#emc | Logs for 2008-01-26

Back
[00:01:49] <seb_kuzminsky> one nice thing with sudo is you get a log of all your superuser commands
[00:02:39] <fenn> where's that?
[00:03:03] <seb_kuzminsky> in the syslog
[00:03:13] <Ziegler> I didnt know that
[00:04:31] <fenn> ah now i see it. /var/log/auth.log
[00:04:31] <seb_kuzminsky> everyone is looking in their syslogs and typing sudo echo hello world
[00:04:39] <archivist> hehe
[00:05:03] <seb_kuzminsky> auth, right fenn, my mistake
[00:06:41] <fenn> hmm are steppers sinusoidal or trapezoidal?
[00:57:14] <GNieport1> hi
[01:10:53] <GNieport1> excellent, the new 7i33 is working 8-)
[01:17:19] <seb_kuzminsky> nice gnieport!
[01:23:52] <GNieport1> Thanks; anyone know if the 5i20 driver supports a higher PWM rate than the 32kHz indicated in EMC User Guide?
[01:24:08] <GNieport1> The card itself does 100kHz+
[01:24:51] <seb_kuzminsky> I'm not sure how the current driver does it. I
[01:25:10] <seb_kuzminsky> I'll be writing a hostmot2 driver for the 5i20 in a bit, then I'll let you know how that works :-)
[01:25:27] <GNieport1> there is a parameter passed to the driver when loadrt is used
[01:26:19] <GNieport1> loadrt hal_m5i20 [foo] [dacRate=32000]
[01:27:08] <GNieport1> is the hostmot2 designed to extend the number of axes available for servo control?
[01:27:16] <GNieport1> step & dir?
[01:27:43] <seb_kuzminsky> brb, gotta give the kids a bath
[01:27:47] <seb_kuzminsky> they smells funky
[01:27:50] <GNieport1> hehe cya
[01:40:04] <archivist_emc_> ew machine running and sending webcam pics of itself
[01:40:31] <archivist_emc_> www.archivist.info/webcam.jpeg
[01:43:57] <skunkworks_> archivist_emc_: I don't see it running...
[01:44:14] <archivist> refresh
[01:44:39] <skunkworks_> oh - I have to refresh
[01:44:42] <skunkworks_> heh
[01:44:46] <skunkworks_> cool
[01:45:05] <archivist> quick and dirty its not forcing your browser to refresh
[01:45:11] <skunkworks_> I really can only see the rotory going
[01:45:36] <skunkworks_> atleast that is all I can retain from refresh to refresh ;)(
[01:45:58] <skunkworks_> and now I am getting sick ;)
[01:46:41] <skunkworks_> How are you liking emc?
[01:46:48] <archivist> well it manages to run a webcam app at the same time!
[01:47:12] <skunkworks_> * skunkworks_ is usually IRC-ing and surfing while cutting..
[01:47:17] <skunkworks_> atleast
[01:47:45] <archivist> I had dependency hell to find a working webcam app
[01:48:26] <skunkworks_> Do you have a picture of the machine?
[01:48:32] <archivist> two more axis to get done
[01:48:54] <archivist> www.archivist.info/cnc
[01:51:46] <archivist> 2 am and now I can think of machine vision over the weekend, hack the cam add lenses and lights, see where the cutters is, maybe
[01:51:46] <skunkworks_> ah - so your building someting almost from scratch
[01:51:58] <skunkworks_> bits of stuff
[01:52:00] <archivist> yup scratch/scrap
[01:52:05] <skunkworks_> nice
[01:52:39] <archivist> base is toolmakers setting jig with nice slides
[01:53:02] <archivist> column likely to be lathe base
[01:53:08] <skunkworks_> I like your z. Are you going to use the angle block and tombstone?
[01:53:17] <skunkworks_> ah
[01:53:39] <archivist> headstock =lathe headstock on ballslides
[01:54:05] <skunkworks_> horizontal or vertical?
[01:54:18] <skunkworks_> (spindle)
[01:54:42] <archivist> erm no pencils or cad programs have been used in the design so far
[01:54:50] <archivist> vertical spindle
[01:55:03] <skunkworks_> heh - those are the best designs :)
[01:55:32] <archivist> that is anormal 3 axis mill and added rotary for gear cutting
[01:57:08] <archivist> just need to learn how to grab single pics when I want them next
[01:57:18] <GNieport1> It's alive! First commanded axis move. The 38 year old iron lives again. :-P
[01:58:19] <skunkworks_> GNieport1: great! Now you need to get over to Gamma-X's place and help him out ;)
[01:58:24] <archivist> hehe
[01:59:09] <archivist> skunkworks, webcam now has proposed table on it
[02:00:08] <Gamma-X> ok i got my axis's to move completely but they dont like to stop when i hit go in manual mode haha.
[02:00:10] <GNieport1> skunkworks: indeed
[02:00:18] <Gamma-X> and i messed with the pid but it didnt realy do anything...
[02:00:49] <skunkworks_> archivist: I am not seeing it.. I think
[02:00:52] <GNieport1> G-X it's probably important that they stop :)
[02:01:17] <Gamma-X> yeah i figured that.... but i have no idea why they dont stop hahaha had to pull the estop, i was like oh shit! its moving! hahaha
[02:01:29] <GNieport1> G-X the 7i33 was burned up, the new one is wiggling my X axis right now
[02:01:43] <archivist> skunkworks, the slab under the rotary
[02:02:19] <skunkworks_> GNieport1: 38? that is the age of this... http://www.electronicsam.com/images/KandT/DSCCurrent.JPG
[02:02:44] <skunkworks_> archivist: same as this? http://www.archivist.info/cnc/stagetwo/P1200008.JPG
[02:03:29] <archivist> yup
[02:03:41] <skunkworks_> nice - diamond core drill?
[02:04:13] <archivist> where what? no
[02:04:34] <skunkworks_> (wondering how your going to mount it/ things to it.)
[02:04:53] <toastydeath> that is a nice milling machine
[02:04:56] <toastydeath> whose is it
[02:05:40] <skunkworks_> this? http://www.electronicsam.com/images/KandT/oldkandt.JPG
[02:06:02] <archivist> the big old iron
[02:06:16] <skunkworks_> we have it in our shop
[02:06:17] <GNieport1> it's huge. The chips could be fearsome.
[02:06:33] <skunkworks_> the 60's vintage controller finally died.
[02:06:44] <skunkworks_> we decided it was time for an upgrade
[02:06:52] <GNieport1> meh. ~50 years isn't too bad
[02:07:30] <Gamma-X> GNieport1 great!
[02:08:12] <GNieport1> G-X: yep, I'm letting the screw warm up now before running some tuning waveforms
[02:08:46] <Gamma-X> now all i gotta do is figure out why my axis's dont stop ... and configure pid and i should be good. atleast in my head that sounds good
[02:09:47] <toastydeath> yeah that machine is very cool guys
[02:12:32] <skunkworks_> archivist: don't you ever sleep?
[02:12:50] <archivist> soon its getting cold in here
[02:13:02] <archivist> got to drive home yet
[02:14:30] <skunkworks_> we got another 2 inches of snow here..
[02:14:39] <skunkworks_> just keeps coming :)
[02:15:17] <skunkworks_> atleast it is supposed to be in the 30's this week end. (been below 0f)
[02:16:19] <Gamma-X> GNieport1 i tried adjusting that pid but it didnt seem to do anything...
[02:24:34] <acemi> is there a plan to remove RTLinux support for the future release of EMC2?
[02:59:31] <GNieport1> Gamma X were you able to command a move?
[03:08:15] <Gamma-X> kinda
[03:08:21] <Gamma-X> i changed the fault ouput thing.
[03:08:30] <Gamma-X> so i could do some moves without messin with pid.
[03:09:13] <Gamma-X> my x servo amp aint hooked up so i cant realy open a file and run it cause it seas theres no x moving at all. but i can move my axis's
[03:09:27] <Gamma-X> i hit the proper axis and then the move button, but it didnt stop! lol
[03:11:47] <jmkasunich> Gamma-X: you will not accomplish much until you tune your PID loops
[03:13:25] <Gamma-X> jmkasunich ive tried readin online how differant people tune pid but im having a hard time in general...
[03:13:37] <Gamma-X> tried hookin up halscope to it but couldnt even get that haha.
[03:14:05] <Jymm> SWPadnos: pg 689 (PDF17), PN: Series 0030 002.23 https://www.relayspec.com/catalogs/009315_Section5b.pdf
[03:14:13] <Gamma-X> tomorow im dedicating my whole day to that machine, i even didnt take OT. Just to get my machine up. Maby someone that isnt doin much can talk me through some stuff on the phone? hahah
[03:16:30] <GNieport1> The EMC Integrator pdf has a whole section on tuning
[03:16:43] <Gamma-X> i know... i tried but i guess not hard enough.
[03:16:48] <SWPadnos> step 0: make sure the e-stop chain works
[03:16:50] <Gamma-X> just confused on where to start with the values.
[03:16:55] <SWPadnos> step 1: make sure encoder feedback works
[03:16:56] <Gamma-X> already done.
[03:17:00] <Gamma-X> it does.
[03:17:06] <GNieport1> Suggest to go thru the HAL tutorial in the EMC User pdf
[03:17:08] <Gamma-X> limit switches work aswell
[03:17:20] <Gamma-X> i did that 2... 4 times so far
[03:17:29] <SWPadnos> step 2: make sure motor command chain works (ie, the torque/velocity/position changes as commanded by HAL)
[03:17:53] <Gamma-X> SWPadnos if ur reffering to making the motors move then yes thats complete.
[03:17:53] <SWPadnos> that gets to step 3: tune
[03:18:05] <SWPadnos> no, I'm referring to actually controlling motion
[03:18:53] <SWPadnos> I can make the motors move with a battery in the right place
[03:18:53] <Jymm> oh baby!!!
[03:18:53] <jmkasunich> making the motors move in the right direction, at approximately the right speed, when you send a command to the dac
[03:18:53] <Gamma-X> ill have to hook up my x servo amp tomorow and test out a example file.
[03:18:58] <jmkasunich> example file?
[03:19:02] <jmkasunich> you mean a g-code file?
[03:19:13] <Gamma-X> jmkasunich well i might have done something wrong but i hit + on the y and it kept going...
[03:19:18] <Gamma-X> jmkasunich correct.
[03:19:26] <jmkasunich> g-code files are about step 8
[03:19:44] <jmkasunich> if you haven't tuned anything, then you should expect it to do strange things
[03:19:53] <jmkasunich> do you even know what the gains are now?
[03:19:57] <Gamma-X> hahaha
[03:20:11] <jmkasunich> do you know how to check what the gains are?
[03:20:17] <Gamma-X> negative.
[03:20:18] <SWPadnos> Jymm, I think I have some of those in my DIN crate actually
[03:20:19] <jmkasunich> do you know how to change gains on the fly
[03:20:30] <jmkasunich> you need to read AND DO the HAL tutorial
[03:20:35] <Jymm> SWPadnos: I have four of them now.
[03:20:45] <SWPadnos> do you know how to test how well a set of gains is working?
[03:20:52] <Gamma-X> nope.
[03:21:17] <Gamma-X> i musta skipped a lot in order to rush, wich i know isnt right after everyone here did a lot of hard work to create emc.
[03:21:28] <Gamma-X> and support and maintain
[03:21:43] <SWPadnos> you really need to read and understand how HAL works to be able to set up a machine
[03:21:49] <toastydeath> YOU SHOULD BE ASHAMED OF YOURSELF FOR SKIPPING THAT
[03:21:52] <SWPadnos> understand is the key word there
[03:21:53] <toastydeath> GOD WILL SMITE YOU
[03:21:58] <SWPadnos> DOG!
[03:22:04] <SWPadnos> oops - dyslexia
[03:22:10] <toastydeath> haha best awful joke ever
[03:22:15] <Gamma-X> hahahahahah
[03:22:16] <SWPadnos> dyslexics - UNTIE!
[03:22:28] <Jymm> SWPadnos: I went back to HSC (the place you bought stuff from) and got two DB25-F to screw term, as well as those 10" DIN rails that had a bunch of screw terms and two relays on.
[03:22:37] <SWPadnos> oh, cool
[03:22:54] <eric_U> why don't I take some time off from working on my machine to help Gamma-X get his running
[03:23:06] <SWPadnos> because yours is more important to you
[03:23:07] <Gamma-X> cause then u would need to buy beers
[03:23:16] <eric_U> that's right
[03:23:21] <Jymm> SWPadnos: I also found the square snap-in outlets. These "seem" better quality than the scrap one I had in hand, but yes, I need to make a cross cutout (for the record)
[03:23:32] <SWPadnos> ok - thought so :)
[03:23:50] <Gamma-X> eric_U u free this weekend?
[03:24:08] <eric_U> funny
[03:24:13] <Gamma-X> I can have my dads wife cook for u! She makes a kilelr meatloaf!
[03:24:14] <SWPadnos> Jymm, you'll still needa buffer to drive those relays from a parport - they need 22mA
[03:24:21] <Gamma-X> ill even buy the beers for U!
[03:24:24] <SWPadnos> err - 18mA
[03:24:37] <eric_U> I would if I was a couple hours closer
[03:24:46] <SWPadnos> so some open-collector thing like a 5803/2803 (whichever ir is)
[03:25:08] <Gamma-X> eric_U u can stay at my place...
[03:25:22] <fenn> Gamma-X wanna help me remove this crank pulley bolt? i'll buy u a beer
[03:25:26] <Gamma-X> haha. but u might have a family to tend too
[03:25:30] <Jymm> * Jymm cleans up the screw terminal strips he got and MAKES SURE that there is no loose single strand of wire remaining - no blowing up any 1200VA xfmrs (again).
[03:25:34] <SWPadnos> it's probably faster for me to fly from Burlington than it is for you to drive
[03:25:43] <SWPadnos> heh - oops
[03:25:43] <Gamma-X> hahaha yes but cheaper to drive...
