#emc | Logs for 2007-03-24

[00:07:25] <eric_u> interesting comment to the review
[00:08:48] <eric_u> I'm used to reading fractured English from spending so much time online and engineering journals have pretty much given up on proper english as a requirement
[00:12:03] <eric_u> there is a free book you can download
[00:47:34] <crepincdotcom> milling in progress ;-)
[00:50:23] <crepincdotcom> god DAMN it
[00:50:43] <crepincdotcom> forgot to tighten the machine straps... the board was getting pushed all over the table lol
[00:50:46] <crepincdotcom> <-- noob
[00:57:48] <ejholmgren> http://www.cartertools.com/jstev01.jpg
[00:58:09] <ejholmgren> interesting motor mounts
[00:59:14] <jtr> yeah, John Stevenson does some neat stuff
[01:02:59] <jtr> http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/
[01:09:20] <jtr> somewhere he posted pics on repairing a torn up woodruff key seat, but I don't see it there.
[01:20:02] <ejholmgren> http://www.dumpstercnc.com/acme_couplers.html
[01:20:34] <ejholmgren> hrmmnn.... I just just have a lathe "magically" show up in the garage
[01:20:41] <ejholmgren> so I don't have to resort to things like that
[01:21:04] <Jymmmm> Well, fuck a duck... Just got done makign the new plates to allow the cables thru and to add 2.5" shopvac for dust collection... Even at half speed, the walls bow in.
[01:21:41] <Jymmmm> ITS JUST CANVAS!!! NOT air/water proof! - and 8 sq ft too
[01:23:05] <jtr> Jymmmm: have you seen the louvered dryer vents? Perhaps one of those, mounted facing in, would let in enough air.
[01:23:49] <Jymmmm> jtr: I think what I'll do is remove all of the canvas and see if I can find A/C filters that will fit
[01:24:16] <Jymmmm> HVAC filters I mean - hopefully that'll let air in and keep the dust in too
[01:25:06] <Jymmmm> jtr but that is a good idea too =)
[01:25:34] <Jymmmm> actually a really good idea, but I think I'll need more area than 4"
[01:26:20] <Jymmmm> Jymmmm is now known as Jymmm
[01:26:33] <jtr> Might want to just put in a small one (HVAC filter) across from the dust collecto port to make sure you have cross flow.
[01:27:04] <Jymmm> hang on, let me get pics
[01:28:06] <Jymmm> http://farm1.static.flickr.com/167/424362247_9bce9d31df_b.jpg
[01:28:16] <Jymmm> the bank top
[01:28:18] <Jymmm> back
[01:28:25] <Jymmm> is all canvas
[01:29:03] <Jymmm> 4' x 2'
[01:29:26] <Jymmm> the whole enclosure is 4' x 4' x 3' tall
[01:32:12] <jtr> Where's the dust collector port? Will the incoming air take a shortcut directly to the exhaust port, or maybe move too slow to carry the dust if you open up that much area?
[01:32:56] <ejholmgren> do you have more pictures of the table?
[01:36:17] <jtr> 'Course, it is a dust collector you're working with.
[01:39:18] <jtr> oops, maybe not - you're back to a vacuum?
[02:10:44] <ejholmgren> hrmmnn... can you use a cuttoff wheel in a regular miter saw?
[02:15:40] <ejholmgren> nevermind, just read several rants about bodily injury etc
[02:33:03] <Jymmm> jtr: You see that white plate at the back bottom? I just replaced that (which has a 2" hole in it to pass cables thru) with a larger plate that has a 2", 2.60", and 4" holes
[02:33:25] <Jymmm> ejholmgren of the machine itself?
[02:33:48] <Jymmm> ejholmgren like this? http://farm1.static.flickr.com/166/424362252_1bd7946026_b.jpg
[02:34:32] <Jymmm> http://farm1.static.flickr.com/149/424362153_e80fc2eb16_b.jpg
[02:35:02] <Jymmm> and the damn sander and all that work is still .017" "cupped"
[02:36:36] <toastydeath> i wonder how fast a roughing endmill would cut through wood
[03:31:15] <ds3> how fast is your spindle?
[03:31:56] <SWPadnos> 3 bazillion Hz
[03:32:40] <ds3> so the SFPM can be measured in terms of C? ;)
[03:33:08] <SWPadnos> oh yeah - and only quantum cooling is efficient enough
[03:34:26] <ds3> the startup energy spikes must be huge
[03:36:58] <SWPadnos> that's where all the dark matter comes from
[03:37:01] <SWPadnos> I use up all the energy
[03:37:09] <jmkasunich> not to mention the synchrotron radiation...
[03:38:51] <jmkasunich> how many zeros is a bazillion anyway?
[03:39:09] <SWPadnos> a gazillion
[03:39:32] <jmkasunich> thats a lot
[03:39:47] <SWPadnos> yeah - 10^gazillion
[03:39:51] <SWPadnos> :)
[03:40:12] <jmkasunich> so what do you use for tooling?
[03:40:20] <SWPadnos> hard shit
[03:40:32] <jmkasunich> I haven't seen any of that
[03:40:39] <SWPadnos> I have actually
[03:40:48] <jmkasunich> I don't want to know where
[03:40:50] <SWPadnos> I've handled a Viking turd - petrified
[03:40:54] <SWPadnos> oops - sorry
[03:41:24] <jmkasunich> must be hard to get it in a collet
[03:41:37] <jmkasunich> I'd expect runout problems too
[03:42:26] <SWPadnos> no runs there
[03:42:59] <SWPadnos> damn. if only I had 480 3ph: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=170094515168
[03:43:04] <SWPadnos> and $1200
[03:43:30] <jmkasunich> what for?
[03:43:36] <SWPadnos> err - no reason
[03:44:42] <Jymmm> I need a low speed spindle
[03:45:12] <Jymmm> like a drill motor, that I can attach my tappign head to
[03:45:46] <jmkasunich> to use on your CNC?
[03:45:51] <Jymmm> yeah
[03:46:04] <jmkasunich> hope you don't want to tap anything particularly hard
[03:46:17] <Jymmm> al being the hardest
[03:48:50] <Jymmm> but finding one that will fit on my machine is the hard part
[03:50:09] <jmkasunich> I'm finding it hard to work up the energy to work on these docs
[03:50:18] <jmkasunich> sleep sounds like a good idea
[03:50:24] <SWPadnos> indeed
[03:50:30] <SWPadnos> hurry - it's still before midnight
[03:50:31] <Jymmm> jmkasunich grab a beer for every 100 hundreds you type =)
[03:50:38] <Jymmm> ^words
[03:51:04] <Jymmm> ok pizza time, laters
[03:52:07] <jmkasunich> Jymmm: its not the typing that is the problem, its the thinking
[03:52:13] <jmkasunich> and the beer won't help that at all
[03:53:25] <eric_u> but the docs will become progressively more interesting from an artistic viewpoint
[03:53:36] <SWPadnos> not in lyx
[03:53:57] <ds3> what is the docs for?
[03:54:30] <skunkman> skunkman is now known as skunkworks
[03:54:33] <jmkasunich> stepgen - its been revised a bit for version 2.2
[03:56:33] <jmkasunich> the man page is up do data.... sometimes I wonder about having duplicate info in man pages and in the lyx documents
[03:56:44] <jmkasunich> man page is up TO date
[03:56:57] <jmkasunich> I must be tired, can't type
[03:57:01] <jmkasunich> goodnight all
[03:57:48] <SWPadnos> heh. good night
[04:32:44] <crepincdotcom> my new mill is coming together nicely
[04:33:04] <crepincdotcom> the small mill died in the middle of making my h-bridge though... *sigh*
[04:33:21] <crepincdotcom> do i get an award for being alone in the shop on friday night?
[04:33:34] <SWPadnos> depends on the shop
[04:34:26] <crepincdotcom> its a pretty nice shop
[04:34:43] <SWPadnos> then no awards ;)
[04:34:55] <crepincdotcom> drat
[04:34:59] <crepincdotcom> my table is sliding nicely
[04:35:00] <eric_u> sounds like its own reward
[04:35:02] <crepincdotcom> 44" by 22"
[04:35:15] <eric_u> that's a pretty big table
[04:35:17] <SWPadnos> rent should be extra when there's no crowd around
[04:35:28] <crepincdotcom> actually its kinda mean... all these huge expensive comercial cnc machines looking at my dinky thing
[04:38:13] <crepincdotcom> oh my god
[04:38:15] <crepincdotcom> life lesson
[04:38:39] <crepincdotcom> NEVER take the nut off of a ball screw unless you want to spend 2 hours finding the balls, cleaning them, and figuring out how to get them back in
[04:40:17] <tomp> get a magnet to help retrieve them. do you know how many you should have?
[04:40:20] <ds3> hahaahahaahahahahahahah
[04:40:39] <ds3> at least that nut is reloadable
[04:41:20] <crepincdotcom> i have no idea how many i should have
[04:41:25] <crepincdotcom> i got about 95%
[04:41:33] <crepincdotcom> i think theres space left for about 4 balls
[04:41:56] <crepincdotcom> ill get some... it just wasnt a good time trying to figure out how they went back in :-/
[04:42:39] <tomp> the size must match, else you change the preload
[04:42:46] <crepincdotcom> yah
[04:42:49] <crepincdotcom> figured as much
[04:43:03] <ds3> think there is suppose to be a loading pattern as the balls may be different sizes
[04:43:12] <crepincdotcom> oh god
[04:43:20] <crepincdotcom> i cant see that theyre different sizes
[04:43:28] <ds3> mic them
[04:43:31] <crepincdotcom> i *really* dont want to measure them all.....
[04:43:39] <SWPadnos> heh - and you need a very good micrometer to see the difference in size - they're specified down to 1 or 10 microns
[04:44:21] <crepincdotcom> oh my god
[04:44:31] <crepincdotcom> i am so screwed
[04:44:30] <ds3> the mititoyos anniversery edition reads down to 0.00001, IIRC
[04:44:38] <crepincdotcom> * crepincdotcom dies
[04:44:42] <tomp> pick them up between 2 parallels and see if some stay and others fall out ( over a deep tray this time! )
[04:44:55] <crepincdotcom> mm now that is a good idea
[04:44:58] <SWPadnos> good parallels
[04:45:03] <crepincdotcom> yeah
[04:45:08] <tomp> lapped parallels :)
[04:45:19] <SWPadnos> parallel parallels
[04:45:20] <eric_u> they alternate the big and small ones, don't they?
[04:45:39] <crepincdotcom> ha
[04:47:10] <ds3> I don't completely see the point of ball screws as my circles are just fine using 60deg V screws
[04:47:45] <tomp> find a deep metal tray (like a bread pan) and put magnets in it, then keep the balls in that tray, and look for the rest of them with another magnet
[04:47:52] <tomp> who made these ball screws?
[04:48:19] <crepincdotcom> no clue, i found the unit next to a dumpster on campus
[04:52:01] <SWPadnos> ds3, ballscrews are nearly twice as efficient as acme or "standard thread"
[04:52:33] <SWPadnos> also, I think they have higher load capacities, though I'm not sure of that
[04:53:13] <ds3> SWPadnos: isn't that at the price of having to deal with back drive?
[04:53:36] <SWPadnos> only on some vertical axes
[04:54:09] <ds3> someone had me convinced that a table on ball screws would like a boat floating in water
[04:54:16] <ds3> is that not true?
