#emc | Logs for 2006-02-26

[00:12:01] <jepler> alex_joni: maybe this? http://ipsysctl-tutorial.frozentux.net/chunkyhtml/tcpvariables.html#AEN477
[00:13:27] <alex_joni> jepler: thanks
[00:13:32] <alex_joni> I'll look further into it
[00:15:33] <cncuser> basically you can echo around a little in /proc/sys/net/ipv4/
[00:15:53] <cncuser> but it may not even be your puter but the router infront
[00:27:42] <cradek> skunkworks: have you tried ubuntu on a SATA drive?
[00:28:27] <skunkworks> no - I only have about 3 motherboards with sata controllers on them
[00:28:35] <skunkworks> they are all in service right now.
[00:28:36] <cncuser> hehe
[00:28:41] <cncuser> oh :(
[00:28:51] <skunkworks> having issues?
[00:28:54] <cncuser> looked funny for a few seconds
[00:28:55] <cradek> ok, just wondered
[00:29:00] <cradek> no, never tried it
[00:29:17] <cradek> I assume it works but don't know for sure
[00:29:25] <alex_joni> all 3 in service? makes you wonder..
[00:29:34] <cncuser> return to sender
[00:29:41] <skunkworks> ;)\
[00:29:56] <cncuser> and get money back
[00:30:40] <skunkworks> that sounded that odd? I had to read it a few times to get your meaning. ;)
[00:30:56] <cncuser> did you put them all ontop of each other in a 1HU case ?
[00:31:21] <skunkworks> just work stations
[00:31:25] <cncuser> else.... -> money back :)
[00:31:57] <cncuser> just barfing
[00:32:24] <jepler> "in service" == working
[00:32:31] <jepler> "in for service" == broken
[00:32:40] <cncuser> aahhhhh
[00:32:50] <alex_joni> "in service" = somehow working
[00:32:54] <skunkworks> right
[00:33:04] <alex_joni> or still somehow working ;)
[00:33:25] <cncuser> then you may savely erase that part of your memory that holds the information of my part of the conversation for the las 3 minutes
[00:34:05] <skunkworks> no problem :) - I on the other hand never have problems understanding others ;)
[00:34:08] <jepler> cncuser: I'm sorry about how crappy a langugae English is.
[00:34:41] <alex_joni> jepler: don't be, you weren't there when they defined the ontology
[00:34:52] <cncuser> jepler: hehe
[00:35:17] <cncuser> hehe: i put your sorry ontop of my dict.leo.org ;)
[00:36:06] <skunkworks> how is the tp going?
[00:36:25] <alex_joni> like duracell
[00:36:36] <alex_joni> it's going and going and going..
[00:36:41] <alex_joni> ROFL
[00:36:55] <skunkworks> ;)
[00:37:18] <alex_joni> don't mind me .. too tired to think straight
[00:37:34] <skunkworks> are you guys trying to fix what is there or start new? Missed that part.
[00:37:47] <alex_joni> skunkworks: both
[00:38:00] <skunkworks> that works ;)
[00:38:00] <alex_joni> in that order
[00:58:34] <cncuser> good night
[01:12:27] <alex_joni> night all
[01:13:11] <skunkworks> night
[01:17:13] <CIA-8> 03cradek * 10emc2/src/emc/kinematics/tp.c: blends were unsatisfactory with previous change, reverting part of it
[02:44:12] <tfmacz> Hello
[03:24:43] <CIA-8> 03cradek * 10emc2/src/emc/kinematics/tp.c:
[03:24:43] <CIA-8> in Nov 2000 I think wshackle meant to block blending for sharp corners.
[03:24:43] <CIA-8> Either it never worked or it got lost somewhere along the line, but it
[03:24:43] <CIA-8> seems like a good enough idea.
[03:35:41] <staggerlytom> hello all
[03:35:48] <cradek> hi again
[03:35:55] <staggerlytom> hi chris
[03:36:06] <cradek> new around here?
[03:36:18] <staggerlytom> tom powderly
[03:36:53] <cradek> ah right
[03:38:16] <cradek> I see you are in the Mazak photos
[03:38:23] <cradek> is that your machine?
[03:41:57] <staggerlytom> sorry, distracted, yes and others too
[03:42:30] <staggerlytom> the machine beongs to Roland Freistad ( -10 spelling )
[03:42:48] <cradek> how's it working by now?
[03:43:13] <jmkasunich> I don't think Roland is using it for much
[03:43:19] <staggerlytom> I dont know, I thought i'd go visit and see before the next fest
[03:43:35] <staggerlytom> i heard it was runnign and was in use
[03:43:53] <staggerlytom> there's too many info sources on these things
[03:44:22] <staggerlytom> email, 2 irc , 2 wiki.. u get conflicting info
[03:44:44] <staggerlytom> Ray knows I 'spose
[03:45:49] <cradek> I didn't see ray's plan for this year yet - I wonder what he's come up with
[03:46:07] <staggerlytom> I was hopin for some input on that too
[03:46:30] <cradek> afraid you missed him
[03:46:36] <staggerlytom> busy guy , you devs have a full plate
[03:48:07] <staggerlytom> about fragmented info... is there a single place to look at the 'weave' solution?
[03:48:26] <staggerlytom> how the sinewave was given to the TP?
[03:48:28] <cradek> all I saw was a post to the list with a screen grab
[03:49:25] <staggerlytom> was it Graham Stabler of the UK ( univ Nottingham or Essex )?
[03:49:33] <cradek> let me see if I can find it
[03:49:38] <staggerlytom> thanks
[04:01:57] <staggerlytom> in Chicago, J&L Supply is selling RongFu table top mills at 700$. (regular 1460.67)
[04:05:05] <jmkasunich> mill-drills, right? (no knee)
[04:05:23] <jmkasunich> is the head on ways or a round column?
[04:06:21] <staggerlytom> round column but 4.5" dia
[04:06:53] <jmkasunich> round means you can't raise or lower without losing position
[04:06:54] <staggerlytom> and dove tail ways but much wider than most
[04:07:08] <staggerlytom> yeah for 700$ tho
[04:08:05] <staggerlytom> maybe you can leave it at one height and use parallels on table
[04:08:36] <staggerlytom> (btw i'm not affiliated)
[04:09:12] <jmkasunich> I already have a mill that needs that - a Shoptask
[04:09:26] <jmkasunich> I also have a knee mill (Van Norman #12)
[04:10:07] <staggerlytom> me too(shoptask), just fingered out the slop on the Z drive worm
[04:10:33] <jmkasunich> pathetic isn't it
[04:10:41] <jmkasunich> I completely replaced that with a ballscrew
[04:10:58] <staggerlytom> all the small mills have dove tail ways, we need one with linear rails
[04:11:15] <staggerlytom> ball screw? outrigger like?
[04:11:31] <jmkasunich> yeah, made a new, stiffer banjo
[04:11:59] <jmkasunich> block of steel where the plate that is supposed to hold the Z motor bolts on, with a pair of bearings and a rotating nut
[04:12:10] <jmkasunich> the screw is attached to the banjo sticking up into the nut
[04:12:56] <staggerlytom> so the z motor axis is vertical with this new mount?
[04:12:59] <jmkasunich> yes
[04:13:13] <jmkasunich> it is outboard of the screw, drives the nut thru a toothed belt
[04:13:36] <staggerlytom> any gas shock or spring counterbalance?
[04:13:45] <jmkasunich> no, the quill isn't that heavy
[04:13:59] <staggerlytom> original quill?
[04:14:03] <jmkasunich> yes
[04:14:21] <staggerlytom> you're the VFD expert, what did you do for the drive?
[04:14:35] <jmkasunich> the spindle drive? that is unchanged so far
[04:14:52] <jmkasunich> Z axis is just a NEMA 34 stepper, gonna use a Gecko
[04:15:16] <staggerlytom> lemme get an url for you, a nice spindle, not too expensive
[04:15:42] <jmkasunich> I do want to replace the spindle motor with three phase and use a vfg, but that is just one item on a very long to-do list
[04:17:01] <staggerlytom> www.5bears.com, this guy hand builkt a nice linear rail mini mill with www.kavo-drives.com spindle
[04:17:25] <staggerlytom> yep, no typos for once
[04:17:36] <staggerlytom> ok, built
[04:17:48] <jmkasunich> they look very fast, and not cheap
[04:18:47] <staggerlytom> 400$ (he found) for 25Krpm is cheap to me, but not for a lot of kinds of milling
[04:19:12] <jmkasunich> at this point I don't need those speeds
[04:19:51] <jmkasunich> $400 gets you drive, motor, and the spindle itself?
[04:20:09] <staggerlytom> noooo just spindle ( used ebay he said )
[04:20:23] <jmkasunich> oh
[04:20:39] <jmkasunich> so normal prices are probably as expensive as they look
[04:20:56] <jmkasunich> multi-kilobucks
[04:22:30] <staggerlytom> I looked for prices, didnt see
[04:23:30] <jmkasunich> I didn't need to look to know it was too rich for me
[04:24:08] <staggerlytom> but this dove tail slide stuff, how are people getting speed & accuracy with such?