[03:25:56] <SWPadnos> depends on how desperate JetBlue is, actually
[03:26:05] <Gamma-X> fenn where u live at?
[03:26:06] <SWPadnos> it's sometimes $100 round trip from here to JFK
[03:26:13] <fenn> indiana
[03:26:14] <Gamma-X> wow!
[03:26:17] <eric_U> on one of my cars, to get the crank pully bolt loose, you put a wrench on the frame and start the engine
[03:26:28] <Gamma-X> SWPadnos find out and let me know ill pay ur airfare! lol
[03:26:30] <Jymm> SWPadnos: Be fore warned... I'm gonna beat you over the head with a wet newspaper the next time I see you! (damned fscking son of bleeeech buffer opto-driver crap) Thank you and have a nice day.
[03:26:56] <Gamma-X> fenn u cant get it off?
[03:27:18] <Jymm> off, on, down, up, same diff.
[03:27:18] <fenn> well, i haven't gotten the pneumatic system set up for the mighty 1" impact wrench yet
[03:27:26] <SWPadnos> Jymm, so that's the thanks I get for being nice to your parrots?
[03:27:39] <Gamma-X> parrots? what?! haha
[03:27:52] <Jymm> SWPadnos: You still have your fingers don't ya? And no blood was spilled was it?
[03:28:02] <SWPadnos> Gamma-X, it's ~$300 for tomorrow/sunday travel - not as good a deal the day before
[03:28:12] <Gamma-X> dammmnnn
[03:28:15] <Gamma-X> how about next weekend?
[03:28:22] <Gamma-X> fly in friday and leave w/e
[03:28:23] <SWPadnos> nope, I'll be in Las Vegas
[03:28:30] <Gamma-X> NOOOOO!!!!
[03:28:36] <Jymm> VIVA LOST WAGES!!!
[03:28:47] <SWPadnos> workin, dont'cha know
[03:28:50] <Gamma-X> u got the life!
[03:28:56] <SWPadnos> heh
[03:29:08] <SWPadnos> actually, I'm only home on alternate weekends for the next couple of months
[03:29:17] <Gamma-X> come on irc... talk the shit, get some from wifey id assume... fly around the u.s. .... make a crap laod of money.
[03:29:26] <Gamma-X> SWPadnos weekend after?
[03:29:29] <SWPadnos> well, one out of 4 ain't bad
[03:29:40] <SWPadnos> nope, that's my home weekend before I head for Wichita
[03:29:45] <Jymm> SWPadnos: Alright, you win... Not now of course, but gimme some ideas on a PCB buffer/opto thingymabob when you get a chance - url's would do.
[03:30:41] <SWPadnos> then I only have one day at home before the wife and I go to Wisconsin fir a week
[03:30:42] <SWPadnos> Jymm, I suspect I can design one pretty easily, once we know what you need
[03:31:04] <fenn> Jymm: look at anders wallin's opto card. it's pretty straight forward
[03:31:31] <SWPadnos> he actually doesn't need or want isolation on the step outputs - the drives are already isolated
[03:31:43] <eric_U> why use optos then?
[03:31:45] <Jymm> SWPadnos: Isn't there something I can just buy to iface paraport with multiple buffer lines?
[03:31:55] <Jymm> eric_U: (I was being funny)
[03:32:00] <SWPadnos> eric_U, that's my point - they aren't needed for the drives
[03:32:26] <Jymm> the drives are already opto, the relay/scontactors aren't.
[03:32:28] <SWPadnos> oh right - well, optos are useful on the input side and possibly for other functions, but relays may take care of that
[03:32:42] <fenn> relays are electrically isolated
[03:32:56] <Jymm> fenn: the coil connected to paraport
[03:32:57] <eric_U> yeah, I use ODC5 for those
[03:32:57] <SWPadnos> sure, though they need buffering
[03:33:12] <fenn> hmm i think i'm starting to understand now
[03:33:15] <SWPadnos> heh
[03:33:22] <fenn> what you want is a relay driver
[03:33:24] <Jymm> fenn: that makes one of us ;)
[03:33:25] <Gamma-X> SWPadnos i guess ill wait until ur in the area being that I dont realy have money to pay u like that but If i can offer anything to u please let me know... favors... w/e
[03:33:26] <SWPadnos> and a charge pump too, can't forget that
[03:33:41] <eric_U> buy a bob
[03:33:43] <Jymm> SWPadnos: LawrenceG have me a chip# for the CB
[03:33:53] <SWPadnos> Gamma-X, I'm happy to help, but only if it seems that you're really trying to get things done on your own
[03:34:01] <SWPadnos> so get back to that HAL manual for the fifth time ;)
[03:34:24] <SWPadnos> Jymm, uh - ok. what are you talking about?
[03:34:27] <fenn> Gamma-X: if it makes you feel worse, hal is really pretty easy
[03:34:44] <SWPadnos> yeah - even fenn gets it, and Jymm is starting to :)
[03:35:15] <Gamma-X> i hate the manual to be honest but i love hal.... once u get to realize what the values are it is simple. once i get pid tuned and my machine right im going to make a hal file wich has commented how to's above the lines of code.
[03:35:15] <Jymm> SWPadnos: LawrenceG made a chargepump using some MAXIM IC
[03:35:20] <SWPadnos> ah, ok
[03:35:35] <SWPadnos> you mistyped CP as CB - I thought you were talking about a circuit breaker :)
[03:35:53] <Jymm> SWPadnos: Nah, 11 meters, "Good Buddy"
[03:36:23] <SWPadnos> jmkasunich had a schematic for one too, using a couple of caps and diodes and a resistor - that's probably the simplest circuit
[03:36:29] <Gamma-X> but i gtg... the lady friend is here. Watchin a movie ill talk to u all later on tomorow. Thanks. and if anyone is free tomorow pm me ur "real" name and number and i can give u a call.
[03:36:34] <SWPadnos> (that's a "normal" charge pump)
[03:36:46] <Jymm> sounds like a RC circuit
[03:36:48] <SWPadnos> enjoy, and read that HAL tutorial again
[03:36:57] <SWPadnos> Jymm, no, it's a charge pump
[03:37:06] <Gamma-X> will do. thanks. - Whitey
[03:37:14] <Jymm> lol, still SOUNDS like a RC circuit =)
[03:38:02] <Jymm> SWPadnos: Ok, seriously though, dont they have pre-made relay interface that are DIN mountable (just curious)?
[03:38:41] <SWPadnos> do you mean a reverse spike diode?
[03:38:41] <SWPadnos> or a buffer?
[03:38:44] <Jymm> buffer
[03:38:48] <SWPadnos> probably not
[03:39:05] <Jymm> Hmmm, seems strange that such a thing doens't exist.
[03:39:09] <SWPadnos> there are DIN-mountable holders for two-terminal devices, and possibly for 3-terminal devices, so you might be able to make one
[03:39:43] <SWPadnos> but it's probably assumed that anything in a control box that's connecting to a DIN-rail device has "industrial" i/o capability
[03:40:43] <Jymm> yeah, ok.
[03:40:48] <SWPadnos> the breakout board would consist of a connector, probably 3 ICs (due to >8 outputs and some inputs), a few discrete components (the most complex of which could possibly be a 5V regulator), and screw terminals
[03:41:14] <SWPadnos> it's not a hard board to design, once you know which bits need to be input and which output (or what you might want to be able to change around)
[03:41:45] <SWPadnos> and it may be that one of the el-cheapo boards from cnc4pc or whatever would be fine, once you know how to define "fine"
[03:42:55] <Jymm> SWPadnos: Ok, I'll look into the BOB's (again).
[03:43:08] <SWPadnos> wait until you know what you need ;)
[03:43:12] <Jymm> k
[03:51:17] <toastydeath> durrr
[04:24:57] <GNieport1> http://pastebin.ca/872985 Is this a suitable file to use for axis tuning?
[04:25:39] <SWPadnos> the first G01 line has 1.5 instead of 0.5
[04:25:54] <GNieport1> I had to add that to avoid a limit fault
[04:25:55] <SWPadnos> there should probably be dwells or pauses in there as well
[04:26:00] <SWPadnos> hmm. odd
[04:26:16] <GNieport1> the first incremental 0.5 would exceed 0.
[04:26:34] <SWPadnos> uh
[04:26:45] <SWPadnos> what are your limits?
[04:26:53] <GNieport1> machine high limit is 0, low limit is -32
[04:27:09] <SWPadnos> ok, I'm not sure that's a common configuration :)
[04:27:12] <GNieport1> remember we wore this out last night :)
[04:27:32] <SWPadnos> I don't remember that from last night - I may have been asleep or ignoring the conversation
[04:27:43] <SWPadnos> I sure hope I wasn't participating, because I have no recollection of it at all :)
[04:27:57] <GNieport1> something about ASME standard coordinates
[04:28:19] <SWPadnos> oh sure, roght hand rule and all - I remember a little of that (lathe-related though)
[04:28:25] <SWPadnos> right hand rule, that is
[04:29:54] <GNieport1> anyhow, I have the EMC dafault PID values, and got an oscillation superimposed on the path. Would it be unusual to reduce the P gain below the default?
[04:30:24] <SWPadnos> not really
[04:30:39] <GNieport1> by 'unusual' I mean 'Scared the crap out of me" :)
[04:30:46] <SWPadnos> I don't think a "good" default can be chosen, given the variety of scales and drives used
[04:30:49] <SWPadnos> heh
[04:31:07] <SWPadnos> are you starting with I, D, and all FF* at 0?
[04:31:12] <GNieport1> yes
[04:31:20] <SWPadnos> and also with all scales correct?
[04:32:00] <GNieport1> the scales are 0, which is what I need
[04:32:11] <SWPadnos> that shouldn't be possible
[04:32:12] <GNieport1> er
[04:32:17] <GNieport1> sorry, 40,000
[04:32:32] <GNieport1> I was thinking of output scaling
[04:32:33] <SWPadnos> ok, stepgen or some kind of servo?
[04:32:35] <SWPadnos> heh
[04:32:46] <GNieport1> servo, rotary encoders
[04:33:07] <SWPadnos> ok. you should also try to make the PWM output "sane
[04:33:10] <SWPadnos> "
[04:33:23] <SWPadnos> it's not strictly necessary
[04:33:27] <GNieport1> by sane, do you mean the ripple a low level?
[04:33:40] <GNieport1> or, limit the max
[04:33:48] <SWPadnos> no, I mean that there's some velocity you'd expect from 100% duty cycle
[04:33:56] <SWPadnos> you should use that as the scale, in user units
[04:34:01] <SWPadnos> what kind of amps?
[04:34:16] <GNieport1> Danaher IDC
[04:34:31] <SWPadnos> analog, PWM, step/dir ... ? :)
[04:35:23] <GNieport1> analog input, 1200 Watt
[04:35:36] <SWPadnos> ok, and velocity or torque mode?
[04:35:40] <GNieport1> V
[04:35:52] <SWPadnos> ok - so 10V out means some number of inches/second
[04:35:57] <GNieport1> yes.
[04:36:08] <fenn> GNieport1: iirc toast had it backwards, and the origin is in the front left corner (all positive coordinates)
[04:36:18] <GNieport1> d'oh
[04:36:35] <toastydeath> yeah, cradek pointed it out
[04:37:02] <toastydeath> i thought we caught you in time but apparently not
[04:39:22] <GNieport1> okay, I fixed it.
[04:39:38] <GNieport1> I misunderstood; i was still in the room
[04:40:49] <GNieport1> so facing the mill, when x table moves left, coordinates decrement?
[04:41:11] <SWPadnos> no
[04:41:26] <SWPadnos> imagine a piece of graph paper on the table, and a pen in the spindle
[04:41:32] <GNieport1> spindle over front left corner is 0,0
[04:41:51] <GNieport1> got it
[04:41:53] <SWPadnos> yes, so X increases as the table moves to the left
[04:42:06] <SWPadnos> (the spindle moves right)
[04:42:07] <GNieport1> sorry for the diversion
[04:42:12] <SWPadnos> no problem
[04:42:25] <GNieport1> I'll try the part program again at P of 50 from 300
[04:42:36] <SWPadnos> one sec
[04:43:07] <SWPadnos> the units thing - it probably doesn't matter, but it would make some things easier to understand when troubleshooting
[04:43:20] <klick> I tried to use M6 to change tools, well basically i wanted it to pause the program to change it manually myself, but I had a line of "T2 M6" and it didn't seem to stop there, it just flew past it, how can I accomplish this?
[04:43:23] <GNieport1> no, I want this to be as 'standard' as possible
[04:43:39] <SWPadnos> GNieport1, ok - hold on one sec
[04:43:58] <klick> I though about using M0 to pause, but you can't move around in pause mode
[04:43:59] <SWPadnos> klick, you can use the hal_manualtoolchange program. it's in one of the sample configs (I don't remember wwhich)
[04:44:19] <klick> SWPadnos: ok i'll look there, i wondered if it had to do with my config
[04:44:32] <SWPadnos> for most configs, the toolchange request output is connected to the "tool changed" input in the HAL files
[04:45:05] <SWPadnos> you can stick the userspace "change the tool" dialog in there though, which will hold until you press the OK button
[04:45:20] <SWPadnos> I don't know if you can jog around at that point though, as the program is still running
[04:45:26] <jmkasunich> you can't
[04:45:40] <jmkasunich> moving the machine manually isn't possible while a program is running
[04:45:53] <SWPadnos> EMC will move the spindle to the toolchange position though, if it's defined
[04:46:05] <SWPadnos> (right?)
[04:46:11] <jmkasunich> yes
[04:46:23] <SWPadnos> ok, thought so
[04:46:26] <klick> why not when it's paused?
[04:46:37] <jmkasunich> because thats how it works
[04:46:41] <SWPadnos> because it's not paused, it's waiting for the toolchanger to finish
[04:46:46] <klick> ok, so i should just split the program into 2 programs
[04:46:51] <jmkasunich> you can't move while paused either
[04:46:52] <klick> because my Z level is going to change
[04:47:08] <SWPadnos> are you using tool length compensation?
