#emc | Logs for 2009-01-18

Back
[00:55:57] <The_Ball> skunkworks, i read vonvection routing, i got a little confused about your finish description
[01:00:44] <skunkworks> The_Ball: what? vonvection?
[03:37:23] <KimK> SWPadnos: Wasn't that a Hardinge that you just bought? Does it have a parting knife? "VERTICAL CUT OFF SLIDE FOR HARDINGE CNC CHNC 2 W/ PUMP Item number: 350153091625" http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=350153091625
[03:38:22] <eric_unterhausen> what does that do?
[03:39:38] <jmkasunich> shoves a parting tool into the work, at a fixed Z location independent of the normal tool turret
[03:39:54] <jmkasunich> it mounts on top of the headstock
[03:44:22] <KimK> Works great when barfeeding. And they make low-cost "manual barfeeders" (more of a bar-grabber-and-puller, really)
[03:45:56] <jmkasunich> he got a barfeeder (or at least parts of one) with the lathe
[03:46:06] <KimK> Excellent!
[03:46:20] <jmkasunich> swp is a hobby machinist, I don't think he anticipates production work
[03:46:34] <jmkasunich> and the barfeeder might stick out the garage door if he tries to use it
[03:48:54] <KimK> "Hey, what are these threaded rods sticking out of your driveway at regular intervals? They look like holddowns for a barfeeder."
[03:49:43] <eric_unterhausen> nuts embedded in the driveway might be more practical
[03:50:21] <KimK> There you go, SWP, problem solved
[03:50:54] <JymmmEMC> SWPadnos: ping
[03:51:03] <jmkasunich> heh - neither swp or I knew it needed to be bolted down
[03:51:12] <jmkasunich> JymmmEMC: I believe swp is on a plane
[03:51:24] <JymmmEMC> oh, damn =(
[03:51:35] <jmkasunich> KimK: the feeder he got has two tripods and a long pipe
[03:51:46] <JymmmEMC> Know of any wifi routers with gigabit and at least 8 ports?
[03:51:57] <jmkasunich> no
[03:53:54] <eric_unterhausen> wired router w/ access point
[03:54:16] <JymmmEMC> I was trying to avoid that scenario if possible
[03:54:48] <JymmmEMC> Gawd linksys website is all flash now, so the dan thing doens't work
[03:55:06] <eric_unterhausen> don't you hate flash
[03:55:16] <eric_unterhausen> I have it blocked, that can be annoying but it's faster
[03:57:36] <KimK> Generally barfeeders need to be held down so they don't wander off of their carefully aligned spindle center due to vibration (and safety from long rod "whipping"). But a temporary install with sandbags and lowered spindle RPM might work? I've never heard of it being tried.
[03:58:48] <jmkasunich> if he decides to use it, sounds like he should bite the bullet and bolt it down
[03:59:02] <jmkasunich> that will be a while though - gotta do the basic conversion first
[03:59:38] <KimK> Did the feet look like they were drilled for bolting? Or just flat plates?
[03:59:55] <jmkasunich> one hole per foot, three feet per tripod
[04:00:14] <KimK> Well, that should do it.
[04:06:30] <KimK> Hey, at least his barfeeder would be pointing *out*, or should I say in. Anyway, the other direction wouldn't be very good at all, LOL
[04:08:16] <jmkasunich> I'm not sure if he's actually decided on a place and orientation in the garage
[04:09:40] <KimK> Still on the trailer? Yeah, sorry to send our cold weather up that way, but not sorry enough to want to keep it here.
[04:09:53] <jmkasunich> he got it off the trailer and into the garage
[04:10:31] <KimK> Oh, well, that's a good thing. Now it won't get all ice-covered.
[04:11:48] <KimK> Did you go along on that adventure? How many went along?
[04:13:55] <jmkasunich> he drove by himself from vermont to cleveland oh where HGR is
[04:14:09] <jmkasunich> I live in a suburb of cleveland, I helped him load
[04:14:13] <jmkasunich> then he drove back
[04:17:36] <KimK> Oh, OK. I guess I thought it was a "road trip".
[04:17:44] <jmkasunich> for him it was
[04:36:28] <KimK> OK, here's a question for the experts: Could EMC2 be adapted (as-is) to a twin-spindle VMC? I suppose you'd have to run one Z as a slave to the other Z. And align the tool lengths "identically". It might be suitable only for two-pass roughing and finishing on the same part, rather than identically-matched double-production. But would it work?
[04:36:42] <tomp> wow, a cnc laser dot-matix engraved 'bone-jar' ( a casket for your bones ) http://scripturn.com/video_5qsjT7Wnc80.html
[04:36:58] <tomp> KimK, like a moog hydra point? master/follower?
[04:37:39] <KimK> And what would it take to add full Z independence (right and left tool tables, etc.) to EMC2?