[04:24:42] <jmkasunich> worked for bridgeports for 80 years or so
[04:24:45] <jmkasunich> works for me
[04:25:19] <jmkasunich> all depends on your needs vs. your wallet
[04:25:23] <staggerlytom> hand cranked, not the 60 ipm everybody talks of with emc, nor the resolution I hear people talking
[04:25:46] <jmkasunich> bridgeport makes cnc machines
[04:25:58] <jmkasunich> just gotta be more carefull about lube
[04:26:28] <staggerlytom> the lube can change the position by as much as 1/2 thou
[04:27:04] <jmkasunich> and there are a _lot_ of folks who can live with +/-0.001 or even 0.005 in most of what they do
[04:27:17] <jmkasunich> not everybody is making jet engine parts
[04:27:57] <staggerlytom> true, my background was mold making
[05:27:04] <jmkasunich> jmkasunich is now known as jmk_sleep
[05:30:30] <asdfqwega> Hm...am I the only one awake at this hour?
[05:34:21] <asdfqwega> Dangit, and I was hoping to tell someone about my gcode -> DXF converter
[10:45:04] <_bart> Is there an emc2 deb for BDI-4 yet ?
[10:45:37] <fenn> maybe
[10:45:55] <fenn> its pretty close.. i've been kinda out of the loop tho
[10:45:59] <_bart> where ?
[10:46:44] <_bart> same here - Last I heard, emc2 was broke badly
[10:47:05] <fenn> they started a branch called "testing" which is more stable than head
[10:47:24] <fenn> its a cvs tag but should be made into a release
[10:48:59] <_bart> released as a usable deb ?
[10:49:11] <fenn> no i dont think its released yet
[10:49:35] <fenn> i see scripts for building a deb
[10:50:05] <_bart> yes... and they are shite
[10:50:22] <fenn> i dunno anything about making debs
[10:50:38] <_bart> I know a little.
[10:51:16] <fenn> you can use chkinstall to make a temporary deb
[10:51:34] <fenn> works for my purposes
[10:51:49] <fenn> er checkinstall
[10:52:32] <_bart> I don't want to risk breaking my BDI-4 box.
[10:53:03] <fenn> why cant you run emc2 in place?
[10:53:54] <_bart> I heard from cradek, I think, that make installs were not recommended.
[10:54:38] <fenn> yeah i mean, just "make" not "make install"
[10:56:30] <_bart> You can run emc2 after doing a make without the install ?
[10:58:43] <fenn> yep just type scripts/emc
[12:39:34] <cncuser> good morning
[13:04:10] <alex_joni> hello
[13:04:22] <cncuser> hi alex_joni :)
[15:35:31] <_bart> Now I see why emc2 is not recommended on a bdi install
[15:35:49] <_bart> got a "kernel oops"
[15:37:51] <rayh> Hi _bart Who recommended not running it?
[15:38:04] <_bart> cradek
[15:38:18] <rayh> 4.30 or later?
[15:38:51] <_bart> 4.38 with 2.6.15 and rtai 3.3
[15:39:19] <rayh> Ah. I've not tried it there but it compiles and runs well with 4.30.
[15:39:45] <_bart> man and the config is a pita
[15:39:48] <rayh> What was the oops message?
[15:39:58] <skunkworks> i ran it on 4.30 for a while - no issues - the wiki install instruction are right on.
[15:40:16] <_bart> looks like a null pointer
[15:40:31] <rayh> emc2 config or what?
[15:41:04] <rayh> Hi Dan. Morning to you.
[15:41:14] <dan_falck> hi ray
[15:42:17] <skunkworks> ray - how is the tp looking?
[15:42:44] <rayh> I'm really impressed with what they'd done yesterday.
[15:44:09] <skunkworks> Good - I probably wont be able to play with it for a week or more. sounds like they are wraping thier heads around it well.
[15:44:47] <rayh> The blends during a circle showed very small squiggles in accel.
[15:45:10] <rayh> Not perfect but a lot closer than we were last week.
[15:45:29] <skunkworks> :)
[15:45:55] <rayh> I ran it with a 2 hour contouring program while watching accel.
[15:46:10] <skunkworks> wow
[15:46:20] <rayh> running about 50 ipm.
[15:46:24] <steves_logging> steves_logging is now known as steve_stallings
[15:46:34] <rayh> Hi steve.
[15:46:55] <skunkworks> I need to play with halscope. Sound and looks really neat.
[15:47:08] <rayh> Awesome
[15:48:00] <steve_stallings> Hi Ray, everybody
[15:48:02] <skunkworks> we just need to get les to test a little ;)
[15:48:04] <rayh> Takes more brain than I've got to get it to work right but with a little help from my friends I can get some good traces.
[15:49:37] <steve_stallings> I responded to your dev list e-mail on blending and thought I should check in here to see if my name was going to get bloodied. 8-)
[15:49:39] <rayh> The guys are talking about building sets of standard configurations with a menu to select.
[15:56:28] <skunkworks> Hi steve - I am sam. I have only been playing with emc2 and linux for a few months (testing). It ha been fun. started with bdi4.30+emc2 and now cradeks ubuntu setup.
[15:56:50] <rayh> Nice job Steve.
[15:57:41] <rayh> phone
[16:08:10] <dmessier> good morming all
[16:09:56] <alex_joni> bon jour
[16:10:38] <dmessier> ca va??
[16:11:36] <alex_joni> qui, mersi
[16:11:44] <dmessier> ma chienne et en challeur... nous avons des chiens nouveaux dans la coure
[16:11:44] <alex_joni> comme d'habitude
[16:12:06] <rayh> Where is i18 when you really need it?
[16:12:07] <alex_joni> what kind?
[16:12:19] <alex_joni> rayh: new dogs in the yard
[16:12:20] <dmessier> brnese mountain
[16:13:07] <rayh> gnau
[16:13:09] <dmessier> bernese mountain dog.. first heat..
[16:13:53] <alex_joni> dmessier: nice
[16:13:59] <alex_joni> are those brown?
[16:14:10] <dmessier> she's actin retarded.. the pug is trying to crawl inside her... there are new boy dogs circling the perimiter
[16:14:20] <dmessier> tri-colour
[16:14:25] <alex_joni> http://www.berner.org/centerpic/rotator.php
[16:14:53] <alex_joni> dmessier: very cute :P
[16:15:09] <dmessier> that looks like my Jewel... ; )
[16:15:33] <giacus> hello
[16:16:09] <alex_joni> hi jacky
[16:16:10] <dmessier> the wife says ours looks cuter..LOL
[16:17:07] <dmessier> almost a yr old about 90 lbs
[16:17:48] <alex_joni> dmessier: then you're set ;)
[16:17:56] <alex_joni> I've seen it gets bigger though
[16:18:50] <dmessier> yes there a male around the corner over 145 lbs
[16:19:20] <dmessier> we may breed her .. on her next heat... we'll see...
[16:19:43] <alex_joni> you need to let her chose ;)
[16:20:22] <dmessier> i know.. they are super sensitive..even to loud scoldings
[16:30:06] <giacus> where can I find latest emc.ini ?
[16:30:33] <alex_joni> giacus: which one?
[16:30:40] <giacus> emc2
[16:30:46] <alex_joni> in CVS?
[16:30:53] <giacus> yeah
[16:31:04] <alex_joni> but there is no emc.ini (there is sim.ini, stepper_inch.ini, stepper_mm.ini)
[16:31:20] <alex_joni> and hardware specific ini's (stg, motenc, univpwm, univstep,etc)
[16:32:37] <giacus> how is possible, time ago I deleted som log file from /var/log, now the ini file in ~emc2/configs/ is blank
[16:32:50] <giacus> it exist, but blank
[16:33:01] <alex_joni> there is no ini in emc2/configs/
[16:33:13] <alex_joni> all the inis are in emc2/configs/<subdir>/
[16:33:17] <giacus> how you start it ?
[16:33:30] <alex_joni> where subdir is sim, stepper, stg, m5i20, etc
[16:33:42] <alex_joni> how did you configure?
[16:33:50] <alex_joni> local (run-in-place) or install ?
[16:34:03] <giacus> I was runnig it from ~emc2 dir
[16:34:19] <giacus> using ./scripts/emc.run
[16:34:27] <alex_joni> ok, now it's scripts/emc
[16:34:39] <alex_joni> but you need to ./configure --enable-run-in-place
[16:34:59] <alex_joni> by default ./configure prepares the software for installing
[16:35:09] <alex_joni> ./configure && make && sudo make install
[16:35:17] <alex_joni> then you run it as: 'emc'
[16:35:52] <giacus> uhm, ok, going to check from there
[16:36:40] <Jack_emc2> ok
[16:36:54] <alex_joni> Jack_emc2: my advice.. use a new folder, get the latest TESTING
[16:37:06] <alex_joni> and go the full way from scrap (lots have changed recently)
[16:37:15] <Jack_emc2> kernel is, is that ok ?
