#emc | Logs for 2010-08-27

[00:01:52] <alex_chally> heh, i just accidentaly shorted out and blew up a 50W wire wound resistor
[00:01:54] <alex_chally> that was fucking loud
[00:03:01] <Eric_K> just don't blow up a capacitor, they stink and aren't too loud
[00:03:14] <alex_chally> Eric_K, heh
[00:03:20] <alex_chally> I killed an oil filled cap a few months back
[00:03:28] <alex_chally> it was a cool plume of smoke
[00:04:34] <theorb> theorb is now known as theorbtwo
[00:05:10] <Eric_K> anyone heard a due date for the mesa brushless motor drives?
[02:18:27] <skunkworks> so - can anyone see a reason why cat6 shielded wouldn't work well for differential encoder cables?
[02:19:07] <skunkworks> 3 pairs for a,b,z and one pair for power
[02:21:40] <atmega> never done it, but it should be fine for any reasonable distance
[02:23:41] <skunkworks> none of the servos physically move - so bending isn't an issue
[02:24:22] <skunkworks> 25ft is probably max
[02:27:54] <skunkworks> what do nomal people use?
[02:28:05] <skunkworks> or normal even
[02:28:32] <atmega> at work, I generally use premade cables.
[02:28:33] <cradek> is this for a moving or stationary motor?
[02:29:19] <skunkworks> all servos are stationary
[02:30:16] <cradek> then heck why not
[02:30:21] <skunkworks> heh
[02:30:45] <skunkworks> question #2 - would it have to be shielded?
[02:37:03] <KimK_> I would recommend that it really should be shielded if you can manage it. Maximum immunity from noise pickup and possible false counts, which is kind of "unforgivable" on a quadrature encoder. You don't want to end up like those stepper guys, wondering if you lost or gained steps somewhere along the way. Just my opinion. OTOH, if you try it without a shield and it works, well, "Good on ya, mate!"
[02:38:45] <Valen> skunkworks: if you have differential signals just make sure you use the pairs in the cabl
[02:38:47] <Valen> cable
[02:39:10] <Valen> also cat6 is lots more expensive than cat5 so you might think about using cat5e
[02:39:55] <Valen> also if you can, given your not running 1000mbit over it, see if stranded cable will work for you, the flex life of standard cat5 is so low that you could well hit its limits in normal use
[02:41:26] <skunkworks> thanks
[02:41:43] <skunkworks> I wonder if anyone local has it
[02:42:01] <KimK_> Yes, Valen has good advice for you there too, "Don't cross the streams", I mean pairs, lol. And Cat5/5e is fine. I wouldn't use Cat3 (phone) though, you want a little more frequent twists than that.
[02:42:43] <KimK_> Especially if you're going to try the unshielded trick. (Kids, don't try this at home, lol.)
[02:42:43] <skunkworks> right - I am a network guy -- I know my pairs :)
[02:43:35] <KimK_> Excellent
[02:45:51] <KimK_> Is your dad working on a Y axis bracket or adapter or whatever is needed there?
[02:46:38] <Valen> we have TTL in a somewhat sheilded cable on our encoders and it seems to work fine
[02:48:01] <KimK_> Valen: OK, thanks, that's good to know.
[02:48:31] <Valen> linear scales rather, ~2m cables
[02:49:27] <KimK_> skunkworks: Oh, and remember to only ground your shield on one end. Probably the control end, but either way would be OK. No connection on the other end.
[02:50:52] <KimK_> You probably have a better ground on the control end (ground wire brought in?), so that would be my choice.
[03:29:20] <skunkworks> KimK: dad was machining the last 2 servos to accept the encoders.
[03:29:43] <skunkworks> logger_emc: bookmark
[03:29:43] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2010-08-27.txt
[03:30:22] <skunkworks> and everything is getting grounded at one spot in the control box.\
[03:30:49] <cradek> shield seems like cheap insurance - KimK_ is right
[03:32:59] <KimK_> Re: grounded in control box: very good, you should be fine. Hi Chris, what's up tonight?
[03:33:02] <skunkworks> yah... I just don't have any at the moment. Maybe I will call around. You would think someone would stock it.
[03:33:34] <cradek> KimK_: making more dominoes - I have two very successful fixtures for it now
[03:34:59] <morfic> did i get this right, you had made a script of sorts to make all the dots for all pieces?