[04:37:40] <jmkasunich> KimK: you could send the Z commands from the motion controller to two PID loops
[04:37:55] <jmkasunich> neither one would be slaved to the other, they'd both follow whatever EMC tells them
[04:38:01] <jmkasunich> homing would be a challenge though
[04:38:29] <jmkasunich> the same homing issues occur with gantries when people wnat to use an motor and screw for each side
[04:39:37] <jmkasunich> you could also use custom kinematics if you want Z to control one spindle and W to control the other
[04:39:51] <KimK> Yes, they would have to be carefully aligned as close as possible for both tool changers to work. At least there wouldn't be any gantry-stress-type problems.
[04:39:55] <jmkasunich> M3 thru 5, and S words, would control both
[04:40:31] <jmkasunich> things like rigid tapping would be a problem too
[04:40:34] <KimK> They might in fact have the same spindle motor, I'm not sure about that part of it.
[04:41:05] <KimK> I'd have to see the machine.
[04:41:12] <tomp> same spindle motor speed for roughing & finishing?
[04:41:38] <jmkasunich> rough on one spindle and finish on the other isn't very attractive
[04:41:48] <jmkasunich> roughing probably has higher feeds, and more passes
[04:41:56] <jmkasunich> finish is probably fewer passes but slower feeds
[04:42:07] <jmkasunich> if X and Y are the same for both spindles, you can't make both happy
[04:42:08] <KimK> Oooh, that would be a problem. But I don't know how it's driven.
[04:42:25] <jmkasunich> two parts at a time would be more feasable
[04:43:40] <jmkasunich> with some very clever HAL work you could even do rigid tapping - set aside two slots in the toolchanger
[04:44:11] <jmkasunich> on spindle A, slot 1 has a tap, 2 is empty, on spindle B, slot 1 is empty, 2 has the tap
[04:44:39] <tomp> tool table trickery
[04:44:43] <jmkasunich> select tool 1, use HAL to select the spindle A encoder, tap the hole, then select tool 2, use HAL to select the spindle B encoder, and tap again
[04:45:07] <jmkasunich> the tapping operation takes twice as long, but if tapping is a small fraction of the total part time, it isn't much of a loss
[04:45:57] <jmkasunich> if the spindle are driven by the same motor, with a non-slip drivetrain (tooth belts for example) then you could probably rigid tap both at the same time
[04:46:05] <KimK> To get full Z independence though, you'd really have to have right and left tool offsets. Could you call for tool #1, and, I don't know, tool #101 ? I'd really rather call for tool #1L and #1R in that situation.
[04:46:29] <jmkasunich> full independence seems impossible
[04:46:35] <jmkasunich> there is only one g-code program
[04:47:04] <SWPLinux> the tool numbers are just numbers once they get to HAL
[04:47:04] <KimK> Yes, and only one X-Y location (in two places) at a time.
[04:47:05] <jmkasunich> when you say T20 M6, emc can't possibly know which spindle you want to change tools in
[04:47:35] <jmkasunich> SWPLinux: in DC?
[04:47:38] <SWPLinux> you could do M106 for spindle 2, and also use an M-code to set the speed
[04:47:40] <SWPLinux> yep
[04:48:00] <jmkasunich> can you read back? Kimk had some input about the barfeed
[04:48:03] <jmkasunich> if not, logs
[04:48:03] <KimK> Hi, SWP
[04:48:09] <SWPLinux> hi KimK
[04:48:56] <SWPLinux> so, I did some of the vim tutorial on the plane ride
[04:49:32] <SWPLinux> I think I've discovered that vim is kind of like a text browser that lets you edit
[04:50:13] <tomp> heh lesson1: how to exit vim
[04:50:23] <SWPLinux> that I knew :)
[04:50:30] <SWPLinux> I even know how to save and exit
[04:51:06] <KimK> * KimK guesses that SWP is an emacs fan?
[04:51:22] <SWPLinux> no, not really
[04:51:30] <SWPLinux> I used to use it, but that was 20 years ago
[04:51:50] <SWPLinux> I was most comfortable with the Borland editor actually
[04:51:55] <eric_unterhausen> how do you live on unixes without vi?
[04:52:10] <SWPLinux> I generally use pico or nano
[04:52:21] <SWPLinux> or some GUI editor, like gedit or kate
[04:52:36] <eric_unterhausen> I like kate, the vertial block mode selection is great
[04:52:51] <SWPLinux> I've tried a couple of IDEs as well - Anjuta is nice, and there's another one that's no longer free I think
[04:53:01] <KimK> Really? So for the famous "vi vs. emacs" question, you pick "none of the above"? There's a "linux freedom" answer for you.!
[04:53:07] <eric_unterhausen> I know people that are crazy fast with vi
[04:53:12] <SWPLinux> but I keep noticing that they get in the way a lot
[04:53:32] <SWPLinux> eric_unterhausen: yeah, you should see jepler
[04:53:52] <SWPLinux> paul C is fast as hell with mc as well
[04:54:12] <tomp> what is the editor in mc ( clone of ne )
[04:54:15] <eric_unterhausen> I was using vi a lot, and I would get frustrated in windows
[04:54:15] <SWPLinux> I used to be that way with BC, but it's been a long time (BC 3.1 for DOS was the end of that era
[04:54:31] <SWPLinux> I don't know what the mc editor is
[04:54:35] <cradek> I watched someone try to edit source code in gedit recently - oh god.