[16:37:26] <alex_joni> yup
[16:37:38] <alex_joni> what system is that? your own kernel?
[16:37:48] <alex_joni> sounds like a BDI kernel though
[16:37:53] <Jack_emc2> this is BDI updated by apt
[16:37:58] <Jack_emc2> yeah
[16:37:59] <alex_joni> ok
[16:38:21] <alex_joni> then it's best if you do './configure --enable-run-in-place'
[16:38:31] <alex_joni> then 'make && sudo make setuid'
[16:38:42] <Jack_emc2> ok.. lets try
[16:38:43] <alex_joni> then 'scripts/emc' to run it
[16:38:47] <jmk_sleep> jmk_sleep is now known as jmkasunich
[16:52:25] <giacus> alex_joni: is this correct ? cvs -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/emc login
[16:52:47] <giacus> to get emc2 Testing ?
[16:53:02] <alex_joni> yes
[16:53:05] <jmkasunich> that is the first line, it simply logs you into the cvs server
[16:53:07] <giacus> can't login
[16:53:20] <jmkasunich> what error do you get?
[16:53:33] <giacus> nothing, return to prompt $
[16:53:39] <alex_joni> giacus: that is OK
[16:53:41] <jmkasunich> then it worked ;-)
[16:53:41] <giacus> as passwd I used enter
[16:53:47] <giacus> work ?
[16:53:50] <giacus> :(
[16:53:51] <alex_joni> yes, that's how it's supposed to work
[16:53:58] <giacus> K
[16:53:59] <alex_joni> it writes a .cvspass file
[16:54:13] <alex_joni> you need to use the next line to actually check out emc2 (testing)
[16:54:14] <jmkasunich> try the next line
[16:54:26] <alex_joni> co -rTESTING emc2
[16:54:28] <giacus> cvs -z3 -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/emc co emc2
[16:54:37] <alex_joni> put the -rTESTING in there
[16:54:46] <giacus> cvs -z3 -d:pserver:anonymous@cvs.sourceforge.net:/cvsroot/emc co -r TESTING -d emc2testing emc2
[16:54:56] <giacus> I read at cncuser log :)
[16:54:59] <alex_joni> perfect
[16:54:59] <jmkasunich> yes, that will put it in the subsdir emc2testing
[17:00:11] <fenn> * fenn nudges Bo-Dick
[17:00:17] <fenn> go on, say something
[17:00:29] <alex_joni> * alex_joni holds his breath
[17:00:33] <Bo-Dick> :D *lol*
[17:00:46] <alex_joni> * alex_joni breathes again
[17:01:07] <Bo-Dick> I'd like to see some image galleries.
[17:01:31] <fenn> Bo-Dick is making a switching power supply to run some steppers, or something along those lines
[17:01:36] <alex_joni> what kind if images?
[17:01:43] <giacus> alex_joni: is enough to move old ~ emc2 dir to emc2_old ?
[17:01:53] <fenn> Bo-Dick: check out the "usage map" in the topic
[17:01:57] <alex_joni> yup, but it won't work without recompiling
[17:02:06] <alex_joni> so I suggest make clean first, then move
[17:02:11] <giacus> ok
[17:02:14] <alex_joni> after you move, configure && make again
[17:02:37] <giacus> ty
[17:04:08] <Bo-Dick> Well I dont have access to proffessional equipment. I'm even compromising to hold the budget low. http://www.carmi.se/misterstarshine/Projects/images.htm
[17:04:37] <Bo-Dick> Can it be more handmade than this?
[17:05:36] <alex_joni> oh my.. don't think so ;)
[17:05:56] <dmessier> oui muchacho...
[17:06:07] <fenn> thats not french
[17:06:17] <dmessier> ci
[17:07:51] <dmessier> does it work???
[17:09:01] <Bo-Dick> Well...
[17:09:50] <dmessier> does it work well???
[17:10:11] <alex_joni> I would be thrilled if it worked..
[17:10:19] <alex_joni> wouldn't dream about 'well' ;)
[17:10:43] <Bo-Dick> Its not assembled yet but I've done a successful testrun. In fact there must be a another card not shown in the gallery between the control part and the driver part.
[17:11:35] <Bo-Dick> The damned common emitter darlington pairs must have full voltage to open and close. This cant be delivered by the TTL control part.
[17:12:01] <Bo-Dick> So I have a card in between to raise the voltage with open collector buffers.
[17:12:20] <Bo-Dick> However that part didn't look so good. I'm gonna scan it anyhow.
[17:13:03] <Bo-Dick> Thats about how my controller is gonna work.
[17:13:39] <fenn> Bo-Dick: i dont get the "stimulus" pin.. most software outputs step and direction signals
[17:14:04] <Bo-Dick> Its present in the image.
[17:15:53] <dave-e> hi ray
[17:16:34] <rayh> Hi Dave.
[17:16:43] <dave-e> ah you are awake!
[17:17:07] <skunkworks> cvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv
[17:17:27] <Bo-Dick> Observe that the heatsinks are custom made.
[17:17:27] <dave-e> I liked the comments on tp
[17:17:48] <skunkworks> sorry - cat walked across the keyboard.
[17:18:03] <dave-e> get a catapult
[17:18:30] <skunkworks> ;)
[17:19:27] <dave-e> did I miss something....why would we change modes for corners at ~ 90?
[17:20:51] <dave-e> hmmmm dead silence
[17:20:52] <rayh> Thanks. Steve's post went a long way to clarifying the potential.
[17:21:01] <dave-e> yes it did.
[17:21:25] <fenn> dave-e: if i had to guess, i'd say it's so you don't blend the corner out of existence
[17:22:11] <dave-e> well, with a given tolerance....one decels to make the corner with the allowed tolerance......re Sonja McFarlane
[17:22:42] <fenn> wouldn't that be nice
[17:22:44] <dave-e> at the accel of the machine
[17:24:04] <dave-e> those of us with 20 in/sec^2 won't do corners like LW's router which should do .5-1 g
[17:26:49] <dave-e> on a related issue.....someone has done a cam package which calculates the immersion angle for the cutter and keeps that constant...of course this won't
[17:27:06] <skunkworks> do some controllers have a tolerance setting? where you can say I want the tool path to be within X thousands
[17:27:06] <dave-e> work for finish cuts but certainly will for roughing
[17:27:18] <fenn> you mean the ramp in angle? what's an immersion angle?
[17:27:19] <skunkworks> it adjusts the speed to fit
[17:27:39] <dave-e> the angle of the cutter that is in the metal.
[17:27:48] <fenn> skunkworks: it's part of the rs274 stanard but emc doesn't implement it
[17:28:15] <anonimasu> hm, that's too bad
[17:28:55] <skunkworks> ok - that would give the end user the control to say what they want.
[17:29:04] <Jack_emc2> alex_joni: emc2 succesful compiled, it work
[17:29:25] <anonimasu> dave-e: I know mastercam will calculate the immersion angle..
[17:29:32] <alex_joni> Jack_emc2: nice
[17:29:39] <anonimasu> hm, dinner..
[17:29:46] <fenn> skunkworks: yep and requires plenty of thought to even figure out how to put it into code
[17:30:25] <skunkworks> fenn: - right - I can't even imagine. (well maybe a little bit)
[17:31:33] <fenn> skunkworks: for example how much headroom do you give yourself for feed override? or do you just blend at the maximum velocity of the machine?
[17:32:33] <Bo-Dick> My project is progressing very slowly right now. The thing is that I've realized how incredibly important it is to protect my circuits. Thats why I would like a overvoltage protection circuit.
[17:32:35] <dmessier> Skunk... Toshiba 888 control has such athing.. called SHAPE
[17:33:08] <fenn> Bo-Dick: protecting your life is more important than your chips
[17:33:16] <skunkworks> I probably think about it too easy then. I would think that if you gave the max feed you wanted to run and how accurate you want the tool to follow - any overide would run the tool path faster - if it could.
[17:33:17] <dmessier> it also has constant surface speed when using a ballnose
[17:33:43] <skunkworks> dmessier: - nice.
[17:33:45] <dave-e> but that implies real-time tp
[17:33:46] <dmessier> yes.. exactly how it works..
[17:34:15] <dmessier> its look ahead calculates angular deviations..
[17:34:34] <dmessier> and maps against machine kinematics
[17:35:00] <fenn> skunkworks: yes you could do tp in realtime for simple stuff.. sorry i usually think about wild edge cases
[17:35:20] <dmessier> this works on the wildest edges...
[17:35:47] <fenn> hmmm
[17:35:51] <dmessier> i tested worse case stuff for the mold & die industry
[17:36:15] <dmessier> should i even mention square corner machining... ; )
[17:36:52] <dmessier> or orbital boring
[17:37:21] <dmessier> there's a 888.2 out now...