[03:35:18] <cradek> yes I generated the gcode with a python program
[03:35:55] <KimK_> I know USDigital has the flex version, 4-pair, shielded, nice grey (silver?) PVC jacket. Very nice, but they get about a buck a foot for it, I seem to recall.
[03:36:06] <cradek> I do that a lot - easier than doing anything complicated in gcode, and it gives me nice straight gcode to actually run, which I prefer for various reasons
[03:36:27] <skunkworks> heh - like restarts :)
[03:36:36] <skunkworks> RFL
[03:36:48] <cradek> yeah it's just simpler to work with
[03:36:54] <morfic> i like a lot of the python stuff floatig around, stuff people often show off on list or here, nice stuff
[03:37:07] <cradek> also I can see at a glance how far along it is, etc
[03:37:15] <morfic> kind of like a "conversational control in building blocks"
[03:37:35] <KimK_> cradek: You should post that on your blog as a how-to, not just for dominos, but as you say, modifiable for lots of things.
[03:38:33] <skunkworks> oh well - I think it is time for bed
[03:38:38] <cradek> KimK_: python.org has a decent tutorial and I'm not sure how I could add more
[03:39:02] <KimK_> skunkworks: Did you get the "cable lump" drivers?
[03:39:06] <KimK_> bah
[03:39:15] <KimK_> too slow
[03:39:27] <morfic> cradek: while i have not touched the formulas i found for getting intersection points, it did help our VB programmer (who does the feature recognition programming for our CAM software) understand that his thinking of finding said point is overly complicated.
[03:39:28] <cradek> that always stinks
[03:39:54] <cradek> KimK_: he said he expected the driver chips in one day
[03:40:00] <skunkworks> KimK: yes
[03:40:18] <skunkworks> actually - we ordered 4 and got 5 - so we should be set :)
[03:40:30] <KimK_> Ha, thanks for coming back just to let us know!
[03:40:52] <skunkworks> no problem. night :)
[03:40:58] <KimK_> goodnight
[03:41:16] <cradek> morfic: it would be truly awesome if you guys could write and contribute it
[03:41:27] <KimK_> That was nice of Sam.
[03:41:45] <morfic> after he understood that he said the API has nothing and he is not sure how long it would take to implement, so i told him it's amazing that they do not expose the functions they use for trimming or "intersecting point", i sat down and opened API docs on my computer, searched "intersect" 3 result topic sounded best to me, and there it had a link to "belong to 'GeoUtil' "
[03:42:26] <morfic> now how do i tell people he needs to make better use of the API docs w/o getting him fired, he has good ideas, but he says a lot of times it can't be done
[03:42:27] <cradek> morfic: I don't really understand what you said there
[03:42:42] <cradek> what API?
[03:42:49] <cradek> what is GeoUtil?
[03:43:05] <morfic> cradek: when i do not get myself in trouble more, i will pick up on G71 again, just because i have not coded in a while and would love to just get back at it
[03:43:37] <cradek> well that would sure be jumping into the deep end :-)
[03:43:37] <KimK_> I'm guessing the API used by your Visual Basic programmer?
[03:43:55] <morfic> cradek: i am talking about the API of our CAM Software, what it exposes of its internal stuff to use in VB for "feature recognition", you give it a model, slice it, tell it what type it is, bam, made morfic jobless by programming itself
[03:44:58] <KimK_> morfic: What CAM software do you make or work on? Unless you'd rather not say.
[03:44:59] <cradek> oh ok, you're just saying your programmer doesn't read docs, I get it!
[03:45:13] <morfic> to me, they expose everything, and i keep finding stuff for him, without really having touched the VB side of it, just by having an idea how all the .NET stuff was layed out last time i did some C#.NET
[03:45:58] <morfic> KimK: we *USE* Esprit, we are users, but to use "feature recognition" to do what i said up there, you need someone making VB code, lots and lots of it
[03:46:12] <morfic> cradek: in my usual long morfic words :/
[03:46:13] <morfic> yes
[03:46:41] <KimK_> morfic: OK, thanks
[03:49:01] <cradek> morfic: can you get to wichita in november?
[03:49:39] <cradek> KimK_: will you too?