[04:47:10] <jmkasunich> if you need to touch off the new tool, yes, thats probably the best way to do it
[04:47:21] <klick> no
[04:47:38] <jmkasunich> if your tools are repeatable in Z when you install them, then tool length comp can make it simpler
[04:47:46] <SWPadnos> there's a macro for setting tool length, if you have a sensor of some sort
[04:47:49] <klick> they aren't repeatable
[04:48:02] <jmkasunich> then two programs is probably best
[04:48:06] <klick> ok :)
[04:48:06] <SWPadnos> I'm not sure how to use it - I think you need to put the macro in the program and call it after each tool change
[04:48:13] <SWPadnos> yes, multiple programs are probably best :)
[04:48:23] <jmkasunich> SWPadnos: since its useless without a sensor, lets not go there
[04:48:23] <fenn> klick: look at skeleton.ngc in /usr/share/emc/nc_files
[04:48:40] <fenn> i think that's where it goes at least
[04:48:43] <SWPadnos> yep - multiple files is the simplest solution
[04:49:22] <fenn> ooh nevermind (reads back)
[04:49:39] <klick> can you start execution of gcode from a certain line?
[04:50:02] <fenn> yes but there's differing opinion on what state the machine should be in
[04:50:08] <SWPadnos> GNieport1, I'm not sure if you need to account for the fact that the DAC output goes to 10V, but you should set the scale so an input of 1 to the DAC gives you 1 inch/second on the motor
[04:50:16] <klick> yea, i wondered about that
[04:50:35] <jmkasunich> klick: run-from-line is one of the less robust features of EMC, I personally avoid it
[04:50:37] <fenn> right now it's sorta useless
[04:50:55] <fenn> since you cant turn the spindle on and start from a line and have it stay on
[04:51:09] <klick> i turn the spindle on with the switch
[04:51:14] <fenn> that works then
[04:51:26] <klick> mine's nothing complicated, homemade eheh
[04:51:36] <GNieport1> SWP: is that scaling covered in the docs?
[04:51:51] <SWPadnos> I hope it's in some example somewhere :)
[04:52:24] <SWPadnos> what's the top speed of your motor (given 10V command voltage)?
[04:52:33] <GNieport1> well, the hardware is capable of much better than 10IPS
[04:52:39] <SWPadnos> heh
[04:52:47] <GNieport1> I set it to max out at 1600 rpm
[04:52:54] <SWPadnos> you can limit it also, but the scaling can still be set correctly
[04:52:57] <GNieport1> which is 300 IPM
[04:53:08] <SWPadnos> ok - so 10V = 300 IPM = 5 IPS
[04:53:26] <SWPadnos> or is 10V still higher speed, but the MAX_OUTPUT limits it to 300 IPM?
[04:53:37] <GNieport1> So I need a 0.2 scale factor?
[04:53:54] <SWPadnos> well, that's where I'm not sure - I think some drivers use the reciprocal and others don't (but I could be wrong)
[04:54:10] <GNieport1> 10V actual output is 1600 rpm. 5 turn/inch screws
[04:54:30] <SWPadnos> ok, that's 320 IPM
[04:54:39] <GNieport1> aye
[04:54:53] <SWPadnos> well now I can't do the math in my head
[04:54:55] <SWPadnos> :)
[04:55:02] <GNieport1> that's fine :)
[04:55:23] <GNieport1> do you know offhand the title of the scale I will be adjusting?
[04:55:33] <SWPadnos> what is your DAC/PWM card?
[04:55:39] <GNieport1> OFFSET_SCALE, etc
[04:55:42] <SWPadnos> no
[04:55:50] <GNieport1> Mesa 5i20
[04:55:52] <SWPadnos> oh, OUTPUT_SCALE probably
[04:56:13] <SWPadnos> it should be setting the parameter m5i20.0.dac-nn-scale, I think
[04:56:18] <SWPadnos> lemme check (or you can :) )
[04:56:19] <GNieport1> right.
[04:56:24] <GNieport1> no, that is it
[04:56:42] <SWPadnos> ok - so whatever ini setting goes to that (probably OUTPUT_SCALE)
[04:56:55] <GNieport1> so the standard is the DAC value 1 = 1 IPS
[04:57:13] <SWPadnos> the default is to not assume anything about the setup ;)
[04:57:43] <SWPadnos> ah - it's called gain
[04:57:53] <GNieport1> Why does the scale matter if I have position feedback?
[04:58:19] <GNieport1> To get the ballpark figure?
[04:58:33] <SWPadnos> ok, you do need 0.2-ish, the value gets multiplied
[04:58:42] <SWPadnos> no, it's more for "correctness" :)
[04:58:51] <GNieport1> btw P of 50 is much better, smooth all the way up to 150% feed
[04:58:55] <SWPadnos> when you look at the PID output, the number is in inches/second
[04:59:08] <SWPadnos> take the derivative, and it's a usable unit - inches/second^2 (accel)
[04:59:19] <GNieport1> i see
[04:59:30] <SWPadnos> if you use a scale of 1, then there's all sorts of scaling you have to do when troubleshooting, for no good reason
[04:59:39] <GNieport1> aha
[05:00:16] <SWPadnos> also, when you set limits (like max pid output), it's in user units
[05:00:34] <fenn> hmm why is the dac output modifier called gain instead of scale?
[05:00:41] <SWPadnos> dunno
[05:00:52] <SWPadnos> gain should be on an ADC I'd think, but whatever
[05:01:28] <SWPadnos> it's possible that gain is multiplied (so you need to use the reciprocal as the setting), but scales are divided (or the reciprocal is done internally)
[05:01:41] <SWPadnos> that may not be a pattern, but I have seen both I think
[05:01:56] <GNieport1> Well, technically isn't gain just the same thing mathematically as a scale
[05:02:12] <fenn> gain to me has to do with signal strengths and voltages and
[05:02:14] <SWPadnos> technically, it's whatever we do with it ;)
[05:02:24] <fenn> but scale is about numbers
[05:02:35] <SWPadnos> well, these are getting converted to voltages, so ...
[05:02:41] <GNieport1> but P gain is input times factor
[05:02:48] <SWPadnos> anyway, as long as it's consistent it should be OK
[05:02:54] <GNieport1> :)
[05:03:03] <SWPadnos> heh
[05:03:07] <fenn> right but PID was originally implemented as an analog circuit
[05:03:40] <GNieport1> With amplifiers, that have mathematical gain factors
[05:03:42] <fenn> i would call it a calibration constant i guess
[05:03:43] <GNieport1> I win lol
[05:03:55] <fenn> but that's a mouthful
[05:04:23] <SWPadnos> so anyway, you need to set the gain to around 0.2 :)
[05:04:34] <GNieport1> fenn, I agree it is a calibration constant, too
[05:04:41] <GNieport1> 0.2. right.
[05:04:59] <SWPadnos> 0.2-ish - actually 6/320
[05:05:05] <SWPadnos> err - 60/320
[05:05:24] <SWPadnos> .01875
[05:05:28] <SWPadnos> argh
[05:05:29] <GNieport1> will EMC solve that eqn for me
[05:05:31] <SWPadnos> 0.1875
[05:05:33] <SWPadnos> no
[05:05:35] <SWPadnos> not yet
[05:05:42] <GNieport1> I was teasing
[05:05:51] <GNieport1> lets not get it bloaty
[05:06:07] <SWPadnos> we've discussed it off and on over the years :)
[05:06:14] <fenn> so, emc outputs values in in/s and dac value is in rev/min?
[05:06:26] <fenn> * fenn feels slow today
[05:06:37] <GNieport1> i have set up servo drives that would solve the gearing equation to the max precision of the drive
[05:06:43] <GNieport1> n/m
[05:06:52] <SWPadnos> the DAC output was calculated using the IPM rate for a given RPM
[05:07:17] <SWPadnos> so it's in IPS (due to using seconds internally)
[05:07:22] <fenn> 5 rev / inch right?
[05:07:28] <GNieport1> yep
[05:07:35] <fenn> what's it gotta do with DAC?
[05:07:47] <SWPadnos> 10V = 1600 RPM
[05:07:52] <GNieport1> DAC scales directly to volts
[05:08:00] <SWPadnos> 1600 RPM / 5 revs/inch = 320 IPM
[05:08:37] <SWPadnos> oh - you want 320 IPM (= 5.333 IPS) to be 10V
[05:08:40] <SWPadnos> so the gain is 1.875
[05:08:49] <fenn> ugh there is throbbing bass notes and i cant think
[05:09:22] <fenn> * fenn moves to a less resonant chamber
[05:09:25] <SWPadnos> heh
[05:10:58] <GNieport1> OUTPUT_SCALE is what I want to modify.
[05:11:10] <SWPadnos> yep
[05:11:16] <SWPadnos> I think you may want 1.875
[05:11:18] <GNieport1> It is read by m5i20.0.enc-0X-gain
[05:11:24] <klick> could the hal_manualtoolchange thing be modified to allow for Z movement up and down, while it's in M6 state, which i'm assuming is somewhat of a pause... Not sure, could that be extended to basically have some kind of "jogging"
[05:11:37] <SWPadnos> hmmm - I need to sleep, so work through the math yourself :)
[05:11:58] <SWPadnos> klick, it is not paused when waiting for a toolchange
[05:12:05] <SWPadnos> the program is still running, but in a state where no motion can occur
[05:12:10] <klick> oh
[05:12:20] <SWPadnos> it's just waiting for a toolchange
[05:12:38] <SWPadnos> (think of it like a long dwell command)
[05:12:43] <klick> I thought someone mentioned that you could use a macro to change the zero level height, if you had a sensor or something, that would involve movement
[05:12:54] <SWPadnos> yes, but it happens after the tool is changed
[05:13:02] <klick> but how is that done?
[05:13:09] <klick> it's done while the program is executing is it not
[05:13:14] <SWPadnos> a toolchange would be TnnM6 then a separate line that calls the measuring macro
[05:13:17] <SWPadnos> yes
[05:13:54] <klick> is hal_manualtoolchange considered a macro?
[05:14:12] <SWPadnos> the macro may actually be a move to some position, then toolchange, then measure, then move back to where it was
[05:14:31] <SWPadnos> no, it's a program that has HAL pins, which you would connect to the tool change hooks in the motion controller
[05:14:40] <SWPadnos> when I say macro, I mean G-code subroutine
[05:14:47] <klick> oh, ok
[05:15:26] <fenn> emc moves to a certain position when doing a tool change (TOOL_CHANGE_POSITION i think)
[05:15:43] <SWPadnos> http://cvs.linuxcnc.org/cvs/emc2/nc_files/tool-length-probe.ngc?rev=1.4
[05:16:01] <klick> the reason i want some level of manual control is different tools are differnet lengths, and i cant just move to Z10.0, cause i usually set Z to where the tool would cut all the way through...
[05:16:29] <SWPadnos> sure - you need a "safe zone" for tool changes and length measurements
[05:16:45] <klick> yea
[05:16:55] <klick> manual control is my "safe zone" :) eheh
[05:16:58] <fenn> [EMCIO] TOOL_CHANGE_POSITION = 0 0 2
[05:17:24] <klick> fenn: my point is though that won't work, my Z will change, i can't set a Z value there that would be worthwhile
[05:17:31] <fenn> why not?
[05:17:55] <fenn> set the Z value to maximum
[05:18:02] <klick> cause some of my bits are 4 inches long, while others are 1/2 inch long, if i switch between the 2, a high Z value on the 4 inch long one will move it up beyond the top
[05:18:07] <SWPadnos> because the program is still running during a tool change, and you can't move the machine
[05:18:25] <fenn> um, turn off tool length offsets when changing tools then?
[05:18:41] <fenn> actually no, it does that already
[05:18:42] <klick> see, perhaps it's because i don't setup my coordinates how i should, i dunno. I just set Z to basically where I would cut all the way through the material
[05:19:01] <klick> i don't use tool length offsets, perhaps i should just read about that, and maybe that's my solution
[05:19:04] <SWPadnos> usually Z=0 is at the top of the part, and Z=-xxx is the table
[05:19:13] <klick> yea, see i don't do that :) eheh
[05:19:17] <SWPadnos> or Z=0 is the "top of the world"
[05:19:20] <SWPadnos> heh
[05:20:02] <klick> really, i guess i read all the wrong example files
[05:20:10] <klick> i never see that many negative Z's
[05:20:23] <klick> like skeleton.ngc is all positive Z's
[05:20:25] <klick> anyway, doesn't matter
[05:20:42] <fenn> z=0 should be at the bottom in a right handed coordinate system, but the part top surface could still be some positive value
[05:20:55] <klick> that's how i do it
[05:21:08] <SWPadnos> that link I posted shows what you need to do if you want to measure tool lengths after changing tools
[05:21:13] <klick> if my material is 1 inch thick, Z1.0 puts me on the top of the material
[05:21:27] <SWPadnos> it may be more complex than you want, and it's certainly harder than using separate files for each tool :)
[05:22:01] <klick> yea, well eventually i plan to have many possible tool changes, so I might as well figure it out now
[05:22:07] <SWPadnos> heh
[05:22:10] <SWPadnos> good plan :)
[05:22:17] <fenn> i'm confusing the issue i think
[05:22:28] <fenn> there's work coordinates and machine coordinates
[05:22:29] <klick> i'm probably confusing it hehe
[05:22:56] <fenn> work coords often have z=0 at the top of the raw stock
[05:23:45] <fenn> machine coordinates often have z=0 at the bottom of z travel (so g53 z0 would crash a tool into the table if tool offsets were off)
[05:25:33] <fenn> the numbers must describe the distance between two datum points eventually
[05:26:31] <fenn> you've got home offset, work offset, tool offset, g92 offset..
[05:29:13] <fenn> gosh the rs274 manual is hard to read
[05:31:00] <fenn> "To make the current point have the coordinates you want (without motion), program G92 X- Y- Z- A- B- C- , where the axis words contain the axis numbers you want. All axis words are optional, except that at least one must be used. If an axis word is not used for a given axis, the coordinate on that axis of the current point is not changed. It is an error if:
[05:31:05] <fenn> all axis words are omitted.