[04:54:39] <SWPLinux> heh
[04:54:40] <eric_unterhausen> midnight commander
[04:54:54] <eric_unterhausen> I thought it was a port of the borland stuff
[04:54:56] <SWPLinux> eric_unterhausen: yeah, I don't know what it uses for editing
[04:55:00] <SWPLinux> could be
[04:55:19] <eric_unterhausen> I think they open sourced that when the realized they couldn't sell it for windows
[04:55:28] <SWPLinux> cradek: I use gedit a fair amount, but watcking you and Jeff navigate around source code made me want something better :)
[04:55:34] <SWPLinux> watching
[04:55:58] <SWPLinux> so I figures I'd try to learn about vi before bad-mouthing it any more
[04:56:02] <SWPLinux> figured
[04:56:28] <eric_unterhausen> I like kdevelop and even eclipse better
[04:56:30] <SWPLinux> I think the esc key is too far out of place on this keyboard to be useful for vim
[04:56:37] <cradek> people are so religious about editors because it takes a long time using one until it doesn't get in the way of your thinking anymore
[04:56:57] <SWPLinux> and there aren't many that can do that, even with a lot of practive
[04:57:00] <SWPLinux> gah
[04:57:02] <SWPLinux> practice
[04:58:49] <SWPLinux> there's supposed to be a symbol browser extension for kate, but I think it doesn't do what vi/ctags do for you
[05:00:30] <tomp> hah seen this link "Return to the Obsolete Technology Homepage " while looking at the editors i used on the Lobo Max-80
[05:01:00] <SWPLinux> there was a great item in the in-flight catalog
[05:01:18] <SWPLinux> something like "new ancient technology for ...."
[05:01:22] <cradek> SWPLinux: tags is one of the most basic features a programmer's editor should support... indenting, split screen, tags, quote/paren matching, ...
[05:01:31] <SWPLinux> I just couldn't quite reconcile that
[05:01:36] <SWPLinux> ye
[05:01:38] <SWPLinux> p
[05:01:42] <cradek> missing any of those => discard editor
[05:02:14] <SWPLinux> that was one thing the CBuilder editor added to BC 3.1
[05:02:14] <cradek> I'm not religious but I know a programmer's editor when I see it :-)
[05:02:21] <SWPLinux> heh
[05:02:29] <SWPLinux> real programmers use a chisel ;)
[05:02:55] <SWPLinux> on the hard drive, you know, to encode the bits
[05:05:42] <jmkasunich> I wonder if kate does tags
[05:05:42] <tomp> where are tags explained? googling 'editor tags' is fruitless. is it a 'vim-enese' word?
[05:06:08] <cradek> emacs and vi both do tags (etags vs ctags)
[05:06:19] <eric_unterhausen> my impression is that kate is the editor in kdevelop
[05:06:19] <cradek> everyone else surely uses one of these two formats
[05:07:15] <tomp> ok, thanks, http://www.linuxjournal.com/article/0153
[05:09:12] <tomp> oh wow, fenn: Cory Doctorow is on cover of LJ
[06:31:41] <JustinXJS2_> JustinXJS2_ is now known as JustinXJS2
[07:03:39] <maddash> is anyone here?
[07:04:17] <eric_unterhausen> no
[07:04:27] <maddash> sweet
[07:04:43] <maddash> unidirectional height limit sensing with a camera
[07:04:59] <maddash> impossible? it's been bugging the hell out of me the past 9 hours straight
[07:05:03] <eric_unterhausen> what's a unidirectional height
[07:06:07] <maddash> basically, i've got my garage door, and i want to detect if an incoming car/truck/hulk is too high
[07:06:38] <eric_unterhausen> that sounds like a tough one with one camera
[07:06:46] <maddash> the old way to do it is to set up two posts on either side, and attach an IR diode on one side and photo sensor on the other
[07:06:56] <maddash> right
[07:07:49] <maddash> with a camera, I sample @ QVGA (320x240). i set my camera so that the the midline of my image (line 119 = 240/2 -1) to line up with the maximum height
[07:08:25] <maddash> so then it's a simple matter of detecting temporal changes of the pixels in line 119
[07:08:43] <maddash> if the vehicle is too high, pixels of line 119 will change, and I set off an alarm
[07:10:00] <maddash> problem is that there's no depth detection, so when a car that is low enough to fit comes in, and a truck (that is too tall) follows it, the system will blindly set off the alarm (and halt the car instead of *just* the truck)
[07:10:35] <maddash> what sort of objects absorb IR?
[07:10:42] <eric_unterhausen> rubber
[07:11:06] <eric_unterhausen> fleece absorbs ultrasonic
[07:11:33] <eric_unterhausen> black paint can screw up ir laser rangefinders
[07:12:08] <maddash> really? this (http://answers.google.com/answers/threadview/id/779697.html) first post says, "but back [black, sic] reflects ir!"