[04:54:26] <SWPadnos> I don't think so
[04:54:33] <SWPadnos> there's less friction, but it's not gone
[04:54:42] <ds3> Oh
[04:54:45] <SWPadnos> my Bridgeport stays still quite nicely, thank you
[04:55:14] <ds3> but if you push the table, does your handles spin freely?
[04:55:16] <SWPadnos> I notice especially when I "rapid" it (by hand) ;)
[04:55:23] <crepincdotcom> im trying to decide whether or not to put servos on this
[04:55:26] <SWPadnos> nope. I have to push pretty hard for that to happen
[04:55:28] <crepincdotcom> ds3: mine does
[04:55:42] <crepincdotcom> thats the issue.... with open loop i'm afraid it will "migrate"
[04:55:50] <ds3> crepincdotcom: how big of a machine?
[04:55:50] <crepincdotcom> SWPadnos's is likely a much higer quality though
[04:55:59] <eric_u> the BP we have at work with ballscrews works fine with handwheels.
[04:56:03] <SWPadnos> if I get the screw rotating, then I can keep it going with pressure on the table
[04:56:21] <SWPadnos> it works, but it's hard to move the table by pushing on it, and making the ballscrews spin
[04:56:41] <SWPadnos> it is easier with the ballscrews than it was with the original Acmes
[04:56:43] <crepincdotcom> ds3: table 44"x22"
[04:57:03] <ds3> crepincdotcom: that is huge!
[04:57:07] <eric_u> we have a microscope on an axis with ballscrews and precision rails, and it will ride down when the power dies. can be a pain
[04:57:24] <SWPadnos> that's a vertical, no?
[04:58:02] <ds3> SWPadnos: can it safely take off 0.5 of material while climb cutting without having to hold the handle to brake it?
[04:58:15] <crepincdotcom> ds3: well my old one was 5"x1.75", so I figured I'd move up
[04:58:21] <eric_u> vertical axis, yes. Other axes no problem
[04:58:22] <SWPadnos> well, if I use the table locks, sure ;)
[04:58:28] <ds3> Oh hehehehe
[04:58:55] <SWPadnos> I can hold it pretty steady though - I think I can pretty easily put ~1000 lbs force on the table
[04:58:57] <eric_u> bp cannot do .5 climb cutting
[04:59:22] <SWPadnos> depends on the material and depth ...
[04:59:24] <eric_u> at least the ones I've used bend like a wet noodle
[04:59:39] <SWPadnos> aluminum at 0.050 or so should be fine ;)
[04:59:43] <ds3> a hogging EM should help
[05:00:04] <SWPadnos> yep. the roughing Alu mills I have can take out pretty good chunks
[05:00:28] <crepincdotcom> 5" face mills are pretty nice ;-)
[05:00:30] <ds3> speaking of which... I missed the Enco free shipping sale :/
[05:00:36] <eric_u> al at 050 is about the limit of what I find to be acceptable
[05:00:47] <SWPadnos> it's OK - there'll be another one next week/month or something
[05:01:00] <eric_u> never used roughing mills on al
[05:01:05] <ds3> this Enco one didn't have the usual $50 min purchase
[05:01:17] <SWPadnos> I've done 1.5" cutters at 0.1" depth
[05:01:22] <SWPadnos> or thereabouts
[05:01:39] <ds3> I have done 0.125 conventional in steel on a BP
[05:01:53] <SWPadnos> what cutter width?
[05:01:58] <ds3> 0.125"
[05:02:02] <eric_u> I hope to get mine running soon, hopefully a series II rigid ram can take heavier cuts
[05:02:04] <ds3> i was slotting
[05:02:29] <SWPadnos> ok - that's a lot less metal (but harder, of course)
[05:02:38] <toastydeath> uh
[05:02:53] <toastydeath> i can climb in a2 tool steel, 3/4 wide cutter, climb milling, .5" deep
[05:02:58] <toastydeath> on a bpt series 1
[05:03:17] <SWPadnos> I'm sure it depends largely on the cutter, coating, coolant ...
[05:03:18] <toastydeath> it just takes some preperation and securing the mill
[05:03:26] <toastydeath> no coolant
[05:03:28] <ds3> is that AFTER you heat treat it? ;)
[05:03:40] <toastydeath> before, but since it's A2, it hardens as you cut it
[05:03:43] <ds3> * ds3 whistles innocently
[05:03:51] <toastydeath> as long as you push it hard enough, it doesn't go over RC40
[05:03:55] <toastydeath> and still cuts
[05:04:18] <tomp> snug up the locks and feel the cut, watch the color
[05:04:34] <toastydeath> using a roughing endmill helps a lot
[05:04:36] <bytecolor> TiCN coating is the shizzle
[05:04:38] <toastydeath> five or seven flute
[05:04:56] <ds3> wish they made 1/4" shank roughers
[05:05:36] <toastydeath> lol
[05:05:45] <ds3> i'm serious
[05:05:55] <toastydeath> i think the philosophy is if you can only put 1/4 of an inch in the collet, you shouldn't be using a rougher?
[05:05:59] <ds3> small machine, limited HP... got to make the best use of it
[05:06:16] <ds3> roughers reduce the HP needed for a given cut
[05:06:17] <toastydeath> small machines can't handle the feed of a rougher
[05:06:38] <toastydeath> most machines aren't limited by stalling spindles
[05:06:42] <ds3> it is all relative
[05:06:45] <toastydeath> rather, something else fails
[05:06:49] <toastydeath> all relative to what?
[05:07:11] <ds3> I can stall the spindle with a normal 1/4"
[05:07:18] <ds3> hoping a rougher can help
[05:07:54] <ds3> so, the same 1/4" may do fine on a bigger machine and wouldn't need a rougher; but for a small machine, a 1/4" rougher would help
[05:08:24] <toastydeath> i'm skeptical of the advantage on a machine that small, but not being able to see it, i'll just say fair enough
[05:08:29] <bytecolor> dont think I've ever used a rougher < .375
[05:08:41] <ds3> they make 1/4" roughers with 3/8 shanks
[05:08:57] <crepincdotcom> whats the difference between a rougher and a standard end mill?
[05:09:11] <bytecolor> ah, have to be a stubbly little bastard
[05:09:14] <ds3> better chip clearance I think
[05:09:21] <crepincdotcom> hm
[05:09:29] <toastydeath> no
[05:09:33] <bytecolor> roughers have staggered 'teeth'
[05:09:34] <toastydeath> the rougher has teeth
[05:09:42] <toastydeath> and only cut on part of each tooth
[05:09:49] <ds3> oh
[05:09:58] <ds3> throught that leads to better chip evacuation?
[05:10:02] <SWPadnos> they're often called "chipbreakers"
[05:10:03] <toastydeath> nope
[05:10:12] <eric_u> seems like I have seen very small roughing mills
[05:10:20] <ds3> o I see
[05:10:19] <toastydeath> i'd never heard them called chipbreakers
[05:10:21] <SWPadnos> http://cgi.ebay.com/Nachi-1-4-x-3-8-Roughing-Hog-End-Mills_W0QQitemZ140012673054QQcategoryZ58222QQrdZ1QQcmdZViewItem
[05:10:24] <bytecolor> you have to regrind the flutes of a rougher
[05:10:34] <tomp> 'swedish' end mill, was a staggered tooth design for roughing but cant find them
[05:10:34] <crepincdotcom> so what is the point then
[05:10:37] <crepincdotcom> faster cut rate?
[05:10:40] <SWPadnos> well, often on eBay ;)
[05:10:41] <toastydeath> yeah
[05:10:44] <toastydeath> less force per tooth
[05:10:44] <crepincdotcom> ah
[05:10:45] <bytecolor> the gullet
[05:10:49] <tomp> thats it
[05:10:56] <ds3> always wonder if you can use a tap as a rougher
[05:11:01] <toastydeath> you can't
[05:11:03] <crepincdotcom> oh god
[05:11:05] <SWPadnos> heh - no way
[05:11:08] <crepincdotcom> that sounds like a horrible idea
[05:11:10] <eric_u> the difference in cutting force is quite stark
[05:11:19] <ds3> a rougher looks like a tap (kind of)
[05:11:34] <toastydeath> and a dremel tool looks like a jig grinder
[05:11:34] <crepincdotcom> oh i've seen those
[05:11:41] <SWPadnos> it has banding on the flutes, and there the similarity ends
[05:11:56] <crepincdotcom> ok i know what we're talking about now
[05:12:12] <ds3> corn cobs is another name I heard of
[05:12:50] <ds3> in case anyone else wants to try... burying a 1" rougher 1" deep in aluminum WILL stall a Haas minimill!
[05:12:50] <toastydeath> what are you cutting on the small mill, ds3
[05:13:13] <ds3> toastydeath: mix of alum and steel... it is a general purpose hobby machine
[05:13:26] <ds3> or you mean the Haas?
[05:13:32] <toastydeath> no, the 1/4 spindle
[05:13:39] <crepincdotcom> ds3: we have haas minimills!
[05:13:43] <ds3> that's a Taig
[05:13:44] <crepincdotcom> i stalled one ;-)
[05:14:05] <ds3> crepincdotcom: they are nice machines though
[05:14:11] <crepincdotcom> ehn passable
[05:14:15] <toastydeath> are you running the spindle at 10k?
[05:14:18] <toastydeath> or whatever its limit is
[05:14:22] <crepincdotcom> i've been using the Hass vf2 and vf4 lately
[05:14:27] <crepincdotcom> I usually run it at 6k
[05:14:30] <ds3> I run it at 10K for engraving
[05:14:34] <crepincdotcom> wicked small bits ill put to 10k yeah
[05:14:43] <crepincdotcom> facemills at 6k are fun :-)
[05:14:51] <toastydeath> get a cubic zirconium endmill
[05:14:56] <toastydeath> and run it at 10k for aluminum
[05:14:59] <crepincdotcom> sounds expensive
[05:15:07] <ds3> sounds like it will cost more then my mill ;)
[05:15:26] <toastydeath> it all depends on what you want to do
[05:15:29] <eric_u> every time I go into the shop, the Haas toolroom mill looks more and more like a VNC
[05:16:09] <toastydeath> the smaller the mill's shank, the faster i'd want to run it
[05:16:17] <toastydeath> hell, i want to run a VF-0 at 10k
[05:16:24] <ds3> problem is lack of coolant
[05:16:29] <toastydeath> you don't want coolant at 10k
[05:16:29] <ds3> stock VF0's don't do 10K
[05:16:35] <toastydeath> you'll crack the endmill
[05:16:47] <crepincdotcom> * crepincdotcom goes to cut a peice
[05:16:47] <ds3> 10K needs some HP
[05:16:51] <crepincdotcom> brb
[05:16:59] <toastydeath> 10k, light depth of cut, light feed
[05:17:10] <toastydeath> like .001 or .002 inches/rev
[05:17:11] <eric_u> what about at 80k?