[05:31:27] <fenn> is it just me?
[05:31:32] <SWPadnos> it's you
[05:33:05] <klick> but can you issue a G92 command while in pause mode, well without variables and macros, well.. hmm
[05:33:22] <klick> or in tool change mode
[05:33:34] <SWPadnos> MDI and run mode are separate, and can't be combined
[05:33:41] <fenn> (unfortunately)
[05:33:43] <SWPadnos> pause is a state in run mode
[05:34:00] <klick> yea
[05:34:22] <klick> why can't you magically add a new line to interpret after pause mode picks back up? :)
[05:34:42] <SWPadnos> you can stop, then select the run-from line, and then run-from-line
[05:34:55] <SWPadnos> and while stopped you can jog and change tools all you want
[05:35:19] <klick> yea, but someone mentioned a state problem with that, what state elements would you lose?
[05:35:21] <SWPadnos> there are issues with the spindle turning off every time you change modes, and there may be others as well
[05:35:29] <SWPadnos> spindle for one ;)
[05:35:31] <klick> granted, i don't use "spindle on", "coolant on"
[05:35:57] <SWPadnos> most other modes are kept, but I'm sure there are some that aren't (probably the ones you want)
[05:36:02] <klick> what about tool offsets, G17 ?
[05:36:23] <klick> i use the simpilest of everything
[05:36:50] <SWPadnos> I don't know about offsets, planes are probably retained
[05:36:54] <klick> G17 G20 G40
[05:36:54] <klick> G64 P0.001
[05:37:02] <klick> i do those 2 commands in the beginning of every program
[05:37:04] <SWPadnos> you could try it and see
[05:37:11] <klick> then it's all G00 and G01 moves
[05:37:13] <klick> that's IT
[05:37:26] <klick> yea, that's true, trying would probably be easier hehe
[05:37:31] <SWPadnos> hmmml. I think the optional stop function works
[05:37:32] <klick> it's downstairs, and i'm upstairs
[05:37:38] <klick> M1 ?
[05:37:48] <SWPadnos> could be :)
[05:38:04] <SWPadnos> you can set the optional stop input true (or wire it to a true constant) in HAL
[05:38:12] <SWPadnos> then issue M1 for tool changes
[05:38:17] <klick> i did notice that you can turn that on and off
[05:38:25] <klick> yea, but that's no different then M0 right?
[05:38:43] <SWPadnos> it will always stop, and the operator can change tools, set offsets (if that works) fiddle with coordinate offsets, etc
[05:38:53] <klick> well, i'm gonna do an M0, stop the program, change my tool, then try the "run from line" which i don't know how to do, but hopefuly will figure out, i use Axis
[05:38:55] <SWPadnos> hmmm - sure, you could use M0 (if that's "stop")
[05:39:11] <klick> M0 is pause
[05:39:13] <SWPadnos> it may be in the axis quick-ref (under help)
[05:39:17] <klick> well, i can pause, then hit the stop button
[05:39:20] <SWPadnos> ok, then M0 will *NOT* work
[05:39:33] <SWPadnos> M1 will, if you hard-wire the "optional stop" input to "yes"
[05:39:40] <klick> M2 is end
[05:39:42] <SWPadnos> I think that's a stop, not a pause
[05:40:00] <klick> M0 program stop
[05:40:00] <klick> M1 optional program stop
[05:40:00] <klick> M2 program end
[05:40:02] <SWPadnos> but I don't know G-code all that well, so I'd consult the hard-to-read manual rather than listen to me
[05:40:13] <SWPadnos> hmmm, then M1 may not do it either
[05:40:27] <klick> M0 makes axis have the "pause" button depressed
[05:40:32] <klick> and you can unclick it and it resumes
[05:40:35] <klick> or hit "S"
[05:41:05] <klick> where is the axis manual online, i can't even find the axis website
[05:41:21] <fenn> http://linuxcnc.org/docs/html/
[05:41:21] <klick> it's hard to search for axis linux emc, when it just shows you info about how to control axis's
[05:41:45] <SWPadnos> since AXIS is now included with EMC, I'm not sure how much "new" information is on the original axis site
[05:42:19] <klick> well that's more documentation that i've ever found before on it, nice, thank you
[05:42:23] <fenn> it seems to be more of a blog now than anything
[05:43:01] <fenn> btw it's axis.unpy.net
[05:43:28] <klick> although i see nothing about run from line
[05:44:48] <klick> axis is the preferred interface right?
[05:45:20] <fenn> yes, for the most part
[05:46:06] <Jymm> axis because it has the neat "see it as you go" inteface, but there is tkemc as an alternative, and another one I can't remember
[05:46:16] <Jymm> mini something or another
[05:46:19] <fenn> tkemc and mini both have backplots
[05:46:26] <fenn> but only axis has preview
[05:46:49] <Jymm> fenn: thats the wording I was looking for =)
[05:47:01] <Jymm> fenn: didn't know the others had backplot though
[05:47:02] <klick> oh yea, preview is nice
[05:47:10] <klick> i like having a preview
[05:47:17] <Jymm> KEYSTICK FTW!!! lol
[05:48:02] <fenn> keystick on an old monochome vfd would be cool
[05:48:02] <klick> did you just have an aneurysm
[05:48:21] <Jymm> fenn: those lil 5" crt's
[05:49:36] <fenn> or a TI-85
[05:49:42] <SkinnypuppY34> What ubuntu is the next EMC going to be worked into ? Hardy Heron is it ?
[05:50:22] <SkinnypuppY34> Or the next ubuntu rather
[05:50:30] <fenn> oh man my brain is really failing tonight.. whats the word for the little portable button panel on a cord
[05:50:53] <fenn> SkinnypuppY34: probably the next LTS (long term service)
[05:51:05] <fenn> i think its scheduled for the next release after gutsy gibbon
[05:51:18] <fenn> in summary: the next release
[05:51:22] <SWPadnos> 8.04 "Hardy Heron" is the next scheduled LTS release
[05:51:33] <fenn> ah H is after G isnt it
[05:51:52] <SWPadnos> in english, yes
[05:51:56] <fenn> i should go take a nap and stop bothering you all
[05:52:15] <SkinnypuppY34> I was just upgrading a gutsy Virtualmachine to 8.04 was why I ask. Thanks SWPadnos
[05:52:31] <SWPadnos> 8.04 alpha something>
[05:52:33] <SWPadnos> ?
[05:53:21] <SkinnypuppY34> Think alpha 3 , I was looking at the blueprint page. .https://blueprints.launchpad.net/ubuntu/hardy
[05:53:32] <fenn> pendant!
[06:12:24] <GNieport1> Halscope
[06:12:37] <GNieport1> what shall I probe to tune the PID?
[06:12:56] <GNieport1> I haven't found a combination that shows the classic step decay
[06:13:04] <GNieport1> to set P
[06:13:21] <GNieport1> or am I going down the wrong track
[06:15:08] <fenn> axis.n.pos-fb
[06:15:32] <fenn> or you can look at ferror
[06:16:26] <GNieport1> what is the easiest way to command a square wave? I have a gcode program loaded that jogs back and forth, but the accel is not aggressive enough to tune
[06:16:31] <GNieport1> with
[06:16:50] <fenn> sorry motor-pos-fb
[06:16:51] <SWPadnos> you could set accel to 5000 or something
[06:17:11] <GNieport1> is siggen a better command source
[06:17:12] <SWPadnos> and make sure ferror is about an inch if you're doing 1/2 inch moves
[06:17:18] <SWPadnos> yes, for square waves
[06:17:34] <fenn> siggen sends an actual square wave as opposed to some cubic spline shape
[06:18:35] <GNieport1> If i link siggen to the x command, will I be able to scope the error? I was able to make siggen jog the axis earlier, but had no way to measure the error
[06:19:02] <SWPadnos> scope the command and the feedback. trigger on the command edge(s)
[06:19:10] <GNieport1> k
[06:19:15] <GNieport1> thaks
[06:19:18] <fenn> axis.N.f-error is the error
[06:19:19] <GNieport1> thanks, too
[06:19:21] <SWPadnos> sure
[06:19:32] <SWPadnos> only if you're using motmod to do the motion
[06:19:38] <fenn> oh, uh, yeah
[06:19:38] <SWPadnos> with siggen, there are no axes
[06:19:43] <GNieport1> brb shop heater is trying to go nuclear =-O
[06:19:44] <fenn> hmm
[06:19:49] <SWPadnos> heh
[06:21:02] <GNieport1> how do i send the square wave through motmod?
[06:21:12] <SWPadnos> you don't
[06:21:26] <fenn> <half baked rant> emc should automatically link signals when certain modules are loaded, such as kins modules, and this would allow you to unlink them and hack what is going through motmod </rant>
[06:21:30] <GNieport1> without the motion controller, how will PID influence the output
[06:21:31] <SWPadnos> you disconnect axis.n.motor-pos-cmd from the PID, and connect siggen output in its place
[06:21:42] <SWPadnos> PID is a separate HAL component
[06:21:59] <GNieport1> O
[06:22:18] <GNieport1> great!
[06:22:41] <SWPadnos> in the mesa config dir, there's a mesa-pidtest.hal file. take a look at that
[06:23:06] <SWPadnos> the scales will be wrong, and some other stuff too, but it should be a good starting point
[06:23:09] <GNieport1> I couldn't get it to to work
[06:23:17] <SWPadnos> I don't know if it's meant to be run using halrun -I
[06:23:22] <GNieport1> hmm
[06:23:41] <SWPadnos> yep
[06:23:41] <GNieport1> possibly i could not get it to work due to my damaged 7i33 analog output card :)
[06:23:43] <SWPadnos> says so
[06:23:55] <SWPadnos> # Run it with:
[06:23:57] <SWPadnos> # halrun -I m5i20_pidtest.hal
[06:24:08] <SWPadnos> # Do not run this script while you are running emc!
[06:24:37] <GNieport1> blech, shop is smokey
[06:24:48] <SWPadnos> heh - kerosene heater? :)
[06:25:29] <GNieport1> used motor oil heater, burning off residues from the firebox exterior
[06:26:03] <GNieport1> best it has ever run :)
[06:30:39] <GNieport1> I believe the info headers in the m5i20_pidtest.hal are out of date. /scripts/realtime start does not work
[06:31:03] <GNieport1> unless I'm a weenie
[06:34:02] <SWPadnos> are you running from an installed EMC2, or a compiled one?
[06:34:57] <fenn> its $emc2_directory/scripts/realtime or /etc/init.d/realtime
[06:35:10] <GNieport1> yeah, I have to do /etc/init.d/realtime start
[06:35:29] <GNieport1> LiveCD install
[06:35:45] <GNieport1> Installed onto hard drive
[06:35:48] <SWPadnos> hmmm. I'd think that halrun would do that for you
[06:36:28] <GNieport1> RTAPI: ERROR: could not open shared memory (errno=2)
[06:36:28] <GNieport1> HAL: ERROR: rtapi init failed
[06:36:28] <GNieport1> halcmd: hal_init() failed: -9
[06:36:28] <GNieport1> NOTE: 'rtapi' kernel module must be loaded
[06:36:41] <GNieport1> nope :)
[06:37:15] <SWPadnos> hmmm. something is wrong then. halrun loads realtime when used on a run-in-place EMC
[06:37:21] <GNieport1> wait... halrun?
[06:37:30] <SWPadnos> it even has a -U option to unload realtime stuff
[06:37:35] <SWPadnos> yes
[06:37:39] <SWPadnos> halrun -I m5i20_pidtest.hal
[06:37:52] <GNieport1> oh, the infor header says use halcmd -f pid...
[06:38:11] <SWPadnos> um - no, I copied that from the version on the CVS web server
[06:38:11] <GNieport1> substitute halrun?
[06:38:38] <SWPadnos> http://cvs.linuxcnc.org/cvs/emc2/configs/m5i20/m5i20_pidtest.hal?rev=1.7
[06:38:45] <SWPadnos> make sure you have the same file
[06:38:54] <SWPadnos> which says to use halrun -I
[06:38:54] <GNieport1> 2.2.2 has an older INI file version than 2.1.7
[06:39:09] <SWPadnos> huh?
[06:39:18] <GNieport1> hey, I wiped the drive and installed the liveCD last week :)
[06:39:20] <SWPadnos> the ini file is unused for this
[06:39:48] <GNieport1> that was a general note about file version mixup
[06:39:49] <SWPadnos> this file hasn't changed during the 2.2 cycle
[06:40:01] <SWPadnos> what version are you looking at?
[06:40:13] <SWPadnos> I mean where is the number you saw?
[06:41:12] <GNieport1> # Version of this INI file
[06:41:12] <GNieport1> VERSION = $Revision: 1.13 $
[06:41:21] <GNieport1> in the sample_configs
[06:41:23] <GNieport1> m5i20
[06:41:28] <SWPadnos> that's the CVS internal version number, and can be safely ignored by almost everybody
[06:41:42] <GNieport1> emc 2.1.7 had version 1.6XXX
[06:41:48] <SWPadnos> 13 is higher than 6
[06:41:56] <GNieport1> crap
[06:41:59] <SWPadnos> heh
[06:42:10] <GNieport1> :-[
[06:42:49] <fenn> we all make that mistake originally
[06:43:21] <SWPadnos> yeah - and when numbers go over 99, people get really confused
[06:43:25] <GNieport1> the CVS pidtest script is different than the one I have :)
[06:43:29] <SWPadnos> "how can you have 2.103?"
[06:44:04] <fenn> SWPadnos: it's just a smidgeon of an upgrade, i only changed 0.14% of the code!
[06:44:29] <fenn> .003/2.1
[06:44:39] <SWPadnos> heh - if this is a config from 2.1.7, the last comment from that branch was something like "fix misleading comments ..." ;)
[06:44:50] <SWPadnos> fenn, yeah - lots of small changes ;)
[06:45:54] <fenn> fortunately we dont have to completely rewrite every line in order to use a new major number
[06:46:05] <SWPadnos> yay!