[17:38:39] <dmessier> i had a small high ticket item growing roots in the show room...had it all and 120 k rpm to boot 8000 rpm minimum
[17:38:56] <fenn> ooh orbital boring.. i knew i didnt make that phrase up
[17:39:01] <dmessier> was hard to tool up..
[17:39:16] <dmessier> Fenn what does yours entrail???
[17:39:26] <fenn> for cutting splines and gears and such on a lathe
[17:39:34] <fenn> with a single point tool in the toolpost
[17:39:47] <fenn> move X back and forth really fast
[17:40:02] <fenn> well, maybe not gears
[17:40:06] <dmessier> we use a horizontal machining center and turn it into a lathe..
[17:40:57] <fenn> like a millturn?
[17:41:24] <dmessier> its controled spindle orientation mapped to an xy plan orbit around a stationary part...using z to in feed..
[17:41:43] <dmessier> single point tooling
[17:42:11] <fenn> uh, X _and_ Y?
[17:42:18] <dmessier> radius of gyration is variable so tapers and thrs are possible..
[17:42:29] <fenn> what do you make with it?
[17:42:50] <Bo-Dick> How can I test this protective device? http://www.carmi.se/misterstarshine/ov_prot.gif
[17:42:59] <dmessier> turn journals on axles... bottle boring
[17:43:17] <Bo-Dick> Of course.... http://www.carmi.se/misterstarshine/Projects/ov_prot.gif
[17:44:00] <fenn> i dont know what bottle boring is either
[17:44:24] <dmessier> into a small hole to open it up like a bottle..
[17:44:24] <fenn> i guess its just making a cavity like inside a vase?
[17:45:01] <Bo-Dick> * Bo-Dick has been puzzled by this for a while now
[17:45:16] <dmessier> imagine carving out the inside of a beer bottle
[17:47:00] <dmessier> with out spinning the part... contron the toolpoin position around an axis.. then add an x-y circle to the dia. you wanna bore to... then move in z..
[17:47:45] <dmessier> rads and tapers are a breeze this way..
[17:48:37] <fenn> Bo-Dick: what's wrong with plugging it into the wall?
[17:48:59] <fenn> Bo-Dick: or smash up a disposable camera and zap it
[17:49:37] <jepler> Bo-Dick: what's that supposed to do? Short the + and 0 rails together when the voltage is above about 4*13V?
[17:50:08] <Bo-Dick> Thats the idea yes.
[17:50:32] <jepler> won't that just melt the SCR?
[17:50:48] <dave-e> so do you put a fuse in the line or just a big resistor?
[17:51:32] <Bo-Dick> I will have a fuse
[17:51:50] <jepler> you depend on the SCR to survive long enough to blow the fuse, then
[17:52:05] <dmessier> depends on whether you want to shut her down or melt her down
[17:52:06] <dave-e> that is within the I^2R rating of the scr
[17:53:06] <dave-e> I once had a system with a magnetic breaker and a fuse .... and the breaker always beat the fuse out.
[17:53:44] <rayh> Sure. Breakers are the weakest link in all but very specialized fuses.
[17:53:54] <dmessier> it was in front of the fuse.. NOT behind
[17:54:13] <rayh> Doesn't matter they were in series.
[17:54:38] <dmessier> all right
[17:54:42] <rayh> Uh I should have asked what was in front of the fuse?
[17:54:56] <rayh> Not the crowbar circuit, I hope.
[17:55:11] <dmessier> basically yes..
[17:55:30] <Bo-Dick> Well the crowbar is supposed to blow the fuse
[17:55:39] <rayh> Yep.
[17:56:03] <rayh> Hopefully before the traces leap off the board.
[17:56:47] <dave-e> this was 12 ga wire....and only a 20 A fuse ... pretty safe
[17:58:16] <dmessier> we hit 1 machine with the pressure washe and it DID NOT LIKE WAking up... we had boards with solder balls on the bottom...we had them retraced and installed fine.. once the machine had dried
[17:59:11] <dmessier> i opened a tacho and cot a cup of water... new kid did do a thorough wash... ; (
[17:59:13] <dave-e> hi steve
[17:59:19] <SWPadnos> hi there
[17:59:23] <rayh> I've got some old Allen Bradley power supplies in the shop 100a +5 volt.
[17:59:33] <rayh> Huge traces and crowbar.
[17:59:49] <Bo-Dick> Got any images?
[18:00:09] <rayh> No. Long old stuff.
[18:00:58] <dmessier> 100a 5v.. is a nice jolt.. i remember those..
[18:01:10] <rayh> The fuse was on the 300 volt primary side of the transformer.
[18:02:01] <Bo-Dick> So it only needed 2 amps fuse on the primary?
[18:02:17] <dmessier> were those the $$$ ceramic fuses?? we had them on almost all the german gear
[18:02:38] <rayh> Right. You should see the mess when the guy stuck a 30 amp automotive fuse in there.
[18:02:52] <dmessier> LOL....
[18:03:09] <rayh> Burned clear through the epoxy board.
[18:03:12] <Bo-Dick> That wasn't a very good idea :)
[18:03:44] <dmessier> let the smoke out BABY
[18:03:51] <alex_joni> any transistor works great asa a fuse
[18:03:56] <rayh> Did it ever.
[18:04:05] <rayh> Gotta run. Nice talking guys.
[18:04:13] <fenn> SWPadnos: would you like to give a lecture on why switching PS are a bad idea for driving motors? i wouldnt mind listening to it again
[18:04:14] <dmessier> l8r ray.. ;)
[18:04:29] <SWPadnos> err... not really ;)
[18:04:41] <alex_joni> fenn: he's busy :P
[18:04:53] <jmkasunich> fuses, who needs em
[18:05:09] <jmkasunich> I'm happy, I just fixed a 18" LCD panel
[18:05:16] <alex_joni> nice.. the backlight stuff?
[18:05:28] <jmkasunich> replaced 4 PCB mounted fuses with pieces of wire, and now it works!
[18:05:29] <dmessier> hard wire it perfectl the first time...
[18:05:54] <alex_joni> jmkasunich: lol ;)
[18:06:15] <jmkasunich> I wonder for how long, before the backlight inverter burn up?
[18:06:56] <dmessier> time it...
[18:07:01] <fenn> when that happens, you have a nice DIY projection screen panel
[18:07:08] <jmkasunich> 5 mins so far
[18:07:14] <jmkasunich> I had that before
[18:07:24] <jmkasunich> I actually have 4 industrial LCDs here
[18:07:38] <jmkasunich> 2 18", with the same symptoms (maybe the same fix will work on the other one)
[18:07:54] <jmkasunich> 2 15", one seems completely dead, the other has smearing
[18:07:54] <alex_joni> jmkasunich: at leas did you use thin wires?
[18:08:06] <jmkasunich> yes, one strand from a small stranded wire
[18:08:19] <alex_joni> then you should be set
[18:08:30] <jmkasunich> about 30AWG
[18:08:53] <dmessier> reminds me of Vietnam...LOL
[18:09:25] <jmkasunich> the fuses were 1.25A, SMT things a little smaller than a 1206 resistor
[18:09:55] <cradek> maybe you should measure the current in those places
[18:10:21] <cradek> if it's below 1.25 in at least steady-state, you wouldn't have to worry
[18:10:39] <giacus> hey guys, any issue using mm as units with latest emc2 vers. ?
[18:10:54] <cradek> giacus: both units work fine
[18:10:55] <alex_joni> giacus: not that we know of
[18:11:03] <giacus> ok. thanks
[18:11:06] <alex_joni> giacus: use stepper_mm.ini as a starting point
[18:11:11] <giacus> yes
[18:13:29] <jmkasunich> and the answer is: ....
[18:13:31] <jmkasunich> 11 minutes
[18:13:44] <jmkasunich> "click", and squealing noise from the main power supply
[18:13:53] <jmkasunich> :-(
[18:13:59] <alex_joni> ouch
[18:15:04] <dmessier> any smoke yet??
[18:15:25] <dmessier> is the main on a power bar??
[18:15:42] <jmkasunich> no smoke, it just stopped working
[18:15:54] <fenn> jmkasunich: wanna sell me a lcd panel?
[18:15:54] <jmkasunich> the AC input goes to a switcher that makes 12V
[18:16:05] <jmkasunich> everything else runs from the 12
[18:16:32] <jmkasunich> the squeal is the main switcher current limiting, it's OK, but I bet the backlight inverter is toasted
[18:16:53] <dmessier> 12 v @ 1.25A ..still gets HOT
[18:16:55] <jmkasunich> fenn: if you don't mind one that doesn't light up....
[18:17:25] <jmkasunich> general warmness on the inverter output transformers
[18:17:32] <jmkasunich> nothing specifically hot
[18:18:40] <dmessier> this place has the highest percentage of DR's of detruction...
