#emc | Logs for 2007-07-18

[00:00:49] <toastydeath> lend it to like, the art class
[00:01:15] <toastydeath> they'll all make bongs on it
[00:01:21] <ds2> if all you are doing is straight and ortogonal cuts on a CNC lathe, why bother with CNC
[00:01:33] <toastydeath> because it's hip and trendy
[00:02:04] <toastydeath> they spent craaap on the manual machines, and that gets 99% of the use
[00:02:11] <toastydeath> they have one good lathe (a Leblond)
[00:02:15] <JymmmEMC> toastydeath: you you commissioned them to =)
[00:02:25] <toastydeath> whut?
[00:03:05] <ds2> leblonds aren't very friendly to use for some reason
[00:03:10] <toastydeath> whaaaaaaaat
[00:03:37] <toastydeath> how ISN'T a leblond easy to use
[00:03:46] <toastydeath> they're beautiful lathes!
[00:03:50] <ds2> not sure why, used it twice in the class then avoid it at all cost
[00:03:53] <toastydeath> you can run them without looking
[00:04:00] <toastydeath> wow dude
[00:04:02] <ds2> for one, their collet closer have the rubber grip wheel
[00:04:05] <toastydeath> they've got a learning curve
[00:04:18] <ds2> the lever ones are so much easier to use
[00:04:29] <toastydeath> i never used a collet closer on the leblond at school
[00:04:40] <toastydeath> it's like a 16x40
[00:04:47] <toastydeath> it's got a 12" three jaw on it
[00:05:00] <ds2> then chuck changing on it requires a spanner as it uses a draw nut
[00:05:08] <toastydeath> yar
[00:05:10] <ds2> the D mounts are easier, IMO
[00:05:26] <toastydeath> even so, a L-spindle is easier than changing an A1 or A2 spindle
[00:05:31] <ds2> just the little things... a clausing with variable speed was my favorite
[00:05:41] <ds2> what's an A1/A2?
[00:05:47] <toastydeath> leblond made some of the most accurate lathes around
[00:06:01] <toastydeath> and were fairly heavy duty compared to other lathes in the same category
[00:06:09] <toastydeath> A1/A2 is what CNC lathes use
[00:06:19] <toastydeath> threaded bolts
[00:06:23] <toastydeath> on a short taper spindle nose
[00:06:29] <toastydeath> with a drive lug
[00:06:36] <ds2> oh like a D mount with nuts instead of the cams?
[00:06:43] <toastydeath> similar, yeah
[00:06:49] <toastydeath> and A1/A2 has a power drawbar
[00:06:52] <toastydeath> usually
[00:07:08] <toastydeath> psssh psssh goes the chuck
[00:07:13] <ds2> but I do agree, the leblonds were all huge lathes
[00:07:40] <toastydeath> and accurate, dude
[00:07:42] <toastydeath> unless you got a really beat up old one
[00:07:42] <ds2> but then I don't like the HLV that much either... it is quite but...
[00:07:51] <toastydeath> i guess it's different strokes etc
[00:08:00] <ds2> this is a JC, they probally were abused by students
[00:08:17] <toastydeath> the leblonde is one of those lathes that students have a hard time breaking
[00:08:24] <toastydeath> but it's got a huge learning curve
[00:08:42] <ds2> ours got cracked covers and broken handles to start with
[00:08:49] <toastydeath> ah
[00:09:21] <ds2> I think one of them somehow didn't have enough X travel to let me face
[00:09:57] <toastydeath> weird?
[00:10:02] <toastydeath> was the cross slide all the way forward?
[00:10:05] <toastydeath> =)
[00:10:19] <ds2> don't remember
[00:10:27] <ds2> was the last straw before I decided to stay away
[00:10:42] <toastydeath> you guys had colchesters?
[00:10:45] <toastydeath> have/had
[00:10:48] <ds2> think so
[00:11:02] <toastydeath> yeah, you had good stuff man
[00:11:06] <toastydeath> even if you don't like the leblond
[00:11:12] <toastydeath> i wish we had that
[00:11:43] <ds2> I think the school used to be a lot bigger in this area as some text books mention them... but donno much else beyond that
[00:12:04] <toastydeath> probably
[00:12:14] <toastydeath> most places really cut back since nobody is going into manufacturing anymore =(
[00:12:34] <ds2> in our case, we it is NASA's fault
[00:12:39] <toastydeath> ?
[00:13:00] <skunkworks> our tech school used to have a whole machining department. mazak with pallet changer, puma lathe.. Now I don't even know if they have manual stuff anymore.
[00:12:59] <ds2> NASA was responsible for 50% of the enrollment back when they were big
[00:13:16] <ds2> then NASA cut back at AMES and that enrollment dried up
[00:14:10] <toastydeath> damn
[00:14:15] <toastydeath> to both items
[00:14:17] <toastydeath> that makes me so sad
[00:14:23] <toastydeath> i love manufacturing
[00:15:32] <ds2> they used tohave classes on gear making, CAM making, etc and those are no longer on the catalog
[00:15:51] <toastydeath> damn
[00:16:13] <skunkworks> I actually thought about going into machining at the time.. But descided on electronics.
[00:16:33] <toastydeath> probably a wise decision
[00:16:40] <ds2> electronics isn't in that much of a better shape
[00:16:51] <skunkworks> I don't do either full time. ;)
[00:17:03] <skunkworks> I am in IT which is hurting also ;)
[00:17:07] <toastydeath> yeah
[00:17:17] <JymmmEMC> ds2: what catalog?
[00:17:22] <toastydeath> class catalog
[00:17:41] <JymmmEMC> yeah, WHICH one, we have lots of JC's around here
[00:18:30] <JymmmEMC> SJCC, DeAnza, Evergreen, Foothill, etc
[00:19:03] <toastydeath> the one that had cam and ger making, clearly!
[00:19:12] <JymmmEMC> they all do
[00:19:31] <toastydeath> well obviously not anymore
[00:19:36] <toastydeath> ZING
[00:19:47] <toastydeath> i am bitter because my college never had those things
[00:19:49] <toastydeath> =(
[00:20:23] <JymmmEMC> toastydeath: they didn't even have eletric typrwriters when you were in skool
[00:20:25] <ds2> JymmmEMC: DeAnza
[00:20:34] <JymmmEMC> ds2: SJCC does
[00:20:40] <ds2> SJCC still has gear/cams?
[00:20:51] <ds2> I must look into that one
[00:20:53] <JymmmEMC> ds2: go look
[00:21:04] <JymmmEMC> they did the last time I got a catalog from them
[00:21:31] <toastydeath> JymmmEMC: i am 21
[00:21:33] <ds2> there is in the catalog and then there is offered...deanza had it listed for a long time but not offered
[00:21:44] <JymmmEMC> toastydeath: I stand by what I said
[00:21:48] <toastydeath> hahahahahaha
[00:21:55] <toastydeath> who am i to argue with that =((
[00:22:13] <JymmmEMC> =)
[00:22:14] <skunkworks> MacMasterMike: hi
[00:22:20] <MacMasterMike> hello gang
[00:22:58] <JymmmEMC> Well, I better go dl the TCNC docs, and see if I can get 120IPM or better
[00:23:10] <MacMasterMike> recently dived head first into the CNC world, I picked up two benchmen VMC 5000 at an auction
[00:23:19] <MacMasterMike> and have a question about it
[00:23:54] <toastydeath> impressive
[00:24:01] <MacMasterMike> system boots up logs into windows NT just dandy and does a diagnostic
[00:24:01] <toastydeath> is it a general question or is it specific to your vmc
[00:24:11] <MacMasterMike> kinda specific
[00:24:14] <JymmmEMC> MacMasterMike: Yes, your wife WILL divorce you because you spent zillions of dollars on your new CNC "hobby".
[00:24:17] <ds2> ewwwwwwwwwwwwwwwwww machines running NT?
[00:24:25] <toastydeath> ask anyway but i don't think anyone here can help
[00:24:29] <toastydeath> what's up
[00:24:45] <toastydeath> what control, also
[00:24:59] <JymmmEMC> ds2: shush you
[00:25:16] <MacMasterMike> firstly i get this error as it does a 1 of 12 program boot up
[00:25:17] <MacMasterMike> http://ramzal.com//upload_files/1020142084_MVC-067F.JPG
[00:25:25] <MacMasterMike> vickers acromatic 2100
[00:25:45] <toastydeath> today is like the day of random controls that no one has heard of ever
[00:26:12] <toastydeath> that looks like uh, a windows NT thing
[00:26:18] <toastydeath> not actually like, an error on the hardware
[00:26:33] <MacMasterMike> http://ramzal.com//upload_files/1059631446_MVC-066F.JPG
[00:26:33] <toastydeath> but again, i've never heard of your control so i can't really help
[00:26:42] <JymmmEMC> MacMasterMike: Um, do yo actually have a P drive?
[00:26:42] <MacMasterMike> there is a picture of them
[00:26:46] <toastydeath> oh, it's a tiny thing
[00:26:51] <toastydeath> i was thinking you got a vMC
[00:27:20] <toastydeath> MacMasterMike: you might ask on the cnczone forum, or the home shop machinist forum
[00:27:39] <toastydeath> if nobody there knows, you might ask on practicalmachinist.com
[00:27:40] <MacMasterMike> ok was thinking of writing up a post
[00:27:54] <toastydeath> but you might get laughed at and have your thread locked
[00:27:56] <toastydeath> so like
[00:27:55] <MacMasterMike> ive seen a few posts on that website from googleing
[00:28:02] <toastydeath> try on cnczone and hsm
[00:28:29] <JymmmEMC> MacMasterMike: http://www.practicalmachinist.com/cgi-bin/ubbcgi/ultimatebb.cgi?ubb=get_topic;f=13;t=004066;p=0
[00:28:52] <toastydeath> gg pm
[00:28:53] <MacMasterMike> yeah ive read that thread over
[00:29:11] <JymmmEMC> MacMasterMike: Just somethign I googled, never actually READ it =)
[00:29:38] <MacMasterMike> http://ramzal.com//upload_files/1013311871_MVC-070F.JPG
[00:29:40] <JymmmEMC> MacMasterMike: do you have the original install doscs?
[00:29:46] <JymmmEMC> discs
[00:29:48] <MacMasterMike> i can get into spectraCAM software but get that error
[00:29:47] <JymmmEMC> OH
[00:29:53] <JymmmEMC> You are missing the dongle
[00:30:21] <MacMasterMike> just to clerify that isnt the pendant?