[03:49:45] <morfic> cradek: and it bothered me i had formulas for him to check out, and i don't have time to truly test it out, like right now i am already past doing useful brainwork, idle chat, no problem
[03:50:51] <cradek> morfic: or are you one of those UKers - I can never remember where pepole are
[03:51:03] <morfic> cradek: i would love to, but it's a little hard to say, especially if we are still just before the moment that we put power to the machine it would be nice to get some last minute "damn, that's a neat idea" moments
[03:51:24] <morfic> cradek: nah, TX, although i could speak to IchGuckLive natively ;)
[03:51:38] <morfic> don't worry, i'm legal :)
[03:51:39] <cradek> ah, from TX is not a bad drive
[03:52:06] <morfic> more of a is wife or me going to be transitioning jobs at the time thing
[03:52:09] <cradek> I think mozmck came from TX last year
[03:52:17] <morfic> sometimes it's hard swallowing my pride
[03:52:25] <morfic> and no machinist is free of too much pride
[03:52:51] <cradek> I don't know what you're talking about again - why would you have to swallow your pride to hang out with us?
[03:52:57] <mozmck> yeah, north texas.
[03:53:15] <cradek> hi mozmck! everyone is reporting success with your latest cd.
[03:53:23] <mozmck> might not make it this time. Some other stuff may get in the way :(
[03:53:23] <morfic> a good machinist is a lot like a prima donna, easily getting their feelings hurt if the director tells them they did the part de deux wrong
[03:53:38] <cradek> darn
[03:53:53] <mozmck> hi cradek! that's good!
[03:53:59] <mozmck> yeah, we'll see.
[03:54:05] <morfic> cradek: difficulty swallowing my pride increasing == holding on to job for long decreasing
[03:54:06] <cradek> unfortunately it's impossible to pick a time everyone can go. it's the inverse birthday paradox.
[03:54:25] <morfic> i love what i do there, but i do not deal too well with 70yr olds
[03:54:26] <cradek> morfic: oh one of those times where the end's in sight? that sucks.
[03:55:50] <morfic> it's more of a soap box car race, not sure yet what's going to happen, even if you think you added controls, it's mostly gravity and hoping you get where you want to be when you want to be there, the finish line
[03:56:31] <morfic> i think i am too worn out to say this in a way i am granted keeping my IQ points though, tonight at least
[03:56:54] <morfic> can i /msg you real quick?
[03:57:22] <cradek> who me? you don't have to ask that.
[04:05:26] <KimK_> cradek: Thanks for your advice, I will look at the Python.org tutorial when I get around to Python, but have other fish to fry first. BTW, you may not have heard this, but I know of a fellow who is porting (forking) Qcad to Qt4. So this is a good thing because it frees Qcad-free from the Ribbonsoft folks, who have (in my opinion, at least) been dragging their feet on improving Qcad-free to promote Qcad-paid. New name not certain yet but the Qcad folks say, "d
[04:05:27] <KimK_> on't use Qcad".
[04:05:47] <KimK_> bah, flooding again.
[04:06:02] <cradek> I never understood the qcad free/pay setup
[04:06:11] <cradek> if someone starts improving the gpl version, that's terrific
[04:07:02] <KimK_> Yes, I think there's an inherent conflict of interest there. So, yes, to be coming out (back?) under GPL.
[04:07:58] <KimK_> Also looking better under Qt4
[04:14:54] <KimK_> I might not be around until late tomorrow, I am supposed to go to the Minnesota State Fair with John and his family tomorrow. It should be fun, I have been told about all kinds of food "on a stick".
[04:15:21] <cradek> yep I hear that's the big attraction
[04:15:39] <cradek> my mother encountered "salad on a stick" at the iowa state fair which was a week or two ago
[04:16:19] <L84Supper> whats the best way to varispeed drive a Bridgeport J head? Replace motor with servo or just use 3phase motor drive controller?
[04:17:44] <KimK_> Salad must take a long stick, lol! I think today was the first day, and it runs through this weekend and next, I believe?
[04:17:57] <cradek> sounds fun
[04:18:03] <L84Supper> http://www.truetex.com/dcdrv.htm
[04:18:11] <cradek> L84Supper: surely vfd is the simple answer
[04:18:29] <L84Supper> the stock motor is a joke
[04:18:35] <KimK_> L84Supper: Why not just use the existing motor? Ha, cradek beat me to it.
