWPLinux> I also hope that I can get the heck out of here on Monday :)
[21:23:33] <gene_> Yeah, that whole state of Kalifornia spooks me anymore.
[21:23:51] <SWPLinux> I went outside this morning to use the hot tub, and I thought they had just extinguished some gas-powered tiki torches or something
[21:23:57] <SWPLinux> I think it was the smog
[21:24:06] <Jymmm> SWPLinux: If you do get that card, I wold test it in the parking lot.
[21:24:15] <SWPLinux> Jymmm: I'll bring the laptop inside ;)
[21:24:51] <Jymmm> SWPLinux: I wouldn't
[21:24:58] <gene_> probably, I've seen both LA and San Jose where you needed to take your lung out and wash them, couldn't see across the damned street.
[21:25:17] <SWPLinux> I'm quite sure they don't sell these, especially "well used"
[21:25:21] <gene_> s/lung/lungs/g
[21:25:24] <SWPLinux> yep
[21:25:42] <Jymmm> SWPLinux: No, the whole testing before buying thing in the store.
[21:26:02] <SWPLinux> ah
[21:26:09] <gene_> Dunno if I'll get the Globes here, no cable.
[21:26:20] <Jymmm> SWPLinux: It's weird, but they dont like ppl opening pckages in the store.
[21:26:35] <SWPLinux> I think it'll be on CBS, but I don't know if they'll have the "pre-game" show
[21:26:42] <SWPLinux> yeah, that makes sense
[21:26:55] <SWPLinux> open package, remove contents, replace package on shelf, Profit!
[21:27:03] <Jymmm> lol
[21:27:34] <gene_> Great, thats _my_ station, I was the Chief at the local CBS from 1984 to mid-2002
[21:28:06] <SWPLinux> here's an interesting (future) tablet computer for CNC control: http://www.mtechlaptops.com/future_laptop_models.html
[21:28:19] <SWPLinux> the T8900M, 3/4 down the page
[21:28:23] <Jymmm> "Who you calling a cutie queen you lint licker!
[21:28:27] <Jymmm> "
[21:28:30] <SWPLinux> gene_: cool
[21:28:43] <dan1mal_garage> i'm glad most of our smog in san diego blows up north
[21:29:01] <dan1mal_garage> i think san bernadino gets it all
[21:29:24] <SWPLinux> we get all ours from the midwest :)
[21:29:29] <SWPLinux> (back home anyway)
[21:29:40] <dan1mal_garage> you have smog in vt?
[21:29:51] <SWPLinux> no, not really
[21:30:00] <dan1mal_garage> i dont think we had smog in CT
[21:30:08] <SWPLinux> we used to get some acid rain kind of stuff, but I think some of the clean air things have actually helped with that
[21:30:08] <gene_> or from WV, we burn some of the coal that lights your lights
[21:30:33] <gene_> I actually think its better here than 20 years ago.
[21:30:36] <SWPLinux> we have coal and wood burning plants in Burlington, but they have pretty good scrubbers on them
[21:30:47] <skunkworks_> gene_: exact stop http://imagebin.ca/view/BDC9xDmn.html
[21:30:54] <dan1mal_garage> the only smog in VT comes from wood burning stoves and all the pot heads
[21:30:59] <SWPLinux> heh
[21:31:06] <skunkworks_> gene_: blended http://imagebin.ca/view/XQY1xE.html
[21:31:35] <SWPLinux> hey skunkworks, you should update that system :)
[21:31:39] <dan1mal_garage> in VT, 2% of the population voted for the green party (the pro-weed party)
[21:31:52] <SWPLinux> no, green isn't pro-hemp
[21:31:52] <dan1mal_garage> most out of any state
[21:32:00] <SWPLinux> that's the "legalize hemp" party
[21:32:05] <SWPLinux> (I'm not kidding)
[21:32:26] <SWPLinux> at least the Vermont Green party isn't about hemp
[21:32:32] <SWPLinux> or pot
[21:32:44] <Jymmm> bummer
[21:33:09] <gene_> Only I'm seeing it 2-4x worse than that. Hemp? Legal? It sure would save the taxpayers a hell of a lot of money wasted on enforcement now.
[21:33:30] <SWPLinux> yeah, tax it instead
[21:33:49] <Jymmm> I prefer pot to be legal over boose - less deaths
[21:33:51] <cradek> what would we do with all the empty space in the prisons?
[21:34:06] <Jymmm> cradek: rent em out like storage
[21:34:06] <dan1mal_garage> cradek: more room for CEO's
[21:34:22] <SWPLinux> CEO storage
[21:34:25] <dan1mal_garage> ha
[21:34:31] <skunkworks_> SWPLinux: I should update it... ;)
[21:34:31] <gene_> Gotta run guys, I'm gonna let this sit and cook a while, yes, I agree on that, CEO's? Best idea I've heard yet!
[21:34:50] <SWPLinux> and mid-level management at all financial institutions
[21:34:53] <SWPLinux> might as well lower the bar a bit for them
[21:34:56] <dan1mal_garage> VT is full of phish fans
[21:34:58] <SWPLinux> or raise the bars
[21:35:07] <SWPLinux> that's because Phish was from there :)
[21:35:12] <dan1mal_garage> i know lol
[21:35:18] <gene_> So they can't climb back out, yes.