[00:30:32] <JymmmEMC> probably paraport lil thing
[00:30:50] <JymmmEMC> about 1.5" long
[00:30:57] <JymmmEMC> DB25 on each end
[00:31:03] <JymmmEMC> or USB
[00:31:30] <JymmmEMC> MacMasterMike: do you have the original install discs?
[00:32:11] <MacMasterMike> dont think so i have a floppy back up disc
[00:32:18] <JymmmEMC> You MIGHT be able to contact the mfg and see what a replacement dongle costs
[00:33:15] <JymmmEMC> or contact the auction house and see if they can find them for you
[00:33:40] <JymmmEMC> discs and dongles
[00:33:59] <toastydeath> wow
[00:34:04] <toastydeath> impressive, JymmmEMC
[00:34:07] <JymmmEMC> MacMasterMike: what version of NT, 4 ?
[00:34:17] <JymmmEMC> toastydeath: ???
[00:34:23] <MacMasterMike> 3.51
[00:34:31] <toastydeath> i didn't think anyone would be able to help
[00:34:35] <toastydeath> you are like ANSWERMAN
[00:34:38] <toastydeath> a+
[00:34:41] <Skullworks-PGAB> * Skullworks-PGAB cringes
[00:35:05] <JymmmEMC> MacMasterMike: Well, I might be able to help you out doing on OS reinstall, if you can get the discs for the software you have.
[00:35:50] <MacMasterMike> i bet that stuff is hard to come by
[00:35:57] <JymmmEMC> Skullworks-PGAB: shut up before I replace your 64bit with a colecovision
[00:36:04] <MacMasterMike> and pretty sure i dont have it in the boxes of stuff i got with them
[00:36:25] <JymmmEMC> MacMasterMike: Well replacement media is ususally available, if you can prove you have a license.
[00:36:31] <Skullworks-PGAB> * Skullworks-PGAB still has NT4 Workstation...
[00:36:36] <JymmmEMC> might have to pay $50
[00:36:50] <Skullworks-PGAB> not that I would ever want to use it
[00:36:52] <JymmmEMC> Skullworks-PGAB: I got em all baby!
[00:37:01] <JymmmEMC> Skullworks-PGAB: nt4 is good
[00:37:09] <MacMasterMike> http://ramzal.com//upload_files/1073042886_MVC-068F.JPG
[00:37:18] <JymmmEMC> rock solid if you DONT install any M$ apps on it =)
[00:37:36] <Skullworks-PGAB> I'm a W2k fan myself
[00:37:47] <MacMasterMike> i can always play some touch screen solitare on it
[00:38:10] <JymmmEMC> Skullworks-PGAB: I ran a MTa on it for 3 yrs with 99.9999% uptime. The downtime was updates.
[00:38:42] <skunkworks> I have had the same uptime at work with nt and 2000 server.
[00:38:53] <JymmmEMC> liar
[00:38:57] <skunkworks> :)
[00:39:07] <Skullworks-PGAB> WXP= W2k with the serial numbers filed off and a fresh coat of M$ Paint - and all the extra user classes removed.
[00:39:12] <JymmmEMC> I KNOW you dont have that with 2k
[00:40:00] <skunkworks> I have never had a 2000 server lock up/stop working. I only have it on 3 machines though. ;)
[00:40:31] <skunkworks> I think a lot of it is hardware.. I only use supermicro hardware.
[00:40:32] <JymmmEMC> try wks, fun stuff =)
[00:40:52] <Skullworks-PGAB> Some USB will hose up W2k at times
[00:41:09] <JymmmEMC> which you won't usually find on server
[00:41:16] <JymmmEMC> usb devices that is
[00:41:44] <Skullworks-PGAB> thats how I hook up the 500GB drives for offline backups
[00:42:42] <Skullworks-PGAB> tape drives are just too damn slow when you are going thru 2 TB
[00:44:09] <Skullworks-PGAB> when the VP ask why we don't just burn it to DVD's ... I ask "OK who wants to slap him first?"
[00:45:19] <ds2> MMmmm 2TB of DVDs
[00:45:39] <ds2> the reply should be, so youjust volunteer to be the disk changer ;)
[00:46:29] <MacMasterMike> if worse comes to worse think it possible to build new controls for the thing?
[00:47:16] <jmkasunich> sure, use EMC2
[00:48:00] <MacMasterMike> EMC2?
[00:48:05] <Ziegler> oh man
[00:48:17] <jmkasunich> see the topic for this channel ;-)
[00:48:43] <jmkasunich> "Welcome! EMC (Enhanced Machine Controller) is a linux-based opensource CNC control."
[00:49:41] <MacMasterMike> ok thanks
[00:49:52] <Ziegler> http://www.linuxcnc.org/
[00:50:09] <MacMasterMike> i think someday a new controller would be needed as dealing with a 12 year old controller might be tough
[00:50:52] <Ziegler> probably not as bad as you think
[00:51:38] <MacMasterMike> yeah
[01:19:12] <toastydeath> we have controls from the 80's
[01:19:16] <toastydeath> they still work great
[01:19:36] <toastydeath> the only problem is program size, if you have complicated contours
[01:20:00] <skunkworks> we had one from the early 70s - designed in the 60s
[01:20:11] <skunkworks> worked until a year or 2 ago
[01:20:21] <toastydeath> haha awesome
[01:21:53] <Skullworks-PGAB> Hurco - designed to fail - used Dallas memory chips which each contained its own lithium battery
[01:22:15] <Skullworks-PGAB> battery dies - machine dies
[01:22:29] <skunkworks> you can still get them - thanks to you. I ordered one for some of our old motherboards.
[01:23:26] <ds2> the dallas chips can have its Li battery replaced
[01:23:36] <skunkworks> not the one I had.. potted
[01:23:44] <ds2> it can
[01:23:51] <ds2> involves dremels, and hacksaws
[01:23:54] <skunkworks> they where only a few dollars.
[01:23:56] <ds2> people were doing it for the NVRAMs on old Suns
[01:24:08] <ds2> that's if you can find them
[01:25:43] <skunkworks> I just bought one a few months ago
[01:26:03] <skunkworks> batteryoutlet.com or something like that.
[01:26:16] <Ziegler> can ya hardwire them?
[01:26:29] <ds2> it was easier (probally not as reliable) to dremel them and replace it with a standard battery
[01:26:30] <Skullworks-PGAB> Well for like $8000 and no warranty Hurco would rebuild (butcher) the PCB and replace the Dallas with CMOS ram and a external batt.
[02:33:00] <Jymmm> yo ho ho
[02:51:45] <ds2> Hmmm
[02:57:51] <CIA-8> 03jmkasunich 07TRUNK * 10emc2/src/hal/drivers/mesa_5i2x/5i2x_gpio.c: determine whether to export out and oe HAL pins from source and mode, not explicit config vars
[02:59:09] <CIA-8> 03jmkasunich 07TRUNK * 10emc2/src/Makefile: add GPIO code to 5i2x driver
[03:39:25] <toastydeath> has anyone ever used a tracer mill, or seen one used
[03:41:24] <Jymmm> toastydeath: Yeah http://tinyurl.com/2shqe3
[03:43:25] <Jymmm> toastydeath: Would you mind doing me a favor, and change your OTHER nick so your current nick tab completes in less than 42,000 keystrokes? Not like there are THAt many in this channel.
[03:44:36] <Jymmm> Gawd, why in the hell do I ALWASY mix up accel and velocity?!?!?!?
[03:51:09] <toastydeath> i can't do that, dave
[03:51:47] <toastydeath> actually i think i can
[03:51:48] <toastydeath> hold on
[03:52:42] <toastydeath> Jymmm: uh, that's not a tracer?
[03:53:21] <Jymmm> toastydeath: Well, it's the closedt thing I cna think of other than munitions
[03:53:37] <toastydeath> a tracer mill is a hydraulically controlled mill
[03:53:55] <toastydeath> that is semi-automatic, it uses a stylus head riding on a template to control it
[03:54:20] <Jymmm> Ah, you need to ask in #low-rider-mods =)
[04:02:34] <tomp> moog hydropath was a mill with a hydraulic touchprobe, it's deflection controlled the xyz motion. a large ball end probe followed a master form, and a 2nd head with a spindle would mill stock accordingly. i cant google up any images, but used 'em & fixed one.
[04:02:55] <tomp> moog made the hydraulic spool valve
[04:03:13] <Jymmm> tomp: (Hey, I didnt want to say this out loud, but your age is showing =)
[04:03:26] <tomp> btdt
[04:03:29] <toastydeath> well i meant like
[04:03:38] <toastydeath> how does the mechanism work - how does it trace an XY pocket or contour
[04:03:38] <Jymmm> tomp: ok, ya lost me there
[04:03:54] <toastydeath> i see some of these listed as automatic
[04:04:10] <toastydeath> so i'm just wondering how the X-Y decision process works
[04:04:34] <toastydeath> like, if i put a plastic part rigid enough to withstand the stylus pressure, could i mirror it in metal
[04:04:52] <toastydeath> or does the template have to be special made to accomodate the stylus partiularities
[04:06:07] <tomp> yes automatic, but no program, a model maker made a form outta wood and epoxy, the one nose would move toward it till it touched, the other ( same machine smaller ball and a cutter ) would follow on the steel. once the 'master' bumped, it nudged around, say ccw, at a set height, following the form.
[04:07:18] <toastydeath> ahhh that's exactly what i wanted to know
[04:07:37] <tomp> the other spindle would cut stock because it just followed , maybe 'tracer mill' will bring up some images
[04:07:43] <toastydeath> does it go the opposite direction if it LOSES pressure?
[04:07:48] <toastydeath> oh, i've seen plenty of tracer mills
[04:08:06] <toastydeath> i'm just trying to see exactly how far one could push them in replicating existing parts
[04:08:42] <tomp> yes, loss of pressure was kindof 'turnleft' like a roomba algortihm. precision is reductin the ball diameters
[04:09:13] <toastydeath> could you kind of add/remove deflection distance necessary to change apparent tool diameter?