[05:17:28] <toastydeath> better get a ceramic endmill at 80k
[05:17:37] <ds3> I would be scared at 80K
[05:17:39] <toastydeath> actually a ceramic endmill would probably work at 10k
[05:17:41] <eric_u> I have some heads off of a moore tool grinder, 20k, 40k and 80k
[05:17:45] <toastydeath> if you got one that was a full 1/4"
[05:17:46] <ds3> (no encloure)
[05:18:28] <crepincdotcom> * crepincdotcom back
[05:18:48] <tomp> mcmaster carr has a nice page describing roughing endmills
[05:19:00] <toastydeath> ds3, you should try ceramic endmills
[05:19:06] <toastydeath> if you can get one 1/4"
[05:19:11] <SWPadnos> here you go: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=220023632956
[05:19:22] <ds3> toastydeath: who sells them?
[05:19:29] <toastydeath> pretty much all the major folks now
[05:19:31] <toastydeath> valenite, kennametal
[05:19:45] <ds3> let me look at MSC
[05:19:48] <tomp> down to 3/16 at mm
[05:19:55] <toastydeath> i don't know if the surface speed will be high enough
[05:20:01] <toastydeath> in your application
[05:20:04] <crepincdotcom> i have a 10 mil endmil
[05:20:17] <toastydeath> note, ceramic, not carbide
[05:20:46] <toastydeath> although if the sfm won't go high enough a hard grade of carbide will work
[05:21:50] <toastydeath> ooh
[05:21:55] <toastydeath> only ~650 sfm
[05:22:01] <toastydeath> at 10k for 1/4"
[05:22:06] <ds3> toastydeath: the inserts?
[05:22:16] <toastydeath> they make solid ceramic, too
[05:22:22] <toastydeath> but inserts are way more common
[05:22:35] <toastydeath> not gonna find a very good insert mill with 1/4 shank though
[05:23:02] <toastydeath> but you could probably run plain HSS at 10k
[05:23:05] <toastydeath> it's under 700 sfm
[05:23:11] <toastydeath> in aluminum
[05:23:17] <ds3> Holy crap.... $11.43 each insert
[05:23:39] <ds3> and that isn't even the shape I want
[05:23:58] <toastydeath> yeah
[05:24:02] <toastydeath> they go fast though
[05:24:11] <ds3> if I was making money....
[05:24:22] <toastydeath> eh
[05:24:43] <ds3> this is a hobby machine!
[05:24:44] <toastydeath> * toastydeath shrug
[05:25:25] <toastydeath> they make good roughers
[05:25:34] <toastydeath> because as the edge wears down it resharpens itself
[05:25:51] <toastydeath> until it cracks
[05:26:45] <toastydeath> i always break inexpensive tools
[05:26:59] <toastydeath> so i guess buying good stuff and not breaking it winds up being less expensive in the long run
[05:27:07] <toastydeath> i should probably just learn to ease up
[05:27:28] <ds3> apparently the shop class instructor got a bit PO'ed at a student breaking a thread mill... $200 cutter!
[05:28:17] <toastydeath> nice
[05:29:11] <ds3> don't understand why he doesn't just grind back some taps... people say that works
[05:29:34] <toastydeath> i don't know why you'd let a student near a thread mill
[05:29:48] <ds3> it was a CNC class and that was one of the exercises
[05:30:04] <ds3> that cutter has survived at least 6+ classes
[05:30:51] <toastydeath> lol
[05:30:58] <toastydeath> everything eventually breaks in a cnc class
[05:31:42] <ds3> no kidding.... seems like every one of the ejector lips on the ER collet nuts are broken
[05:32:40] <crepincdotcom> heh here people break the ruby probes
[05:32:42] <crepincdotcom> those arent cheap
[05:32:54] <ds3> ruby probes? for the CMM?
[05:33:02] <crepincdotcom> for the hass machines
[05:33:07] <crepincdotcom> to probe the part
[07:32:10] <anonimasu> morning
[07:45:54] <anonimasu> brb
[08:02:23] <CIA-18> 03petev 07TRUNK * 10emc2/src/emc/nml_intf/emcglb.h: -Deleted unused type (nameval) and variable declarations.
[11:53:22] <cncjunior> hi !
[12:34:05] <CIA-18> 03alex_joni 07TRUNK * 10emc2/tcl/bin/ (halconfig.tcl halshow.tcl): internal errors shouldn't be translated, they are only usefull for the initial author
[13:12:03] <CIA-18> 03alex_joni 07TRUNK * 10emc2/src/po/ro.po: updated translation
[13:21:30] <roel1> Hi all
[13:39:27] <maddash> I just figured out why my parport isn't receiving signals
[13:40:00] <maddash> it was in ECP
[13:40:05] <maddash> mode
[13:41:52] <CIA-18> 03jepler 07TRUNK * 10emc2/src/hal/utils/halsh.c: like the tcl command [exec], arrange for any trailing newline in [hal] to be removed
[13:42:43] <CIA-18> 03jepler 07TRUNK * 10emc2/src/hal/utils/ (halcmd.c halcmd_commands.c halcmd_commands.h): add new commands 'stype' and 'ptype' to retrieve the type (e.g., bit) of HAL items
[13:47:42] <CIA-18> 03jepler 07TRUNK * 10emc2/tcl/bin/halshow.tcl:
[13:47:42] <CIA-18> - get rid of a few items that keep halshow from working "standalone"
[13:47:42] <CIA-18> - prefer [hal show] to [exec halcmd show]
[13:47:42] <CIA-18> - use new [hal stype] and [hal ptype] to make watching more robust
[13:47:42] <CIA-18> - prevent an error pop-up when clicking non-leaf pins or signals to watch
[13:50:53] <CIA-18> 03jepler 07TRUNK * 10emc2/tcl/bin/halshow.tcl: - [string trim [hal]] is no longer necessary
[13:54:42] <anonimasu> *yawn*
[13:56:46] <CIA-18> 03jepler 07TRUNK * 10emc2/tcl/bin/halshow.tcl: make watch area scroll when a large number of items are watched
[13:57:58] <maddash> does hal_parport change the transmission mode of the parallel port?
[13:58:24] <maddash> ie, ECP, EPP, SPP, CPP, OPPS, etc?
[14:06:14] <jepler> no. it may set the port to input mode (outb to ioaddr+2) but other than that it assumes the port is already configured for SPP
[14:42:50] <alex_joni> yay, my internet link has been upgraded
[14:42:53] <alex_joni> 4Mbit now
[14:47:31] <cncjunior> hi !
[14:50:36] <roel1> hi junior
[14:55:03] <cncjunior> i tested m51 p0 / p1 spindle override. m51 p0 don't work
[15:01:27] <Jymmmm> PORN RULES!!! http://www.cnn.com/2007/TECH/internet/03/22/internet.blocking.ap/index.html
[15:32:38] <CIA-18> 03jepler 07TRUNK * 10emc2/src/emc/motion/control.c: fix bug reported by cncuser: spindle override disable (m51p0) had no effect
[15:32:48] <jepler> oops, that should have said "cncjunior"
[15:32:51] <jepler> cncjunior: thanks for your report
[15:34:21] <cncjunior> jepler : i think thant is not difficult to fix it
[15:42:33] <anonimasu> *yawn*
[15:43:08] <cncjunior> is there a driver and / or application for a dspace 1104 card under linux ?
[15:46:25] <jepler> no.
[15:47:30] <cncjunior> ok
[15:55:12] <cncjunior> jepler : i look at pwm signal which came out from pasrport with halscope and a real scope.
[15:56:52] <cncjunior> jepler : it' the same shape
[16:00:21] <cncjunior> jepler : what alghoritm is it use to generate this ?
[16:03:00] <jepler> cncjunior: the algorithm for the PWM waveform generation?
[16:03:14] <anonimasu> http://youtube.com/watch?v=r7tT8iEXVM4
[16:03:57] <jepler> cncjunior: The pwm generator works by adding a positive value proportional to the desired duty cycle to accumulator. When the accumulator is greater than zero, the output goes high (if permitted by the maximum frequency).
[16:04:01] <jepler> When the output is high, a larger value is subtracted from the accumulator. As a result, it oscillates around zero to generate PWM or PDM, based on the values that are added and subtracted.
[16:04:15] <jepler> that's a comment from the source file, which you can view online: http://cvs.linuxcnc.org/cgi-bin/cvsweb.cgi/emc2/src/hal/components/pwmgen.c?rev=1.12;content-type=text%2Fplain
[16:08:37] <cncjunior> jepler : thank you
[16:10:03] <Jymmmm> Random thought... Ya know, if I make a 4" flange for my 2.5" ShopVac, I could get lots of volume instead of buying a dedicated dust collector.
[16:10:42] <cradek> I think shop vacs can't run continuously, they'll overheat
[16:11:20] <Jymmmm> I bought a speed control, so I can run it at half speed.
[16:12:08] <cncjunior> jepler : any way, on parport at 50 % i read 2.49 V, at 25 % there are 1.24 V and all the way the function is linear.
[16:13:13] <Jymmmm> cradek: I Have a non-contact thermometer, I'll let it run adn measure it.
[16:14:26] <maddash> http://www.linuxcnc.org/EMC_news_history/IRC-logs/IRC-chat-2004-07-25.txt does not exist. anyone know where it's supposed to be? I wanna check out the homing support of emc2
[16:15:45] <cradek> maddash: http://linuxcnc.org/docs/2.1/html/config/ini_homing/index.html
[16:16:01] <cradek> I can't help with the IRC logs but there's good documentation for homing
[16:16:12] <maddash> grazie, cradek
[16:17:03] <cradek> beware anything written years ago :-)
[16:17:38] <maddash> http://linuxcnc.org/docs/EMC2_User_Manual.pdf --- this link is...dead? clicking it resulted in 3 timeouts...
[16:17:54] <maddash> can anyone confirm the above?
[16:17:56] <maddash> cradek?
[16:18:18] <Jymmmm> works fien here
[16:18:19] <Jymmmm> fine
[16:18:40] <cncjunior> jepler : what i don't understant is why the total period, duty cicle + non duty cicle is different ? it varies with the amount of override .
[16:19:06] <cradek> works for me
[16:19:32] <cradek> cncjunior: sounds like you are in PDM mode, not strict PWM
[16:19:52] <maddash> sounds like it's time that my isp and i had a chat...
[16:20:37] <cradek> bbl
[16:20:47] <cncjunior> jepler : i want to use pwm not pdm
[16:23:14] <maddash> hm, so homing requires the use of axis?
[16:23:25] <maddash> erm, ignore that
[16:45:23] <jepler> cncjunior: have you set pwmgen.N.pwm-freq to the PWM frequency you desire? The default value of 0 selects pdm mode
[16:45:46] <cncjunior> jepler : no
[17:08:49] <MrSunshine> ahh
[17:08:52] <MrSunshine> so there is a channel
[17:09:00] <MrSunshine> having huge compilation problems with emc :/
[17:09:46] <skullworks-PGAB> this is the place
[17:09:46] <MrSunshine> http://paste.cplusplus.se/paste.php?id=5476
[17:12:27] <MrSunshine> anyone know what can be wrong? :)
[17:13:56] <Jymmmm> How do most adjustable air regulators work, by blocking or by leaking?
[17:14:27] <skullworks-PGAB> when they get back to the kyb I'm sure they can figure it out.
[17:14:51] <MrSunshine> kyb?
[17:14:53] <skullworks-PGAB> pressure differential diaphram valves
[17:15:02] <skullworks-PGAB> keyboard
[17:16:00] <SWPadnos> MrSunshine, what version of hal_lib.c do you have?
[17:16:26] <skullworks-PGAB> spring pressure vrs diaphram pressure - determines if valve opens or closes.