[07:12:31] <eric_unterhausen> my experience has been different
[07:13:02] <maddash> ok, i believe you over the crazy guy who can't spell correctly
[07:13:44] <maddash> does IR work during the daytime?
[07:15:07] <eric_unterhausen> never really tried it
[07:22:29] <maddash> so I'm screwed.
[07:22:40] <eric_unterhausen> depends on your budget
[07:22:59] <eric_unterhausen> pretty sure there are ways to make ir work outside
[07:24:20] <maddash> yeah,, but then how would ir help?
[07:24:31] <eric_unterhausen> you brought it up :)
[07:25:05] <eric_unterhausen> all you need is a sick lm200 and you're set
[07:25:34] <maddash> is that LM200?
[07:25:44] <eric_unterhausen> lms200
[07:26:06] <maddash> http://www.pages.drexel.edu/~kws23/tutorials/sick/sick.html ?
[07:26:14] <eric_unterhausen> yeah
[07:26:27] <eric_unterhausen> there are safety versions too, more likely to show up on ebay
[07:26:28] <maddash> lasers are evil -- they burn retinas
[07:26:35] <eric_unterhausen> hope not
[07:26:52] <eric_unterhausen> I've been working in a room with about 10 of them for many years
[07:26:56] <maddash> how big is that thing anyway? it looks like the size of a blender
[07:27:07] <eric_unterhausen> small coffee pot
[07:27:08] <maddash> or a coffee maker
[07:27:12] <maddash> christ
[07:27:16] <eric_unterhausen> that's what I meant
[07:27:19] <maddash> why so big?
[07:27:31] <eric_unterhausen> there is a japanese company that makes smaller ones
[07:27:33] <eric_unterhausen> lower range
[07:30:04] <eric_unterhausen> Hokuyo makes laser range finders that fit in the coffee cup
[07:31:19] <maddash> hm, but do they come in line laser form? otherwise, I'd have to know where to aim the laser
[07:31:32] <eric_unterhausen> they measure the distance in a plane
[07:31:59] <eric_unterhausen> aiming is a little tricky
[07:33:20] <maddash> 'they measure the distance in a plane' <--- could you explain?
[07:33:29] <maddash> which plane?
[07:34:16] <eric_unterhausen> they tell you the first return on a plane if something is in range
[07:34:45] <eric_unterhausen> basically, they have a wheel that sends a laser beam out
[07:35:26] <eric_unterhausen> the sick lms200 takes a measurement ever 1/2 degree over 180 degrees
[07:35:37] <eric_unterhausen> every 1/2 degree
[07:41:18] <eric_unterhausen> the ld-oem does 270 degrees https://www.mysick.com/saqqara/pdf.aspx?id=im0027485&lang=en&page=1
[07:41:48] <eric_unterhausen> I like those because the top part spins
[07:44:56] <maddash> i think the top part in all of us spins
[07:45:26] <maddash> hm, how accurate is the ultrasound range finder?
[07:45:49] <eric_unterhausen> that's a tough one because the beam spreads
[07:46:00] <eric_unterhausen> I'm sure if you spend tons of money they get better
[07:46:05] <maddash> right, that's my question -- what's the angle of "vision"?
[07:46:42] <eric_unterhausen> the ones I've used have around ten degrees or so
[07:47:11] <eric_unterhausen> now that I think about it, maybe more like 20 degrees
[07:47:16] <eric_unterhausen> IR is better
[07:47:26] <maddash> as in IR range finder?
[07:47:34] <eric_unterhausen> yes
[07:47:45] <eric_unterhausen> pretty sure sick has some that work outdoors
[07:47:48] <maddash> but IR sends out a single narrow beam, right?
[07:47:50] <maddash> see, what I need is a line
[07:48:05] <eric_unterhausen> so you need something like the lms
[07:48:34] <eric_unterhausen> you are going to have problems with the car/truck scenario unless you do a lot more
[07:49:23] <maddash> you mean add more stuff to my system to handle special cases?
[07:49:51] <eric_unterhausen> or you can just stop traffic until you identify the guy that's going to stop traffic
[07:49:55] <maddash> hahah
[07:50:08] <maddash> could be the fucktard all the way in the back with a monster truck
[07:50:32] <eric_unterhausen> rangefinder will help some with that problem
[07:54:31] <maddash> so use rangefinder to measure how far away the portion of the object that triggers line 119 is?
[07:54:52] <eric_unterhausen> could do that, or just forget the camera
[07:55:22] <maddash> but the rangefinder can't measure just a
[07:55:25] <maddash> line, right?
[07:55:28] <maddash> fucking camera
[07:55:41] <eric_unterhausen> not sure what you mean by a line
[07:55:57] <eric_unterhausen> but if there is a line in range, you can do segmentation to find it
[07:56:28] <maddash> e.g., i'm sensing height, so it helps to be sensitive to the midline only
[07:57:11] <maddash> if the rangefinder sends out a wide angle beam, then objects that are short but far away could also trigger a fault
[07:57:35] <maddash> segmentation? as in image segmentation, or what?