[04:19:05] <KimK_> Tell us what problems you have had with the motor?
[04:19:30] <L84Supper> after i plug it in I'll give you a list ;)
[04:19:55] <KimK_> J head, is that 2 HP?
[04:20:11] <L84Supper> nope 1958 1 HP
[04:21:25] <KimK_> Well, that's a little low, but is it a regular-sized Bridgeport? (Not a "2", or similar?)
[04:21:57] <L84Supper> nope, full size, only from the old school edition
[04:23:35] <L84Supper> I think it only works with horn rimmed shop glasses
[04:24:00] <cradek> and a white lab coat?
[04:24:21] <KimK_> Well, 1 HP isn't terrible. If you can find a 2 HP <TimAllen>or more</TimAllen>, then put that in, if it's cheap and easy. Otherwise, I'd say, "It is what it is."
[04:24:54] <KimK_> It's still R8, right?
[04:25:17] <L84Supper> that link above has a simple dc motor drive howto
[04:26:44] <L84Supper> with DC motor and the pulleys you could get some real torque on the edge of a bit
[04:27:30] <L84Supper> I might cnc this one
[04:28:16] <KimK_> Yes, well, he lists four benefits of DC motors (only three really, since a correctly-selected VFD can run on single-phase too), but he neglects to tell you the the two biggest disadvantages of DC motors.
[04:28:56] <L84Supper> really big batteries
[04:28:57] <moopy> i prefer brushless
[04:29:42] <KimK_> One: The evil commutator. Enough said. Two: Permanent magnets to get hot, crack, or de-magnetize by some other method.
[04:29:55] <L84Supper> single phase motor is an easy swap as well
[04:30:16] <moopy> i never have a magnet crack
[04:30:48] <KimK_> A single phase-motor? Ha, I'd go back to the DC motor first, lol.
[04:31:21] <L84Supper> steam and a flywheel?
[04:31:26] <Eric_K> I spent 3 months trying to demagnitize a motor, couldn't do it
[04:31:41] <moopy> ever used a brushless universal motor?
[04:32:02] <L84Supper> Erik_K: did you try grinding the magnets into powder?
[04:32:25] <moopy> if you use neodymium then heat will demag them
[04:32:35] <Eric_K> we ran the thing as hot as we could
[04:32:40] <Eric_K> ran it stalled
[04:32:40] <L84Supper> melting work as well
[04:32:49] <KimK_> OK, no cracking. How about heat? Vibration? Got any of that? Is the "magnetic strength" gauge on the outside of your DC motor reading as high as it did when it was new? (Yeah, my DC motors don't have a gauge like that either).
[04:34:36] <Eric_K> we used back emf as the measure of demagnitization
[04:35:10] <KimK_> The great thing about induction motors is when you shut them off the "magnets" are gone. When you turn them on again, the "magnets" come back, good as new.
[04:35:22] <L84Supper> 3HP lawnmower engine with centrifugal governor
[04:35:32] <KimK_> Hi Steve. Up late?
[04:35:39] <SteveStallings> SteveStallings is now known as steves_logging
[04:35:58] <KimK_> Ha, just starting the logger. Bye Steve.
[04:36:21] <L84Supper> why don't they make permanent magnets that you can turn off?
[04:36:30] <KimK_> There you go!
[04:36:32] <steves_logging> Hi Kim, just restarting a crashed system.
[04:36:40] <L84Supper> we'll make millions!!
[04:38:00] <Eric_K> how do indicator holder bases work? They obviously are permanent magnet, but you can turn it on and off
[04:39:33] <L84Supper> "I only have 3 phase wye, do I need a converter for single phase 120?"
[04:39:50] <L84Supper> some people shouldn't be working with machinery
[04:40:50] <KimK_> Ha
[04:41:54] <elmo40> http://gearjournal.com/
[04:42:04] <elmo40> L84Supper: only millions? that isn't enough
[04:42:35] <elmo40> Eric_K: they 'work' by moving the magnet away from the bottom.
[04:43:12] <elmo40> a few millimeters from the edge is all it takes to make the magnetic field weak enough to be able to move it.
[04:43:17] <elmo40> it never turns off.
[04:43:32] <L84Supper> anti-magnets!
[04:44:22] <L84Supper> absorbs your excess magnetism
[04:47:54] <elmo40> well... there are thousands of youtube vids on people claiming to bend magnetism to create motors that use no electricity.