[17:16:33] <MrSunshine> took down 2.1-branch from cvs
[17:16:38] <MrSunshine> tried with the released tarball
[17:16:46] <MrSunshine> before that
[17:16:47] <MrSunshine> same error
[17:16:53] <MrSunshine> but looks like it comes from the actual linux headers :/
[17:17:04] <SWPadnos> ok - I don't see an include on line 2097 in the CVS version of hal_lib.c
[17:17:22] <SWPadnos> there is one on line 2112 now
[17:17:50] <MrSunshine> got it compiling
[17:17:55] <MrSunshine> no
[17:17:57] <MrSunshine> i didnt :P
[17:18:03] <MrSunshine> but ALOT less errors
[17:19:33] <jepler> MrSunshine: http://cvs.linuxcnc.org/cgi-bin/cvsweb.cgi/emc2/src/hal/hal_lib.c.diff?r1=1.52;r2=1.53
[17:19:44] <jepler> MrSunshine: apply this patch
[17:19:54] <MrSunshine> ahh cool
[17:19:56] <MrSunshine> thanks :)
[17:22:14] <cncjunior> jepler: now it works fine, thank you for the tip
[17:23:19] <Jymmmm> 9ga steel = .114" ???
[17:24:47] <skullworks-PGAB> I think its thicker
[17:25:09] <skullworks-PGAB> * skullworks-PGAB goes to get his machiner's handbook
[17:25:24] <MrSunshine> Application initialization failed: this isn't a Tk applicationunknown color name "Black"
[17:25:25] <MrSunshine> gah :)
[17:25:48] <MrSunshine> anyone know how many cuttning blades there should be on a milling tool for aluminium? :)
[17:27:03] <skullworks-PGAB> depends on the tool, depth of cut, type of cu
[17:27:36] <skullworks-PGAB> I assume you mean teeth of flutes
[17:28:48] <skullworks-PGAB> 9ga = .1398 to .1495"
[17:28:55] <Jymmmm> ty
[17:28:58] <asdfqwega> logger_aj, bookmark
[17:29:14] <asdfqwega> er
[17:29:19] <MrSunshine> well there are 1 2 3 and 4 and all the way up to 6 "blades" on the piece right? .. tried with a 4 one and it just pumps it full of alu and it stops cutting and just drives throght the material insted :P
[17:29:23] <MrSunshine> thats a 1mm deep cut
[17:29:24] <asdfqwega> logger_emc, bookmark
[17:29:24] <asdfqwega> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-03-24.txt
[17:29:50] <asdfqwega> wow, a 'bot with a 'tude
[17:30:05] <skullworks-PGAB> ry a 2 flute, use cutting oil, WD40 works in a pinch
[17:30:09] <skullworks-PGAB> try
[17:30:45] <Jymmmm> MrSunshineL From what I understand, the more flutes, the finer the finish. So yousuing something like 2 flutes for rough cuts, then more flutes for finer finish work.
[17:33:08] <MrSunshine> Jymmmm, ahh :)
[17:33:53] <skullworks-PGAB> what size cutter you trying to use?
[17:34:12] <jepler> MrSunshine: MrSunshine the "unknown color name "black"" message indicates that your X server configuration is incorrect: the location of the 'rgb.txt' database is unspecified or incorrect
[17:34:24] <jepler> MrSunshine: probably any 'wish' application will fail with the same error
[17:37:10] <jepler> http://ubuntuforums.org/showthread.php?p=858323
[17:38:31] <MrSunshine> where the heck should it be then :/
[17:38:45] <MrSunshine> on my distro its in /usr/share/X11/
[17:40:45] <MrSunshine> and it doesnt help wherever i put it
[17:40:46] <MrSunshine> in those locations
[17:42:51] <skullworks-PGAB> MrSunshine : What Distro?
[17:43:42] <MrSunshine> gentoo
[17:44:03] <asdfqwega> MySunshine: http://www.moldmakingtechnology.com/articles/030401.html
[17:44:58] <skullworks-PGAB> Ouch, I remember gentoo very well. I got lazy and shifted to ubuntu
[17:45:29] <asdfqwega> whoops...MR. Sunshine
[17:45:32] <MrSunshine> ive tried ubuntu, installed wrong drivers 1!! time and blam it all got destroyed, X wouldnt start anymore uninstalled the driver etc but problem still persited :P
[17:45:46] <skullworks-PGAB> I did become very good at compiling kernals during my Gentoo stage though.
[17:47:54] <MrSunshine> stupid rgb.txt
[17:48:13] <MrSunshine> i have it in all the dirs i can find on the net, STILL it doest work
[17:48:46] <jepler> ! grep -i rgb /var/log/Xorg.0.log
[17:48:46] <jepler> (==) RgbPath set to "/etc/X11/rgb"
[17:48:57] <jepler> MrSunshine: try looking in the X server log to find the location it is trying to use
[17:49:06] <jepler> MrSunshine: it is possible that the file is not re-read without restarting X
[17:49:15] <MrSunshine> and there isnt any "Black" in it
[17:49:17] <MrSunshine> its a black
[17:49:20] <MrSunshine> but no Black :P
[17:49:39] <jepler> MrSunshine: well that could be the problem too
[17:49:55] <MrSunshine> cant try to restart X atm either :P
[17:50:02] <MrSunshine> or i can ... humm :)
[17:50:02] <MrSunshine> brb
[17:51:08] <MrSunshine> yey :)
[17:51:14] <MrSunshine> but it crash :/
[17:52:02] <MrSunshine> ImportError: No module named Tkinter
[17:52:02] <MrSunshine> :
[17:52:18] <jepler> MrSunshine: you haven't installed all the necssary packages
[17:52:21] <MrSunshine> python :/
[17:52:27] <MrSunshine> yeah found a page on it :)
[17:54:35] <MrSunshine> how many diferent script languages is used in emc? :P
[17:54:38] <MrSunshine> tcl, python ...
[18:00:51] <MrSunshine> heh, this program will be a hell to make work right? :P
[18:01:23] <skullworks-PGAB> No, not really.
[18:01:39] <skullworks-PGAB> its just your starting point
[18:01:44] <MrSunshine> ok got a question atleast, i can run it out the paralellport of the computer right? :)
[18:02:13] <skullworks-PGAB> Gentoo builds are stripped down - lean builds that are great for speed
[18:02:30] <MrSunshine> i ment, setting up the program when i have it started
[18:02:40] <MrSunshine> im looking at the parts i can run now and im .. confused :P
[18:02:47] <skullworks-PGAB> most use the para port
[18:03:24] <skullworks-PGAB> some people use PCI cards like the mesa 5i20
[18:03:28] <CIA-18> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: too many people thought this message was some kind of error
[18:03:36] <MrSunshine> hal_parport: dlopen: /home/sys2/tmp/hora/emc2.1-branch/rtlib/hal_parport.so: cannot open shared object file: No such file or directory
[18:03:37] <MrSunshine> doh :/
[18:03:38] <skullworks-PGAB> or STG
[18:04:09] <skullworks-PGAB> but most stepper setups run from the para port
[18:04:19] <skullworks-PGAB> many servo setups too
[18:04:37] <MrSunshine> why isnt the parport.so files built? :/
[18:05:24] <jepler> MrSunshine: you built with --enable-simulator -- no hardware can be controlled
[18:05:38] <MrSunshine> ahh
[18:05:38] <jepler> MrSunshine: only with a real-time kernel is emc useful for controlling hardware of any kind
[18:05:41] <skullworks-PGAB> that I don't know - my last Gentoo build was in 2004
[18:05:46] <MrSunshine> but i cant start the parts cause that is missing :P
[18:06:14] <SWPadnos> you can only use the simulator configs
[18:06:22] <SWPadnos> which you can modify, but not far enough to actaully control any hardware ;)
[18:06:31] <skullworks-PGAB> Does your PC have enough power to run the live CD?
[18:06:44] <MrSunshine> amd64 :P
[18:07:07] <SWPadnos> are you running a 64-bit kernel?
[18:07:19] <skullworks-PGAB> mem was the key issue - needs a good 512 to work smooth I here
[18:07:19] <MrSunshine> nah
[18:07:26] <MrSunshine> to much didnt work with 64 bit when i tried last
[18:07:29] <SWPadnos> ok - and single or multiple cores?
[18:07:37] <SWPadnos> yeah - I've noticed that
[18:07:57] <SWPadnos> hmmm. I should boot the liveCD in my big tower, just for fun
[18:08:40] <SWPadnos> but it would be too annoying to waste 75% of the CPU, half the memory, and 99% of the video capability
[18:08:46] <skullworks-PGAB> Question: could your run from a Live CD with your config settings on a USB thumb drive?
[18:08:57] <jepler> skullworks-PGAB: skunkworks does that
[18:09:09] <skullworks-PGAB> an Install less test run?
[18:09:15] <SWPadnos> yes
[18:09:26] <SWPadnos> you can even save some configuration and documents with it
[18:10:34] <MrSunshine> so i have to have a realtime kernel to run it at all ? :)
[18:10:58] <SWPadnos> to run any hardware, yes
[18:11:27] <SWPadnos> skullworks-PGAB, https://wiki.ubuntu.com/LiveUsbPendrivePersistent
[18:11:29] <skullworks-PGAB> building an RT kernal is real fun
[18:11:45] <skullworks-PGAB> or can be
[18:11:58] <jepler> it's no fun -- that's why the emc2 developers provide a pre-built one for dapper
[18:12:45] <eric_u> after you build it about 15 times, it isn't so hard any more
[18:13:35] <SWPadnos> heh
[18:13:54] <SWPadnos> I probably went through 50 kernels during the 2.5 development cycle
[18:14:03] <SWPadnos> not many of them RT though
[18:14:25] <skullworks-PGAB> cool (homework reading assignment)
[18:14:51] <SWPadnos> I'm halfway through that process. skunkworks followed it (and needed to use lilo, FYI)
[18:15:16] <eric_u> about the 15th time was when I realized you could just copy the config from the boot directory and edit that
[18:15:19] <jepler> I may not have built my own kernel post-2.2, except to tweak and rebuild a distro kernel for this or that
[18:15:34] <SWPadnos> yep. I had the tweaker's distro - gentoo ;)
[18:15:44] <eric_u> I do it a lot because we have some robots that have weird hardware
[18:16:01] <SWPadnos> Gentoo could be fun on the Opteron, but I haven't bothered since installing Ubuntu
[18:16:23] <eric_u> I like ubuntu
[18:16:35] <eric_u> we were using redhat, so then we started using fedora
[18:16:42] <eric_u> but the instability is a pain
[18:16:47] <skullworks-PGAB> * skullworks-PGAB thinks ALL robots ARE weird hardware.
[18:17:12] <skullworks-PGAB> but I try to work with them anyway.
[18:17:13] <eric_u> ours have embedded pc's inside
[18:17:57] <SWPadnos> what brand of embedded PC, out of curiosity?
[18:18:01] <SWPadnos> or brands
[18:18:05] <eric_u> advantech
[18:18:15] <skullworks-PGAB> Its real fun when they decide to try and change a part while the machine is running.