[06:46:14] <fenn> though maybe it's a good idea sometimes
[06:46:36] <fenn> i'm sure chris would disagree
[06:47:42] <fenn> spaceship two has a painting of richard branson's mom wearing a swimsuit and bubble helmet :0
[06:47:48] <SWPadnos> well, he did write the simple_tp
[06:47:57] <SWPadnos> ewwww
[06:48:08] <SWPadnos> unless it's from like 50 years ago
[06:48:10] <SWPadnos> :)
[06:48:37] <fenn> 2.5MB: http://www.virgingalactic.com/pressftp/content/Illustrations/In Colour/Galactic Girl/Galactic Girl.jpg
[06:49:23] <fenn> lets try that again: http://www.virgingalactic.com/pressftp/content/Illustrations/In%20Colour/Galactic%20Girl/Galactic%20Girl.jpg
[06:49:56] <SWPadnos> heh - I got it after a failed attempt
[06:50:07] <SWPadnos> cool logos actually
[06:55:17] <fenn> to be fair, the carrier plane is named after her, and that's what the space suits will look like (sorta)
[07:00:27] <SkinnypuppY34> Galacticgirl spacesuit kewl
[07:04:03] <fenn> i thought suborbital was just a gimmick until i started reading about skyhooks/tethers/rotovators
[07:04:42] <fenn> you just need to get to mach 12 or so at 100km, and then the orbiting rotating tether swoops down and swings you into orbit like a 1000km long trebuchet
[07:04:46] <GNieport1> I just did this; do you expect that I will still be able to probe error with halscope?
[07:04:48] <GNieport1> halcmd: linksp testsigone <= siggen.0.square
[07:04:48] <GNieport1> halcmd: unlinkp pid.0.command
[07:04:48] <GNieport1> halcmd: linksp testsigone => pid.0.command
[07:08:21] <SkinnypuppY34> Anyone know what I need to playback .flv's in totem? Mpg,mp4,and avi all work fine.
[07:08:22] <fenn> no, f-error only works if axis.0.motor-pos-cmd is hooked up
[07:08:46] <fenn> SkinnypuppY34: gnash? or flashplugin-nonfree
[07:09:00] <SWPadnos> you can use a sum block with one of the input scales set to -1
[07:09:08] <SWPadnos> and probe the output of that
[07:09:16] <fenn> there should be a hal math block :P
[07:09:38] <SWPadnos> or, just align the zero levels of the command and feedback traces (and set the scales the same)
[07:09:56] <GNieport1> I think I'll use axis.0.motor-pos-cmd
[07:10:38] <fenn> loadrt math pins="error,fb,command" equation="error=fb-command"
[07:11:00] <fenn> is that hopelessly naive?
[07:11:13] <SWPadnos> yes
[07:11:17] <GNieport1> lol
[07:11:31] <SWPadnos> you'd need to put an infix-to-postfix parser into a kernel module
[07:11:40] <SWPadnos> what happens if you wnt two equations?
[07:11:46] <SWPadnos> what happens if there's an error in one of them
[07:12:05] <SWPadnos> (or worse, you want 10 equations, and number 3 has an error
[07:12:06] <SWPadnos> )
[07:12:27] <fenn> well, mach 3 has it!! *ducks*
[07:12:42] <SWPadnos> except for the "loadrt" part ;)
[07:14:27] <fenn> ok then, whats an easy way to do what i want and compile it on the fly?
[07:14:43] <SWPadnos> well, I have no idea
[07:14:53] <fenn> rtlab is something like this right?
[07:15:01] <SWPadnos> I wouldn't mind having the functionality, mind you
[07:15:08] <SWPadnos> I don't think so
[07:15:43] <SWPadnos> there is probably a "math function" block, which may get JIT-compiled, but I'm not sure it's also RT / kernel code
[07:16:21] <fenn> ew what's this biological experiment stuff.. i thought it was MATLAB in realtime
[07:16:43] <SWPadnos> scilab?
[07:17:44] <SkinnypuppY34> hmm I installed flashplugin nonfree but totem is still complaining about a decoder on flvs
[07:18:15] <fenn> scicos-rtai maybe? it certainly doesnt have the pizzaz of 'rtlab'
[07:18:20] <SWPadnos> heh
[07:18:23] <SWPadnos> night night guys
[07:18:35] <SkinnypuppY34> night SWP
[07:21:56] <SkinnypuppY34> Think I'll do the same, thanks again fenn I'll try the flv later
[08:20:10] <cnc_engineer> Guide me, how can I add big DRO's in axis interface that can display relative / machine coordinates?
[08:29:58] <fenn> is this what you want? http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PyVCP
[08:30:26] <fenn> also there is a way to modify the size of coordinates displayed in the preview window with the X resources database
[08:30:54] <fenn> i can never find the docs for that though
[08:32:03] <cnc_engineer> I am interested in modifying the size of coordinates displayed in preview window
[08:32:24] <cnc_engineer> What is this X resources database?
[08:33:56] <fenn> it defines things like what color to use for the background, what fonts, etc
[08:34:13] <fenn> there is an option that tell axis what font to use for the coordinates
[08:34:45] <cnc_engineer> Please guide me to change that..
[08:35:08] <fenn> i am looking..
[08:47:15] <fenn> cnc_engineer: i can't find it
[08:47:43] <fenn> cnc_engineer: ask jepler when he's around
[08:57:53] <fenn> well. this does *something*: echo '*Axis*font: -adobe-courier-bold-r-normal--32-*-*-*-*-*-*-1' | xrdb ; emc
[09:04:04] <fenn> aha i have to hack axis to do it
[09:04:45] <fenn> cnc_engineer_: still there?
[09:06:06] <fenn> anyway, for google posterity, here's what i did; change coordinate_font = o.option_get("font", "Font") or "9x15"
[09:06:14] <fenn> to coordinate_font = o.option_get("font", "Coordinate_Font") or "9x15"
[09:06:33] <fenn> then do: echo '*Axis*Coordinate_Font: -adobe-courier-bold-r-normal--32-*-*-*-*-*-*-1' | xrdb ; emc
[09:13:24] <fenn> you can use appres to see what has been set
[09:14:05] <fenn> and change 32 to some other number for bigger or smaller font size
[09:17:45] <cnc_engineer_> fenn: Sorry I was disconnected. You were telling me how to cahnge the font size and background of "AXIS" interface using x resource database..
[09:17:58] <cnc_engineer_> fenn: Sorry I was disconnected. You were telling me how to change the font size and background of "AXIS" interface using x resource database..
[09:18:31] <fenn> http://fennetic.net/pub/irc/axis-dro.png
[09:18:48] <fenn> logger_emc: bookmark
[09:18:48] <fenn> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-01-26.txt
[09:22:53] <cnc_engineer_> Fenn: Thank you very much...
[10:01:50] <cnc_engineer_> fenn: Please explain me again, the procedure of changing the size of DRO in "axis". What I an doing is changing the code coordinate_font = o.option_get("font", "Coordinate_Font") or "9x15" in /usr/bin/axis file and
[10:02:41] <cnc_engineer_> fenn: then on command prompt I issue the command echo '*Axis*Coordinate_Font: -adobe-courier-bold-r-normal--32-*-*-*-*-*-*-1' | xrdb ; emc
[10:03:16] <cnc_engineer_> fenn: but it gives an error message
[10:06:13] <fenn> what's the error message?
[10:07:06] <fenn> try replacing adobe with *
[10:08:42] <cnc_engineer> Fenn: Sorry I am having serious problem with myy internet Connection
[10:24:33] <cnc_engineer> fenn: Did you guide me anything..
[10:28:22] <fenn> cnc_engineer: what's the error message?
[10:30:44] <micges> logger_emc: bookmark
[10:30:44] <micges> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-01-26.txt
[10:32:44] <fenn> well, i am going to bed, good luck
[14:28:15] <jensor> Hello. I need help to enable a new ethernet card in my emc computer - any available?
[14:29:45] <jmkasunich> this isn't a general linux forum - your best be is to google it
[14:34:05] <jensor> John: I am not using general linux. If I go elsewhere I am afraid of lousing up my specialized kernal that has been tailored to EMC.
[14:34:36] <jmkasunich> unless you have one VERY strange ethernet card, the kernel is not involved
[14:34:45] <jmkasunich> you are running Ubuntu, right?
[14:38:00] <jmkasunich> hello? is this thing on?
[14:38:11] <BigJohnT> yes
[14:38:30] <jmkasunich> I was wondering what happened to jensor - I asked him a simple yes/no question
[14:42:52] <jmkasunich> maybe he's googling.... I hope so
[14:43:06] <jlmjvm> jmkasunich:have you been working on the screw mapping for emc
[14:44:34] <jmkasunich> "working" as in lately, no. "worked", yes - to the best of my knowledge its been working fine for a year or more
[14:44:47] <jmkasunich> I more recently did the actual mapping for my own machine
[14:45:03] <BigJohnT> what is "screw mapping"?
[14:45:12] <jmkasunich> I haven't touched the code in a long time tho
[14:45:38] <jmkasunich> BigJohnT: you can prepare a file that tells EMC what your screw error is at different places on the screw
[14:45:42] <jmkasunich> and it will compensate
[14:45:46] <BigJohnT> ok
[14:45:50] <jlmjvm> is there any info on how to use it,or a sample file perhaps
[14:45:50] <BigJohnT> cool
[14:47:45] <jmkasunich> http://www.linuxcnc.org/docview/html//config_ini_config.html#sub:%5BAXIS%5D-Section
[14:47:57] <jmkasunich> see the part that says "COMP_FILE="
[14:49:11] <jlmjvm> yep
[14:50:02] <jlmjvm> COMP_FILE = file.extension
[14:50:15] <jmkasunich> that is the documentation for screw comp
[14:50:24] <jmkasunich> kinda sparse, sorry
[14:51:14] <jlmjvm> could that be, COMP_FILE =joe.comp,if thats what i wanted to call it?
[14:51:19] <jmkasunich> yes
[14:51:40] <jlmjvm> cool,just making sure,still a little green
[14:51:46] <jmkasunich> I use names like xscrew.comp, yscrew.comp, etc
[14:51:48] <BigJohnT> is the seperator for the values in a triplet a space?
[14:52:01] <jmkasunich> yes
[14:52:10] <jmkasunich> I think, lemme check
[14:52:16] <BigJohnT> ok
[14:52:21] <BigJohnT> lol
[14:52:35] <jmkasunich> I used tabs, but space or tab is OK
[14:52:39] <jlmjvm> could you pastebin a sample .comp file for us to use as a guide
[14:53:37] <jmkasunich> my file is long, because I attempted (not very successfully) to use it to fix a periodic error
[14:53:50] <jmkasunich> typically you'd measure the error every inch or so
[14:53:57] <alex_joni> if (3 != sscanf(buffer, "%lf %lf %lf", &nom, &fwd, &rev)) {
[14:54:23] <jmkasunich> say your axis runs from -6" to +6", and you measured every inch
[14:54:36] <alex_joni> /* expecting nominal-forward-reverse triplets, e.g.,
[14:54:36] <alex_joni> 0.000000 0.000000 -0.001279
[14:54:36] <alex_joni> 0.100000 0.098742 0.051632
[14:54:37] <alex_joni> 0.200000 0.171529 0.194216 */
[14:55:28] <jlmjvm> brb
[14:55:31] <jensor> jmkasunich:maybe I am too paranoid, but when I looked I see instructions like the following it scares me into thinking that the kernal is somehow involved
[14:55:33] <jmkasunich> I prefer type 1, which is "nominal forward_error reverse_error", but that is up to you
[14:55:49] <alex_joni> heh
[14:55:57] <alex_joni> there goes the pasting
[14:56:06] <alex_joni> jensor: use pastebin.ca for longer texts to paste
[14:56:15] <jmkasunich> jensor: you can't dump lots of stuff into the channel, it will kick you out
[14:56:16] <alex_joni> the paste only the URL in here
[14:56:45] <jmkasunich> I asked you a simple yes-no question 21 minutes ago
[14:56:52] <jmkasunich> are you or are you not using Ubuntu?
[14:56:58] <jensor> yes
[14:57:20] <jmkasunich> you are using "our" version, with the RT kernel, right?
[14:57:35] <jensor> yes
[14:58:05] <jensor> how do I paste to pastebin.ca?
[14:58:17] <archivist> then swithcoff plug in network card and reboot, test
[14:58:24] <jmkasunich> in _almost_ every way, "our" ubuntu IS standard ubuntu. 99.9% of non-emc questions will have the same answers whether you are using our ubuntu or "standard" ubuntu
[14:59:07] <jensor> Okay I'll go reboot and see what happens
[14:59:11] <jmkasunich> hang on a sec
[14:59:33] <jmkasunich> what is the problem? you never even told us that
[14:59:56] <jmkasunich> you wrote: "Hello. I need help to enable a new ethernet card in my emc computer - any available?"
[15:00:06] <alex_joni> jmkasunich: maybe he didn't plug it in yet?
[15:00:30] <alex_joni> http://www.trilobyte.ch/temp/MOV05502.MPG
[15:01:45] <jensor> I am rebooting th emc machine now
[15:04:35] <jensor> the router light doesn'tdoesn't respond
[15:05:17] <jensor> whati s the signifcance of the url trilobyte etc?