[04:53:09] <L84Supper> wow $5K for a converted CNC Bridgeport
[04:54:36] <L84Supper> are new ones really ~$25K
[05:01:57] <ds2> how much are proper CNC mills (used)?
[05:02:55] <elmo40> need more info then that.. bed size, HP of motor, rapid IPM, spindle taper... brand/make/model
[05:04:13] <elmo40> a used 4-axis Mazak with a 15" pallet, CAT40 15HP motor, rapid ~2200IPM would go for $55k, if it was around 10-12 years old.
[05:04:21] <ds2> compareable to the bridgeport refits
[05:04:41] <ds2> 2200IPM? :)~~~~~~
[05:37:49] <L84Supper> http://www.smfmachinetools.com/warner3.htm Warner & Swasey Turret Lathe
[05:38:46] <L84Supper> not worth moving
[06:30:06] <The_Ball> had to upgrade the old ups hehe http://wigen.net/upsupgrade.jpg
[07:38:42] <IchGuckLive> good morning from germany
[07:39:20] <IchGuckLive> i just finisht the last python code for grid_generator and will now upload it to the wiki
[07:42:32] <IchGuckLive> Done!
[07:43:45] <awallin_> if anyone can get this pdf I would be interested: http://www.crcnetbase.com/doi/abs/10.1201/9780203859476.ch93
[07:46:34] <IchGuckLive> awallinthis is a fake
[07:46:55] <IchGuckLive> they want to pay you for this document as hi res
[07:47:28] <awallin_> the whole book is 259 dollars, I don't want to pay that much...
[07:49:11] <IchGuckLive> this is only a listing to well nown things so it is not worth 10$
[07:51:55] <IchGuckLive> http://www.google.de/#hl=de&source=hp&q=fully+GPU-based+volume+representation&aq=f&aqi=&aql=&oq=&gs_rfai=&fp=23ab45853b9e610e
[07:52:12] <IchGuckLive> take the second it is also the same
[07:52:26] <IchGuckLive> so it is grapt from another source
[07:52:48] <IchGuckLive> http://www.google.de/url?sa=t&source=web&cd=2&ved=0CCEQFjAB&url=http%3A%2F%2Fwww.mate.tue.nl%2Fmate%2Fpdfs%2F9881.pdf&rct=j&q=fully%20GPU-based%20volume%20representation&ei=NW53TMLvK9DIswbt6JjqBQ&usg=AFQjCNEukX7Uj2ZRoCOwLKllYf3y6RpXmA&cad=rja
[07:53:52] <awallin_> that paper might be interesting too, but it's not the same paper...
[07:56:15] <IchGuckLive> http://morpheus.pte.hu/~tuxi/
[07:56:58] <awallin_> yes, the videos are here: http://www.youtube.com/user/BalazsTukora
[07:57:03] <IchGuckLive> its a artikel over that thing not the code or things to use
[08:00:16] <IchGuckLive> http://www.youtube.com/watch?v=t2XLnpgrGP4 <- this is quit amazing
[08:01:06] <IchGuckLive> http://gpgpu.org/
[08:02:59] <IchGuckLive> awallin also available for 10.04 now
[08:06:14] <awallin_> throwing more hardware at the problem might be one solution
[08:06:23] <awallin_> improving the algorithms is another...
[08:06:49] <IchGuckLive> it uses VDPAU
[08:07:23] <awallin_> what?
[08:13:14] <IchGuckLive> oh i think im on the wrong path it is only a speeder for the logical agorithems
[12:37:10] <JT-Hardinge> * JT-Hardinge guesses the internet connection on the router is fried or dead...
[12:40:29] <Dave911> JT- I can see your message ...
[12:41:06] <JT-Hardinge> yea, I plugged into the switch that the antenna hits first here in the shop
[12:41:41] <JT-Hardinge> it's funny as my all my lan connections work on the router
[12:42:47] <JT-Hardinge> I have two cables running down to the router and swapped them with the same results lan ok internet doa
[12:43:18] <JT-Hardinge> * JT-Hardinge wonders if a router can be half borked like that or should I keep looking
[12:47:56] <Dave911> Network gear for me has always physically worked or been entirely dead..... but that is just my experience... usually for me it turns out to be operator error.. ;-)
[12:49:17] <JT-Hardinge> another funny thing is neither cable running down to the beer cave will pass the internet straight to a computer from the switch but here in the shop I can plug into the switch and it works
[12:49:42] <JT-Hardinge> but both cables will pass the lan with no problems...