[04:09:23] <toastydeath> like, is there fine adjustment
[04:09:35] <toastydeath> or if i have a .495" endmill, i need a .495" ball
[04:09:40] <toastydeath> no fine adjustment from a .5" ball
[04:09:58] <tomp> http://www.hardymachinery.com/Parts/Large/MIL10.jpg
[04:10:13] <toastydeath> yes, i have seen plenty of them, just never used one
[04:10:24] <toastydeath> the cinci hydrotels look interesting
[04:11:01] <toastydeath> also i like the adjustments on that one
[04:11:03] <toastydeath> that you linked
[04:11:16] <toastydeath> other mills i've looked at don't have XY tracer head fine adjustment
[04:11:37] <Jymmm> heh... "Trace-O-Matic" =)
[04:12:05] <tomp> fine adjust?... only thing i remember was slower feed, smaller diameter, higher rpm
[04:13:20] <toastydeath> so if you re-ground an endmill
[04:13:34] <toastydeath> like, in my example, reground a .500" endmill and it winds up .495"
[04:13:45] <toastydeath> you have to also have a new stylus ground to .495"
[04:13:47] <toastydeath> or use a brand new endmill
[04:14:14] <tomp> oh yeah, of course ( but actually you had a .505 finger for stock .500 ball mills)
[04:14:29] <toastydeath> yeah of course for which
[04:14:36] <toastydeath> Yes there is wear offset
[04:14:38] <toastydeath> or yes you need a new ball
[04:14:57] <toastydeath> also ty very much, you are the first person i've talked to who has ever laid hands on one
[04:15:04] <tomp> if the follower leaves stock you're ok
[04:15:26] <toastydeath> why?
[04:15:32] <toastydeath> you can adjust out, but not in?
[04:15:46] <tomp> because you can get rid of it, but cant put it back ;)
[04:15:55] <toastydeath> but how would you get rid of it using the tracer mill
[04:16:21] <tomp> i just meant it was safe, and a new full sized follower would clean it up
[04:16:24] <toastydeath> oh
[04:16:30] <toastydeath> cool
[04:16:46] <Jymmm> toastydeath: tis okey, I thought the very same thing.
[04:17:00] <Jymmm> toastydeath: you noob! =)
[04:17:04] <toastydeath> lol
[04:17:15] <toastydeath> that is kind of interesting, i'd have thought someone would have come up with a wear offset on the machine
[04:17:18] <toastydeath> oh well
[04:17:31] <toastydeath> still, grinding a new peg is not hard with a small cylinderical grinder
[04:17:36] <Jymmm> toastydeath: actually, I didn't even think about that aspect of the probe needing to be the safe size as the cutter.
[04:17:48] <Jymmm> s/safe/same/
[04:18:15] <toastydeath> lol
[04:18:15] <Jymmm> toastydeath: they did, it's called CNC =)
[04:18:25] <toastydeath> cnc doesn't do the same thing a tracer mill does
[04:18:33] <toastydeath> they still use them in aerospace for some parts
[04:18:34] <Jymmm> toastydeath: but it compensates
[04:18:41] <toastydeath> indeed
[04:18:49] <Jymmm> no wonder the shuttle blew up
[04:19:01] <Jymmm> jk
[04:19:15] <Jymmm> am I hungry yet?
[04:19:20] <tomp> make one with a 3d probe and emc ;) no part program
[04:19:26] <toastydeath> lol
[04:20:04] <Jymmm> tomp: I thought about making a touch probe, but not sure how I'd set it up if it has to be the same size as the cutter.
[04:20:06] <toastydeath> digitizing heads seem like a big waste of time compared to doing things on a tracer
[04:20:15] <toastydeath> if you only have one or two parts
[04:20:36] <toastydeath> i guess it will depend on the specific application
[04:20:36] <Jymmm> oh, have you guys seen the open source prototyper?
[04:20:41] <tomp> or use it to create the part program ( put that 47 Indian up there and get the prints asutomagicly )
[04:22:10] <toastydeath> hahah
[04:22:35] <Jymmm> http://reprap.org/bin/view/Main/WebHome
[04:23:21] <tomp> i'm trying to remember, i think the probe WAS the hydraulic valve driving the axis ( deflection of tip let more or less hyd pressure into one side of x or y piston )
[04:25:15] <tomp> gotta go, nite all
[04:25:24] <toastydeath> my understadning is that it's a set of spool valves
[04:25:26] <toastydeath> cya
[04:26:50] <wholepair> HELLO!!!!
[04:27:30] <wholepair> I am drunk... But I think I can still type well enough to ask a question - is anyone here?
[04:27:52] <toastydeath> sort of
[04:28:11] <wholepair> sort of is good enough I guess -
[04:28:52] <wholepair> I want to change the back ground color of tkemc - do you(toasty death) know how to do this?
[04:29:03] <toastydeath> no
[04:29:09] <wholepair> sh@#$^
[04:29:19] <cradek> you edit the resources file
[04:29:27] <cradek> (after you sober up)
[04:29:38] <cradek> otherwise it'll end up even worse :-)
[04:29:36] <wholepair> why after I sober up?
[04:29:41] <toastydeath> seconded
[04:30:19] <cradek> actually you might get decent results just deleting the resources file - the default colors may be better than the blue (probably wasn't true when it was written - tk used to be much uglier)
[04:30:40] <wholepair> I just finished mitering another steel cross frame - I think I can edit a txt file -
[04:30:53] <wholepair> where is the resources file - then I will go delete it -
[04:30:56] <toastydeath> what does mitering a cross frame have to do with editing text files
[04:31:26] <wholepair> nothing - just trying to communicate how drunk I am -
[04:31:34] <wholepair> not enough to cut off a finger
[04:32:09] <toastydeath> that has more to do with luck
[04:34:59] <wholepair> ok - I agree, I am lucky... I am running UBUNTU, where is the resources file, in root/etc/emc2...
[04:36:04] <wholepair> I don't like the blue because when I am writing gCodes in MDI mode I can't read the txt - it gives me a headache!
[04:36:53] <cradek> /etc/X11/app-defaults/TkEmc
[04:38:50] <wholepair> found it - and should I just delete it you say -
[04:38:54] <toastydeath> rename it
[04:38:59] <toastydeath> tkemc.old
[04:39:01] <wholepair> good idea
[04:39:12] <toastydeath> then, just because
[04:39:18] <toastydeath> i usually do a touch
[04:39:25] <toastydeath> just to bring another file in in case something freaks out
[04:42:21] <wholepair> AWSOME- everything is grey now - I like it -
[04:42:35] <wholepair> thanks :-)
[04:43:43] <wholepair> see ya later - http://bikegeeks.com/PlainTiddlyWiki/
[06:49:13] <Jymmm> are there opto isolated SSR's? Or are they by default?
[07:45:22] <x3rox> cradek: Are you here?
[08:05:00] <x3rox> I think about inserting isolator-IC's into the lines from/to the PC's parallel port (i.e. the ISO721 - IC). Which lines of a parallel port are usually used for a CNC mill? I guess, D0..D7 are used as outputs, but which lines are used as inputs?
[08:05:33] <Jymmm> as many as we can use and abuse =)
[08:05:58] <Jymmm> even some of the outputs are used as inoputs =)
[08:06:10] <Jymmm> but I was speaking of SSR's specifically
[08:06:13] <x3rox> Fine. And what is really needed? I have to put an IC in for every particular signal line!
[08:06:48] <Jymmm> well, unless that IC happens to have more than one opto in it
[08:07:02] <Jymmm> and two power sources too
[08:07:48] <x3rox> This is no opto, it is based on a thick xxx-oxide layer which does the same.
[08:08:29] <x3rox> I assume it will be a good idea to use the parallel port as SPP (standard parallel port)?
[08:09:07] <Jymmm> as an initial test at least,
[08:11:32] <x3rox> Ok. Let's put it together: 3 steppers need 6 outputs. Switching on/off the spindle and coolant needs additional two, results in 8. This is exactly what a parallel port offers with D0...D7. Is this ok?
[08:12:00] <Jymmm> do you need limit or home switches?
[08:12:40] <Jymmm> (inputs)
[08:12:45] <x3rox> This are inputs, so I can use here Paper-End, Error, Acknowledge?
[08:13:01] <x3rox> What is usually used?
[08:13:08] <Jymmm> whatever you like, really.
[08:13:41] <Jymmm> there are some singlas that are iin and outputs, outs only, and in's only
[08:14:16] <x3rox> This requires EPP or ECP port mode, yes?
[08:15:19] <Jymmm> I dont recall the specifics, I *THINK* ECP doens't allow you to alter the in/out assignments, but I could be wrtong which mode allows what
[08:15:21] <x3rox> Don't want to experience what happens when one port pin is set as output and I force some voltage from outside onto it...
[08:16:02] <x3rox> So I prefer standard parallel port, this is the one equal since 1970?
[08:16:12] <Jymmm> thus opto isolation is a good thing =)
[08:17:35] <x3rox> I think so. Strobe and reset are normally also outputs. They can be used like D0..D7? Really?
[08:17:58] <x3rox> Would like to use one to power the PC-side of the isolator.
[08:18:28] <Jymmm> not a good thing to do, use the PC PS lead instead.
[08:19:32] <x3rox> You mean an adapter between keyboard plug and PS/2 connector, or a plug in USB-port?
[08:19:51] <Jymmm> No, like a HDD or FDD power cable
[08:20:05] <Jymmm> red is 5VDC, yellow is 12VDC
[08:20:26] <x3rox> I know. (assemble and sell PCs since >15 years.
[08:20:43] <x3rox> But USB also offers +5V?
[08:20:58] <x3rox> (without opening the computer)
[08:21:03] <Jymmm> and iirc max of 60mA
[08:21:36] <Jymmm> HDD/FDD offer at least 4A minimum (typically)
[08:21:39] <x3rox> Well, should be sufficient for 12 couplers, if they are NO OPTOS? ;-)
[08:22:21] <Jymmm> Try it, if you blow up something becasue too much power draw, you'll know for sure.
[08:22:56] <x3rox> :D USB allows up to 500mA per port.
[08:23:44] <Jymmm> if you want to trust a connector that is made to EASILY be removed, go for it
[08:30:02] <x3rox> I know this disadvantage. But it allows me to put the whole stuff into a little patchbox directly at the PC and use then a normal centronics cable to the mill which is already isolated.
[08:30:28] <x3rox> Are USB -> Parallel adapters supported by EMC?
[08:30:42] <x3rox> I guess, not.
[08:32:30] <x3rox> Jymmm: Can I collect all end switches to one input line of the printer port? Must they be parallel or serially connected to each other?
[08:35:32] <Jymmm> You can, but it's tricky wiring
[08:37:31] <x3rox> It would save some inputs for better usage, therefore I ask. The homing must then always keep all switches unpressed and home every axis until the signal is triggered, then return until this one switch is released again, then do the next axis,... Can EMC do this by default?
[08:38:05] <Jymmm> have you looked at the wiki?
[08:38:50] <x3rox> I have printed out the user manual and I am still reading it, but I saw nothing about that until now.