[21:35:18] <dan1mal_garage> and ben and jerry
[21:35:21] <SWPLinux> yep
[21:35:36] <SWPLinux> I used to get ice cream from Ben and Jerry, in the old gas station where they started
[21:35:37] <dan1mal_garage> and fort ticondaroga (sp)
[21:35:42] <SWPLinux> that's in New York
[21:35:52] <dan1mal_garage> isnt it on the boarder?
[21:35:58] <Jymmm> Chunky Monkey to Hemp Hippo!
[21:36:02] <SWPLinux> nope, there's a lake in between
[21:36:10] <dan1mal_garage> champlain?
[21:36:15] <SWPLinux> not very wide there, but very noticeable
[21:36:17] <SWPLinux> yes
[21:36:24] <dan1mal_garage> i forgot, it's been a long time since i've been there
[21:36:40] <dan1mal_garage> all i remember is visiting the fort when i was at lake champlain
[21:36:43] <SWPLinux> oh, and I think they just blew up that bridge, it was closed about 6 months ago because it was in bad shape
[21:36:59] <dan1mal_garage> we took a ferry i think
[21:37:10] <dan1mal_garage> but i was like 8
[21:37:11] <SWPLinux> yep, there's one of those too
[21:37:12] <SWPLinux> heh
[21:41:16] <dan1mal_garage> how fast is safe for the classicladder thread to be set at?
[21:41:28] <dan1mal_garage> it's at 50ms, that seems a bit slow
[21:41:39] <SWPLinux> servo period is pretty common, I think
[21:41:41] <SWPLinux> so 1 ms
[21:41:49] <dan1mal_garage> ok thats what i was thinking
[21:41:58] <dan1mal_garage> i cant see any reason for 50ms
[23:23:51] <jt-plasma> a buddy of mine just gave me two large boxes of terminal blocks :)
[23:23:51] <jt-plasma> fuse blocks, double level terminals, ground terminals, triple level terminals, ends, dividers, jumpers HOLY COW
[23:37:05] <andypugh> Freaky networks stuffs
[23:37:08] <kanzure__> kanzure__ is now known as kanzure
[23:37:08] <kanzure> kanzure is now known as kanzure-
[23:38:43] <Jymmm> SWPLinux: I picked up a couple LM317TZ's yesterday, gonna try and mess around with the CC if I have the right resistors =)
[23:43:50] <andypugh> I have started to always use SMT resistors on stripboard and matrix board. I am not sure why.
[23:44:04] <andypugh> But I have always had a thing about miniaturisation, I used to practice writing very, very small too.
[23:44:12] <ds3> SMT resistors are easier on those boards
[23:44:17] <andypugh> I am rather liking matrix board and Verowire with SMT components.
[23:44:58] <ds3> I do something similar and use copper tape for traces on those boards
[23:45:52] <andypugh> Verowire is insulated, so you can just go point-to-point and the little pen makes it easy to handle.
[23:46:44] <danimal_garage> ok i'm having intermittent issues with my toolchanger
[23:46:49] <danimal_garage> i had it working pretty good earlier, but every once in a while it doesnt settle properly
[23:47:13] <danimal_garage> the issue is that there's a switch that tells the control what position its in, but it goes past that switch one position before it stops
[23:47:47] <danimal_garage> so it's actually one position further than where my ladder told it to stop
[23:48:13] <danimal_garage> so when i change one position, it doesnt move because it's already in the next position
[23:48:47] <danimal_garage> so when i do t5, it's actually in t6, so when i do t6, it doesnt move
[23:51:01] <andypugh> Move the switch?
[23:52:24] <danimal_garage> that doesnt help, because it always settles one position further than the point that triggers it to stop
[23:56:50] <andypugh> That seems, errr, eccentric?
[23:56:50] <livid> livid is now known as alxm
[23:56:53] <danimal_garage> so basically if i move the switch, it's just gunna go the same difference further past the switch as before
[23:56:53] <andypugh> No way to make it 0.99 of a position?
[23:56:56] <danimal_garage> no
[23:57:07] <danimal_garage> it has to be 8 even points
[23:57:12] <andypugh> Two switches? A "you need to stop soon" and a "yes, this is it"?
[23:57:18] <danimal_garage> it's a turret, so it spins, and has 8 positions
[23:57:40] <kanzure-> kanzure- is now known as kanzure
[23:58:53] <danimal_garage> cant really do that
[23:58:56] <andypugh> How is is driven?
[23:58:56] <danimal_garage> air
[23:58:57] <andypugh> One direction only?
[23:58:57] <danimal_garage> yes
[23:58:57] <danimal_garage> cradek: what voltage are you running your selaniods at for the turret?
[23:58:58] <andypugh> So _why_ does it always go one position too far? That seems wrong.
[23:58:58] <andypugh> Can't you slow it right down? Lets be honest, if a tool change takes 10 seconds it's not the end of the world.
[23:59:48] <danimal_garage> it's air