[12:52:05] <JT-Hardinge> oh well time to head out to the other shop and make some parts
[12:53:55] <Dave911> wierd..
[13:02:44] <Valen> routers can often half die
[13:26:38] <Dave911> Really... I have never heard of that. Good thing they are pretty cheap now.
[13:40:44] <Jack> Jack is now known as Guest53367
[13:44:51] <IchGuckLive> hi all
[13:46:54] <JT-Work> hi
[13:50:31] <IchGuckLive> heavy weather today in germany
[13:52:19] <Valen> Dave911: ADSL ones especially are prone to sucking
[13:53:12] <IchGuckLive> Valen: DSL16000 B)
[13:54:19] <Valen> ?
[13:56:14] <IchGuckLive> lucky how has a good Inet connect
[13:56:42] <Valen> i get 6-8 mbit roughly
[14:00:47] <IchGuckLive> thats way enoph for good download
[14:11:57] <Jymmm> Mornin
[14:12:29] <IchGuckLive> :DD good evening
[14:13:23] <skunkworks> morning (not quite as much morning as jymmm)
[14:13:43] <Jymmm> heh
[14:14:06] <awallin_> if I have one of four bits true (from a pyvcp radibutton), how do I convert that to a 2-bit binary number?
[14:14:48] <awallin_> radiobutton
[14:14:54] <skunkworks> hmm - I did that in ladder - but I would guess you could do it in hal. (converted 15 bits to a number)
[14:15:42] <skunkworks> here
[14:15:44] <skunkworks> http://www.linuxcnc.org/docs/2.4/html/man/man9/weighted_sum.9.html
[14:15:52] <Jymmm> awallin: if it's bits, it already is binary, you want hex or decimal instead?
[14:16:26] <skunkworks> awallin: weighted_sum − convert a group of bits to an integer
[14:16:36] <awallin_> I have one of four bits true, and I want that to map to 00, 01, 10, 11
[14:16:54] <awallin_> so 1000, 0100, 0010, 0001 should map to 00, 01, 10, 11
[14:17:08] <skunkworks> oh
[14:17:15] <skunkworks> make a comp
[14:17:20] <skunkworks> :)
[14:17:40] <awallin_> yeah, this is supposed to show a newbie how easy and flexible HAL is...
[14:17:46] <skunkworks> heh
[14:18:07] <skunkworks> I can't think of anything... but others might have a solution
[14:18:22] <awallin_> I'm using a mux4 to set axis.N.jog-scale
[14:18:31] <awallin_> maybe there are better ways of selecting the jog-scale ?
[14:18:44] <Jymmm> 1000 is 16bit, you would need to make the conversion too
[14:18:59] <skunkworks> awallin: what about http://www.linuxcnc.org/docs/2.4/html/man/man9/mux4.9.html
[14:19:11] <skunkworks> I don't quite understand what it does - but sounds close
[14:19:36] <IchGuckLive> Jymmm: 4bits not the nummber
[14:20:05] <Jymmm> IchGuckLive: He's doing radio (one of many)
[14:20:09] <IchGuckLive> number is 8
[14:20:12] <Jymmm> but doens't want 1100
[14:20:37] <Jymmm> is it divide by 2?
[14:20:39] <awallin_> four match8 components would do it, but it's not exactly elegant
[14:21:06] <awallin_> who designed pyvcp and hal this way anyway... :|
[14:21:40] <awallin_> can pin values be set directly?
[14:21:45] <awallin_> or did that go away?
[14:21:48] <Jymmm> 1000 = 8d, wants it to be 11 = 3d, so 1000 /2 - 1, is that right ?