[18:18:51] <dmessier> why so??
[18:19:45] <fenn> cant break an omelet without making a few eggs
[18:20:28] <dmessier> i like that one Fenn
[18:20:48] <fenn> i actually typed it before i noticed
[18:23:39] <dmessier> ;)
[18:27:32] <anonimasu> :)
[18:33:49] <Bo-Dick> Is this a good construction? http://www.carmi.se/misterstarshine/Projects/fuse.gif
[18:34:44] <jmkasunich> depends on the rating of the transformer and rectifier
[18:35:06] <jmkasunich> 3A * 230V = 690VA
[18:35:32] <jmkasunich> if its a 500VA xfmr, good
[18:35:47] <jmkasunich> if its 100va, the transformer isn' protected very well
[18:36:03] <jmkasunich> if its 1000VA, the fuse will blow prematurely
[18:36:27] <Bo-Dick> The most intresting is if the overvoltage protection part really protects the load...
[18:37:11] <Bo-Dick> Each zener has a 13v breakdown voltage
[18:38:18] <jmkasunich> 35*1.414 = 49.5V
[18:38:23] <jmkasunich> 4 * 13V = 52V
[18:38:36] <jmkasunich> not much margin against false trip on high line or transient
[18:38:46] <jmkasunich> what is the max safe voltage for the load?
[18:39:27] <Bo-Dick> In fact I've written the label wrongly. Its only 24 vdc from the transformer.
[18:39:52] <jmkasunich> so about 17V AC?
[18:40:19] <Bo-Dick> Sorry again... uuh 24 vac of course!
[18:40:50] <Bo-Dick> And slightly under 35 vdc from after the smoothing cap.
[18:40:52] <jmkasunich> so about 34VDC
[18:40:58] <Bo-Dick> Yeah!
[18:41:28] <Bo-Dick> The technique of connect zeners in serial like this was my own idea.
[18:41:45] <Bo-Dick> I haven't been able to test this concept in real life though.
[18:42:18] <jmkasunich> you'll need to use a beefy SCR
[18:42:56] <Bo-Dick> Do you think this one is hardcore with a heatsink? http://www.carmi.se/misterstarshine/Projects/device.jpg
[18:43:10] <Bo-Dick> TO-66 package...
[18:43:32] <jmkasunich> no clue, pictures don't tell the tale, datasheets do
[18:43:49] <jmkasunich> when that thing fires it will discharge your main filter/smoothing cap
[18:44:08] <Bo-Dick> I haven't been able to find any sheets for this one.
[18:44:09] <jmkasunich> possibly a couple hundred amps peak, depending on the size of the cap and wire resistance/inductance
[18:44:56] <Bo-Dick> Hmmm. Do you think it would be better to have a fuse just before the overvoltage protection circuit?
[18:45:18] <jmkasunich> that is a good idea
[18:45:53] <Bo-Dick> What I was thinking was that the primary fuse would blow in a crowbar situation but I'm probably mistaken.
[18:45:59] <jmkasunich> the fuse can be sized closer to the load that way - it doesn't see the inrush of the transformer, or the charging of the smoothing cap when you turn it on
[18:46:33] <jmkasunich> the primary fuse would blow eventually (you hope), but it would let thru a lot more energy than a properly sized fuse later
[18:46:42] <Bo-Dick> The problem is that i'm gonna need a 20 amp fuse. Where can one get these?
[18:47:12] <jmkasunich> shouldn't be too hard
[18:47:14] <jmkasunich> where are you?
[18:47:32] <Bo-Dick> It must be a swift one also.
[18:47:44] <Bo-Dick> So that the scr wont die
[18:49:23] <Imperator_> Hi all
[18:49:24] <Bo-Dick> Does anyone recognize this logo? http://www.carmi.se/misterstarshine/Projects/device.jpg
[18:50:00] <Imperator_> what about a EMC Fest this year ???
[18:50:01] <jmkasunich> looks vaguely familiar, but I can't place it
[18:50:18] <Bo-Dick> It isn't samsung is it?
[18:50:53] <Bo-Dick> Not even www.alldatasheet.com has a match for this one. Grrr.
[18:50:57] <jmkasunich> don't think so
[18:51:03] <jmkasunich> that part is very old
[18:51:16] <Bo-Dick> * Bo-Dick is devastated
[18:52:01] <jmkasunich> Bo-Dick: where are you? in the USA?
[18:52:09] <Bo-Dick> No in Sweden
[18:52:13] <SWPadnos> .se = non USA ;)
[18:52:25] <Bo-Dick> Its Sweden
[18:52:35] <jmkasunich> oh... so finding a suitable fuse on the digi-key catalog won't be much help
[18:53:00] <Bo-Dick> What time is it in the U.S now?
[18:53:15] <jmkasunich> 1:52 pm
[18:53:31] <Imperator_> www.rs-components.com I think they are all over the world
[18:53:32] <Bo-Dick> Here it is around 8:00 pm
[18:53:33] <SWPadnos> depends on where you are, but the east coast is 1:53 PM
[18:56:45] <Bo-Dick> How can I search for logotypes
[18:57:21] <Bo-Dick> I'd like to know the manufacturer for the device
[18:58:50] <jmkasunich> I wonder if it is Harris Semiconductor?
[18:59:01] <jmkasunich> their logo is differnet now
[18:59:09] <jmkasunich> but that could be an H and S
[18:59:21] <jmkasunich> and I think they made SCRs long ago
[18:59:35] <jmkasunich> that part is truly ancient, I haven't seen a TO-66 in ages
[19:00:16] <Bo-Dick> Are they considered obsolete?
[19:01:20] <jmkasunich> yes
[19:01:45] <jmkasunich> TO-66 was replaced by TO-220, and TO-3 was replaced by TO-247
[19:01:55] <jmkasunich> plastic packeges with metal tabs
[19:06:05] <Bo-Dick> In fact TO-3 and TO-66 was somewhat more difficult to mount and definitly more difficult to connect.
[19:19:11] <Bo-Dick> It still puzzles me how to test an overvoltage protection circuit.
[19:19:27] <Bo-Dick> I would have to construct some kind of test bench for it.
[19:20:23] <Bo-Dick> First normal voltage... ...then a sudden attack of mains voltage... ...fuse blows... ...circuits damaged or not?
[19:21:06] <Bo-Dick> OR just I just trust this device and assume it'll work when the time eventually comes?
[19:23:21] <Bo-Dick> Do scr's usually have a high voltage rating?
[19:24:04] <jmkasunich> varies, from as low as 50V to as high as 2000
[19:24:40] <Bo-Dick> I guess it also depends on the type of packages they come in?
[19:28:58] <Bo-Dick> Wonder where one could get 20 amp fuses. The largest I've got is 8 amps I think.
[19:30:28] <SWPadnos> a local electrical supply store should have (or be able to get) fuses
[19:32:08] <jmkasunich> damn, seems a terrible shame to throw this panel away
[19:32:28] <Bo-Dick> I use to salvage fuses from scrapped stuff but I've never found a fuse over 4 amps.
[19:33:34] <fenn> jmk i'll try to make use of the busted one too
[19:35:14] <jmkasunich> my problem is storage space
[19:36:05] <fenn> i could pay for shipping if you wanted
[19:36:37] <jmkasunich> lemme see how big the pile is when I'm done, and get back to you
[19:38:37] <jmkasunich> I've already removed the one with the blown 12V supply from its chassis, the panel itself is intact
[19:38:49] <jmkasunich> (needs backlight inverter and/or bulbs)
[19:38:56] <jmkasunich> the other one I will leave assembles
[19:39:00] <jmkasunich> the other one I will leave assembled
[19:39:11] <jmkasunich> this one would also need a +12 supply
[19:39:55] <jmkasunich> the panel is a NEC 128102AC28-01A, for whatever that is worth
[19:53:35] <giacus> I'm using BASE_PERIOD = 0.000025 with a celeron 800 mhz
[19:53:44] <giacus> what you think abut that value ?
[19:53:55] <giacus> it seems work good
[19:55:30] <giacus> INPUT_SCALE = -400 0
[19:55:36] <giacus> OUTPUT_SCALE = -400
[19:55:43] <giacus> it seem wok ok too
[19:55:48] <giacus> work*
[19:55:55] <steve_stallings> steve_stallings is now known as steves_logging
[19:56:03] <giacus> same value for input - output scale
[19:58:17] <dave-e> if you want to spend a few minutes (hours) searching for a logo http://www.uspto.gov/
[20:00:37] <giacus> jepler: cradek around ?
[20:00:47] <cradek> yes
[20:00:50] <giacus> hi
[20:00:56] <cradek> hi
[20:01:10] <giacus> can I proceed to install axis from apt ?
[20:01:18] <giacus> its updated ?
[20:01:19] <cradek> are you using ubuntu?
[20:01:25] <giacus> nope, BDI
[20:01:38] <cradek> you probably need to compile it then.