[15:05:34] <jmkasunich> its a movie of somebody using emc to run an axis
[15:05:50] <jmkasunich> has nothing to do with you, just something interesting that alex posted for people in the channel to se
[15:05:58] <archivist> jensor is the router set to dhcp and giving your ubuntu an ip address
[15:06:20] <jensor> I'll check its in the opther room
[15:07:34] <jmkasunich> archivist: if his router works with another PC (probably the one he's talking to us from, which is probably windows), then its probably fine
[15:07:55] <alex_joni> and if the light doesn't light up on the router it's probably a faulty cable
[15:08:00] <archivist> probably he just needs to try something
[15:08:11] <alex_joni> the light goes on even if the card isn't configured/driver loaded
[15:08:25] <jensor> Scuse me the router is right here and this is a new setup a first for me and I don't really know how to detrermine if is is set to dhcp
[15:08:40] <jmkasunich> jensor: don't worry about router settings yet
[15:08:41] <archivist> normally automatic jensor
[15:08:52] <jmkasunich> first just answer questions
[15:09:08] <jensor> The router is communicating find with the windows box that I am now on
[15:09:16] <jmkasunich> ok
[15:09:23] <jmkasunich> that means the router is almost certainly fine
[15:10:01] <jmkasunich> if you plug in the cable from the Linux PC, and the router light for that port doesn't come on almost immediately, I'd suspect the cable
[15:10:25] <jmkasunich> even if the driver for the network card is completely missing, the light should come on as soon and the cables are connected
[15:10:36] <jensor> oh
[15:11:02] <jmkasunich> do you have another cable?
[15:11:18] <alex_joni> maybe you can try with the cable from the windows pc for a couple of seconds
[15:11:31] <jmkasunich> I'd suggest using the known good cable from the PC you are talking on, but that will break your net connection and probably drop you out of the channel
[15:12:07] <jensor> It will take some time to continuity test the cable but I suppose I'll need to do that
[15:12:21] <jmkasunich> continuity test isn't very good
[15:12:33] <jmkasunich> timeconsuming, and not always valid
[15:12:37] <jmkasunich> better to seap
[15:13:05] <jmkasunich> put the suspect cable on the working PC, and the working cable on the suspect PC, note which light(s) come on, and draw the appropriate conclusions
[15:13:11] <archivist> just use the ubuntu cable on the windows box to test it
[15:13:12] <cradek> if your hub/router has a special uplink port be sure you're not plugged into it
[15:13:45] <jmkasunich> oops - not "seap", I meant "swap"
[15:13:45] <jensor> another thing I could try is to bring the emc box over here next to the router and use the short known good cable
[15:13:57] <jmkasunich> yes
[15:15:01] <jmkasunich> I have some chores I have to do - back in an hour or two
[15:26:31] <fretless85> hey guys
[15:27:00] <alex_joni> hi
[15:27:56] <cradek> fretless85: I think you need a smaller mill. we could trade if you want. you pay shipping though
[15:29:31] <archivist> hehe
[15:30:38] <fretless85> cheers! (i got a beer right next to me ;) )
[15:30:48] <alex_joni> what's it doing there?
[15:30:55] <alex_joni> seems such a waste
[15:31:12] <fretless85> im going to drink it, i guess...
[15:31:29] <cradek> ha
[15:31:35] <cradek> too early for beer here, coffee is better
[15:32:01] <archivist> Ive just sat down for 7up time
[15:32:36] <fretless85> its never to early for a beer ;)
[15:32:54] <fretless85> cuz im german ;) lol
[15:32:56] <archivist> milling table fitted this morning, rotary next
[15:33:00] <BigJohnT> I have 25 gallons on tap
[15:33:53] <Gamma-X> i hope SWPadnos can allow me to call him hahaha.
[15:34:02] <fretless85> hmmm yummy! diebels ;)
[15:34:17] <BigJohnT> Gamma you make it move yet?
[15:34:26] <archivist> cradek I may have another manifestation of the joint 3 following but 2.2.3 may fix that as well
[15:34:29] <Gamma-X> BigJohnT yes!
[15:34:38] <BigJohnT> sweet
[15:34:54] <cradek> archivist: I bet it will
[15:35:36] <cradek> archivist: did you have linear (xyz) G91 moves before a G90 rotary move?
[15:36:00] <Gamma-X> BigJohnT just gotta tune it.... wich im finaidng impossible att
[15:36:33] <archivist> cradek no not done a g91
[15:36:35] <cradek> Gamma-X: is it moving?
[15:37:35] <Gamma-X> cradek yes sir
[15:37:56] <alex_joni> fretless85: I like Koelsch better :P
[15:37:57] <jensor69> Hey - just moved the emc box over next to the router and used the known good ethernet cable about 4 feet long and it worked fine. Therefor I need to find out what is wrong with the long cable -about 30 feet long
[15:38:14] <alex_joni> jensor69: did you mount the connecters?
[15:38:27] <cradek> excellent
[15:38:34] <jensor69> yes
[15:38:38] <archivist> or using cat5 not cat5e
[15:38:50] <jensor69> not sure one or the other
[15:39:08] <archivist> it should be cat5e for 100mb
[15:39:13] <jensor69> I think its cat 5
[15:39:21] <cradek> cat5 is fine
[15:39:35] <jlmjvm> alex_joni:should my x position show "nan" in the display when using a comp file?
[15:39:40] <cradek> it's very easy to get the ends on wrong :-)
[15:39:43] <Gamma-X> whats the differance its all copper...
[15:39:44] <jensor69> the cable is marked 5e
[15:39:52] <archivist> thats ok
[15:40:03] <BigJohnT> Koelsch is my favorite
[15:40:17] <alex_joni> jlmjvm: nope
[15:40:22] <jensor69> the ends match the known good cable
[15:40:51] <archivist> check continuity next then
[15:41:05] <jensor69> the unknown is the connectors how does one tell if they are for strranded or single conductor
[15:41:24] <fretless85> kölsch taste like water
[15:41:48] <cradek> I think cat5 is never stranded
[15:42:25] <cradek> well if it is I haven't seen it (I've made a lot of cables)
[15:42:30] <BigJohnT> I also like hefeweizen and witbier
[15:42:32] <jensor69> mine is stranded , but I don;t know about the connectors I used
[15:43:16] <jlmjvm> 0.0000000.0000000.000000
[15:43:16] <jlmjvm> 1.0000000.9900001.010000
[15:43:16] <jlmjvm> 2.0000001.9800002.020000
[15:43:34] <jlmjvm> thats what i have for a test file
[15:43:49] <fretless85> hefeweizen is okay but what the heck is witbier?
[15:44:42] <fretless85> you mean weißbier?
[15:46:23] <Gamma-X> hey anyone with a workin machine with servos, can u give me an example of what ur pid looks like?
[15:46:31] <BigJohnT> its pale malt, 30% wheat malt, 10% flaked oats with Hallertau and Saaz hops and bitter orange peel and corriander seed added with 10 minutes left in the boil
[15:47:02] <cradek> Gamma-X: you can't copy someone else's tuning and have it work
[15:47:04] <Gamma-X> im awaear of this
[15:47:06] <Gamma-X> I didnt say i wanted to do that hahaha
[15:47:13] <cradek> P,I are higher, D is pretty small
[15:47:24] <fretless85> sounds interesting bigjohn
[15:47:36] <Gamma-X> so would p most likely be at 100?
[15:47:52] <BigJohnT> it is nice
[15:47:52] <cradek> that depends on the scaling of your pid output
[15:47:59] <Gamma-X> starting values, p100 I 75 d 10?
[15:48:12] <BigJohnT> I normally make 10 gallon batches
[15:48:14] <cradek> starting values everything 0 except P
[15:48:32] <cradek> carefully increase P to barely oscillate
[15:48:43] <cradek> damp oscillation with D
[15:48:46] <Gamma-X> well p was at 100 to begin with
[15:49:03] <fretless85> but i guess i stay with my "altbier"
[15:49:05] <Gamma-X> im assuming that the highest value in pid is 100
[15:49:05] <alex_joni> jlmjvm: what's your travel?
[15:49:15] <alex_joni> Gamma-X: nope, you can have 10000 if you want
[15:49:25] <cradek> they are not percentages
[15:49:26] <Gamma-X> damn.
[15:49:42] <Gamma-X> whats that number actualy reffering to. not a voltage right?
[15:49:47] <cradek> you will have to read about pid and understand what each number means!
[15:49:55] <cradek> you cannot tune unless you know what the numbers do
[15:50:13] <Gamma-X> cradek i do understand just not what the number do... in terms of if its not percent what is it?
[15:50:32] <alex_joni> Gamma-X: do you know what P is?
[15:50:32] <cradek> once you know you will know what to try changing when you see a certain behavior in halscope
[15:50:45] <cradek> Gamma-X: read about pid on the web, you will probably have to venture away from linuxcnc stuff to find it
[15:51:57] <BigJohnT> Gamma start on page 27 of the integrater manual and study that
[15:52:14] <jlmjvm> alex_joni:0 to -17.5 inches
[15:52:49] <Gamma-X> ok
[16:22:08] <alex_joni> jlmjvm: so you have only negative travel
[16:22:20] <alex_joni> how come your comp table has compensation values for positive travel then?
[16:23:41] <jlmjvm> because it was my first try
[16:24:09] <jlmjvm> if i change values to negative it gives alarm
[16:25:05] <jlmjvm> says compensation values must increase
[16:25:44] <alex_joni> indeed
[16:25:54] <alex_joni> you need to start from your minimum value
[16:25:58] <alex_joni> and go up
[16:26:08] <alex_joni> so -17.5, -16.5, ... 0
[16:26:46] <jlmjvm> lemmee try that
[16:50:36] <jmkasunich> that nan surprises me
[16:51:09] <jmkasunich> I thought it I had it coded to simply use the end values if you go to a position that is outside the range in your comp data
[16:59:30] <jlmjvm> it was actually moving around with the nan showing
[17:00:23] <jmkasunich> your axis limits are -something and zero, and your comp file had only positive values in it?
[17:00:37] <jmkasunich> (I'm trying to figure out how to reproduce that result)
[17:00:54] <alex_joni> jmkasunich: yeah, that was how it was set up
[17:02:03] <Jymm> Anyone interested is waxed lacing cord?
[17:02:29] <archivist> ew and some rubber and....
[17:02:53] <Jymm> archivist was that to me?
[17:03:56] <archivist> I see your not awake yet
[17:05:59] <Jymm> When I worked at one place I did virbration testing to 7g's. It was on a 4x4ft magnesium table, powered by a 220VAC 3ph coil about 3' across. The amplifiers were in three full height cabinets.......
[17:08:28] <Jymm> In the cabinets, You could trace each and every wire visually. All hand done, 90deg bends, all lased, no twists at all in any 15 wire bundles. It was such a beautiful thing. cables were self suspended. The hoses for water cooling didn't use any clamps at all. I have NEVER seen anything done so well and beautiful in my life. it all used waxed lacing too.
[17:09:59] <jmkasunich> cost a fortune I'm sure, and nobody would pay for it nowadays
[17:10:56] <Jymm> jmkasunich: Oh, I bet. but just to see such a thing at least once in life. I was very impressed.
[17:11:58] <Jymm> To visisually trace a 8ft wire in a bundle from end to end... wow!
[17:12:55] <archivist> they use the crap plastic trays now, the arms break off, but looms are repairable in plastic trays where lacing was not easily repaired
[17:14:04] <Jymm> Yeah, I have some panduit here
[17:16:58] <jlmjvm> http://pastebin.com/d383555c8
[17:17:37] <jlmjvm> does that look ok for a comp file
[17:19:36] <jmkasunich> yes
[17:19:50] <jmkasunich> won't actually do anything, since all the values on each line are the same
[17:20:27] <jmkasunich> I'm not sure if comments are supported or not - I'm unfamiliar with the code that parses the file, I only work with the numbers
[17:20:56] <jmkasunich> are you still getting nans?
[17:21:02] <jlmjvm> i cant tell any difference yet when i change the values
[17:21:12] <jlmjvm> no,the nans have been gone
[17:21:54] <alex_joni> I think it'll just skip over lines it doesn't understand
[17:22:01] <jlmjvm> before i made this file i just had 2 values,middle and end
[17:22:03] <alex_joni> understand meaning 3 floats on a line
[17:24:16] <jlmjvm> should you be able to see a difference between the actual and commanded machine number if its working?
[17:24:33] <jmkasunich> no
[17:25:09] <jmkasunich> the only way to see what it is doing is to go poking around with halmeter, or to actually measure machine movement with indicators, etc
[17:25:20] <jlmjvm> no wonder i cant see a diff when i change the values
[17:25:35] <jmkasunich> from the GUI, it "just works", as if you have perfect screws
[17:26:03] <jlmjvm> ahh,should be able to see it with the counters on my bridgeport screws
[17:27:24] <jlmjvm> it appears to be working,just couldnt see any diff in the readout
[17:29:06] <jlmjvm> i just met a guy that has a laser he uses to set machines with
[17:30:19] <jlmjvm> always wondered how that was done
[17:36:10] <jmkasunich> jlmjvm: do you happen to have a saved copy of the ini and comp files that caused the nan problem?
[17:36:24] <jmkasunich> I'd like to fix it, but I can't seem to duplicate it here
[17:40:37] <jlmjvm> ive got the ini,i changed the comp file
[17:40:55] <jlmjvm> i should be able to do it again
[17:41:52] <jmkasunich> if you can pastebin the files I'd appreciate it
[17:42:03] <jmkasunich> nan shouldn't happen, ever, IMO
[17:55:23] <jlmjvm> got it,lemmee paste
[17:56:35] <jlmjvm> http://pastebin.com/d695cfc1
[17:57:13] <jmkasunich> thats the comp file
[17:57:16] <jmkasunich> do you have the ini?
[17:57:24] <jlmjvm> http://pastebin.com/d5f2ebe1e
[17:57:30] <jmkasunich> thanks
[17:57:34] <jlmjvm> yw
[17:58:04] <jlmjvm> i found that if you add a blank line to the top of the comp file the nan goes away
[17:58:30] <jmkasunich> hmm. very interesting
[17:58:39] <jmkasunich> thanks, that detail might be important
[17:58:40] <jlmjvm> but thats not the case with my present comp file
[17:58:55] <jlmjvm> it doesnt have to have a blank line
[17:59:20] <jlmjvm> just wanted ya to know i stumbled upon that
[17:59:44] <jmkasunich> could you pastebin joe.hal too?