[11:25:54] <cradek_> cradek_ is now known as cradek
[11:40:11] <martin_lundstrom> Dallur: Are you there,
[11:40:14] <martin_lundstrom> ?
[11:52:53] <maddash> dual xeon board baby!!
[12:13:09] <skunkworks> So - (not that I need to do it..) for now do you have to program the xy/uv separatly for now? for a foam cutter and edm?
[12:13:37] <cradek> yes there's no magic offsetting
[12:14:00] <skunkworks> I was wondering :) NIce work
[12:14:08] <cradek> one day we might want pluggable interpreters to handle those kinds of things
[12:14:29] <skunkworks> did you see the picture of barry? :)
[12:14:30] <cradek> I understand lathes treat UVW in particular strange ways too
[12:14:37] <cradek> yes he looks very content, though partly naked
[12:14:43] <skunkworks> no experience with that.
[12:14:56] <skunkworks> we call him frankenbarry
[12:15:16] <cradek> aww be nice :-)
[12:15:27] <skunkworks> ;)
[12:15:58] <archivist> heh that name conjures up a "look"
[12:16:25] <skunkworks> http://www.electronicsam.com/images/house/barry.JPG
[12:16:40] <skunkworks> he needed to have a absess removed.
[12:17:27] <cradek> he'll have a lot of work to do, growing all that fur back and getting the colors in the right places
[12:17:35] <maddash> why are foam cutters so popular?
[12:17:50] <cradek> for lost foam casting I assume
[12:18:06] <archivist> my cat had a similar arse end for the same reason
[12:18:09] <skunkworks> I think he is up for it. ;)
[12:18:17] <cradek> or I think you can coat the foam with fiberglass? to make model airplane parts
[12:18:31] <cradek> coat it with something anyway
[12:18:40] <cradek> maybe it's epoxy
[12:20:06] <cradek> for relative UV in foam it seems so utterly trivial to 'fix' the generating program to work with emc
[12:20:18] <cradek> too bad it seems like none of it's open source though
[12:20:26] <skunkworks> I made model of airplane wings with just a hot wire. coated it with fiberglass and epoxy
[12:21:16] <maddash> you mean you coated your wire with fiberglass?
[12:21:27] <skunkworks> No - the foam wings. sorry
[12:21:30] <cradek> ha
[12:22:04] <maddash> how'd you get rid of the foam?
[12:23:37] <skunkworks> had metal air foil patterns on each end of the foam block. ran a wire between the head and tail stock of the lathe - ran current through it to make it hot. Pushed it thru the foam.
[12:23:56] <skunkworks> following the metal patterns
[12:24:51] <maddash> but it's just foam -- foam is useless!
[12:27:33] <skunkworks> not for model airplanes.. Never got into the hobby though. made the plane - had someone else test fly it. Then I think I got a girlfriend ;)
[12:28:06] <maddash> 'think'? you mean she's a man?
[12:28:14] <maddash> j/k.
[12:29:27] <archivist> he found himself under a thumb
[12:29:44] <skunkworks> ;)
[12:49:02] <skunkworks> model airplanes didn't seem as important anymore at the time.. ;
[12:49:05] <skunkworks> ;)
[12:50:50] <archivist> so you were told
[13:01:35] <awallin> the program that shows a lot of emc status, is that emctop.py ?
[13:01:42] <awallin> or emcshow.py ?
[13:05:53] <jepler> emctop I think
[13:06:53] <jepler> emcshow was the name of axis a long, long time ago when it was only a previewer
[13:07:34] <awallin> thanks. I am putting down some random thoughts on VCPs and sending to the dev-list right now
[13:10:36] <awallin> in addition to HAL-VCP (control panels that display or set HAL pins) I'd like to see something like NML-VCP which would enable setup the screens a lot of people like (those who want emc to be more like a commercial cnc control)
[13:16:40] <cradek> when you want something for free, first, always start by insulting the developers.
[13:16:43] <cradek> * cradek grumbles
[13:17:07] <awallin> huh? If I have it's not on purpose
[13:17:12] <cradek> no, not you
[13:17:30] <cradek> sorry
[13:18:07] <awallin> the part-changer guy?
[13:18:13] <cradek> yeah
[13:20:20] <cradek> if I try hard not to take offense, his suggestions are probably good
[13:21:11] <cradek> (well maybe not the keyboard stuff)
[13:22:17] <cradek> bbl
[13:23:03] <maddash> what part-changer guy?
[13:23:18] <awallin> on the user-list there is a feature request
[13:23:28] <maddash> link?
[13:23:48] <awallin> don't know if the archives work...
[13:24:04] <awallin> they used to be broken, try the emc sourceforge page and the user-list archives there
[13:25:26] <maddash> 'emc-users,' right?
[13:25:56] <awallin> yeah
[13:26:58] <awallin> he was asking for a screen where the operator could easily view and set all offsets
[13:27:16] <awallin> not a bad idea. I wonder if it can be done with an NML extension to pyVCP
[13:27:40] <awallin> or if it needs to be hard-coded...
[13:55:14] <jlmjvm> Im getting an unexpected realtime error when i start emc now
[13:56:53] <cradek> awallin: I don't think there's currently a way to get those values with NML. that would have to be the first step.
[13:58:34] <skunkworks> jlmjvm: did you change something recently in your configuration? Has it always done it?
[14:00:04] <jepler> the offsets are internal to the interpreter and there's no way to inspect them from outside
[14:00:05] <cradek> awallin: at the same time we should make sure the same functionality for the tool table is there (I think it was partly there before, not sure what the state is today)
[14:00:47] <jepler> (well, in practice you can and don't necessarily get burned every time -- tkemc and mini look at and rewrite the var file)
[14:01:34] <jlmjvm> skunkworks: i put in a different video card a few weeks ago
[14:02:35] <jlmjvm> dont remember seeing the message after the install
[14:05:18] <skunkworks> the install of the video card or emc?
[14:05:38] <jlmjvm> video card
[14:05:47] <skunkworks> when did it start?
[14:05:50] <jlmjvm> emc was working
[14:06:10] <cradek> what's the new video card (and why did you change it?)
[14:07:04] <jlmjvm> gforce4 4800 titanium,my nasa buddie gave it to me
[14:07:28] <cradek> does that mean nvidia?
[14:07:48] <jlmjvm> yes
[14:07:57] <cradek> try taking it back out...
[14:08:14] <jlmjvm> k,will do
[14:08:49] <jlmjvm> thanks
[14:09:02] <cradek> welcome
[14:09:25] <cradek> * cradek flexes his troubleshooting muscles
[14:09:41] <cradek> haha
[14:11:18] <kanzure> NASA buddy
[14:11:19] <kanzure> Hmm
[14:11:36] <skunkworks> ah - yah. I think you should have walked him thru how to change the video driver to visa
[14:11:38] <skunkworks> visa?
[14:11:41] <skunkworks> vesa
[14:12:23] <cradek> maybe so - but I figured if he changed the card once, he can change it back - maybe a bad assumpution
[14:12:30] <skunkworks> ;)
[14:12:30] <cradek> assumpution?
[14:12:43] <skunkworks> oh - I am rubbing off
[14:12:59] <skunkworks> sorry
[14:13:05] <cradek> ha
[14:13:14] <awallin> cradek: is there a list of supported NML command somewhere in the docs?
[14:13:53] <cradek> I suspect all the documentation for that is written in C
[14:13:57] <cradek> all the documentation that's up to date anyway
[14:14:36] <cradek> but jepler is right - I don't know how to get that internal stuff out of the interpreter.
[14:15:08] <awallin> I think emctop shows the current offset that is in use?
[14:15:12] <awallin> but not all of them
[14:15:27] <cradek> sort of. it shows the sum of all the offsets that are in effect. that's in the stat buffer
[14:15:40] <awallin> right.
[14:16:05] <cradek> I'm scared to wade through task and see how that gets there
[14:16:22] <awallin> so before an offset-panel can be created there needs to be NML messaging for doing that added to interp ?
[14:16:24] <cradek> must be a CANON call that sets it
[14:16:51] <cradek> yeah there's an infrastructure problem to solve before any gui problem
[14:17:17] <awallin> lerman is the interp guy, right?
[14:17:42] <cradek> he added the O words
[14:21:21] <cradek> to do this it's almost like we'd have to have a canon call whenever a variable is written to
[14:21:33] <cradek> I don't know all the troubles involved with that
[14:21:47] <cradek> the gui can set a variable with MDI of course
[14:22:24] <awallin> what about reading the offsets?
[14:22:49] <cradek> that's what I mean, I think the interp would have to tell task whenever they're changed
[14:23:28] <awallin> ok. not that easy then?
[14:23:35] <cradek> I don't know how this works well enough to say
[14:24:20] <cradek> I'd also like the gui to show all the variables and how they change throughout the program, but with readahead that's hairy
[14:24:38] <Unit41> is Ubuntu really all that and a bag of chips ?
[14:25:02] <cradek> Unit41: Ubuntu is by far the most popular Linux distribution today
[14:25:35] <Unit41> better than SuSE ?
[14:26:06] <cradek> which distribution is best is a matter of opinion, not fact
[14:26:14] <cradek> (what I said is a matter of fact)
[14:26:29] <Unit41> im asking for your opinion
[14:26:38] <Unit41> << using suse atm
[14:26:51] <Unit41> its alpha though and buggy
[14:27:03] <cradek> I haven't used suse and I think it's dirty because of the microsoft deal
[14:27:12] <cradek> ubuntu is a great distribution
[14:27:12] <Unit41> ??
[14:27:14] <skunkworks> heh
[14:27:27] <Unit41> microsoft deal ?
[14:27:32] <cradek> the novell-microsoft patent wankage
[14:27:48] <Unit41> ew
[14:28:46] <Unit41> im going to dl ubuntu then
[14:28:55] <Unit41> its really new though ?
[14:29:00] <Unit41> seemed to spring out of nowhere
[14:29:17] <cradek> it's been around for a few years now, steadily gaining in popularity
[14:29:29] <cradek> not sure when the first release was
[14:29:31] <Unit41> aslong as its got kde
[14:29:43] <cradek> there is a kde version, but it's not the default
[14:30:11] <cradek> if you want to run EMC2, be sure to get the LTS (long term service) release
[14:30:25] <cradek> that's the one we build packages for (version 6.06)
[14:30:48] <cradek> brb
[14:52:16] <skunkworks> hey ray
[15:04:10] <rayh> Hi skunk
[15:08:18] <skunkworks> You know I am going to ask about the wireless internet.. ;)
[15:09:00] <rayh> The guy is supposed to come out this week with his boom truck and see how high we have to go with the antennas.