[18:18:20] <SWPadnos> ok. seen their stuff before (and I think I still get mail from them)
[18:20:31] <MrSunshine> gah there is supposed to be a rt-sources package for gentoo but it seems its gone with the wind :/
[18:21:29] <skullworks-PGAB> don't think its avail via emerge
[18:21:45] <MrSunshine> aparently :/
[18:21:52] <MrSunshine> tho the gentoo wiki said it was :)
[18:24:33] <asdfqwega> Oh, MAN, that is so cool
[18:24:58] <asdfqwega> I didn't know you could build AXIS as a simulator without an RT kernel
[18:25:15] <asdfqwega> Now I have a way of previewing stuff on my workstation and laptop!
[18:26:07] <asdfqwega> And the OpenGL works better, too
[18:27:56] <skullworks-PGAB> Without RT OpenGL can grab more clock cycles for smoother renditions
[18:32:50] <danielbr> jepler: I'm testing a joypad. in my hadware joypad.axis.1 and joypad.axis.3 seems inveted
[18:33:11] <asdfqwega> skull: It's not just that. Getting OpenGL hardware accellerated with RT is difficult
[18:33:17] <danielbr> up arrows 1
[18:33:23] <danielbr> down -1
[18:33:45] <danielbr> this is the correct i think
[18:33:58] <danielbr> but is inverted here
[18:36:53] <jepler> danielbr: if the direction of some axis is reversed from what you want, use the scale parameter (if you're using hal_input) or mult2 (if you're using hal_joystick, which I don't think has a scale parameter)
[18:37:16] <jepler> danielbr: hal_input and hal_joystick just pass along information from the linux input device drivers to HAL pins
[18:38:32] <jepler> danielbr: on the only joystick I have handy, pulling the stick towards me gives higher counts, and pushing it away gives lower counts
[18:38:34] <danielbr> i'm testing hal-joystick
[18:38:48] <feoc> ello
[18:39:27] <danielbr> thanks jepler i'll test hal-input
[18:39:40] <feoc> any one here know much about the mesa board ?
[18:40:05] <cradek> depends on the question - just ask
[18:40:22] <feoc> well on the pin out wiring
[18:40:35] <feoc> all the even number pins are GND
[18:40:51] <feoc> can i just connect them to 1 juction ?
[18:43:49] <feoc> just seems like a lot of wires all do the same thing if thats the case
[18:45:36] <jepler> yes, there's no reason you can't just connect all the GND pins together on whatever board you're connecting to the mesa
[18:46:39] <jepler> signal integrity is better on a ribbon cable when there is a GND between two signal wires, rather than having signal wires adjacent.
[18:46:56] <jepler> that's one reason for the number of GND signals on the pinout
[18:48:27] <jepler> "Alternating ground lines, paired with each output channel's signal lines improves noise immunity and reduces cross talk."
[18:51:57] <eric_u> nothing worse than having too few grounds
[18:53:52] <eric_u> feoc, how are you using the mesa board?
[18:54:05] <eric_u> I'm setting it up for servos/analog control
[18:57:24] <ds3> system boards are cheap, why not just have a second board connected over ethernet to do the accelerated OpenGL?
[18:57:51] <eric_u> that's not a bad idea
[18:58:07] <eric_u> I'd like to do that
[19:00:40] <eric_u> but I don't think just running emc over the network doesn't keep the emc computer from rendering the opengl
[19:01:08] <ds3> thought that's the whole point of a remote X session
[19:01:12] <ds3> (not VNC)
[19:01:18] <eric_u> I could be wrong
[19:05:01] <cradek> opengl is rendered in the X server itself nowadays
[19:05:15] <eric_u> so it would be done on the remote machine?
[19:06:17] <cradek> it's rendered in the machine nearest your eyes
[19:06:43] <cradek> (I say it like that because "remote" is ambiguous)
[19:07:00] <cradek> in X speak, that's called the X server (the thing with the monitor attached)
[19:07:09] <cradek> the X client is EMC/AXIS
[19:07:50] <eric_u> remote was ambiguous
[19:09:40] <eric_u> obviously it's worth trying before using the nml trick
[19:14:00] <ds3> no EMC specific intelligence required for this experiment ;)
[19:36:36] <MrSunshine> Linux Gothica 2.6.16-rt29 #1 PREEMPT Sat Mar 24 20:24:08 CET 2007 i686 AMD Athlon(tm) 64 Processor 2800+ AuthenticAMD GNU/Linux
[19:36:40] <MrSunshine> :P
[19:43:14] <MrSunshine> what is RT dir supposed to be? :/
[19:48:25] <anonimasu> MrSunshine: the dir where your realtime kernel is installed.. I think..
[19:50:34] <SWPadnos> probably something like /usr/realtime, but it depends on how you install RTAI
[19:55:56] <Twingy> there's a new ver of gcam available if anyone is interested
[19:57:43] <skullworks-PGAB> any good?
[19:59:50] <feoc> eric_u: analog servos also
[20:04:10] <ds3> does gcam support file imports yet?
[20:05:52] <skullworks-PGAB> Twingy: Link?
[20:12:05] <maddash> page25 of the EMC usr man differentiates between a "soft" limit and a "hard" one. I know that soft ones are set using the "emcAxisSetMin*Limit" calls, but what about the latter?
[20:13:21] <SWPadnos> hard limits are physical switches that turn off the motor drivers, usually
[20:13:28] <skullworks-PGAB> Hard limits are switches which directly cut power to drives and trigger an E-stop.
[20:13:43] <anonimasu> yep
[20:14:02] <maddash> now I'm confused. what about soft ones?
[20:14:05] <jmkasunich> if we want to be accurate, there are three kinds of limits (although most machines don't have all three)
[20:14:18] <jmkasunich> soft limits exist only in the software
[20:14:23] <SWPadnos> emc will tell you that a move would go outside the soft limits, and won't perform the move
[20:14:27] <jmkasunich> they can only work after the machine has been homed
[20:14:35] <jmkasunich> (because before homing, it doesn't know where it is)
[20:14:52] <anonimasu> hard limits on the other hand will always work
[20:14:57] <jmkasunich> emc's hard limits are external inputs that make EMC stop with a fault
[20:15:11] <maddash> ah, so it's like "soft"-ware and "hard"-ware
[20:15:14] <anonimasu> the idea is that even if emc craps out.. they should kill your machine instead of crashing
[20:15:18] <jmkasunich> and "true" hard limits turn off the drives
[20:15:49] <anonimasu> with a good setup you shoudlnt ever touch thoose switches if something dosent break
[20:16:08] <maddash> so clearly, the only way hard limits interact with the parport (and hence emc/hal) is via E-stop, no?
[20:16:41] <maddash> this would imply why "emc would stop but can't tell you which axis has faulted."
[20:16:52] <SWPadnos> I don't think that's necessary, but it's common to put limit switches in the E-stop chain
[20:16:55] <MrSunshine> gah .. hal patch ...
[20:16:56] <MrSunshine> etc
[20:17:06] <MrSunshine> where the HELL do i get the damn hal patch :P
[20:17:19] <SWPadnos> what HAL patch are you talking about?
[20:17:25] <jmkasunich> what hal patch?
[20:17:50] <maddash> hal patch? do I need this?
[20:17:56] <MrSunshine> for the kernel
[20:17:59] <MrSunshine> rtai wont install without it
[20:17:59] <MrSunshine> :P
[20:18:10] <jmkasunich> you are making your own rtai kernel?
[20:18:11] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?RtaiSteps
[20:18:12] <MrSunshine> if i need rtai ?
[20:18:20] <maddash> rtai patch?
[20:18:36] <SWPadnos> maddash, don't listen. this very likely doesn't apply to you
[20:18:59] <maddash> :)
[20:19:08] <skullworks-PGAB> Sunshine is building EMC2 on Gentoo
[20:19:19] <jmkasunich> maddash: if you are using ubuntu with a realtime kernel supplied by the EMC team, or using the EMC live CD, then all this stuff is already taken care of
[20:19:35] <SWPadnos> another wiki page about RTAI: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?RtaiInstall
[20:19:40] <MrSunshine> well ther eis a rt kernel source .. ive built it .. but then that isnt enough? :/
[20:19:39] <maddash> jmkasunich: oh dear
[20:20:10] <maddash> jmkasunich: i compiled 2.6.17 de-debian-etchified kernel and only applied an rtai patch....
[20:20:12] <SWPadnos> MrSunshine, you may be getting confused by the fact that EMC has something called HAL, and the kernel does too
[20:20:25] <jmkasunich> complete different things though
[20:20:34] <SWPadnos> right, hence the confusion ;)
[20:20:45] <MrSunshine> SWPadnos, not realy .. but it bugs the hell out of me that nothing ive stubled on so far says that i need a hal patch for the kernel :P
[20:20:56] <MrSunshine> then when i finaly get everything i have to start over form the freakin start
[20:21:11] <SWPadnos> EMC HAL has no kernel patches
[20:21:22] <SWPadnos> RTAI has some patches, which should be called ADEOS
[20:21:28] <SWPadnos> or IPipe
[20:21:45] <SWPadnos> those specifics are explained on the RTAISteps wiki page
[20:21:58] <MrSunshine> i think its you that is confused my friend :P
[20:22:11] <MrSunshine> ive never ever stated that emc hal has patches for the kernel :P
[20:22:14] <SWPadnos> I certainly am confused by what you're having trouble on
[20:22:31] <SWPadnos> but I don't have all the information, and you do so ...
[20:22:47] <MrSunshine> :)
[20:22:54] <MrSunshine> il test following that wiki insted
[20:23:02] <SWPadnos> good plan ;)
[20:23:14] <maddash> MrSunshine: well, there's only one "HAL" here in emc-world...
[20:23:58] <jmkasunich> got to split "getting EMC working (which involves using EMC's HAL) from "getting RTAI working" (which most of us don't mess with, we use a pre-patched kernel)
[20:24:14] <jmkasunich> the latter may involve messing with some kernel-y thing that is also called HAL
[20:25:00] <maddash> "kernel-y thing"?
[20:25:22] <maddash> HAL maybe kernel mode, but it's hardly touches the kernel source
[20:25:47] <jmkasunich> which HAL are you talking about?
[20:26:06] <maddash> isn't there only one HAL?
[20:26:35] <jmkasunich> there is EMC's HAL, which is the only one that most EMC uses ever have anything to do with
[20:26:51] <jmkasunich> and then there is some other HAL that is very low level, related to kernel and rtai stuff
[20:27:15] <SWPadnos> also stands for "Hardware Abstraction Layer", unfortunately
[20:27:19] <jmkasunich> this whole conversation has been about the difference between the two (and the confusion caused by them both having the same name)
[20:27:27] <SWPadnos> I guess it's abstract enough a concept that everyone can use it ;)
[20:27:41] <jmkasunich> MrSunshine is dealing with the "other" HAL, I think - he hasn't been very clear about that
[20:28:46] <maddash> isn't "the other HAL" usually called "RTAI extensions"?
[20:29:19] <jmkasunich> the other hal doesn't extend RTAI, it is used by RTAI
[20:29:29] <MrSunshine> argh someone shoot me :/
[20:29:35] <jmkasunich> I think - like I said, most people (including me) don't have to deal with it
[20:29:58] <jmkasunich> MrSunshine: it would be helpfull if you would actually communicate a little about the nature of your difficulties
[20:30:25] <MrSunshine> jmkasunich, 1. i can only find rtai etc for OLD kernels
[20:31:18] <jmkasunich> where are you looking?