[18:00:06] <jmkasunich> oh, you have three hal files
[18:00:40] <jmkasunich> its times like this when I could agree with the folks who would like all of EMC's config info to be in one file
[18:05:02] <jlmjvm> which do you need
[18:05:12] <jmkasunich> all of them I think
[18:05:21] <jmkasunich> I'm trying to run your configuration exactly as you are running it
[18:06:33] <jmkasunich> you have joe.hal, custom.hal, and custom_postgui.hal
[18:06:51] <jmkasunich> (unless the last two are dummies created by stepconf?)
[18:07:39] <jlmjvm> they are being used
[18:13:19] <jlmjvm> http://pastebin.com/d262342e
[18:14:03] <jlmjvm> http://pastebin.com/d6d8b8da
[18:14:45] <jlmjvm> http://pastebin.com/d75c88783
[18:14:49] <jmkasunich> thanks
[18:30:26] <usman> Can some one help me out changing the size of DRO in "AXiS" preview section?
[18:31:58] <alex_joni> usman: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PyVCP <-look for big DRO
[18:32:05] <alex_joni> maybe that suits you
[18:32:17] <SWPadnos> fenn just made a change that adds x resources for the DRO in axis
[18:33:16] <alex_joni> SWPadnos: I am aware of that.. but that's only in CVS
[18:33:22] <SWPadnos> yep
[18:33:31] <alex_joni> and will be in for 2.3.0 when it'll happen
[18:34:16] <jmkasunich> well crap
[18:34:41] <jmkasunich> after twiddling things to get jlmjvm's config to run in simulator mode, I can't get the NAN to occur
[18:34:47] <jmkasunich> jogged all axes, nothing
[18:34:56] <usman> fenn: I was having severe problem with my internet connection in the morning, but now I am here. Please guide me the procedure for changing the DRO Size again..
[18:34:59] <jlmjvm> it did it in mdi
[18:35:54] <Jymm> When you list a PN on your website and I click on the link, don't tell me item not found - ya bastards!
[18:36:08] <alex_joni> jmkasunich: maybe it only works in RT ?
[18:36:29] <jmkasunich> maybe it only happens after homing?
[18:36:29] <SWPadnos> usman, first, you need to get the latest CVS version of EMC
[18:36:30] <alex_joni> logger_emc: bookmark
[18:36:30] <alex_joni> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-01-26.txt
[18:36:36] <SWPadnos> and you'll need to compile it
[18:36:40] <alex_joni> usman: maybe you can find there what fenn said
[18:37:05] <alex_joni> SWPadnos: sounds like fenn explained what he needed to do, then realized it's too complicated, then added the stuff to CVS
[18:37:09] <SWPadnos> heh
[18:37:45] <SWPadnos> hmmm - my wife just called down that she's ready. I guess that means we're leaving now
[18:37:46] <alex_joni> hmm.. not really
[18:37:46] <SWPadnos> bbl
[18:43:37] <jmkasunich> jlmjvm: I still can't get it to happen, even after homing all axes, and doing MDI moves on X
[18:46:58] <jmkasunich> this is annoying - even though it only happens with an invalid comp file, I consider anything that results in NAN on the display as a bug - it ought to tell you your comp file is invalid, not print NANs
[18:47:12] <usman> coordinate_font = o.option_get("font", "Coordinate_Font") or "9x15" in /usr/bin/axis file and
[18:47:12] <usman> 10:02:41 <cnc_engineer_> fenn: then on command prompt I issue the command echo '*Axis*Coordinate_Font: -adobe-courier-bold-r-normal--32-*-*-*-*-*-*-1' | xrdb ; emc
[18:57:28] <jmkasunich> oh well, there is a limit to how much time I can afford to spend trying to fix an un-reproducable bug, and I think I just reached it
[18:58:37] <jlmjvm> sorry for the trouble
[18:58:46] <jmkasunich> not your fault
[18:58:54] <jmkasunich> I just wish I could have made it happen here
[18:59:15] <jmkasunich> the fix is probably something trivial, but without having the error occur, it can be impossible to find
[18:59:20] <alex_joni> jlmjvm: can you make ti error again?
[18:59:27] <jlmjvm> yes
[18:59:44] <alex_joni> * alex_joni tries it too
[18:59:46] <jlmjvm> i saved the comp file
[19:00:16] <jmkasunich> alex_joni: may be my mistake was trying to reproduce it on sim
[19:01:12] <jmkasunich> on a real box with a real parport, you won't have to comment out as much stuff
[19:01:21] <jmkasunich> still need to fake out homing, if homing is involved in any way
[19:03:11] <alex_joni> yay.. I got nan
[19:03:19] <eric_U> that's wonderful
[19:03:41] <alex_joni> jmkasunich: I used stepper_inch, added the comp jlmjvm pastebined
[19:03:49] <alex_joni> with stepper_inch it worked ok
[19:04:09] <jlmjvm> brb
[19:04:17] <alex_joni> as soon as I changed travel from -10..10 to -10..0, I get nan on the display after homing and g0x0
[19:04:35] <alex_joni> I now see 2.3729 in the readout and I got a ferror
[19:06:00] <jmkasunich> lemme try that
[19:06:05] <jmkasunich> emc
[19:06:07] <jmkasunich> oops
[19:06:27] <alex_joni> I kept using it.. and it gets worse
[19:06:42] <alex_joni> now the X readout says: 25.1329 (after g0x0)
[19:06:45] <jmkasunich> * jmkasunich fires up a RT computer
[19:06:51] <alex_joni> I noticed the nan appeared after jogging
[19:28:54] <lerman> alex_joni: are you here?
[19:29:07] <alex_joni> yes
[19:29:35] <lerman> I just saw your commit that changed ad message to "Unable to open parameter file..."
[19:30:00] <alex_joni> yes
[19:30:02] <lerman> A downside of that change is that it breaks "easy" internationalization.
[19:30:19] <alex_joni> lerman: that "easy" internationalization doesn't work anyway
[19:30:51] <alex_joni> but I agree it should probably be something like _("") or something like that
[19:31:03] <lerman> OK.... As long as we aren't breaking things that could be fixed.
[19:31:32] <lerman> I'm out of here. The general idea of improving error messages is real good. Thanks.
[19:31:34] <lerman> Ken
[19:31:37] <alex_joni> bye
[19:46:48] <micges> alex_joni: error showing soft limit error when home=min_limit=0 on axis was fixed ?
[19:47:00] <micges> I can find
[19:47:41] <alex_joni> micges: I don't understand
[19:47:43] <Jymm> Note to self: When mfg of router tells you that DMZ will isolate rest of LAN and you think they're full of shit, go with your gut instinct!
[19:48:36] <micges> sorry for thinking shortcut
[19:49:38] <micges> few days ago (I think) someone tell that when in ini home_offset =0 and min_limit=0 then on homing that axes soft limit error occurs
[19:51:30] <alex_joni> micges: that's not an error.. it's a "feature"
[19:51:37] <alex_joni> you need to have it slightly larger
[20:13:32] <jmkasunich2_> yay, no nans
[20:13:43] <alex_joni> wrong channel :P
[20:13:48] <BigJohnT_> I've been out what caused the nans?
[20:14:11] <alex_joni> a screwy #define
[20:14:13] <jmkasunich2_> IEEE floating point pecularities
[20:14:27] <jmkasunich2_> HUGE_VAL is supposed to be the largest possible number
[20:14:35] <jmkasunich2_> it is defined as infinity
[20:14:42] <jmkasunich2_> infinity times zero = NAN
[20:15:04] <eric_U> l'Hopital might disagree
[20:15:07] <jmkasunich2_> if you use DBL_MAX, which is the largest number that is less than infinity, then it works - DBL_MAX * zero = zero
[20:15:16] <Gamma-X> i think i may have something incorrect on my scales cause i told me machine to move 1.0 and it was trying to go 1.0 but the scales were reading small incriments as it moved a lot.
[20:15:33] <eric_U> interface problems
[20:16:07] <BigJohnT_> I see, glad you solved it
[20:16:48] <alex_joni> ain't seen one yet that can't be solved :P
[20:17:30] <jmkasunich2_> once you can reproduce it
[20:17:35] <BigJohnT_> all you have to do is be able to reproduce it
[20:18:06] <BigJohnT_> jmkasunich looks like we are on the same wave lenght
[20:20:48] <alex_joni> BigJohnT_: cool, next time you can save him the trouble of hunting a bug :D
[20:20:51] <alex_joni> *grin*
[20:21:11] <BigJohnT_> LOL
[20:21:57] <jmkasunich> anyone who has to add "To be sure that you receive our email in your inbox (not in a bulk folder), please add the following to your address book" to their emails should probably look in the mirror and say "I'm adding this because my messages ARE SPAM!!!"
[20:22:35] <alex_joni> heh
[20:23:08] <BigJohnT_> I usually hunt things at least as big as a squirrel and bigger except when I'm hunting ticks!
[20:27:34] <BigJohnT_> I did not quit
[20:28:04] <eric_U> dunno what you're talking about , freenode says you quit
[20:28:22] <BigJohnT_> * BigJohnT_ says i'm here
[20:28:50] <eric_U> I moved my mill about a foot so far today
[20:29:06] <eric_U> sorta like the pyramids
[20:29:17] <BigJohnT_> as in under power...
[20:29:26] <eric_U> no, as in across the room
[20:29:34] <BigJohnT_> ok
[20:29:43] <jmkasunich> 10,000 slaves and some logs underneath?
[20:29:51] <eric_U> black iron pipe
[20:29:55] <BigJohnT_> * BigJohnT_ nap time zzzzzzzzzzzzz
[20:29:56] <eric_U> only slave is me
[20:30:16] <BigJohnT_> bye bye
[20:30:18] <alex_joni> * alex_joni introduces eric_U to the art of cloning
[20:30:29] <archivist> 3-5 pipes/bars and you move anything
[20:30:45] <eric_U> I didn't realize how rough the casting is under there
[20:30:58] <alex_joni> archivist: given that you can lift them enough to put the bars underneath it
[20:31:12] <eric_U> lifting can be a little entertaining
[20:31:13] <jmkasunich> alex_joni: I've tried cloning - the problem is that the clones are just as lazy as I am
[20:31:20] <alex_joni> haha
[20:31:22] <jmkasunich> when I told them to help me, they told me to go to hell
[20:31:24] <archivist> builders planks below and above rollers for smoothe running
[20:31:29] <alex_joni> who said you should clone yourself?
[20:31:52] <eric_U> the thing weighs 5000 lbs, don't know if builders planks would work
[20:32:08] <eric_U> I use the 2 roller method, much safer
[20:32:35] <archivist> I add ropes chain to stop run away
[20:33:08] <eric_U> two roller method can only move roller to just past cg, and then you stop
[20:33:50] <Juanjo> Juanjo is now known as juanjo
[20:35:01] <juanjo> hi folks
[20:35:16] <alex_joni> hi
[20:38:24] <archivist_emc> eric_U: working on your own with loads like that I find safer as "helpers" often get in the way/push when they shouldnt etc
[20:38:24] <juanjo> how arwe you? in Tucuman rain heavy
[20:39:25] <juanjo> i now studing the HAL manual. due i cant do outdoors!!!
[20:39:43] <juanjo> you have a EMC running?
[20:39:56] <juanjo> or better working?
[20:40:22] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Videos
[20:41:08] <jmkasunich> archivist_emc: I agree 100%
[20:41:40] <jmkasunich> when you have helpers you are more inclined to try to rush things, or to brute force it with many muscles instead of using ramps, levers, rollers, etc
[20:42:12] <archivist> its a like chess, a heavy move on your own
[20:46:46] <Gamma-X> do u think my feedback on the servos is coming from the shielding not being grounded or possibly the scales not being grounded to chasis?
[20:46:58] <Gamma-X> i read that on a forumn...
[20:47:21] <alex_joni> Gamma-X: first you need to see how the stuff that comes from the scales looks
[20:47:25] <jmkasunich> Gamma-X: what exactly do you mean by "feedback on the servos"
[20:47:37] <alex_joni> maybe your SCALE is simply wrong
[20:47:41] <jmkasunich> do you mean your encoders are counting when the machine isn't moving?
[20:49:11] <Gamma-X> jmk...
[20:49:19] <Gamma-X> they are but the servos are moving realy slowly
[20:49:30] <LawrenceG> Gamma-X: suggestion... remove motor to table coupling, hand crank axis 1" and see if emc says it has moved 1", then get motor drive path working... from hal, make sure you can run drives at 10,20,...100% in each direction
[20:49:33] <Gamma-X> like incredibly slow. cant see it with ur eye but with the scale u can see it
[20:49:54] <jmkasunich> Gamma-X: there is one and only one thing that matters with the encoders
[20:50:02] <Gamma-X> ?
[20:50:15] <jmkasunich> EMC's perception of how far the machine has moved MUST match how far it really did move
[20:50:28] <jmkasunich> don't even THINK about running motors before you know that is correct
[20:50:42] <jmkasunich> like LawrenceG said - decouple, and crank it 1"
[20:50:48] <Gamma-X> well im trying to figure out why my motors move without me hitting anything haha.
[20:50:57] <jmkasunich> if EMC doesn't say you moved 1", figure out why and fix it
[20:51:17] <jmkasunich> they are moving without you hitting anything because you don't UNDERSTAND what you are doing
[20:51:27] <Gamma-X> ill give u that hahaah
[20:51:27] <jmkasunich> you are just following instructions from people
[20:51:35] <jmkasunich> this isn't baking where you can follow a recipie
[20:51:48] <Gamma-X> i wish
[20:51:49] <alex_joni> baking a CNC machine for dummies
[20:52:06] <jmkasunich> you can bake simple stepper machines, thats what stepconf is for
[20:52:17] <GNieport> I'm pretty sure the internet has almost reduced CNC to that level...
[20:52:18] <jmkasunich> servo machines you actually have to know a thing or two
[20:52:25] <LawrenceG> Gamma-X: you should not attempt to close the feedback loop until position feedback and drive systems are working
[20:52:43] <Gamma-X> i gotta configure pid.