[15:09:00] <skunkworks> (when I got satellite internet I had to wait 3 months before an installer was in the area.
[15:09:25] <rayh> yuck
[15:09:38] <skunkworks> they came up from texas.
[15:09:49] <skunkworks> (directway)
[15:11:49] <rayh> This is a "local" microwave guy.
[15:12:05] <skunkworks> Cool. That is good for support then :)
[15:12:17] <rayh> He says I should be able to get between 2-3 meg/sec
[15:12:31] <skunkworks> nice. that is more than I get at home.
[15:12:46] <skunkworks> 1mbs iirc
[15:12:53] <rayh> I'd be happy with 56k+
[15:12:55] <skunkworks> ;)
[16:55:57] <x3rox> cradek: Hello! I am just developing a board with Digital Isolators (ISO 7220M) to protect my printer ports. I actually plan 10 outputs and 6 inputs. Sorely a good idea, isn't it?
[16:57:27] <x3rox> Can pins 10-13,15,16 be used as Inputs?
[16:58:34] <x3rox> 15=/Error 16=Printer Reset (both should be open collector and so usable as inputs, too?
[17:01:20] <x3rox> Anybody else (if cradek is too busy) who can answer my question, please?
[17:10:41] <skunkworks> x3rox: look at page 126
[17:10:45] <skunkworks> http://www.linuxcnc.org/docs/EMC2_Integrator_Manual.pdf
[17:12:18] <x3rox> skunkworks: Thank you very much.
[17:12:48] <skunkworks> no problem
[17:22:03] <x3rox> skunkworks: Strange... It appears to me that both listed modes are _not_ standard LPT modes. One has almost only outputs, the other one inputs. But Pin 16 is normally "Printer Reset" == Output ?
[17:23:00] <alex_joni> x3rox: there is also an 'x' mode
[17:23:26] <x3rox> And this "x" mode is "standard LPT port mode" ?
[17:23:37] <alex_joni> no, standard is output
[17:23:55] <alex_joni> but 'standard' is hard to define
[17:24:29] <alex_joni> input mode is also a standard, ut for ports wchhich support bidirectional mode
[17:24:37] <alex_joni> nowadays all so..
[17:24:44] <alex_joni> I mean do
[17:28:10] <x3rox> Well, I want to make the board in such a way that it is almost 100% "centronics compatible", so it will work with every LPT port. (This is my goal). I would like to use pins 1...9+14 as outputs, and 10...13+15+16 as inputs. Is this possible?
[17:31:45] <alex_joni> x3rox: if you want it to be as compatible as possible, then you need to look at the 'o' mode (output)
[17:33:13] <x3rox> alex_joni: (I would like to get the couplers fully used, so I can get multiples of 2 in each direchtion. The normal "output mode" only shows me 5 inputs. Surely enough for a mill, but if I could get 6 inputs, it would be optimal usage of couplers. Therefore I ask. ;-)
[17:34:35] <alex_joni> x3rox: you can't get more then there are there
[17:34:50] <alex_joni> unless you use the input mode (using D0..D7 as inputs)
[17:35:02] <alex_joni> but hten I suspect you don't have enough outputs..
[17:39:41] <x3rox> alex_joni: Ok, I forget this pin. Docs say that they are open collector and usually read/writable (so it should be possible), but this is pherhaps not implemented in the driver. Doesn't matter, 5 inputs are also sufficient, 10 outputs are more important.
[17:40:40] <x3rox> My layout has a protection resistor, so I can leave it away and then Pin16 is not connected.
[17:40:58] <alex_joni> x3rox: as I said, there is the 'x' mode with the OC stuff in mind, but it doesn't work on all parports
[17:41:20] <x3rox> I must read the manual, it just prints out...
[17:41:56] <x3rox> What is the default way of connecting end-switches to EMC?
[17:42:13] <x3rox> Are they daisy chained and go to one pin?
[17:43:03] <alex_joni> x3rox: EMC2 is quite flexible about that.. you can hook them to individual pins, or chain them together, whatever suits your mood
[17:43:40] <x3rox> :-| But something will be configured by default in the config files, or not?
[17:44:19] <alex_joni> there are various config samples
[17:44:33] <alex_joni> the most common ones don't have limit switches at all
[17:44:52] <x3rox> I see...
[17:48:45] <x3rox> Another thing is: My mill has a static bridge along the X-Axis, moving the spindle (Z-Axis) left/right). The Y-Axis is moving the workpiece forward/backward. So it would be great to have the parking position in the rear left corner to have free access to the workpiece, but the machine-origin (0/0) should be front-left. Is this possible?
[17:49:18] <alex_joni> sure
[17:49:34] <x3rox> A lot of configuring, or easy?
[17:49:57] <alex_joni> you have to read on the HOMING parameters of any axis
[17:50:06] <alex_joni> only 5-10 minutes
[17:50:11] <x3rox> Great!
[17:50:21] <alex_joni> but I'm not sure what you mean under 'parking' position
[17:52:26] <x3rox> When the machine "homes", it shall move up to maximum first, then move to the max. Y, then to minimal X coordinate. In each of this positions, a limit-switch will be triggered. This ends with free access to the workpiece (table id in the front).
[17:53:09] <cradek> you can home an axis in either direction and end at any location along it
[17:53:26] <x3rox> The Zero-Point for working shall be in front-left to have the coordinate system startng in left front.
[17:54:18] <x3rox> Does EMC differentiate between a physical zero point and a zero point relative to the workpiece?
[17:54:33] <alex_joni> yes
[17:54:34] <cradek> that's what the different coordinate systems are for
[17:54:47] <x3rox> (physical == machine's most left/front coordinate).
[17:55:40] <alex_joni> x3rox: lets start by assuming there is only one axis
[17:55:49] <x3rox> Wahh - All this is new to me. I used a borrowed mill with some strange Windoze software (WinPCNC) for a while, where all this is not possible or hidden.
[17:56:03] <alex_joni> you have a min-travel limit and a max-travel limit
[17:56:12] <x3rox> alex_joni: Yes, please explain.
[17:56:17] <alex_joni> the zero position can be anywhere between the two
[17:56:21] <alex_joni> you define it
[17:56:41] <x3rox> Stop.
[17:56:59] <alex_joni> you can have up to 3 different switches on an axis (home switch, max limit switch and min limit switch)
[17:57:12] <alex_joni> * alex_joni stops
[17:57:32] <x3rox> When I switch the machine on, I must somehow synchronize logical zero with physical zero. How's that done?
[17:57:57] <x3rox> This is what I don't understand.
[17:57:58] <skunkworks> Homing..
[17:58:14] <skunkworks> you home the machine.
[17:59:01] <x3rox> Ok. So I home to the rear-left corner. Then I know how much to subtract from Y-Axis to get the front-left and set this value as 0/0. This works?
[17:59:23] <alex_joni> x3rox: hang on a bit..
[17:59:31] <alex_joni> x3rox: for homing you usually push a button
[17:59:42] <alex_joni> then you can have a choice (config-wise)
[17:59:53] <x3rox> Where's this button?
[17:59:57] <alex_joni> either it uses the current position and says this is 0, or whatever value makes you happy
[18:00:07] <alex_joni> (either on the user-interface, or a physical button)
[18:00:32] <alex_joni> or it can determine the 'homing' to take place: start moving in a predefined direction, until a limit switch is hit
[18:01:03] <alex_joni> then more advanced things can happen(if you need it to.. check again for athe same switch with a slower speed, etc)
[18:02:44] <alex_joni> following so far?
[18:03:23] <x3rox> So I can first move Z-Axis up to the limit switch, move down then until the switch is released. This is the max. possible Z-Value. Then I do the same with Y-Axis, then with X-Axis.
[18:04:19] <x3rox> All this I can do automated at startup of EMC?
[18:04:22] <alex_joni> x3rox: not quite.. you don't 'move' it to the max then the lmin
[18:04:22] <alex_joni> you know what the max travel is, no need to f
[18:04:23] <alex_joni> ind it out every time..
[18:05:27] <x3rox> So the machine can actually run into mechanical limits if I program something wrong in the milling job?
[18:05:42] <alex_joni> you just need to hit 'home-all' and it starts moving finding all 3 switches
[18:05:53] <alex_joni> then it will move to a predefined location (which can be 0,0,0 or some other location
[18:07:11] <maddash> hm, I've an incremental encoder attached to my parport, but halscope doesn't show any signals when I twirl the shaft.
[18:07:58] <x3rox> So Home-All syncs EMC's coordinates with the physical ones of the machine, but this is NOT the origin of my tool.
[18:08:35] <x3rox> I mean... origin of my workpiece.
[18:08:54] <skunkworks> your workpiece changes.
[18:09:48] <skunkworks> you would reset your position using your coordinate offset to make 0 where you want it to be at that point. (after the machine is homed)
[18:10:28] <x3rox> So when I want to mill a workpiece, I first move the tool to the workpiece's origin, set this as "working origin". Did I get it?
[18:11:36] <alex_joni> x3rox: yes
[18:11:50] <skunkworks> Emc has a neat button called 'touch off' which does this.
[18:11:58] <alex_joni> maddash: did you add the read and write functions of the parport to a thread?
[18:12:22] <x3rox> Phew. ... What's "touch off"?
[18:13:18] <skunkworks> It is an easy way to set your coordinate system to where you want. (where the cutter is now I want he coordinates to be 4",2",5"
[18:13:51] <x3rox> Ok.
[18:14:18] <martin_lundstrom> Hello everyone
[18:15:36] <x3rox> What's about tool changes? If I have to change the tool during a milling job, the next tool will almost sure have a different length. Therefore I think about mounting a sensor at a fixed place and have the mill sensing the new length after a tool change. Is this possible?
[18:16:16] <x3rox> So the different length shall be automatically compensated...
[18:16:46] <alex_joni> x3rox: yup, it's been done with emc
[18:16:48] <skunkworks> that is what tool tables are for ;)
[18:17:03] <alex_joni> what you also can do is have a tool table (if you have fixed size tools)
[18:17:20] <alex_joni> but if you mount them into a collet, then that won't work.. obviously
[18:17:24] <maddash> alex_joni: yep. twirling doesn't do anything, but when I connect a wire between pin 20 and pin 13 (which is where one of the outputs from my incremental encoder lead to), halscope fires off a spike
[18:17:29] <alex_joni> (at least I think it's called a collet)
[18:17:46] <alex_joni> maddash: then I guess it either doesn't work, or it's not connected right
[18:18:32] <alex_joni> does it stay always low? it might need some pull-ups on the lines
[18:18:35] <skunkworks> is the output open collector? do you maybe need pullups?