[20:31:34] <MrSunshine> well im searching google as i cant find any site they have or anything
[20:31:42] <MrSunshine> just random links to random pages that has some files to download here and there
[20:32:01] <jmkasunich> https://www.rtai.org/
[20:32:16] <SWPadnos> http://download.gna.org/rtai/testing/v3/
[20:32:24] <SWPadnos> according to RtaiSteps
[20:32:48] <SWPadnos> they don't have any patches for 2.6 later than ~2.6.17 or so
[20:32:51] <MrSunshine> what a good page jmkasunich ... totaly locked up my firefox :/
[20:32:59] <SWPadnos> (haven't looked - that's from memory)
[20:33:18] <jmkasunich> my firefox handles it just fine
[20:33:22] <alex_joni> same here
[20:33:24] <alex_joni> hi guys btw
[20:33:30] <SWPadnos> same here
[20:33:31] <maddash> MrSunshine: install dillo
[20:33:32] <SWPadnos> hi Alex
[20:33:44] <alex_joni> use a proper distro :D
[20:34:05] <alex_joni> SWPadnos: got an advice for a new laptop?
[20:34:29] <MrSunshine> yey there is for my current kernel
[20:34:32] <SWPadnos> I've been thinking about that recently, and I can't decide whether lightweight is more important than fast
[20:35:04] <alex_joni> any feelings about Lenovo?
[20:35:09] <SWPadnos> I suspect that any modern laptop will be "fast enough" for most things, and I can wait a few extra minutes for an FPGA to get placed/routed
[20:35:20] <SWPadnos> not really. I use http://www.mtechlaptops.com
[20:35:37] <SWPadnos> they sell rebranded Clevo or Kapok machines, and I've had very good luck with them
[20:35:47] <SWPadnos> both for stability, hardiness, and Linux drivers
[20:36:14] <SWPadnos> I still use a 6 year old laptop I bought there - just considering replacing it now ...
[20:36:28] <SWPadnos> (a laptop I bought there 6 years ago ;) )
[20:38:38] <ejholmgren> maddash: install what?! O_o
[20:38:39] <alex_joni> SWPadnos: I was thinking about getting a HP one
[20:38:58] <alex_joni> ejholmgren: some exotic vegetable
[20:39:00] <SWPadnos> unfortunately, I have no direct experience with any brand name laptops
[20:39:33] <skullworks-PGAB> re: HP - if possible get a Compaq branded laptop.
[20:39:46] <maddash_> ejholmgren: dillo
[20:40:14] <asdfqwega> I've stuck with IBM thinkpads thus far
[20:40:14] <alex_joni> skullworks-PGAB: why is that?
[20:40:15] <maddash_> maddash_ is now known as sudo_maddash
[20:40:18] <skullworks-PGAB> better support than the HP side of the comapny
[20:40:28] <alex_joni> asdfqwega: yeah, but those are gone nowadays :)
[20:40:37] <alex_joni> skullworks-PGAB: I'm not sure that applies to europe though
[20:40:46] <sudo_maddash> hm, wouldn't it be cool if you could home multiple axes simultaneously?
[20:40:55] <asdfqwega> alex_joni: They can still be found on ebay
[20:40:57] <alex_joni> sudo_maddash: you can
[20:41:08] <alex_joni> asdfqwega: I'm shopping for a Core 2 Duo
[20:41:18] <skullworks-PGAB> My friend was manager of service dept at a computer store - loved the Compaqs
[20:41:33] <sudo_maddash> alex_joni: got a doc for that?
[20:41:43] <skullworks-PGAB> Europe - why not - they all come from asia anyway.
[20:41:43] <SWPadnos> skullworks-PGAB, is that good? ;)
[20:41:59] <asdfqwega> If I was going to get another laptop, to actually take places - I'd get a Toughbook or a Gobook
[20:42:26] <asdfqwega> Rugged, military grade laptops - drop 'em, kick
[20:42:30] <skullworks-PGAB> * skullworks-PGAB would just LOVE a toughbook!
[20:42:30] <asdfqwega> 'em
[20:43:12] <sudo_maddash> ...
[20:43:20] <sudo_maddash> crud, did I disconnect again?
[20:43:33] <skullworks-PGAB> anyway the service dept loved compaq's because they were easy to work on, fast fix, well supported online
[20:43:45] <alex_joni> asdfqwega: yeah, I know.. but 4k$ is too much
[20:43:50] <asdfqwega> sudo_Maddesh: If you did, how are we still reading you?
[20:44:09] <asdfqwega> 4k$? That's only for the newest
[20:44:12] <alex_joni> sudo_maddash: nope, you're still there
[20:44:32] <alex_joni> asdfqwega: don't think there are cheaper rugged laptops (new ones)
[20:44:40] <sudo_maddash> sudo_maddash is now known as maddash
[20:45:02] <asdfqwega> I guess it depends on what you want
[20:45:06] <skullworks-PGAB> the hates Sony
[20:45:45] <skullworks-PGAB> Sony uses security screws, so unless you have the special tools...
[20:46:09] <SWPadnos> that's too funny
[20:46:12] <skullworks-PGAB> Torx with a center post
[20:46:28] <SWPadnos> I don't use screws unless I want to change the CPU (or battery /: )
[20:46:32] <SWPadnos> :/
[20:46:35] <asdfqwega> chuck a torx in the lathe and drill it out :)
[20:46:49] <SWPadnos> I can swap hard disk, floppy, and CD-ROM without tools
[20:46:51] <toastydeath> ^ signed
[20:46:57] <toastydeath> (re: torx)
[20:47:13] <SWPadnos> there are also extra battery packs that can go in one or two of the drive bays
[20:47:15] <skullworks-PGAB> not on most Sony's - HDD is inside - not in a caddy mount
[20:47:25] <SWPadnos> right - that's what's so funny
[20:48:34] <SWPadnos> I paid $2000 or so for mine 6 years aog - P3 1GHz, 512M, 20G, DVD-ROM, floppy, 15" 1400x1050 screen - very reasonable specs for the era and $1000-$2000 less than a comparable Dell, Toshiba, Compaq ...
[20:48:35] <skullworks-PGAB> * skullworks-PGAB contemplates mounting a sony to a mill table to test EMC fixed cyles.
[20:48:50] <SWPadnos> and I can upgrade however I want with off-the-shelf parts
[20:48:56] <SWPadnos> (except the CPU module)
[20:51:16] <asdfqwega> This may seem silly, but my attachment to Thinkpads revolves around one reason: The thumbpoint ("eraser tip")
[20:51:39] <SWPadnos> I guess if you hate those, then Thinkpads wouldn't be such a great choice then ;)
[20:51:50] <SWPadnos> actually, a Mac laptop is a pretty good thing these days
[20:52:20] <SWPadnos> they're pretty comparable in price to similarly configured Dells
[20:52:40] <SWPadnos> and for the most part, a Mac is a Unix terminal with a pretty UI
[20:53:03] <MrSunshine> haha freakin shoot me, ive applied the hal patch to the kernel, /usr/src/linux poiints to the patched tree but when installing rtai i STILL get that the damn patch isnt applied :P
[20:53:48] <SWPadnos> MrSunshine, you may want to ask about that on the RTAI mailing list
[20:54:18] <SWPadnos> A couple of us have compiled RTAI-patched kernels, but we certainly don't do it all that often
[20:57:32] <MrSunshine> ahh now emc compiles with realtime setting atleast :)
[20:59:20] <skullworks-PGAB> progress
[21:00:56] <MrSunshine> what is the usual motors in bigger cnc machines? .. commercial .. some kind of steppers that also or what? :)
[21:01:37] <SWPadnos> usually servos for bigger machines
[21:01:55] <SWPadnos> oops - time to run. bbl
[21:01:56] <skullworks-PGAB> Commercial machines nowdays tend to use AC servos
[21:02:05] <MrSunshine> NOTE: 'rtapi' kernel module must be loaded :/
[21:02:27] <skullworks-PGAB> like Fanuc "Red cap" motors.
[21:03:41] <skullworks-PGAB> Fanuc "Yellow cap" motors are PMDC brush servos
[21:04:19] <skullworks-PGAB> but those are usually only found on machines at least 10 yrs old or older
[21:08:53] <MrSunshine> why didnt i get any rtapi modules? :/
[21:11:07] <MrSunshine> why the hell do you need realtime kernel anyways? :P
[21:13:30] <skullworks-PGAB> well - unless you want to offload the real control of the motors to a seperate real time processor ( such as Berts USB cnc controller) you need a realtime O/S
[21:13:35] <MrSunshine> and stupid live cd, if i do any settings on it, how the hell do i save it? :P
[21:13:55] <skullworks-PGAB> USB pendrive
[21:14:45] <jepler> MrSunshine: ubuntu can also be installed from the live CD
[21:14:46] <MrSunshine> 600 freakin megs :/
[21:14:54] <MrSunshine> i do not want that crappy distro :/
[21:15:07] <jepler> then you have more work to do than someone who will install it
[21:15:37] <jepler> it is rude to complain when you will not take advantage of the work we have done. use it or do not, but don't complain.
[21:15:51] <MrSunshine> or the devs can stop making ubuntu versions and make howtos to get it running in real distros that you dont have to install 200 extra packages to be able to compile a hello world program in
[21:16:09] <MrSunshine> :)
[21:16:14] <jepler> MrSunshine: it is kind of you to volunteer.
[21:16:35] <MrSunshine> ubuntu isnt made for people that do other stuff then browse the web :/
[21:16:36] <jepler> MrSunshine: just so you understand, I think your favorite distro is a sack of manure
[21:17:28] <MrSunshine> i dont realy care, like i said before ive tried ubuntu, installed 1 package wrong and it fucked up the whole system
[21:17:31] <MrSunshine> now thats fun
[21:17:43] <MrSunshine> not even the ubuntu people knew what i should do about it :P
[21:18:28] <jepler> an incompetent user can always make a system unusable.
[21:19:16] <MrSunshine> jepler, i installed the freakin nvidia drivers
[21:19:24] <MrSunshine> and BLAM everything just went to hell
[21:19:28] <MrSunshine> so im incompetent? :P
[21:19:53] <asdfqwega> Hey, I've had a LOT of fun breaking linux
[21:19:57] <MrSunshine> ive never ever managed to **** up gentoo :)
[21:19:58] <asdfqwega> learned alot, too
[21:28:54] <asdfqwega> * asdfqwega fights the temptation to tell MrSunshine to try a 'cat /dev/random >' gag
[21:29:37] <MrSunshine> and you think i started using linux yesterday? :P
[21:29:41] <toastydeath> echo /dev/urandom
[21:29:52] <toastydeath> or cat, rather
[21:43:18] <eric_u> if you cad /dev/random, can't you just stop it with control c?
[21:43:17] <MrSunshine> yes
[21:43:17] <asdfqwega> eric_u: There's more to the joke, and I really don't want to type it out lest someone actually try it [cringe]
[21:43:17] <jepler> bash: cat /dev/random >: No such file or directory
[21:43:17] <jepler> hmph, doesn't work for me
[21:43:17] <eric_u> isn't it better to have them redirect stdin or something?
[21:43:17] <asdfqwega> I think the incantation I'm thinking of involves 'dd if=/dev/urandom'...
[21:43:17] <asdfqwega> ...and I'm going to stop before I say something stupid
[21:43:17] <maddash> does 1 pulse = 1 step?