[07:57:57] <eric_unterhausen> basically the same thing as image segmentation
[07:58:17] <eric_unterhausen> you have the distances to the first object on a geometric plane
[07:58:29] <eric_unterhausen> sorta like pixels on a line
[07:58:38] <maddash> teardrop shape?
[07:59:14] <eric_unterhausen> not sure what you mean?
[07:59:50] <maddash> ah, disregard that. go on
[08:00:41] <eric_unterhausen> I'm thinking you aim the plane of the rangefinder at the height that you want to fault, and you're done
[08:01:00] <eric_unterhausen> sick has some that will do the whole deal for you, they use it for guarding robots and the like
[08:01:03] <maddash> er, range finder has a plane?
[08:01:11] <eric_unterhausen> geometric plane
[08:01:12] <maddash> sorry, it's 3AM and I'm stupid
[08:01:18] <eric_unterhausen> defined by 3 points
[08:01:31] <maddash> do you have a link or something that can illustrate this?
[08:01:38] <maddash> it's 3AM and I'm dumb
[08:04:36] <eric_unterhausen> https://www.mysick.com/saqqara/pdf.aspx?id=im0012759
[08:04:54] <eric_unterhausen> look on page 12
[08:07:32] <maddash> thanks a bunch
[08:09:29] <maddash> where can I buy a lidar?
[08:09:59] <eric_unterhausen> we have a place that gives us a good price because we are educational
[08:10:21] <eric_unterhausen> if you are serious, I can probably hook you up, but it would take a while
[08:11:03] <maddash> how long? what price?
[08:11:06] <eric_unterhausen> I think acroname sells some
[08:11:18] <eric_unterhausen> I think we get lms200 for $3500
[08:11:24] <maddash> holy crap
[08:11:36] <eric_unterhausen> ebay then :)
[08:12:14] <maddash> hm can't use ultrasound, because it emits cone
[08:12:43] <maddash> but i can use a beam to detect along a plane by spinning it around and around at high speeds
[08:12:58] <maddash> so what sort of options do I have for beam range finder?
[08:12:58] <eric_unterhausen> I've always wanted to do that
[08:13:17] <maddash> (that doesn't cost so fucking much. $3600? seriously?)
[08:13:33] <eric_unterhausen> http://cgi.ebay.com/Sick-Optic-Sensor-PLS101-112-Made-in-Germany_W0QQitemZ350139839655QQcmdZViewItemQQptZBI_Automation_Components?hash=item350139839655&_trksid=p3286.c0.m14&_trkparms=72%3A1205|66%3A2|65%3A12|39%3A1|240%3A1318|301%3A0|293%3A1|294%3A50
[08:14:21] <eric_unterhausen> sick kicks butt, look at all the darpa challenge vehicles, they have 10 of them on there
[08:15:11] <maddash> don't you think sick is overkill for height sensing
[08:15:12] <maddash> ?
[08:15:22] <eric_unterhausen> depends on the application
[08:15:23] <maddash> hell, I'll go with sharp ir
[08:15:43] <eric_unterhausen> if someone was paying me for my time, it would be the cheapest possible option
[08:16:52] <eric_unterhausen> I want one so I can guide my wife into the garage
[08:17:19] <maddash> haha
[08:17:20] <maddash> nice
[08:17:49] <maddash> eric, if the system works right, you'll get first pickings
[08:17:53] <eric_unterhausen> you see them go for ~$350 occasionally
[08:18:21] <eric_unterhausen> actually, a camera would work well for the wife parking guidance system
[08:18:50] <maddash> WPGS, sweet
[08:19:11] <eric_unterhausen> my first ambition is the DMBA system
[08:19:17] <maddash> the trick is to center the camera so that the midline sits just on the maximum height
[08:19:17] <eric_unterhausen> daughter missed the bus again
[08:19:57] <maddash> do the Sharp IR sensors emit a beam, or a narrow cone? or just plain old wide cone?
[08:20:08] <eric_unterhausen> it must be a cone
[08:20:19] <eric_unterhausen> the ones we use have a hood
[08:20:28] <eric_unterhausen> but the range is tiny
[08:22:06] <maddash> 35-40ft?
[08:22:38] <eric_unterhausen> which one goes that far
[08:22:39] <eric_unterhausen> ?
[08:24:03] <maddash> my camera
[08:24:06] <maddash> it goes to 60 ft
[08:24:20] <eric_unterhausen> thought we were talking about sharp ir
[08:24:30] <maddash> oh
[08:24:32] <maddash> right
[08:24:40] <maddash> no idea. checking now, anyway.
[08:24:55] <maddash> 3m in ft
[08:25:01] <maddash> whoops
[12:38:05] <The_Ball> I'm designing a trunnion table, is there any reason not to use car wheel bearings for stuff like this? They are easy to get and very strong
[12:41:18] <archivist> The_Ball, Im using cheap tapers as well, probably good enough for run of the mill stuff
[12:43:46] <The_Ball> I'm thinking of getting this rolled ballscrew; http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ssPageName=STRK:MEWAX:IT&item=230319100895
[12:43:56] <The_Ball> but is it normal to have grease nipples on these?