[20:01:57] <giacus> apt-cache search says python2.3-axis - AXIS front-end for emc
[20:02:01] <cradek> I do not make packages for bdi.
[20:02:10] <giacus> ok..
[20:02:19] <cradek> I don't know how old that is.
[20:02:38] <cradek> you are using emc2 right?
[20:02:43] <giacus> I just installed the latest emc2 right now
[20:02:47] <giacus> from cvs
[20:02:51] <cradek> did you use run-in-place?
[20:02:54] <giacus> should be Testing
[20:02:59] <giacus> yeah
[20:03:05] <giacus> I used it
[20:03:21] <cradek> then get a snapshot of axis from axis.unpy.net and install it
[20:03:41] <cradek> http://axis.unpy.net/downloads
[20:03:50] <giacus> yeah, thanks, found it
[20:03:53] <cradek> I recommend 1.2a2
[20:04:19] <cradek> or the latest cvs snapshot
[20:04:32] <giacus> http://axis.unpy.net/files/downloads/nightly/axis-latest.tar.bz2
[20:04:38] <cradek> yes
[20:04:39] <giacus> should be correct
[20:04:41] <giacus> ok
[20:10:47] <tfmacz> Hi Chris, Ted here
[20:11:08] <tfmacz> I sent you an email about the kernel problem
[20:11:14] <cradek> hi
[20:11:19] <cradek> wireless?
[20:11:38] <tfmacz> no wired.
[20:11:46] <cradek> ok
[20:12:26] <tfmacz> the kernel problem is not that big a deal for me. where I have the machine set up there is no network.
[20:13:00] <cradek> so the ubuntu stock kernel doesn't work with your network card?
[20:13:33] <tfmacz> but I tried to compile emc2 against the kernel and got a bunch of can't find "something.d" files
[20:13:45] <tfmacz> no
[20:14:05] <cradek> so now you are trying to build emc2 from cvs
[20:14:08] <tfmacz> 2.6.10-adios and do.
[20:14:39] <tfmacz> I am a raw naovice when it comes to this compile stuff
[20:15:00] <cradek> at the beginning of the emc2 build you will get messages about .d files, but they are not errors
[20:15:26] <cradek> if you get an actual error that stops the compile, let's have a look at that one
[20:15:53] <cradek> so is your OS install itself BDI or Ubuntu?
[20:16:18] <tfmacz> I am starting up the machine I am trying to compile on. it willt ake a minute
[20:16:34] <tfmacz> the os is ubuntu5.10
[20:16:57] <cradek> ok
[20:17:23] <cradek> do you want to do development on emc2 or just run it?
[20:17:46] <tfmacz> I isntalled ubuntu and have a working emc2 system using your .deb packages.
[20:17:51] <tfmacz> I am just a user.
[20:17:56] <cradek> ok
[20:18:05] <cradek> why are you compiling it now then? I still don't quite understand what's going on
[20:19:07] <tfmacz> the current .deb packages install and run 2.6.12-magma. This has the broken network card module
[20:19:22] <cradek> ok so this is all about the network card
[20:19:40] <tfmacz> I am trying to get emc2 running on the kernel which work ok on my network card.
[20:19:50] <cradek> http://ubuntuforums.org/archive/index.php/t-91263.html
[20:20:23] <giacus> cradek: I' missing something
[20:20:26] <cradek> this is the problem you're having? have you found anything online about what has to be done to fix it (under ubuntu)?
[20:20:55] <tfmacz> If I can get this woking then I will run the network out to my shop. Makes it easier to transfer gcode files out there.
[20:21:49] <giacus> cradek: here's my issue : error: invalid Python installation: unable to open /usr/lib/python2.3/config/Makefile (No such file or directory)
[20:22:35] <cradek> giacus: looks like a clear error message - your python 2.3 installation is bad
[20:22:50] <cradek> giacus: maybe there's a -dev package you need to add?
[20:23:01] <giacus> checking ..
[20:23:44] <tfmacz> yes the url you sent discribes my problem. If someone built the emc2 .deb packages to use the kernel my problem would go away.
[20:24:08] <cradek> tfmacz: you would have to talk to the BDI maintainer for that - I do not support the BDI kernel packages
[20:24:17] <jmkasunich> the kernel is the BDI kernel?
[20:24:23] <cradek> yes
[20:24:38] <cradek> unforunately the names are similar - the Ubuntu kernel is 2.6.12-magma
[20:24:55] <jmkasunich> what is it about that kernel that emc2 doesn't like? (or is answering that non-trivial?)
[20:25:27] <cradek> about the bdi kernel? probably nothing - it's just that tfmacz would prefer to not have to compile emc2
[20:25:39] <jmkasunich> oh, I se4
[20:25:40] <jmkasunich> see
[20:25:54] <jmkasunich> he's trying to do a BINARY install on an incompatible kernel
[20:25:59] <jmkasunich> thats a no-no
[20:26:16] <cradek> tfmacz: I would rebuild the Ubuntu kernel with a fix for your network card, but I don't know what the fix is
[20:26:23] <cradek> tfmacz: I'm trying to find information on google
[20:26:32] <cradek> tfmacz: have you found anything about getting that card to work in ubuntu?
[20:28:24] <tfmacz> the bdi kernel works fine. the ubuntu kernel does not. This is all about getting a machine running.
[20:28:44] <tfmacz> yes. use a different kernel.
[20:28:44] <cradek> well we have different goals. for me, it's about fixing a problem with the ubuntu kernel.
[20:29:23] <tfmacz> If I can help you do that it wil solve my problem.
[20:29:43] <cradek> yes, it would.
[20:30:23] <cradek> if you want up-to-date binary packages of emc2, you will need to use the ubuntu kernel and rtai. The bdi maintainer has no interest in building emc2 packages.
[20:30:29] <tfmacz> for me it is either fix the ubuntu kernel or figure out hoe to compile emc2 in a ubuntu installation running the bdi kernel.
[20:30:42] <giacus> cradek: sorry, I've some other issue, maybe gcc, here the latest part of ouput if you can take a look : http://www.rafb.net/paste/results/3vbsiW31.html
[20:31:51] <cradek> giacus: I think you cut off the error message which would have been earlier
[20:32:12] <tfmacz> Chris, what do you need from me to trace the problem with the 3com pcmcia network card?
[20:32:16] <giacus> I just seen the latest part the terminal saved
[20:32:26] <giacus> is there a log ?
[20:32:39] <cradek> giacus: can I help you in a little bit?
[20:32:50] <giacus> yeah
[20:33:00] <cradek> tfmacz: do you get error messages or anything else that would help in a google search?
[20:33:16] <giacus> a moment..
[20:33:51] <giacus> looking how to extend the buffer
[20:34:57] <tfmacz> where might I look for pertinent error messages?
[20:35:01] <cradek> dmesg
[20:35:05] <cradek> or /var/log/messages
[20:36:03] <dave-e> gotta go....time to take the machine with ubuntu5.10 and emc2 down to the shop and try to do some good. :-)
[20:38:25] <cradek> tfmacz: and once more just to be sure, your card also does NOT work with the Ubuntu distributed kernel which is called "Ubuntu, kernel 2.6.12-10-386" on the boot menu?
[20:40:26] <jmkasunich> ie the stock ubuntu kernel
[20:41:10] <tfmacz> I am rebooting to try that kernel an make sure.
[20:42:19] <Jack_emc2> cradek: now should be complete: http://www.rafb.net/paste/results/VcX2r257.html
[20:42:39] <cradek> thirdparty/togl.c:34:66: X11/Xmu/StdCmap.h: No such file or directory
[20:42:46] <cradek> you are missing more -dev packages
[20:43:10] <tfmacz> The distribution kernel is 2.6.12-9-386. I will try 2.6.12-10-386
[20:43:22] <giacus> cradek: ok, thanks :)
[20:44:00] <cradek> giacus: I thought there was a package list on the wiki - looking for it
[20:44:19] <giacus> looking
[20:44:33] <cradek> it's probably libXmu-dev or something like that
[20:44:51] <giacus> ty
[20:45:37] <jepler> these are the packages that the axis package requires to build: Build-Depends: debhelper (>= 4.1.67), emc2-dev, python2.4-dev, libgl1-mesa-dev, libglu1-mesa-dev, libxmu-dev, tk8.4-dev
[20:45:48] <jepler> (package names are for ubuntu, may be a bit different on bdi)
[20:45:51] <alex_joni> apt-get build-dep axis-emc2
[20:45:53] <giacus> jepler: thank you
[20:46:00] <jepler> alex_joni: yeah, that would work too
[20:46:05] <cradek> not on bdi
[20:46:16] <giacus> yes, I'm using BDI
[20:46:20] <alex_joni> cradek: how come?
[20:46:29] <alex_joni> different package names?
[20:46:43] <giacus> mmm
[20:46:48] <cradek> alex_joni: bdi users don't even have that repository, nothing there will work
[20:46:50] <giacus> let me check ..