[18:18:41] <x3rox> Yes. "Spannzange" in German.
[18:18:41] <skunkworks> uh - yah
[18:19:06] <x3rox> skunkworks: Speaking german, too?
[18:19:21] <alex_joni> x3rox: I do
[18:19:32] <skunkworks> I barely can do english..
[18:19:40] <alex_joni> * alex_joni agrees
[18:19:56] <alex_joni> :P
[18:20:00] <x3rox> So it's easier for both of us... :-D
[18:20:23] <alex_joni> x3rox: nur die anderen werden ziemlich bloed in dir Roehre schauen
[18:20:24] <x3rox> How precisely works a microswitch for sensing the tool-tip?
[18:20:30] <alex_joni> depends on the switch
[18:20:37] <x3rox> Therefore I continue in english ....
[18:20:50] <alex_joni> if you have some kind of optical arrangement.. you can get quite accurate
[18:21:04] <x3rox> Hah ... I am no millinaire!
[18:21:09] <alex_joni> I think cradek used a mechanical switch.. maybe he can shed some light on accuracy
[18:21:38] <alex_joni> x3rox: together with a friend of mine we built 4 encoders 2 weeks ago.. about 30$ overall cost
[18:21:46] <alex_joni> using IR emitters and receivers
[18:22:10] <x3rox> But aren't this sensors sensitive on dust and dirt?
[18:22:23] <alex_joni> sure they are.. but you can build a cover :P
[18:22:25] <skunkworks> mine is within .0007" (microswitch)
[18:22:29] <cradek> my basic mechanical switch worked very well, .0005 repeatability
[18:22:44] <alex_joni> sounds like enough to me
[18:22:44] <cradek> maybe it's better, that's about one full step of my Z motor
[18:22:53] <x3rox> By directly pressing the switch with the tooltip?
[18:22:56] <cradek> yes
[18:23:25] <x3rox> Whow. Much better than I thought. A normal micro switch?
[18:23:53] <cradek> yes I used a basic lever switch
[18:23:59] <skunkworks> yes - from the junk box
[18:24:06] <alex_joni> x3rox: are you a member at cncecke.de ?
[18:24:16] <x3rox> No.
[18:24:38] <alex_joni> you should.. it's a german forum on CNC and with an emc specific place
[18:25:34] <x3rox> So putting together: EMC can do an automated measuring of the tool-length during a tool-change and correct hereby the Z-offset?
[18:25:52] <skunkworks> yes
[18:25:55] <cradek> yes
[18:25:59] <skunkworks> yes
[18:26:04] <alex_joni> yes
[18:26:23] <cradek> http://cvs.linuxcnc.org/cvs/emc2/nc_files/tool-length-probe.ngc?rev=1.4
[18:26:35] <x3rox> :D That's what I need. And this is done with one G-Code, or must I insert multiple commands for that?
[18:26:39] <alex_joni> * alex_joni dissapears for a while..
[18:26:49] <cradek> x3rox: see the example link
[18:26:54] <alex_joni> (correct spelling on that..)
[18:29:20] <x3rox> I really must learn G-code... :-/ so g30 moves above the sensor, but what is "g38.2 z0 f15 (measure)" ?
[18:29:54] <cradek> x3rox: start with the gcode quick reference, it's great
[18:29:59] <x3rox> ... and "g91 g0z.2 (off the switch)" ?
[18:30:22] <cradek> http://www.linuxcnc.org/docs/devel/html/gcode.html
[18:30:29] <cradek> you can click each gcode for detailed information
[18:30:40] <cradek> asking us what each gcode does is silly
[18:31:22] <x3rox> I think I will make my mill working and (so I hope) do not drive you crazy hereby, because I will surely have a lot of questions for the first steps...
[18:31:58] <x3rox> Then I will start with milling in a foam block, this doesn't hurt my tools...
[18:37:22] <x3rox> Though, I must learn G-code to understand what they do. It's really impressive how many codes there are... 8-)
[18:37:48] <cradek> pick 4 or 5 favorites and just stick to those :-)
[18:39:07] <x3rox> Ok, Thanks a lot for your kindness. Now I feel better because I know that what I initially wanted is possible. Really great.
[18:39:21] <cradek> welcome
[18:40:52] <x3rox> Final question:What's the best way to navigate the tool to the workpiece's origin?
[18:41:17] <x3rox> Joystick? Mouse? Buttons (I only have 5 inputs!)?
[18:42:08] <cradek> keyboard or jogwheel
[18:42:18] <x3rox> Where is this connected to?
[18:42:41] <cradek> keyboard is just normal, a jogwheel can be connected to any inputs
[18:42:54] <cradek> (keyboard jogging works well)
[18:43:17] <x3rox> You mean using the cursor keys for X/Y?
[18:43:23] <cradek> yes the arrows
[18:43:31] <cradek> page up, page down jog Z
[18:44:02] <x3rox> and can I move here fast and slow and single step?
[18:44:11] <cradek> yes
[18:44:19] <x3rox> How's this done?
[18:44:37] <cradek> you need to install emc and play with it!
[18:45:58] <x3rox> I used a thin plastic film on top of the workpiece and moved down until I feeled it gets squeezed. This was my way to set Z-origin. In I move down too fast, this costs a tool.
[18:46:19] <x3rox> How do you do taht?
[18:46:28] <cradek> you can jog continuously (as long as the key is held) at any speed, or you can jog incrementally (a set distance for each key press)
[18:47:50] <x3rox> Is a joystick no good option? Could be at the machine and be connected by USB or so...
[18:48:03] <cradek> yes you can use a joystick
[18:48:48] <x3rox> both, analog and digital?
[18:48:55] <cradek> yes
[18:49:41] <x3rox> Fine. I'll end now, it's late here, and read the "Integrator's handbook" at home. Thanks for all your kindly help.
[18:50:03] <cradek> goodnight
[18:54:56] <ds2> does a quadrature input jog wheel require a real time input?
[18:55:09] <cradek> yes
[18:55:28] <ds2> oh @$#!#@%$
[18:56:49] <ds2> has anyone made the joystick trigger input (tradional, ISA card) realtime?
[18:57:28] <cradek> buttons the user pushes don't have to be read in realtime - people are slow
[18:57:52] <cradek> jogwheels are special because you can spin them and generate a lot of edges fast, and you don't want to lose position
[18:58:07] <ds2> I was thinking of using those pins as a quad input
[18:58:21] <ds2> have grand plans to use up the parallel port for other things
[18:58:27] <cradek> we do have a driver that reads the status pins on the serial port
[18:58:40] <ds2> is that drive RT?
[18:58:43] <cradek> that would be fine with a mechanical encoder
[18:58:44] <cradek> yes
[18:58:54] <ds2> ah okay, that would do
[18:59:05] <cradek> (I have never tried it)
[18:59:14] <cradek> beware serial ports are +- 9 or 12 V, not logic level
[18:59:20] <ds2> kind of... just need to rig up a +5V source unlike the nice and convenient joystick ports ;)
[18:59:32] <ds2> *nod*
[18:59:34] <cradek> 0-5 will not work for serial
[18:59:46] <ds2> 0-5V is what most quad encoders want
[18:59:54] <cradek> yes I know
[19:00:00] <ds2> I'll worry about what the serial need later on
[19:00:22] <ds2> but I suspect 0-12V might be enough for the serial port, just donno if the encoder is OC
[19:01:05] <cradek> it's easy to add another parallel port if you have pci slots
[19:01:30] <ds2> was hoping it wouldn't come to that
[19:01:43] <ds2> the +5V on the game port made it very very tempting
[19:02:34] <cradek> how many digital inputs are on a standard game port? can you read them directly with inb?
[19:03:11] <cradek> if so it would be very easy for you to write a realtime driver for it
[19:03:40] <skunkworks> hmm - I wonder how fast the analog inputs to the joystick ports are..
[19:03:51] <cradek> slow slow slow
[19:03:52] <ds2> 4
[19:03:53] <skunkworks> probably slow..
[19:03:53] <ds2> yes
[19:04:18] <ds2> the analog inputs aren't really analog inputs
[19:04:33] <ds2> they are a resistor as part of a 555 wired as a one shot mode
[19:04:41] <skunkworks> Quote <cradek> 'people are slow'
[19:04:52] <cradek> ds2: http://cvs.linuxcnc.org/cvs/emc2/src/hal/drivers/serport.comp?rev=1.6
[19:04:53] <ds2> read procedure is to trigger the 555 and wait for it to fire
[19:05:08] <ds2> cradek: my problem is I don't have a kernel dev enviroment to hack the RT side
[19:05:10] <cradek> a realtime driver can be this simple
[19:05:37] <cradek> mostly documentation and a list of the HAL pins
[19:06:13] <ds2> does the Ubuntu CD give me a build enviroment?
[19:06:32] <cradek> no but you can get it using one command
[19:06:37] <cradek> sudo apt-get build-dep emc2
[19:06:56] <ds2> but that means I need to find a machine around with a HD ;)
[19:07:10] <ds2> got boards, got CDROMs, but no spare HDs lying around
[19:07:13] <cradek> I can only help so much - that's up to you :-)
[19:07:27] <cradek> (actually you could do that while booted live if you really want)
[19:07:28] <ds2> *nod*
[19:07:42] <ds2> was hoping it was there already
[19:07:43] <cradek> I think it's a large download though
[19:07:56] <cradek> no, not possible to put all that on one CD
[19:08:04] <ds2> a DVD image?
[19:08:07] <ds2> =)
[19:08:27] <cradek> get a hard drive and shush
[19:08:53] <cradek> (I think a lot of people don't even have DVD equipment, I don't)
[19:09:03] <ds2> it'll get pushed...
[19:09:24] <ds2> DVD readers are showing up free trash grade equipment so...
[19:09:36] <cradek> not writers
[19:09:45] <ds2> good point
[19:09:49] <cradek> I think they still cost actual money
[19:09:54] <cradek> and you need both
[19:10:02] <cradek> we're not switching to a DVD anytime soon
[19:10:19] <ds2> live netboot image? =)
[19:10:26] <cradek> you definitely could set that up
[19:10:36] <ds2> anyways...