[12:47:48] <archivist> look on Hiwin's site for specs
[12:48:03] <The_Ball> oh, I ment in general
[12:48:35] <archivist> I have not seen enough to say
[12:50:17] <fragalot> is it rick rolled?
[12:51:37] <archivist> its at a price thats likely a rolled screw
[12:52:50] <The_Ball> yes, it says it is
[13:22:03] <christel> [Global Notice] Hi all, As some of you may have noticed, we lost our ipv6 servers for a extended period of time in the last day. I am terribly sorry about this, it would appear that we've unmasked yet another peculiar hyperion bug. They're back up now and we shouldn't experience any further problems. Thank you for using freenode and have a great day.
[15:00:12] <JanVanGilsen> Hi, we moved the PUMA today,
[15:00:19] <JanVanGilsen> http://imagebin.org/36120
[15:00:35] <JanVanGilsen> http://imagebin.org/36122
[15:01:05] <JanVanGilsen> http://imagebin.org/36121
[15:01:26] <JanVanGilsen> http://imagebin.org/36123
[15:02:42] <JanVanGilsen> http://imagebin.org/36125
[15:06:13] <archivist> lotso wire and bits:)
[15:08:24] <archivist> JanVanGilsen, found out when I was doing the terminal stuff, (and who for) and what I still have in backups
[15:10:03] <archivist> I can say the terminal was VT220 compatible
[15:10:05] <JanVanGilsen> I asked my father about the terminal, he wants to keep it.. I think he's a bit nostalgic about it..
[15:10:33] <archivist> I would love pics of it at some time if possible
[15:13:49] <JanVanGilsen> I could fix that, but the terminal is still at my fathers workplace, where the robot used to be
[15:14:12] <archivist> no rush :)
[15:15:16] <archivist> I still have a first prototype board, board cad files and the assembler code for it
[15:17:25] <JanVanGilsen> I do have a user manual
[15:20:18] <archivist> I had to write something I presume they did a proper job, I just threw the basics together for the terminal doc
[15:21:26] <archivist> electronics was done by a subcontractor called Instem
[15:24:05] <JanVanGilsen> The cabinet b4 getting it upstairs: http://imagebin.org/36128
[15:24:39] <archivist> upstairs!
[15:25:00] <JanVanGilsen> http://imagebin.org/36128
[15:25:34] <JanVanGilsen> had to strip the whole thing because it was to heavy :)
[15:32:33] <archivist> I wonder about floor loading :)
[15:35:24] <JanVanGilsen> I think it will hold, seems quite ridgid to me
[15:39:13] <archivist> I had loads of test equipment upstairs for a few years :)
[16:47:20] <JymmmEMC> SWPLinux: ping
[17:08:47] <motioncontrol> ok in irc with linux no more windows.
[17:30:22] <motioncontrol> good evening.i write with my pc on linux and emc , for resolve my problem of configutation the my card m5i20 with hostmot2 driver.please help?
[17:33:01] <eric_unterhausen> what is the problem?
[17:34:11] <motioncontrol> thank.i have use emc2.2.8 with m5i20 old version driver, but today i want use new version hm2_pci driver.
[17:34:55] <motioncontrol> i have cpy and rename the hm2_servo directory an remame in M5i20 directory
[17:35:10] <motioncontrol> in configs directory
[17:35:51] <motioncontrol> i have rename the all file 7i34.ini ecc in M5i20.ini , M5i20.hal ecc
[17:37:23] <motioncontrol> i have open the M5i20.hal file and i have change line: loadrt hm2_7i37 config="firmware=hm2/7i37/SVST8_4B.BIT num_encoders=3 num_pwmgens=3 num_stepgens=1" in , loadrt hm2_pci config="firmware=hm2/5i20/SVST8_4B.BIT num_encoders=3 num_pwmgens=3 num_stepgens=1" is ok?
[17:39:35] <motioncontrol> now when i start my emc2.2.8 compile whit -run in place option the terminal have error: insmod: error inserting '/home/emc/Desktop/emc2-2.2.8/rtlib/hostmot2.ko': -1 Unknown symbol in module
[17:39:35] <motioncontrol> M5i20.hal:41: exit value: 1
[17:39:35] <motioncontrol> M5i20.hal:41: insmod failed, returned -1
[17:39:55] <jepler_> did you look in 'dmesg' for the actual error?
[17:40:10] <alex_mobile> dmesg
[17:40:21] <motioncontrol> hostmot2: Unknown symbol release_firmware
[17:40:21] <motioncontrol> hostmot2: Unknown symbol request_firmware
[17:40:21] <motioncontrol> RTAI[math]: unloaded.
[17:40:21] <motioncontrol> RTAI[malloc]: unloaded.
[17:40:21] <motioncontrol> RTAI[sched]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[17:40:22] <motioncontrol> I-pipe: Domain RTAI unregistered.
[17:40:24] <motioncontrol> RTAI[hal]: unmounted.
[17:41:22] <motioncontrol> i dont't understand the problem.