[20:46:50] <alex_joni> it should pull the Build-Depends: list and install those packages
[20:46:58] <giacus> libxmu-dev: Dipende: libxmu6 (= 4.3.0.dfsg.1-14sarge1) ma 6.8.2.dfsg.1-7 sta per essere installato
[20:46:58] <alex_joni> ahhh.. sorry, yes
[20:47:34] <tfmacz> Chris, half a network connection on 2.6.12-10-386. same as 2.6.12-magma
[20:47:42] <cradek> ok, that tells us a lot
[20:47:54] <cradek> my kernel is based on the latest version of 2.6.12-10-386
[20:48:03] <cradek> this tells us the problem wasn't in my build
[20:48:27] <cradek> but ... unfortunately not what it takes to fix it
[20:48:35] <cradek> anything in dmesg or /var/log/messages?
[20:48:42] <tfmacz> looking
[20:48:57] <jmkasunich> cradek: remember my problems with noapic?
[20:49:04] <cradek> jmkasunich: yes
[20:49:04] <jmkasunich> could this be similar?
[20:49:15] <cradek> jmkasunich: the 2.6.12-magma kernel is now built without apic
[20:49:28] <jepler> giacus: my bdi system has libxmu6 4.3.0.dfsg.1-14sarge1 so the libxmu-dev package matches it. If you installed this other version of libxmu6, then you have to install the matching -dev package
[20:50:05] <giacus> jepler: tryng.. now installing python2.4 too
[20:50:41] <cradek> tfmacz: is that usb or pcmcia?
[20:51:38] <SWPadnos> there should be a gtk dependency as well, for halscope and vcp
[20:54:00] <cradek> tfmacz: I'm also asking on #ubuntu
[20:55:38] <tfmacz> pcmcia, there is some stuff in var/log/messages that might be relevant. trying to figure out how to send it to you. right now the box haas no network???
[20:56:19] <cradek> tfmacz: I guess save it to another file and reboot?
[20:56:29] <tfmacz> doing that.
[20:56:45] <cradek> https://wiki.ubuntu.com/HardwareSupportComponentsWiredNetworkCards?highlight=%283c574%29
[20:56:51] <SWPadnos> oh - I have an error from an attempted kubuntu install as well
[20:56:54] <cradek> this shows that card as supported and working...
[20:57:37] <cradek> tfmacz: do you have other pcmcia cards that work in this kernel, or is this the only card you have tried?
[20:57:47] <cradek> tfmacz: I'm thinking maybe it's a pcmcia problem in general
[20:58:51] <tfmacz> I think I have a zircom card but I never tried it cause I seem to remember they wern't supported in linux
[21:00:15] <cradek> filename: /lib/modules/2.6.12-magma/kernel/drivers/net/tulip/xircom_cb.ko
[21:00:15] <cradek> description: Xircom Cardbus ethernet driver
[21:00:22] <cradek> description: Xircom CBE-100 ethernet driver
[21:00:28] <cradek> there are a few xircom modules
[21:00:49] <jmkasunich> probably an irrelevant data point, but I've used a 3Com Megahertz wired ethernet PCMCIA card with both the stock ubuntu kernek and with cradek's
[21:01:10] <giacus> jepler are you using Kde in Bdi ?
[21:08:35] <jepler> giacus: no, I can't stand kde.
[21:08:55] <tfmacz> Chris http://ve7tfm.dyndns.org/upload/kernel-errors.txt
[21:09:23] <tfmacz> not sure what all that means but maybe you cna decifer it???
[21:11:13] <cradek> does anyone here read hu?
[21:11:18] <cradek>!%22&hl=en&gl=us&ct=clnk&cd=4&client=firefox
[21:11:25] <tfmacz> The card in question is a 3CCFEM556B
[21:12:37] <jepler> cradek: I think the last message is suggesting to change the pcmcia configuration to exclude an IRQ. "/etc/pcmcia/config.opts file-ban meg hozzaadtam, hogy exclude irq 5, es igy irq 10-es kent jon fel a PCMCIA"
[21:12:53] <jepler> tfmacz: do you have any other pcmcia cards? Do they work?
[21:13:21] <cradek> tfmacz: maybe you should exclude irq3 to make it pick a different one, and see if that helps (in /etc/pcmcia/config.opts)
[21:13:23] <alex_joni> jepler: http://lists.debian.org/debian-laptop/2002/12/msg00141.html
[21:13:47] <cradek> aha
[21:13:52] <cradek> that gives the same advice
[21:13:54] <tfmacz> The other card I have is a Zircom card. My son has it and is bringing it over in a while. I will try it when he gets here.
[21:13:55] <cradek> exclude irq3
[21:13:57] <giacus> * giacus is going to remove kde and install Ion
[21:14:45] <alex_joni> tfmacz: can you look at the current kernel, (the one that works), and see what irq it's using?
[21:15:22] <tfmacz> irq3 0x300
[21:15:23] <alex_joni> cat /proc/interrupts ?
[21:15:37] <alex_joni> odd
[21:16:03] <alex_joni> but do like cradek suggest, tell it not to try irq3D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[D[C[C[C
[21:16:18] <alex_joni> sorry.. that wasn't me
[21:18:39] <tfmacz> kittycat problems??? how do I make the change take effect???
[21:19:44] <tfmacz> to answer my own question I have to reboot to try it so I assume that will do it???
[21:20:00] <alex_joni> yeah
[21:20:02] <jepler> tfmacz: The easiest way is to reboot after editing /etc/pcmcia/config.opts. It should also work to run 'sudo /etc/init.d/pcmcia restart' from a command prompt.
[21:20:20] <alex_joni> jepler: yeah, but he'll still be trying the same kernel (the one that works)
[21:21:00] <jepler> oh, forget it then
[21:21:11] <cradek> sounds best to reboot
[21:24:57] <tfmacz> irq10 0x300 still no joy
[21:25:18] <cradek> which kernel?
[21:26:14] <cradek> you might still want to try 2.6.12-magma with this irq10 setup if you haven't yet
[21:26:27] <tfmacz> 2.6.12-magma
[21:26:34] <cradek> darn
[21:26:44] <tfmacz> trying 2.6.12-10-386
[21:27:04] <cradek> did you still get that remote fault detected message?
[21:27:43] <tfmacz> will look one it finishes rebooting
[21:33:22] <tfmacz> looks like the same errors in both the 2.6.12-magma and 2.6.12-10-386
[21:34:14] <cradek> tfmacz: not sure what else to try... I hope your other network card works
[21:36:11] <alex_joni> what's the kernel that works?
[21:36:25] <cradek> the bdi kernels
[21:36:28] <tfmacz> rebooting the bdi kernel to see what is in the log files.
[21:36:34] <cradek> good idea
[21:36:40] <tfmacz> yes
[21:42:45] <giacus> now I should have anything installed
[21:42:54] <giacus> what about emc2-dev package ?
[21:43:04] <giacus> cant find it
[21:43:16] <cradek> giacus: if you're using bdi, you cannot use the emc2 packages
[21:43:31] <giacus> thats ok, I used the source
[21:43:33] <giacus> then
[21:43:34] <tfmacz> interesting, the "eth0: remote fault detected" is there with the kernel that works??? all the stuff starting with the "[" is not.
[21:43:47] <cradek> strange!
[21:43:55] <alex_joni> yeah, the remote fault detected is not an actual error that the HW doesn't work
[21:44:05] <cradek> does it also get irq10 0x300?
[21:44:06] <alex_joni> it's an info about the HW link
[21:44:14] <cradek> ah
[21:44:19] <tfmacz> yes irq10 0x300
[21:44:21] <giacus> libxmu6 is now installed
[21:48:18] <tfmacz> this is not an end all situation...I do have a working emc2 system. Just no network for now. If I have to I will look for another card. my network switch will handle a mixed speed network so I might eb able to find a 10meg card cheep...
[21:48:53] <cradek> tfmacz: wish I knew what else to do to help you
[21:48:53] <alex_joni> tfmacz: glad you take it that easy :)
[21:49:00] <giacus> uh.. error: /usr/lib/python2.3/site-packages/_tk_seticon.so: Permission denied
[21:49:04] <alex_joni> cradek: maybe the next kernel will work
[21:49:11] <cradek> sure could be
[21:49:13] <alex_joni> giacus: thought you said 2.4 ?
[21:49:29] <giacus> error: /usr/lib/python2.4/site-packages/_tk_seticon.so: Permission denied
[21:49:57] <giacus> should I compile it as root ?
[21:50:08] <alex_joni> you need sudo for setup.py install
[21:50:16] <giacus> oh.. ok
[21:50:54] <tfmacz> The big thing is the great work being done to get the emc2 package together. With my hardware I am seeing a 300% improvement in performance over emc1
[21:50:55] <giacus> did it ..