[19:10:57] <ds2> its all a matter of convenience... the most convenient things get finished first
[19:11:16] <ds2> and right now the closest to that is getting my home built digitizing probe tested
[19:23:52] <skunkworks> * skunkworks just bought a dual layer dvd burner for <50 dollars from buy.com
[19:25:51] <skunkworks> with light scribe.
[19:27:05] <skunkworks> TM.
[20:26:37] <maddash> :( inc enc doesn't work.
[20:27:25] <JymmmEMC> try enc inc
[20:36:14] <skunkworks> do you have a scope to look at it?
[20:37:05] <maddash> yeah
[20:37:24] <skunkworks> nothing?
[20:37:49] <skunkworks> we have gotten a few used non-working encoders.
[20:41:39] <maddash> ih510
[20:43:22] <ds2> how much backlash should one expect if the machine is using the lowest grade of ball screws?
[20:43:33] <maddash> http://www.globalencoder.com/main.html
[20:43:56] <maddash> shitloads
[20:44:12] <ds2> 0.100? 0.010? 0.001? which ballpark
[20:44:30] <cradek> .010
[20:44:34] <JymmmEMC> 1.0
[20:44:54] <ds2> so in a worse case scenario, 60deg V screws can be better?
[20:44:56] <maddash> you mean, "lowest," "shitty," or "butt-ugly"? which ballpark?
[20:45:11] <cradek> yeah which "lowest"?
[20:45:21] <cradek> I've seen some really crap ballscrews (the nut rattles around)
[20:45:24] <ds2> lowest numerical backlash figures
[20:45:50] <ds2> I'm going to be looking at an old (80's era) CNC lathe this weekend so I'm doing a crash course on reasonable figures for stuff
[20:45:57] <JymmmEMC> the lowest you can get is the nut literally falls off the screw, so take it from there.
[20:46:12] <ds2> JymmmEMC: assume "new"
[20:46:25] <JymmmEMC> I was =)
[20:46:54] <ds2> I wish I had a frame of reference with something like the ones at Triangle
[20:46:59] <JymmmEMC> mfg defect as example
[20:47:07] <ds2> okay
[20:47:58] <skunkworks> look at mcmaster-carr.. I think they give those specs.
[20:48:02] <JymmmEMC> ds2: obviously you WANT none, but expect some, and if that's within a reasonable value for your application and/or if it can be "fixed"
[20:48:19] <skunkworks> for non-preloaded ball screws
[20:48:35] <ds2> JymmmEMC: from what I have read, there is some on all nonpreloaded screws
[20:48:48] <ds2> to get none, you need a preloaded screw with 2 back to back nuts
[20:49:11] <ds2> skunkworks: mcmaster sells "good" stuff... don't think that'll tell me the range
[20:49:27] <JymmmEMC> ds2: Okey, thomson cheap ballscrews and nut has a spec'ed backlash of .01" per foot
[20:49:35] <JymmmEMC> max
[20:49:51] <ds2> okay, that is useful info
[20:49:52] <skunkworks> JymmmEMC: that is most likely error per foot. Not backlash
[20:50:03] <skunkworks> lead error
[20:50:25] <JymmmEMC> well, it'll serve his needs either way
[20:50:46] <ds2> now in practical terms, does that mean if I grab the carridge on the lathe and push/pull it, the max it should move is 0.010"?
[20:51:10] <JymmmEMC> ds2: cumlitive
[20:51:21] <ds2> hmmmm
[20:51:27] <JymmmEMC> check out their website
[20:51:34] <JymmmEMC> might explain it better
[20:52:07] <archivist> ds2 note endplay on the screw as well not just the nut
[20:52:21] <ds2> oh :S
[20:52:33] <cradek> backlash is not cumulative
[20:53:08] <cradek> so 'backlash per foot' makes no sense
[20:53:16] <cradek> those must be the wrong numbers
[20:53:44] <ds2> okay, maybe I should zoom out ask a different question - is there a good field test to evaluate the condition of a ball screw that is still on the machine?
[20:54:40] <JymmmEMC> Standard Screw Accuracy: in / ft (microns / 300 mm)
[20:55:23] <archivist> dti and jog the machine, dti and lean on the carriage, machine stationary
[20:55:39] <JymmmEMC> Standard Screw Accuracy (in/ft):0.005 per foot
[20:56:33] <JymmmEMC> danaher motion 190-9098
[20:56:37] <ds2> okay
[20:57:16] <JymmmEMC> http://www.danahermotion.com/frame.php?pcid=210&url=http%3A%2F%2Fproducts.danahermotion.com%2Fv41%2FCheckOff.asp%3FUser%3DScrew%26Q%3DProduct+Family%7CBall+Screws%EF%BF%BD%5ERegion%7CNorth%2BAmerica
[20:59:00] <JymmmEMC> ds2: okey, that was accuracy, not backlash
[20:59:25] <ds2> JymmmEMC: figured that's the case
[21:13:50] <JymmmEMC> What would be the generic name for the USC board?
[21:18:51] <JymmmEMC> alex_joni:
[21:19:23] <JymmmEMC> "pulse generator" ?
[21:19:30] <anonimasu> JymmmEMC: hardware pulse generator
[21:19:35] <anonimasu> or hardware step generator
[21:19:38] <JymmmEMC> ty
[21:20:42] <JymmmEMC> is that what the pluto is too?
[21:22:08] <skunkworks> It is a work in process.. but yes. http://axis.unpy.net/index.cgi/01183862490
[21:22:53] <JymmmEMC> nad this is what PCI iface?
[21:23:12] <skunkworks> parallel port
[21:23:58] <JymmmEMC> Ok, this is what I don't quite understand.... if we have a limitation on the paraport, how can "something" drive 300kstep/s via the paraport?
[21:24:07] <skunkworks> If your looking pci - then mesa is your card. But that is something you would want to talk to jmk about.. He is the one adding step generator functionallity to it.
[21:25:00] <skunkworks> because - emc is controlling it as a servo in effect - instead of actually having to send the individual pulses to the stepper.
[21:25:16] <skunkworks> emc can say - go at this velocity.. and the pluto output that step rate.
[21:25:22] <JymmmEMC> Pause a sec...
[21:25:57] <JymmmEMC> I understand step pulses, what is different in the way servos are commanded?
[21:26:11] <skunkworks> * skunkworks is trying to act like he knows what he is talking about.
[21:26:49] <JymmmEMC> skunkworks: continue
[21:27:47] <JymmmEMC> is servo a continous stream? no pulses at all?
[21:28:22] <JymmmEMC> ____|-------------------------------------------------------------|____________
[21:28:31] <JymmmEMC> (make that a solid lines)
[21:29:01] <skunkworks> I am trying to think how to explain this. I terms I understand ;). Normally servo period in emc is 1ms give or take. So emc comands a velocity every 1ms and the pluto correct and sends back its actual position (I think)
[21:29:51] <JymmmEMC> skunkworks: No, I'm still trying to understnad the differnce between step/dir cs servo
[21:29:58] <JymmmEMC> s/cs/vs/
[21:30:41] <skunkworks> I am digging my self into a hole.. (the developers are laughing) ;)
[21:30:43] <JymmmEMC> how are servos driven, I guess is the question
[21:31:15] <JymmmEMC> skunkworks: pretend you are driving a servo with a uC
[21:31:45] <JymmmEMC> is it the cw/ccw thing?
[21:32:33] <archivist> velocity actual v commanded and the servo corrects the error
[21:33:33] <JymmmEMC> archivist: how is the velocity "varied" ?
[21:34:04] <skunkworks> I am saying - with external step generators emc can control them like servos. Servos take velocity command in effect and give back postion. Emc tells the servo to run at a certian speed - and the servos feedback device (like encoder) tell emc where it is. Then emc corrects the servos velocity every ms or so. With steppers - emc has to output a stream of pulses - (much faster than the servo loop) to make the stepper move.
[21:35:06] <skunkworks> like into the .05ms or faster.
[21:35:38] <JymmmEMC> skunkworks: That I understand, treat the stepper like servos, and let the ext step gen deal with the overhead. What I'm trying to distinguish is the differnce in how steppers and servos are driven typically.
[21:36:26] <JymmmEMC> steppers: 200 pulse = one full rotation (as exmaple)
[21:36:31] <JymmmEMC> servo: ???
[21:37:03] <skunkworks> steppers are driven by 2 signals - direction, step. Each on/off step makes the stepper move whatever micro stepping amount.. Full stepping yes - 200 pulses per rev
[21:37:27] <skunkworks> servo normally is controlled by a +/-10v analog signal.
[21:37:31] <skunkworks> but that is just one way.
[21:37:45] <skunkworks> I am controlling mine by a pwm signal
[21:37:45] <JymmmEMC> ah, analog.
[21:38:14] <JymmmEMC> how are you getting analog from the paraport?
[21:38:37] <skunkworks> directly from the printer port - PWM works great.
[21:39:00] <skunkworks> I am using the pluto - with outputs a pwm signal in relation to the command given to it.
[21:39:20] <JymmmEMC> single pin per motor?
[21:40:01] <skunkworks> 2 pins. one pwm pin for clock wise - one pwm pin for CCW
[21:40:36] <JymmmEMC> ok, but both pins can't be "on" at the same time can they? Or is that like a braking thing?
[21:40:47] <skunkworks> No.
[21:42:42] <skunkworks> internal to emc - the pluto has a float input for the pwm signal. positive numper make one pin output pwm - negative number makes the other pin output pwm
[21:42:59] <skunkworks> if that makes sense
[21:43:44] <skunkworks> unless you offset the signal.. but that is for another day ;)
[21:44:28] <skunkworks> (output pwm in relation to the size of the number also)
[21:45:05] <skunkworks> Ok - I have to go home.. bbl
[21:45:21] <jepler> JymmmEMC: if you hook a DC voltage across a servo motor, it turns at a rate that is roughly proportional to the voltage
[21:45:26] <jepler> to reverse it, you reverse the voltage
[21:46:14] <jepler> if you hook the DC voltage in the same direction, but disconnect it half the time, it turns at a rate that is roughly proportional to half the voltage
[21:46:34] <jepler> (similar to hooking half the voltage up all the time)
[21:46:45] <jepler> that's why the "duty cycle" of the PWM is roughly proportional to the speed
[21:47:41] <JymmmEMC> jepler: Ok, to the two pins required to drive a servo are to just control the CW (neg lets say) and CCW (positive) ?