[17:41:32] <jepler_> request_firmware is an API provided by the linux kernel. if you don't have it in /proc/kallsyms or /proc/ksyms, then you've improperly configured your kernel.
[17:43:11] <jepler_> it may be CONFIG_FW_LOADER but I'm not sure
[17:44:10] <motioncontrol> excuse.i dont understand.you thing the problem is one option select when i compile rtai menuconfig?
[17:45:00] <jepler_> no, it is a kernel option, not an rtai option.
[17:45:11] <motioncontrol> ok
[17:45:41] <motioncontrol> the file /proc/kallsyms is emply
[17:46:22] <motioncontrol> the file /proc/ksyms not exist
[17:47:16] <motioncontrol> you thing i recompile the kernel and add the option CONFIG_FW_LOADER
[17:48:03] <jepler_> hostmot2 requires request_firmware. on my system, the kernel provides it. you can draw your own conclusion.
[17:49:01] <motioncontrol> ok thanks jepler i ceck my .config
[17:54:58] <motioncontrol> jepler i search in .config makemenuconfig , kernel 2.6.22 but i don't see the option.in this menu you have the optiion?
[17:56:09] <jepler_> $ grep CONFIG_FW_LOADER /boot/config-2.6.24-16-rtai
[17:56:09] <jepler_> CONFIG_FW_LOADER=y
[17:57:28] <jepler_> bbl
[17:58:11] <motioncontrol> this my config-2.6.22-rtai CONFIG_FW_LOADER=m is module
[17:58:51] <motioncontrol> i change in CONFIG_FW_LOADER=y
[17:59:42] <motioncontrol> or is necessity recompile?
[18:03:37] <acemi> yes
[18:05:14] <motioncontrol> excuse acemy i change only in config-2.6.22-rtai the option CONFIG_FW_LOADER=y or is necessity recompile?
[18:05:38] <acemi> you need to recompile
[18:06:09] <motioncontrol> more thanks at all.
[18:26:46] <jepler_> jepler_ is now known as jepler
[19:13:21] <jmkasunich> sunday newspaper + snowblower = confetti ;-)
[20:18:12] <alex_mobile> Good night all.
[20:18:13] <alex_mobile> Good night all.
[20:33:59] <motioncontrol> good evening one problem for driver hm2_pci for m5i20.i have recompile de kernel and i have add option config_fw_loade, the message error in terminal is: M5i20.hal:47: parameter or pin 'hm2_5i20.0.pwmgen.pwm_frequency' not found
[20:34:18] <motioncontrol> what is the problem?
[20:35:06] <motioncontrol> dmesg is hm2: loading Mesa HostMot2 driver version 0.15
[20:35:06] <motioncontrol> hm2_pci: loading Mesa AnyIO HostMot2 driver version 0.6
[20:35:06] <motioncontrol> hm2_pci: discovered 5i20 at 0000:05:07.0
[20:35:06] <motioncontrol> hm2/hm2_5i20.0: firmware hm2/5i20/SVST8_4B.BIT not found
[20:35:06] <motioncontrol> hm2/hm2_5i20.0: install the package containing the firmware, or link your RIP sandbox into /lib/firmw are manually
[20:35:08] <motioncontrol> hm2_5i20.0: board fails HM2 registration
[20:35:10] <motioncontrol> hm2_pci: probe of 0000:05:07.0 failed with error -2
[20:35:12] <motioncontrol> hm2_pci: driver unloaded
[20:35:14] <motioncontrol> hm2: unloading
[20:35:16] <motioncontrol> RTAI[math]: unloaded.
[20:35:20] <motioncontrol> RTAI[malloc]: unloaded.
[20:35:22] <motioncontrol> RTAI[sched]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[20:35:24] <motioncontrol> I-pipe: Domain RTAI unregistered.
[20:35:26] <motioncontrol> RTAI[hal]: unmounted.
[20:55:27] <motioncontrol> please i have one problem for use the hma2_pci driver for mesa 5i20 card is possible help?
[20:56:01] <eric_unterhausen> did you build the previous kernel too?
[20:56:32] <eric_unterhausen> seems like it didn't find your firmware
[20:56:50] <motioncontrol> yes i have the compile the kernet and rtai patch the problem is diffetent now
[20:57:20] <eric_unterhausen> SVST8_4B.BIT do you have that?
[20:57:54] <motioncontrol> i have correct now but the problem is : hm2: loading Mesa HostMot2 driver version 0.15
[20:57:54] <motioncontrol> hm2_pci: loading Mesa AnyIO HostMot2 driver version 0.6
[20:57:54] <motioncontrol> hm2_pci: discovered 5i20 at 0000:05:07.0
[20:57:54] <motioncontrol> hm2/hm2_5i20.0: firmware hm2/5i20/SVST8_4.BIT not found
[20:57:54] <motioncontrol> hm2/hm2_5i20.0: install the package containing the firmware, or link your RIP sandbox into /lib/firmware manually
[20:57:55] <motioncontrol> hm2_5i20.0: board fails HM2 registration
[20:57:57] <motioncontrol> hm2_pci: probe of 0000:05:07.0 failed with error -2
[20:57:59] <motioncontrol> hm2_pci: driver unloaded
[20:58:01] <motioncontrol> hm2: unloading
[20:58:03] <motioncontrol> RTAI[math]: unloaded.