[21:50:57] <giacus> :D
[21:51:15] <cradek> tfmacz: that's great to hear
[21:51:39] <cradek> tfmacz: it's much more flexible too, for different hardware setups
[21:53:16] <alex_joni> tfmacz: what kind of hardware?
[21:55:34] <tfmacz> hardware is homemade. essentially a heavy duty engraver. 12" x 12" x-y and 4" z
[21:55:57] <tfmacz> stepper motors on microstepping drivers.
[21:56:48] <tfmacz> The computer is a Toshiba 233 meg PII laptop
[21:56:58] <alex_joni> nice
[21:57:14] <alex_joni> tfmacz: glad emc2 works so well with it
[21:57:19] <cradek> me too
[21:57:42] <cradek> I had troubles using realtie on my laptop - when the processor fan would turn on or off, the steppers would stall (realtime deadlines missed)
[21:57:45] <cradek> realtime
[21:58:03] <alex_joni> I imagine more than one deadline missed
[21:58:26] <cradek> yeah I think the latency test shows something like .2 seconds!
[21:58:54] <Jymmm> bummer, he's not here =(
[21:58:56] <SWPadnos> I've had ~800 ms latencies on my laptop
[21:59:08] <tfmacz> I have been playing with carving 2.5d pictures from photographs. It works very well.
[21:59:15] <cradek> some laptops see good, some bad
[21:59:19] <cradek> grr seeM
[21:59:24] <Jymmm> tfmacz using what?
[21:59:35] <SWPadnos> or was that ~8 ms? I can't remember
[21:59:46] <cradek> SWPadnos: either is way too long
[21:59:51] <SWPadnos> yes
[22:00:02] <SWPadnos> just a question of how many orders of magnitude
[22:01:19] <tfmacz> I have been using a windoz shareware called Image to G-Code. The free trial just ran out and he wants more than I want to spend to register so am looking for something else to try.
[22:01:32] <cradek> good kvm do not let the computer see the stuff being disconnected
[22:01:50] <cradek> wrong channel!
[22:01:52] <Jymmm> tfmacz: I think I know what your talking about... do yo have any photos of the results?
[22:03:09] <cradek> tfmacz: a simple image-to-gcode program on my website http://timeguy.com
[22:03:25] <cradek> tfmacz: if you can program a bit, you may be able to get it to do what you want
[22:03:38] <cradek> tfmacz: it's just a simple scanline converter; don't know what you had before
[22:05:13] <tfmacz> I am putting together a webpage with my machine and results so far. I just finished cutting a picture of my granddaughter. I have to stain it yet. http://ve7tfm.dyndns.org/upload/carving1.jpg
[22:05:57] <alex_joni> nice
[22:06:24] <Jymmm> tfmacz: what tool did you use on that one?
[22:06:28] <alex_joni> tfmacz: it would be nice if you would add yourself to the EMC usage map (http://www.frappr.com/emctheenhancedmachinecontroller)
[22:07:04] <Jymmm> alex_joni can't someone rename that to something a lil shorter?
[22:07:24] <alex_joni> Jymmm: don't think rename works..
[22:07:27] <tfmacz> Chris, thanks, Lawrence pointed me to it. I will have alook at it.
[22:07:39] <Jymmm> alex_joni bummer.
[22:07:49] <fenn> jymmm how's this: tinyurl.com/deyjq
[22:08:24] <Jymmm> fenn: Thanks, but I dont do tinyurls - you never know what you're going to get.
[22:08:42] <cradek> must be small, I don't see scanlines
[22:09:32] <Jymmm> there is BMP2DIN for free at majosoft too (M$)
[22:09:42] <Jymmm> but I haven't tried it yet
[22:10:17] <Jymmm> asdf? said last night he created a gcode2dxf program, but left before I could ask about it.
[22:11:01] <alex_joni> night all
[22:11:07] <Jymmm> nite alex_joni
[22:13:41] <tfmacz> sorry....the cutter is a 1/16" end mill
[22:14:00] <Jymmm> oh, not ballend huh?
[22:15:02] <Jack_emc2> was too much beautiful ..
[22:15:18] <Jack_emc2> Starting emc...
[22:15:19] <Jack_emc2> insmod: error inserting '/lib/modules/': -1
[22:15:19] <Jack_emc2> Operation not permitted
[22:15:25] <Jack_emc2> what happen ?
[22:15:26] <Jymmm> tfmacz: This program? http://www.imagetogcode.com/
[22:17:13] <jmkasunich> Jack_emc2: need more info
[22:17:35] <jmkasunich> what OS? ubuntu, bdi, other? what emc?
[22:17:45] <Jack_emc2> jmkasunich: BDI
[22:17:49] <tfmacz> Jymmm: yes that is it.
[22:17:57] <jmkasunich> the emc that is on the bdi, or emc2?
[22:18:10] <Jack_emc2> emc2 compiled from source (cvs)
[22:18:20] <jmkasunich> installed or run-in-place?
[22:18:27] <Jack_emc2> yeah
[22:18:33] <jmkasunich> one or the other?
[22:19:06] <Jack_emc2> ./configure run-in-place if I remember
[22:19:08] <Jymmm> tfmacz: Yeah, I wish he documented things a bit more... like specifics on how he created his samples. I emailed him once, never got a response back.
[22:19:24] <Jack_emc2> it was working few minuts ago
[22:19:29] <jmkasunich> for run-in-place (recommended on bdi): cd src; ./configure --enable-run-in-place ; make ; sudo make setuid
[22:19:49] <Jack_emc2> yeah. it what I did before
[22:19:57] <Jack_emc2> I can do it again
[22:20:03] <jmkasunich> slow down
[22:20:11] <jmkasunich> you didn't tell us it was working before
[22:20:15] <Jack_emc2> I just upgrade BDI using apt and reboot
[22:20:25] <Jack_emc2> yeah, it was working
[22:20:39] <jmkasunich> which bdi? 4.38 I bet
[22:21:03] <Jack_emc2> probably.. cat remember
[22:21:08] <Jack_emc2> james@localhost:~/emc2$ uname -r
[22:21:08] <Jack_emc2>
[22:21:42] <Jack_emc2> I also did depmod -ae
[22:21:45] <jmkasunich> before the upgrade you could do "scripts/emc" and it would work, now it doesn't?
[22:21:47] <Jack_emc2> update-modules
[22:22:02] <Jack_emc2> exactly
[22:22:11] <Jack_emc2> it was working before
[22:22:17] <jmkasunich> I'm not sure what paul is doing with BDI anymore
[22:22:31] <jmkasunich> try redoing the configure/make/make setuid
[22:22:37] <Jack_emc2> I just used apt to upgrade, I seen some upgrade on kernel and modules
[22:22:44] <Jack_emc2>
[22:22:56] <Jack_emc2> now, strange things are happening
[22:23:18] <cradek> if you changed your kernel or rtai modules you must recompile emc2
[22:23:20] <Jack_emc2> I sucesful compiled Axis too
[22:23:29] <Jack_emc2> mmmm
[22:23:33] <Jack_emc2> that why
[22:23:36] <Jack_emc2> then ..
[22:23:39] <cradek> maybe
[22:23:45] <Jack_emc2> now tryng
[22:23:47] <Jymmm> tfmacz: Curious, how long was the machien time on your image?
[22:24:35] <Jack_emc2> I'm doing make clean now
[22:25:10] <tfmacz> took about an hour.
[22:25:23] <tfmacz> the image is 6" x 8"
[22:25:38] <Jymmm> tfmacz ah, ok.
[22:26:36] <Jack_emc2> james@localhost:~/emc2/src$ sudo make setuid
[22:26:36] <Jack_emc2> chown root ../bin/emc_module_helper
[22:26:36] <Jack_emc2> chown: impossibile accedere a `../bin/emc_module_helper': No such file or directory
[22:27:28] <Jack_emc2> cant get acces to..
[22:27:40] <jmkasunich> make clean ; ,/configure --enable-run-in-place ; make ; sudo make setuid
[22:27:45] <jmkasunich> don't skip any steps
[22:27:57] <Jack_emc2> sorry.. right
[22:28:20] <Jack_emc2> I forgot the more important :/
[22:28:28] <jmkasunich> ;-)
[22:33:03] <Jack_emc2> nice
[22:33:25] <Jack_emc2> now emc start with some insmod error but seems work
[22:33:35] <Jack_emc2> insmod: error inserting '/lib/modules/': -1 File exists
[22:33:43] <Jack_emc2> lets try Axis now
[22:35:11] <Jack_emc2> well, finally all seems work
[22:35:15] <Jack_emc2> :)
[22:35:19] <Jack_emc2> thanks guys
[22:36:31] <tfmacz> cradek: tried the Zircom cards with no joy. CE-10/A and CEM2
[22:42:56] <giacus> :-)
[23:36:55] <Jymmm> tfmacz: I tried a few zircom cards I have, doens't seem like a lot of support for them, even under M$. At least nothing like intel or 3com.