[20:52:44] <Gamma-X> im workin on it
[20:52:57] <jmkasunich> gotta configure feedback first
[20:53:11] <jmkasunich> move the machine BY HAND, and make sure EMC knows how far it moved
[20:53:16] <Gamma-X> ok
[20:53:21] <alex_joni> then after you finished configuring, you gotta test feedback
[20:53:56] <Gamma-X> jmkasunich i think i may just need to put in differant values for my scales.
[20:54:20] <jmkasunich> probably true
[20:54:40] <jmkasunich> my point is - do that, and test it, before you jutmp ahead to the next step
[20:54:51] <Gamma-X> u got it
[20:55:23] <acemi> is there a plan to remove RTLinux support for the future relase of EMC2
[20:59:16] <ALS> is this the baking class?
[20:59:46] <alex_joni> what'cha baking?
[20:59:55] <alex_joni> we're baking CNC machines..
[21:00:05] <ALS> half baked
[21:00:15] <alex_joni> obviously
[21:01:38] <Gamma-X> anyone ever hear of a dude with the name regor?
[21:02:03] <lerman> Must be Roger's backward brother. :-)
[21:02:15] <Gamma-X> roger neal...
[21:02:50] <Gamma-X> his irc name is regor, i beleive
[21:03:10] <alex_joni> might have lost an 'g'
[21:12:22] <archivist> * archivist is happy having baked some cnc machine today
[21:12:53] <alex_joni> what did you fry again?
[21:13:42] <archivist> sprinkle of table with a topping of rotary table
[21:14:14] <archivist> drilling holes and fitting together
[21:14:23] <alex_joni> no lube with that?
[21:14:32] <jmkasunich> archivist what is your table made of
[21:14:32] <alex_joni> seems a bit dry :P
[21:14:36] <jmkasunich> doesn't look metal
[21:14:50] <archivist> ally table
[21:14:57] <jmkasunich> ah
[21:14:57] <alex_joni> biscuits
[21:15:05] <jmkasunich> it almost looks like concrete
[21:15:31] <archivist> was a bit corroded from the scrap yard
[21:15:42] <archivist> but a nice lump
[21:16:00] <dmes1> fly cut the top off
[21:16:15] <archivist> I wont finish machine the table till later
[21:16:51] <archivist> getting it functional first
[21:17:43] <archivist> now in the state where adding a headstock, it could cut a gear
[21:17:57] <dmes1> large TOSHIBA installs always cut the table a few thousanths.. to square it up NUTS on..
[21:18:33] <archivist> table really matches the slides then
[21:18:48] <dmes1> yup
[21:19:22] <dmes1> some were floor type boring mills
[21:19:43] <archivist> been contemplating that on this one, but my travel is a bit small
[21:19:56] <archivist> well a lot small
[21:20:00] <juanjo> juanjo is now known as Juanjo
[21:20:05] <dmes1> tables arent even attatched to the machine..
[21:20:39] <archivist> building mounted
[21:21:13] <dmes1> yup floor plates
[21:24:23] <dmes1> there was a sweetie.... 80 feet on x axis... 2 26 foot tables on the ends and a 12 foot rotary in the middle.... was a very versatile machine... 5 face head on the spindle.. or a 12inch x 48 inch quill to dig into stuff...
[21:25:02] <archivist> oo machine porn
[21:25:48] <dmes1> yes.. i guess... ; |)))
[21:26:28] <dmes1> but i only tells it like i sees it
[21:28:48] <archivist> machining is not supposed to be described as "dig into stuff", which reminds me of the 3" by x ft cut slabmill on youtube that someone posted a link to
[21:29:00] <fretless85> 80feet=???mm?
[21:29:09] <archivist> lots
[21:29:25] <fretless85> 1inch=25.4mm
[21:29:28] <fretless85> 1foot=?
[21:29:37] <fretless85> let me google^^
[21:29:44] <archivist> better described as meters
[21:30:25] <fretless85> 24384mm thats not that long
[21:30:33] <alex_joni> 1 foot =~ 30cm
[21:30:50] <fretless85> 304.8mm
[21:31:50] <alex_joni> depends on the foot
[21:31:59] <alex_joni> :P
[21:32:04] <alex_joni> http://www.youtube.com/watch?v=WALIARHHLII&feature=related <- ROFLMAO
[21:32:04] <archivist> whats your longest fretless85
[21:37:09] <toastydeath> ysssss
[21:37:16] <toastydeath> asymptote works now
[21:38:46] <fretless85> we delivered an machine to a 80000mm(262.5foot) to a costumer near the ocean,,,it even respond to the tides caused by the movement of the grounding :D
[21:39:04] <Gamma-X> alex_joni wow that bitch was dumb! lol
[21:40:02] <fretless85> a service technician told me that the mill with 2 different programms depending of the tides
[21:40:55] <fretless85> there is an ebb and an flood programm
[21:41:06] <Gamma-X> http://www.youtube.com/watch?v=lqzio0avXB8&feature=related
[21:41:07] <Gamma-X> hahahaha
[21:42:20] <Gamma-X> http://www.youtube.com/watch?v=sAJUMYGgfZA&feature=related alex_joni
[21:43:44] <Ziegler> !help
[21:43:50] <Ziegler> oops sorrt
[21:44:15] <alex_joni> good night all
[21:45:06] <fretless85> good night
[21:45:20] <fretless85> or "gute nacht" alex
[22:13:58] <dmes1> neat.... i saw a vertical lathe capable of tuning 66 feet ( 20 M ) diameters up to the height of the ceiling... they added 6 ft (2M) riser blocks under the column... the cnc pendant was on a reach truck ARM and bucket..( so you could watch the tool..
[22:17:16] <archivist> hehe
[22:19:30] <archivist> I was told once about some of the early "gun" barrel boring machines for very large sizes that were vertical but I have not yet found proper facts on them
[22:20:31] <dmes1> that place had a part on a mill for OVER 1 year... while the program ran... they would feed it 1 week of data in a 24 hr sunday offshift
[22:21:25] <archivist> hehe
[22:22:09] <dmes1> after seeing it ther for 3 months... i asked if the machine was busted..??? they said no.. it runs around the clock xcept sundays..
[22:23:27] <dmes1> ther must have bee 300 flanges on this pressure vessel.. ALL with 8-36-72 holes in them.. ALL on different angles and mangles..
[22:24:19] <fretless85> what the heck?
[22:24:32] <archivist> boiler header or a chemical industry vessel
[22:24:35] <dmes1> 20M tall 4 M diameter tank
[22:24:50] <dmes1> i dunno... but VERY intense
[22:25:01] <fretless85> nice :)
[22:25:07] <fretless85> i like big parts
[22:25:15] <archivist> they need big anealing cookers for them
[22:26:06] <dmes1> i seen the main fitting forgings for the a380 this morning... have pics.. pitty the poor bastage that scraps 1 of those..
[22:26:30] <dmes1> we finish machine HARD... 54-56 Rc
[22:27:29] <fretless85> i did a fitting forging for a a380 tubrine a few month ago
[22:28:00] <dmes1> about 4000 lbs ( 1900 or so KG) raw... 2100 lbs finished wieght..
[22:28:37] <fretless85> the fitting forging i did was 2 halfes one half wieght about 5t
[22:28:40] <dmes1> this is the main landing gear
[22:28:55] <fretless85> we do a lot for airbus
[22:29:07] <dmes1> where do you work??
[22:29:43] <fretless85> the got a boring machine for the main empennage from us
[22:30:13] <dmes1> cool any TOshiba machine in the shop??
[22:30:14] <fretless85> fooke germany we mostly build milling machines and spacial machines
[22:30:40] <fretless85> no toshiba
[22:31:01] <dmes1> what controls are they ordering>>???
[22:31:14] <dmes1> 840 D
[22:31:40] <fretless85> i guess siemens 840d
[22:32:04] <fretless85> does toshiba make big mills?
[22:32:58] <dmes1> its still the finest.. yes toshiba makes LARGER mills than i ever installed for them.... but none came to Canada..
[22:33:30] <fretless85> we got TOS boring mills
[22:33:39] <fretless85> soraluce and zayer bed mills
[22:33:47] <fretless85> and soraluce and shw moving column
[22:34:19] <dmes1> our bread and butter b/mill was a 3m x2.5m x2M....with full rotary
[22:34:38] <fretless85> same here but a little bigger
[22:34:46] <fretless85> but the soraluce mills sucks
[22:35:02] <dmes1> bed mills were all hydrostatic ways.... real slick and smooth
[22:35:35] <fretless85> all thk rails... not good at all
[22:35:47] <dmes1> that was the r26 the R22 was another 1.5 M on all axes...
[22:35:57] <fretless85> to wobbly
[22:37:14] <fretless85> what rails do the toshiba mills use?
[22:37:19] <dmes1> i hate linear guide rails.... had CHIRON DZ16's in 1 place.... what a BIATCH to re-alighn after a crash.... axis upon axis...
[22:37:45] <dmes1> Toshiba have WAYS... no rails....
[22:38:00] <dmes1> more robust
[22:38:03] <fretless85> right
[22:38:19] <fretless85> the tos/shw/zayer machines use rails to
[22:38:25] <fretless85> cant beat it
[22:38:53] <dmes1> tos??? not the one we had... and we had a few models
[22:39:07] <fretless85> TOS boring mills
[22:39:24] <dmes1> yes
[22:39:42] <fretless85> we got a newer model...the table sucks really bad...to thin
[22:39:56] <fretless85> when u put lets say 30 tons on it it bends
[22:40:32] <fretless85> the old model we got takes it without any complaining ;)
[22:40:46] <dmes1> breathing...
[22:40:55] <fretless85> ?
[22:41:42] <dmes1> the machine table is breathing....
[22:42:09] <fretless85> does i have to carry it? :D
[22:42:14] <fretless85> do
[22:42:49] <dmes1> is it a rotary???
[22:43:26] <fretless85> no we got a normal table and beside it there is a rotary
[22:43:35] <fretless85> its a moving column boring mill
[22:44:16] <dmes1> and the floor mounted tabl flexes???
[22:44:46] <fretless85> yea sad but true
[22:45:13] <dmes1> did they pour a proper foundation??
[22:45:22] <fretless85> yes they did
[22:45:38] <dmes1> unreal...
[22:45:41] <fretless85> but the rotary table also flexes when u clamp it really hard
[22:45:45] <skunkworks> logger_emc: bookmark
[22:45:45] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-01-26.txt
[22:46:56] <fretless85> the new models arnt that solid...when i look back to the old model
[22:47:19] <fretless85> on the old model the rotary table is about 400mm thick on the new its 250mm
[22:47:37] <dmes1> dia??
[22:47:43] <fretless85> no
[22:47:50] <fretless85> thick
[22:48:06] <dmes1> but what DIA??
[22:48:18] <fretless85> the dia is 3000
[22:48:48] <fretless85> its not full round more square with rounded edges
[22:48:54] <fretless85> 3000x3000
[22:49:22] <archivist> ribbed or solid 250mm
[22:49:40] <dmes1> well we needed that .. and put a 1.5 M on a bmill recently....the new yr... put a new drive in last week too... 2 M subplates on top with off camber loads...
[22:49:41] <fretless85> solid
[22:49:55] <fretless85> i guess
[22:50:20] <archivist> still with 30 tons on it will bend
[22:50:27] <dmes1> the wrong ppl do the buying...
[22:50:33] <fretless85> not on the rotary
[22:50:49] <fretless85> the rotary is max rated to 20t
[22:51:20] <dmes1> nice
[22:51:24] <archivist> one needs thicknes for stiffness
[22:52:12] <fretless85> yea they buyed a machine with no inner cooling and no spindle orientation
[22:52:24] <fretless85> great buy
[22:56:56] <dmes1> we have a FIDIA being installed that was purchased sight unseen by anyone at our site..
[22:57:56] <fretless85> nice lol
[22:58:23] <dmes1> i figure we should be puttind a spindle in her by the end of march
[22:59:44] <dmes1> we are planning on runnin high strenght alumunum forgings under it.... IM freightened....
[23:00:17] <dmes1> we are going to end up with pretzels...
[23:00:30] <fretless85> haha
[23:01:30] <dmes1> this stuff distorts wickedly under relaxed machining.... under HSM techniques... i think were facked..
[23:02:23] <fretless85> i wish u good luck :D
[23:02:36] <dmes1> you ever flown on a DHC-8??
[23:02:46] <fretless85> nope
[23:04:06] <dmes1> its a high wing turboprop with legs that are 1.75-2.0 m tall - 1 big triangle and a shock
[23:04:25] <fretless85> nice
[23:04:46] <dmes1> average section of the drags is 10mm
[23:05:24] <fretless85> sounds neat!
[23:07:55] <dmes1> landing gear is great to make as long asa they work... not like goodrich's dhc-8q400 gears from a few months back...
[23:13:45] <fretless85> dmesl what are u exactly doing for a living?
[23:15:21] <fenn> he's the big fish in a little pond
[23:15:56] <dmes1> im a methods( manufacturing) engineer for a major landing gear mfg'er
[23:16:17] <fretless85> oh okay
[23:16:20] <dmes1> lil' fish among sharks is more like it
[23:16:46] <fretless85> haha
[23:17:22] <dmes1> and the boeing 787 is coming to a theater near US... GREAt... another mess to contend with
[23:18:02] <fretless85> ;)
[23:18:51] <dmes1> we do f18, v22 osprey,harrier,all BOMBARDIER aircraft... hawker h4000 , 850xp... yadda yadda
[23:19:19] <fretless85> so mostly millitary airplanes?
[23:19:54] <dmes1> there are 6-8 bombardier L/G in house
[23:20:33] <dmes1> business jets.. as well as regional jets
[23:20:57] <dmes1> the Global Express is a hanger queen
[23:23:12] <dmes1> this company did the original gear on the avro arrow... the test rig is still there for inspiration...
[23:24:03] <fretless85> cool ;)
[23:25:20] <dmes1> brb