[21:43:17] <eric_u> of=/dev/hda
[21:43:17] <maddash> anyone?
[21:43:17] <jepler> maddash: yes, if you're talking about steppers
[21:43:17] <maddash> i'm using emc_user_manual.pdf terminology here
[21:43:17] <maddash> jepler: yes, steppers. how does emc know whether it's outputting to a servo or a stepper?
[21:43:17] <jepler> it depends how you hook up HAL
[21:43:17] <jepler> emc doesn't know
[21:43:17] <maddash> ok good
[21:43:17] <skullworks-PGAB> 1 pulse = 1 incrament
[21:43:17] <maddash> so, 22 steps/inch means that my axis should have INPUT_SCALE = 22, right?
[21:43:17] <maddash> because I'm getting joint errors for anything less than ~900
[21:43:17] <jepler> yes, if you are using inch units in your ini file
[21:43:17] <skullworks-PGAB> if your using a stepper driver it could be what ever the driver is set for
[21:43:17] <jepler> the default settings for "following error" are probably inappropriate for INPUT_SCALE = 22
[21:43:17] <maddash> jepler: isn't it 22 steps/inch = 22 pulses/inch?
[21:43:17] <alex_joni> maddash: right
[21:43:17] <alex_joni> although that seems like a fairly low resolution
[21:43:17] <eric_u> you must have big honkin' motors for 22 steps per inch
[21:43:17] <jepler> maddash: The default ini files have MIN_FERROR = 0.010. This means that when the axis is not moving, the commanded and actual position must be within .010 of each other. but 1/22 is .045, so a position 1 step from the desired position is bigger than the permitted following error
[21:43:17] <jepler> (MIN_FERROR = 0.020 from sample_configs/stepper/steper_inch.ini)
[21:43:17] <jepler> (er, .010
[21:43:17] <jepler> )
[21:46:24] <maddash> that would imply that a) *ferror ought to be an integer multiple of input_scale and b) my motors are *not* 22 steps/inch
[21:47:52] <alex_joni> maddash: describe your setup please
[21:48:03] <alex_joni> motor connected to screw, screw moving table, etc..
[21:48:40] <alex_joni> btw, a) should be : "ferror should be at least as large as 1/INPUT_SCALE)
[21:50:11] <maddash> alex_joni: whaaa? 'FERROR is the maximum allowable following error, in user units' -- it would stand to reason that FERROR > MIN_FERROR, no?
[21:50:52] <alex_joni> yeah, meant min_ferror above
[21:50:59] <alex_joni> but both should be bigger than one step
[21:51:17] <maddash> why? and how much?
[21:51:35] <alex_joni> lets assume you can only move 1" in a step
[21:51:57] <alex_joni> say your stepper moves the table 1" on each step
[21:52:12] <maddash> it actually does
[21:52:19] <maddash> erm, not each step
[21:52:21] <alex_joni> if someone writes a program to go to X0.5 then it will only go to 0 or 1
[21:52:21] <maddash> in total
[21:52:34] <alex_joni> I said each step (talking theoretical here)
[21:52:48] <alex_joni> it can't actually reach 0.5 (physically impossible)
[21:53:04] <alex_joni> following so far?
[21:55:29] <maddash> yes.
[21:55:56] <alex_joni> ok.. in this case min_ferror needs to be 0.5" I think
[22:02:40] <maddash> ideally it'd be smaller
[22:02:48] <robin_sz> I always find that a weird parameter ..
[22:02:57] <maddash> why?
[22:03:06] <robin_sz> minimum following error ...
[22:03:22] <alex_joni> maddash: anyways.. I find it hard to believe that your setup is 22 pulses/inch
[22:03:23] <robin_sz> "always make sure we have a following error of at elast blah"
[22:03:32] <maddash> alex_joni: it's not. 1000/inch.
[22:03:38] <robin_sz> its not that, but it confuses my brain from time to time
[22:03:40] <alex_joni> maddash: that sounds more like it
[22:03:49] <maddash> robin_sz: no, that's not what it means. cf. user manual.
[22:03:57] <robin_sz> this I know
[22:04:03] <alex_joni> maddash: he's easy to confuse :P
[22:04:07] <maddash> robin_sz: in fact, it's a bit of a malapropism --- should be "low_speed_ferror"
[22:04:15] <robin_sz> quite
[22:04:28] <alex_joni> not low_speed.. stop_ferror
[22:04:33] <alex_joni> or something like that
[22:05:07] <alex_joni> http://www.youtube.com/watch?v=XJvA5S76ywo
[22:05:11] <maddash> alex_joni: it's 1000, b/c a) I wasn't wearing my glasses when reading off the fine-printed stepper manual and b) i slapped on a gear to increase the steps/inch
[22:05:12] <jepler> joint->ferror_limit =
[22:05:11] <jepler> joint->max_ferror * fabs(joint->vel_cmd) / joint->vel_limit;
[22:05:16] <jepler> if (joint->ferror_limit < joint->min_ferror) {
[22:05:16] <jepler> joint->ferror_limit = joint->min_ferror;
[22:06:01] <maddash> "max_ferror"?
[22:07:06] <alex_joni> I think that refers to max velocity
[22:07:17] <alex_joni> ferror at max velocity
[22:07:23] <maddash> alex_joni: "MIN_FERROR...This is the value by which the axis is permitted to deviate...at very low speeds."
[22:08:07] <maddash> crap - the E-stop isn't working
[22:09:43] <jepler> http://emergent.unpy.net/index.cgi-files/sandbox/ferror.png
[22:10:07] <maddash> jepler: icewm, eh?
[22:10:21] <jepler> for the default stepper config (max velocity 1.2 inch/second, ferror 0.05, min_ferror 0.01) this graph shows the following error at each speed
[22:10:42] <jepler> maddash: yes, you must have a keen eye to identify it with only a small part of the toolbar showing :-P
[22:11:49] <jepler> so you can see that from 0 to some point, the permitted error is 0.01 (min_ferror) then it rises to 0.05 (ferror) at the top speed
[22:12:29] <maddash> yes, this certainly clears it up for me
[22:13:03] <jepler> probably I should only show the first quadrant of the graph
[22:13:33] <maddash> no, it's rather elegant the way it is
[22:13:44] <maddash> how'd you plot this thing?
[22:14:12] <jepler> http://emergent.unpy.net/index.cgi-files/sandbox/ferror2.png
[22:14:14] <jepler> gnuplot
[22:14:41] <jepler> I just fed it a short data file I generated by hand from my understanding of the ferror algorithm I showed part of above
[22:14:44] <jepler> 0 .01
[22:14:47] <jepler> .24 .01
[22:14:49] <jepler> 1.2 .05
[22:21:32] <alex_joni> * alex_joni yawns
[22:21:52] <anonimasu> /usr/local/etc/dhcpd.conf
[22:22:22] <alex_joni> strange place ot hide that
[22:23:34] <maddash> man emcsvr
[22:25:26] <danielbr> jepler: i'm testing hal_input, seems impossible set a parameter or hook some pin with halui adding lines in the hal file
[22:25:55] <anonimasu> alex_joni: it's on bsd..
[22:25:57] <danielbr> emc do not start pin not foud
[22:26:24] <danielbr> or parameter not found
[22:27:16] <alex_joni> danielbr: did you start halui?
[22:27:20] <jepler> danielbr: make sure you use 'loadusr -W' to wait for the component to become ready.
[22:27:22] <danielbr> yes
[22:27:26] <alex_joni> [HAL] HALUI = halui in the ini?
[22:27:54] <danielbr> jepler: i'ii try that
[22:28:02] <danielbr> i'll
[22:33:30] <maddash> HAL: ERROR: pin 'parport.pin-10-in' not found ------wtf??
[22:34:15] <maddash> this occurs when I uncomment the "XLimits" and relevant lines
[22:34:40] <alex_joni> maddash: there's a .0. missing there
[22:34:46] <alex_joni> parport.0.pin-10-in
[22:34:54] <maddash> *yelp*
[22:35:04] <jepler> bbl
[22:35:05] <maddash> hmm, typo, then.
[22:36:07] <maddash> alex_joni: http://cvs.linuxcnc.org/cgi-bin/cvsweb.cgi/emc2/configs/stepper/standard_pinout.hal?rev=1.4;content-type=text%2Fplain
[22:36:26] <danielbr> with loadusr -W all is perfect! thanks again!
[22:36:28] <maddash> alex_joni: I'd change it, but i don't have write access to the cvs...
[22:38:37] <alex_joni> maddash: you can surely change your copy
[22:39:21] <maddash> alex_joni: I changed mine, would it be possible for me to change the cvs one? it'd be nice to make a contribution, albeit a miniscule one...
[22:40:02] <alex_joni> maddash: even reporting it is enough ;)
[22:40:02] <danielbr> seems hal user manual is lacking information about hal_input and hal_joystick
[22:40:06] <alex_joni> I'll fix it in a second
[22:41:07] <danielbr> loadusr -W must be in that doc
[22:41:08] <CIA-18> 03alex_joni 07TRUNK * 10emc2/configs/stepper/ (standard_pinout.hal stepper.tbl stepper_inch.ini): fix parport incantation (noticed by maddash)
[22:41:18] <alex_joni> argh
[22:41:23] <alex_joni> didn't mean the tbl and ini too
[22:41:46] <maddash> alex_joni: er, i hate to be a pain in the ass, but....the last line in that file, too
[22:41:55] <alex_joni> yeah, I changed both
[22:42:47] <maddash> great. now we're one angstrom closer to perfection.
[22:44:47] <maddash> ...that was a joke, fyi...
[22:45:54] <alex_joni> got it
[22:46:07] <alex_joni> but it's not far from the truth :D
[22:48:59] <skullworks-PGAB> " fix parport incantation " - makes it sound like its magic...
[22:49:16] <skullworks-PGAB> and in a way it is
[22:49:18] <alex_joni> skullworks-PGAB: close ..
[22:54:29] <feoc> how do i tell if my encoder inputs are ttl or RS422
[22:54:31] <feoc> ?
[22:55:16] <feoc> mesa manual says The 7I33 also conditions the encoder input signals with input filters for TTL inputs
[22:55:37] <eric_u> do they have 2 lines per signal or one?
[22:55:40] <feoc> or a RS-422 recievers for differential encoder inputs
[22:55:43] <cradek> feoc: does it have A+ A- B+ B-?
[22:55:54] <feoc> one
[22:55:59] <eric_u> then they are ttl
[22:56:01] <feoc> 6 wires come from the encoder
[22:56:12] <eric_u> no index?
[22:56:19] <feoc> it has an index
[22:56:19] <eric_u> oh, sorry
[22:56:36] <eric_u> so what's the extra wire?
[22:57:35] <feoc> A, B ,Index, +vdc, common, gnd
[22:57:48] <feoc> eh
[22:57:55] <feoc> hang on thats not right
[22:59:05] <feoc> yes it is im being a dip
[22:59:53] <skullworks-PGAB> what is the diff between common / gnd?
[23:00:07] <eric_u> not much probably
[23:00:17] <feoc> not a lot i dont think
[23:00:26] <feoc> says case ground
[23:00:35] <skullworks-PGAB> 1 signal gnd / 1 chassis gnd?