[21:48:36] <jepler> JymmmEMC: right. if UP=0 then the "left" side of the motor is connected to GND; if UP=1 then the "left" side of the motor is connected to VDRIVE. DOWN does the same thing but for the "right" side of the motor
[21:49:29] <jepler> these logic-voltage pins actually connect to transistors or mosfets which do the switching
[21:49:45] <jepler> one servo motor is controlled by the same arrangement of transistors that control 1/2 of a stepper motor -- the "h-bridge"
[21:50:14] <JymmmEMC> so, both cw and ccw signals will never be active at the same time?
[21:50:43] <jepler> generally not
[21:51:03] <jepler> UP,DOWN=0,0 and UP,DOWN=1,1 are both stop/brake configurations
[21:51:13] <jepler> UP,DOWN=1,0 and UP,DOWN=0,1 are both "full speed" configurations
[21:51:22] <JymmmEMC> Ok, so TECHNICALLY, you could use one pin in a -128 0 +128 configuration, but not have near the resolution
[21:51:24] <jepler> UP,DOWN=.5,0 and UP,DOWN=0,.5 are both "half speed"
[21:52:10] <JymmmEMC> if the drive supported that config that is
[21:52:13] <jepler> besides resolution, it's nice to be sure you can output something that is sure to mean "stop/brake"
[21:52:29] <JymmmEMC> Ok, I get that =)
[21:52:50] <jepler> but yes -- the "10V servo drives" are just one wire (analog) to transmit clockwise, stop, and counterclockwise
[21:53:01] <jepler> you could do something similar with digital servo drives
[21:53:21] <jepler> but I don't know of a servo driver chip that takes that kind of input
[21:53:44] <JymmmEMC> Side thought... if paraport is digital out, converting to PWM (analog), isn;t the game port (midi) analog out already?
[21:54:03] <jepler> I don't know anything about that hardware
[21:54:35] <Jimmybondi> the game port is analog in only
[21:54:46] <jepler> MIDI is actually a serial protocol
[21:55:38] <JymmmEMC> Jimmybondi: oh, no ananlog out , huh?
[21:58:02] <JymmmEMC> well shit, wheres the ethernet to stepper driver =)
[21:58:03] <jepler> bbl
[21:58:22] <Jimmybondi> btw: to test a cheap PWM2analog converter:
[21:59:00] <Jimmybondi> i've build it with a transistor, 2 resitors and a cap
[21:59:31] <Jimmybondi> thought only for test but working since 2 years
[21:59:38] <Jimmybondi> *xD*
[21:59:55] <JymmmEMC> http://www.globalspec.com/FeaturedProducts/Detail/Acromag/Ethernet_analog_output_modules/15446/0
[22:00:17] <JymmmEMC> I think I spoke too soon... there she be!
[22:00:58] <toastydeath> pew pew pew
[22:01:30] <toast_> er
[22:01:32] <toast_> pew pew pew
[22:01:54] <JymmmEMC> what are you whining about now
[22:02:35] <jepler> Jimmybondi: yeah they can be pretty simple, particularly if you don't require exact linearity. (and if it's a uC output you might not even need the transistor)
[22:03:31] <Jimmybondi> of course - thats right
[22:05:22] <EMC_Jimmy> now on ubuntu here - first time :-)
[22:05:45] <JymmmEMC> $500 http://www.acromag.com/parts.cfm?Model_ID=306&Product_Function_ID=28&Category_ID=26&Group_ID=2
[22:06:43] <archivist> JymmmEMC, analog out yes, servo no
[22:08:05] <archivist> servo includes feed back from the encoders and the ability to correct the signal
[22:08:10] <JymmmEMC> archivist: I was thinking of steppers. It just bugs me that we have CHEAP ethernet cards that can do 1gbps, but can't get 20K from a paraport
[22:09:00] <EMC_Jimmy> have you ever heard about the "UHU" ?
[22:09:12] <EMC_Jimmy> EMC_Jimmy is now known as Jimmybondi
[22:10:15] <Unit41> Jymm was it you with the plasma cutter ?
[22:10:30] <JymmmEMC> no, I think rugabega
[22:11:19] <Unit41> I need someone to plasma cut the top and bottom off a alu can and see how they fit together for a solar collector
[22:12:02] <Unit41> the seal might be good enoguh to heat water with
[22:12:31] <JymmmEMC> you dont have a razor blade?
[22:13:10] <Unit41> I tried one of those steel circular bits
[22:13:12] <Unit41> no good at all
[22:13:24] <JymmmEMC> you dont have a razor blade?
[22:13:34] <Unit41> the bottom is thick though
[22:13:40] <Unit41> I'll try
[22:13:54] <JymmmEMC> you want to cut a can vertically?
[22:14:52] <Unit41> Hor
[22:15:56] <JymmmEMC> the walls are thin, fitting them back together requires cutting fins, or folding a part over to get a snug fit
[22:15:56] <Unit41> dam a razor works great
[22:16:04] <ds2> VGA is analog out
[22:16:21] <JymmmEMC> ds2: yeah, mariss thought of that
[22:16:57] <JymmmEMC> but game card has been hacked a lot more, but Jimmybondi said it's analog in, not out =(
[22:17:07] <ds2> game card is not really an analog in either
[22:17:18] <ds2> at least not one you canjust apply a voltage and read it
[22:17:34] <ds2> the sound card is another analog in/out
[22:18:39] <jepler> VGA would be interesting if you could program "timings" without blanking periods. otherwise you're stuck with spans of 0V interrupting the output you wanted
[22:19:00] <jepler> also lag induced by the frame buffer
[22:19:05] <ds2> add a sample and/hold?
[22:19:16] <ds2> s#/##
[22:23:11] <JymmmEMC> http://www.acromag.com/pdf/972EN_973EN.pdf
[22:23:50] <Jimmybondi> has anyone configured 5 or 6 axis in emc and 2/3 of them angular ?
[22:24:46] <JymmmEMC> pg 3
[22:32:28] <JymmmEMC> Is changing from 8x to 10x microstepping THAT dramatic of a change?
[22:33:25] <cradek> doubt it makes any difference
[22:33:53] <JymmmEMC> cradek: I mean in the ability to drive that pulse train
[22:34:14] <JymmmEMC> 200SPR motor + 10tpi screws + 10x microstepping
[22:34:22] <JymmmEMC> vs 8x
[22:34:25] <cradek> uh, it changes the required pulse speed by 20%...
[22:35:00] <JymmmEMC> oh, so besides the scale, is there something else I should be changinng too?
[22:35:17] <JymmmEMC> xylotex is 8x, these parker's are 10x
[22:35:41] <ds2> can a stepper actually do 1/8 or 1/10 steps?
[22:36:24] <Jimmybondi> yes
[22:36:34] <Jimmybondi> and more
[22:37:03] <JymmmEMC> cradek: I changed the scale from 16000 to 20000 and got joint following error. Is there something els eI missed in the config? max lat is 17500 ns
[22:37:10] <archivist> whether its acurate in the 8/10 microsteps thats another thing
[22:38:27] <JymmmEMC> the parker doesn't do 8x. it's 200,400,2000,5000,1000,128000,18000 ... 50800
[22:38:43] <JymmmEMC> s/1000/10000/
[22:40:12] <JymmmEMC> and the "default" is 25x lol - I wish
[22:41:43] <JymmmEMC> archivist: This isn't an accuracy thing, just the diff between the xylotex and the parker's is all and the ability of the paraport to drive them
[22:42:21] <JymmmEMC> the math always gets me
[22:42:52] <ds2> JimmyBondi: but accurately?
[22:43:45] <JymmmEMC> ds2: It's the step form... more than one coil are energized to achive that
[22:43:58] <ds2> I know how it is done
[22:44:27] <ds2> just didn't think the motion vs proportion of magnetic forces is linear enough to do it more accurate then 1/2 steps
[22:51:14] <Jimmybondi> the accurancy is not so imposrtant at me - and when talking about:
[22:51:51] <Jimmybondi> what are a difference at a step-error of 5 or 10% ? (in 1/10 or 1/20 step)
[22:52:11] <Jimmybondi> i thing a low-cost ballscrew has a bigger fault
[22:54:15] <Jimmybondi> the advance of microstepping iss the smoothness of the steppers
[23:11:36] <centaur> hello all - i am a linux newbie. what is the path to the file pluto_servo? Ubuntu 6.06
[23:13:10] <JymmmEMC> su;updatedb;locate pluto
[23:13:27] <JymmmEMC> oh, ubuntu, nm
[23:13:42] <JymmmEMC> sudo updatedb;locate pluto
[23:13:46] <centaur> thnaks jy - i'll try it - be back
[23:14:43] <JymmmEMC> jepler: Ok, so how far along are you with the pluto?
[23:15:20] <skunkworks> servo seems to work.. Stepper needs to be tested.
[23:15:59] <JymmmEMC> no stepping test done at all so far?
[23:16:37] <skunkworks> http://axis.unpy.net/index.cgi/01183862490
[23:16:52] <JymmmEMC> yeah, I laready saw that
[23:17:35] <ds2> whoa this VFD's electronics is not transformer isolated!!
[23:17:44] <JymmmEMC> what kind of pulse rate compared to paraport driving?
[23:18:14] <skunkworks> New for emc2 and the pluto-p board: a 4-axis step generator capable of 300,000 steps per second.
[23:18:35] <JymmmEMC> and what is paraport?
[23:18:42] <skunkworks> vs if your lucky - 75k
[23:18:50] <JymmmEMC> ah
[23:19:56] <skunkworks> 25k-50k is pretty easy now though
[23:20:20] <ds2> what's the bottle neck for 75K?
[23:22:00] <JymmmEMC> the turn off point
[23:22:11] <JymmmEMC> iirc 2uS delay
[23:24:23] <JymmmEMC> Step Pulse Requirements
[23:24:24] <JymmmEMC> Operate the step pulse input within the following guidelines:
[23:24:24] <JymmmEMC> • 200 nanosecond pulse – minimum
[23:24:24] <JymmmEMC> • 40% – 60% duty cycle (2 MHz maximum pulse rate)
[23:24:37] <JymmmEMC> *sigh*
[23:26:27] <JymmmEMC> Hmmm, I wonder if I can just enable the gear shift??? 8x multiplier
[23:29:44] <ds2> wow a VFD is cheaper then a transformer
[23:30:08] <skunkworks> 1000spr - probably all you need
[23:51:43] <JymmmEMC> skunkworks: The math always gets me on the numbers. why do you say 1000spr is "good enough"?
[23:53:11] <JymmmEMC> I'd be happy if I could do 150IPM on my machine. 200spr motors, 10tpi screws.
[23:56:20] <JymmmEMC> I really should writeup a calculator for all this someday.