[20:58:05] <motioncontrol> RTAI[malloc]: unloaded.
[20:58:07] <motioncontrol> RTAI[sched]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[20:58:09] <motioncontrol> I-pipe: Domain RTAI unregistered.
[20:58:11] <motioncontrol> RTAI[hal]: unmounted.
[20:58:22] <motioncontrol> i have compile emc with --run in place
[20:58:30] <eric_unterhausen> hm2/hm2_5i20.0: firmware hm2/5i20/SVST8_4B.BIT not found
[20:58:33] <eric_unterhausen> that is the error
[20:59:16] <eric_unterhausen> hm2/hm2_5i20.0: install the package containing the firmware, or link your RIP sandbox into /lib/firmw are manually
[20:59:17] <motioncontrol> i have correct in SVST8_4.BIT but the error is :hm2: loading Mesa HostMot2 driver version 0.15
[20:59:17] <motioncontrol> hm2_pci: loading Mesa AnyIO HostMot2 driver version 0.6
[20:59:17] <motioncontrol> hm2_pci: discovered 5i20 at 0000:05:07.0
[20:59:17] <motioncontrol> hm2/hm2_5i20.0: firmware hm2/5i20/SVST8_4.BIT not found
[20:59:19] <motioncontrol> hm2/hm2_5i20.0: install the package containing the firmware, or link your RIP sandbox into /lib/firmware manually
[20:59:22] <motioncontrol> hm2_5i20.0: board fails HM2 registration
[20:59:24] <motioncontrol> hm2_pci: probe of 0000:05:07.0 failed with error -2
[20:59:26] <motioncontrol> hm2_pci: driver unloaded
[20:59:28] <motioncontrol> hm2: unloading
[20:59:30] <motioncontrol> RTAI[math]: unloaded.
[20:59:32] <motioncontrol> RTAI[malloc]: unloaded.
[20:59:34] <motioncontrol> RTAI[sched]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[20:59:36] <motioncontrol> I-pipe: Domain RTAI unregistered.
[20:59:38] <eric_unterhausen> did you read any of that?
[20:59:38] <motioncontrol> RTAI[hal]: unmounted.
[20:59:52] <eric_unterhausen> because I just pointed out the problem
[21:00:15] <eric_unterhausen> it's just a path problem
[21:00:29] <motioncontrol> i don't understand. i have compile with run in place option
[21:00:52] <eric_unterhausen> apparently that doesn't fix up the firmware path
[21:01:16] <jmkasunich> you need to make a symlink from lib/firmware to your rip firmware directory
[21:01:52] <motioncontrol> ok i prove
[21:11:08] <motioncontrol> ok I have create in lib/frmware/hm2/5i20/ and i have copy in directori 5i20 the all 5i20 directory frmware. no error in hal file , but the emc not load the error is : hm2/hm2_5i20.0: registered
[21:11:08] <motioncontrol> hm2_5i20.0: initialized AnyIO board at 0000:05:07.0
[21:11:08] <motioncontrol> hm2_5i20.0: dropping AnyIO board at 0000:05:07.0
[21:11:08] <motioncontrol> hm2/hm2_5i20.0: unregistered
[21:11:08] <motioncontrol> hm2_pci: driver unloaded
[21:11:10] <motioncontrol> hm2: unloading
[21:11:12] <motioncontrol> RTAI[math]: unloaded.
[21:11:14] <motioncontrol> RTAI[malloc]: unloaded.
[21:11:16] <motioncontrol> RTAI[sched]: unloaded (forced hard/soft/hard transitions: traps 0, syscalls 0).
[21:11:18] <motioncontrol> I-pipe: Domain RTAI unregistered.
[21:11:22] <motioncontrol> RTAI[hal]: unmounted.
[21:11:54] <motioncontrol> the terminal error at start emc is emc/task/emctask.cc 312: interp_error: Unable to open file
[21:11:55] <motioncontrol> Unable to open file
[21:11:55] <motioncontrol> emc/task/emctaskmain.cc 2629: can't initialize interpreter
[21:11:55] <motioncontrol> emc/usr_intf/emcsh.cc 3813: can't connect to emc
[21:11:55] <motioncontrol> Shutting down and cleaning up EMC2...
[21:11:55] <motioncontrol> Cleanup done
[21:21:28] <motioncontrol> the problem is ok . In ini file the load m5i20.var is no correct thanks for all
[21:37:40] <motioncontrol> excuse one question in the axis interface , when it load , load the axis.ngc part program.what is the path for this part program?
[22:22:04] <motioncontrol> excuse one question in the axis interface , when it load , load the axis.ngc part program.what is the path for this part program?
[22:23:01] <motioncontrol> excuse one question in the axis interface , when it load , load the axis.ngc part program.what is the path for this part program?