[23:00:42] <eric_u> they aren't connected at the encoder then
[23:00:42] <feoc> yah
[23:00:47] <eric_u> connect them both to ground
[23:01:02] <feoc> k
[23:01:48] <eric_u> too bad they aren't differential
[23:01:55] <feoc> whys that
[23:02:09] <eric_u> if you have unexplained movement, you may want to get a converter for the encoder end
[23:02:25] <feoc> what u mean?
[23:02:33] <maddash> good night, guys
[23:02:37] <eric_u> you can get extra counts from noise
[23:02:58] <feoc> oh
[23:03:03] <eric_u> saves on wire though
[23:03:29] <feoc> if its a issue i can always upgrade encoder i guess
[23:04:31] <alex_joni> use twisted pair cables if you get noise issues
[23:04:43] <alex_joni> one ground on each pair
[23:05:05] <feoc> i wouldnt have hought it to be an issue
[23:05:13] <feoc> the old cnc computer use them
[23:05:55] <eric_u> just mentioned it because lots of people have that problem
[23:06:01] <eric_u> then again, lots don't
[23:06:02] <feoc> ok
[23:06:12] <feoc> ill see how it goes i guess
[23:06:31] <feoc> im just trying to get my head around it all
[23:15:26] <danielbr> alex_joni: have you read my email from March,13?
[23:16:24] <danielbr> about tool change
[23:17:06] <toastydeath> why would you need tool change
[23:17:12] <toastydeath> everyone knows one tool DOES IT ALL
[23:18:28] <danielbr> is about a stationary rack tool change
[23:19:31] <alex_joni> danielbr: yes
[23:19:35] <skullworks-PGAB> toasty!
[23:20:00] <skullworks-PGAB> comment from days past
[23:20:09] <toastydeath> ?
[23:20:21] <skullworks-PGAB> I have done threadmilling with an EDM
[23:20:26] <toastydeath> orly!
[23:20:31] <danielbr> ok alex
[23:20:30] <toastydeath> a+
[23:20:52] <danielbr> toastydeath: http://www.youtube.com/watch?v=UJRVX6HyNTU
[23:21:02] <skullworks-PGAB> old Agie CNC sinker with 3R orbiting head
[23:21:11] <toastydeath> that video is clearly a photoshop
[23:21:35] <feoc> how difficult are toolchangers to run in emc?
[23:21:38] <skullworks-PGAB> Poco electrode - #6-32 threads 1/2" deep
[23:21:54] <toastydeath> danielbr: that is a cool toolchanger
[23:22:02] <feoc> mine has moving carousel rack
[23:22:18] <feoc> http://www.youtube.com/watch?v=UJRVX6HyNTU
[23:22:21] <feoc> oops
[23:22:38] <danielbr> yes, emc2 can not do that yet!
[23:23:54] <feoc> its got peumatics in the spindle to lock the head
[23:24:05] <feoc> then the carousel moves in and changes tool
[23:25:09] <danielbr> yes, this is the way that emc2 can do today
[23:25:24] <feoc> i it difficult?
[23:25:25] <danielbr> another video http://video.google.com/videoplay?docid=6680491859906584476
[23:26:41] <skullworks-PGAB> I think EMC can to ATC if most of the ATC is managed by its own PLC
[23:27:24] <feoc> my ATC has no outside contro
[23:27:29] <feoc> so emc would have to do it all
[23:27:46] <skullworks-PGAB> going to use CL then?
[23:28:12] <feoc> CL ?
[23:28:26] <skullworks-PGAB> Classic Ladder
[23:28:31] <jtr> I thinks they handled the Mazak toolchanger in CL (Classic Ladder)
[23:28:36] <danielbr> today is not possible a ATC like i want also with CL
[23:28:44] <feoc> sorry your gonna have to explain
[23:28:59] <alex_joni> feoc: it's a software PLC
[23:29:00] <feoc> iv not really read any further than just getting the basics runing yet
[23:29:08] <alex_joni> programmable logic controller
[23:29:21] <alex_joni> it's a software for defining sequences & actions
[23:29:53] <jtr> s/thinks/think/ - geez - gotta fire the guy that does my typing..
[23:29:55] <feoc> oh?
[23:30:06] <danielbr> this is what i think about this http://sourceforge.net/tracker/download.php?group_id=6744&atid=356744&file_id=220253&aid=1679514
[23:30:12] <skullworks-PGAB> Its an area in the machine control that handles all the little steps - and verifies each prior requirement has been met
[23:30:23] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?LadderProgramming
[23:30:31] <alex_joni> danielbr: I know..
[23:30:39] <feoc> http://img.photobucket.com/albums/v280/rallyslag/31012007175.jpg
[23:30:45] <feoc> those are my machines
[23:31:02] <feoc> http://img.photobucket.com/albums/v280/rallyslag/02012007172.jpg
[23:31:12] <skullworks-PGAB> Ohhh a Hurco with a MaxII
[23:31:14] <feoc> clse up of the mill you can see the too change just about
[23:31:43] <feoc> sorry bou my spelling eyboard batt is dying
[23:32:01] <skullworks-PGAB> what is that a early BMC40?
[23:32:56] <feoc> dont really know a huge lot bout them
[23:33:20] <feoc> the mill isa hurco and thelathe is churchhill with siemens system 8 controls
[23:33:53] <skullworks-PGAB> * skullworks-PGAB used to run a KM3 with Ultimax II - Its dead now - and I want to buy it...
[23:34:49] <skullworks-PGAB> If that Mill uses the CMC Servomate controller you have a great find.
[23:34:54] <feoc> from whati can tell it seems like a farl good machine just needs new computer
[23:35:05] <skullworks-PGAB> yup
[23:35:18] <feoc> yes it does have a servomate
[23:35:30] <skullworks-PGAB> they used Dallas memory chips - got burned by that
[23:35:31] <feoc> also has a brand new eurotherm spindle drive
[23:36:12] <skullworks-PGAB> the Dallas chips had an internal Litium batt good for 10yrs - then they goto crap
[23:36:13] <feoc> yup it was the memor board that went
[23:36:31] <feoc> so the servomate is a good un?
[23:36:40] <skullworks-PGAB> and Dallas quit making them with no pin compatable replacements
[23:36:50] <skullworks-PGAB> yep
[23:36:58] <skullworks-PGAB> tach feed back
[23:37:01] <feoc> whats good about it?
[23:37:07] <skullworks-PGAB> real smooth
[23:37:19] <feoc> cool
[23:37:26] <skullworks-PGAB> you have the servomate manual?
[23:37:40] <feoc> yup
[23:37:40] <skullworks-PGAB> I have the PDF
[23:37:44] <skullworks-PGAB> OK
[23:37:49] <feoc> its on there website
[23:38:02] <skullworks-PGAB> yup - if you dig for it
[23:38:15] <feoc> found it via google
[23:38:20] <feoc> :)
[23:39:28] <skullworks-PGAB> you could used either Jon's PPMC or a Mesa 5i20 with the 0-10v add on and be back up in fairly short order
[23:39:39] <jtr> re: the Dallas chip - There's a company that makes adapter sockets for changing pinouts.
[23:40:12] <feoc> i got the mesa board with the 7133 board :)
[23:40:39] <skullworks-PGAB> Hurco came up with a board mod - adds a backup battery - but the cost to rebuild the board was like $4000
[23:41:01] <jtr> But you've gotta tell them the pinouts you need.
[23:41:18] <jtr> $4000 painful...
[23:41:54] <skullworks-PGAB> Hurco's fix was a change to CMOS mem - and provide battery power
[23:41:54] <feoc> the mill im hoping il have running soo
[23:42:06] <feoc> then i can try lookig at the lathe
[23:42:27] <feoc> thats gonna be a pain everything is hydraulic
[23:42:37] <skullworks-PGAB> I have no clue about the lathe - but I think that Hurco will clean up nice.
[23:42:48] <feoc> it should do :)
[23:43:17] <skullworks-PGAB> just for giggles - how much did it cost you?
[23:43:30] <feoc> 3000 pounds
[23:43:55] <feoc> seen a few other but they were in a real state
[23:44:15] <skullworks-PGAB> does it use CAT40?
[23:44:35] <feoc> think so yeah
[23:44:42] <feoc> got a load of tool with it
[23:44:48] <skullworks-PGAB> great
[23:44:59] <skullworks-PGAB> think you should for that price
[23:46:02] <feoc> bloke i got it from has been helping me out with the cam stuff aswell so it wasnt a bad deal
[23:46:39] <skullworks-PGAB> that really should rock - the original control has a CPU card - with an 8086 and a 8087 side by side running at IIRC 4.077MHz.
[23:47:00] <feoc> oh?
[23:47:33] <feoc> think the computer in it now is about .5 ghz
[23:47:41] <feoc> 1.5
[23:47:41] <skullworks-PGAB> I think it was good for 400IPM rapids
[23:47:44] <feoc> lol
[23:48:27] <feoc> that sounds alright
[23:48:31] <skullworks-PGAB> the KM3 knee mill with the MaxII was good for 250IPM rapid
[23:48:54] <feoc> im onlly gonna be doing ally with it for the most part anyway
[23:49:17] <skullworks-PGAB> the knee mill used 25mm x 5mm pitch ball screws
[23:50:19] <feoc> are the old controls i have worth anything ?
[23:50:26] <feoc> or just bin fodder
[23:51:02] <skullworks-PGAB> working parts could fetch a premium on ebay - as Hurco quit stocking spares for these
[23:51:27] <feoc> the computer did boot up
[23:51:32] <skullworks-PGAB> that consol - I saw one sell for $800
[23:51:39] <danielbr> skullworks-PGAB: ball screws with preloaded nuts?
[23:51:54] <feoc> took pics of he screens ect before i ripped it out
[23:52:17] <skullworks-PGAB> and its just the dual mono CRTs , kyb, E-stop and jog wheel
[23:53:08] <skullworks-PGAB> those are fine ground screws - multi circuit
[23:53:28] <danielbr> i see
[23:53:33] <skullworks-PGAB> prolly less that .0002"
[23:53:56] <feoc> hurco's fairly acurate then ?
[23:54:02] <feoc> even as old machine ?
[23:54:04] <skullworks-PGAB> they were
[23:54:47] <skullworks-PGAB> I made grinding tracer templates for the US Olymic speed skating team on one
[23:55:01] <skullworks-PGAB> 100 inch radius
[23:55:35] <skullworks-PGAB> a step/direction control would have thrown a fit
[23:55:46] <feoc> ah right
[23:55:50] <skullworks-PGAB> tach feedback - smooth as silk
[23:56:55] <feoc> hopeflly ill get the axis's moving tomomrow
[23:57:07] <feoc> best go bed tho
[23:57:29] <skullworks-PGAB> read up on the servo mates smart hard limits
[23:57:39] <feoc> oh?
[23:58:03] <skullworks-PGAB> stops movement towards the limit - but allows power away from limit
[23:58:14] <feoc> ah yeah
[23:58:21] <feoc> best get tehm the righ way around
[23:58:43] <skullworks-PGAB> that is likely allready hard wired
[23:59:05] <skullworks-PGAB> but who knows the total history
[23:59:20] <feoc> i took out all he wiring
[23:59:56] <CIA-18> 03jepler 07TRUNK * 10emc2/docs/man/man1/halcmd.1: document new commands 'ptype' and 'stype'
[23:59:57] <feoc> figued it beter to wire via the manual than to retrace wires and mods ect