[14:23:17] <Jymmm> too early ti e thinking of bitwise operations =)
[14:23:21] <Jymmm> to be
[14:24:11] <Jymmm> awallin: shift left
[14:24:21] <IchGuckLive> >>
[14:24:32] <IchGuckLive> this will fid a solution
[14:29:49] <Jymmm> awallin: Heh I found your solution
[14:30:48] <Jymmm> awallin: if tou take the 0001, 0010, 0100, 1000 and treat them as decimal, then its power of ten
[14:31:01] <Jymmm> 10^3 == 1000
[14:31:11] <Jymmm> 10^2 = 0100
[14:31:20] <Jymmm> 10^1 == 0010
[14:33:06] <skunkworks> 10^0 is 1
[14:33:20] <Jymmm> 10^0 == 0001
[14:33:23] <Jymmm> yep
[14:33:39] <Jymmm> 0001 1d == 00 0d
[14:33:39] <Jymmm> 0010 2d == 01 1d
[14:33:39] <Jymmm> 0100 4d == 10 2d
[14:33:40] <Jymmm> 1000 8d == 11 3d
[14:33:51] <Jymmm> and that's exactly what he wanted
[14:34:13] <alex_joni> awallin_: setp
[14:34:16] <Jymmm> just convert it to binary
[14:34:28] <awallin_> it's not exactly the math that is rocket science, it's more how to do it nicely in HAL
[14:34:40] <alex_joni> Jymmm: yours is backwards
[14:34:53] <Jymmm> alex_joni: then invert it
[14:35:31] <Jymmm> alex_joni: that's just a truth table of what he descrobed
[14:35:46] <Jymmm> I just added the decimal values
[14:36:09] <alex_joni> still probably easiest/nicest to write a small comp
[14:36:27] <Jymmm> compensation?
[14:36:43] <Jymmm> ppl REALLY need to stop using "comp" as an abbr.
[14:36:57] <Jymmm> computation, computer, compensation, etc
[14:37:51] <alex_joni> http://linuxcnc.org/docs/html/hal_comp.html
[14:38:22] <Jymmm> computation, computer, compensation, component, etc
[14:39:44] <awallin_> http://www.electronics-tutorials.ws/combination/comb_4.html
[14:39:53] <awallin_> too much to ask to have that as standard in emc2 ?
[14:41:43] <IchGuckLive> im off next thunderstorm coming fast my way
[14:42:20] <Jymmm> Ah good old BCD
[14:44:50] <Jymmm> awallin_: I'm sure if you write it up, they'll include it
[14:46:59] <Jymmm> Kingston 64GB SSD == $99.99
[14:52:46] <awallin_> hmh, comp --install seems to install my new component into the system directory, but I am "running in place"
[14:53:13] <alex_joni> maybe you used the wrong comp?
[14:53:20] <alex_joni> did you . scripts/emc-environment ?
[14:53:38] <awallin_> yes, before that the shell did not find "comp"
[14:54:14] <awallin_> hm it goes to emc2/rtlib
[14:54:45] <awallin_> eh, now its working
[14:54:52] <awallin_> but addf fails
[14:55:13] <skunkworks> did you loadrt the new comp?
[14:56:28] <awallin_> yes now it works I am just struggling with the name of the function
[14:57:52] <awallin_> addf mycomp.0
[14:57:54] <awallin_> that fails...
[14:58:44] <skunkworks> if loadrt works - can't you do a show and get the exact funtion name?
[14:59:48] <awallin_> eh I get mycomp.0.mycomp
[14:59:56] <skunkworks> odd
[15:03:55] <awallin_> yeh, now it works
[15:07:46] <awallin_> the 4-input priority encoder: http://pastebin.ca/1926878
[15:09:24] <Jymmm> Jymmm is now known as Red70sShow
[15:09:34] <Red70sShow> Red70sShow is now known as Jymmm
[15:15:32] <awallin_> if anyone can make this more compact or simple, let me know: http://www.anderswallin.net/2010/08/axis-with-pyvcp-jogwheel/
[15:16:41] <pcw_home> Actually thats an encoder but not a priority encoder
[15:16:42] <pcw_home> priority encoder would only test a single bit at once (so a code like 1011 encodes to 11)
[15:17:07] <awallin_> right.
[15:17:17] <awallin_> it would only detect the first 1
[15:17:24] <ries> it's funny to see that every other month people want to have progress bars...
[15:18:32] <pcw_home> A plain encoder may want to indicate an error if more than one bit is set
[15:21:19] <pcw_home> (or have a defined return code in that case)
[15:23:24] <awallin_> have spammed this to the list now...
[15:26:23] <Eric_K> pcw_home: any due date on the brushless drivers?
[15:29:49] <pcw_home> We've been fussing with the EMC interface firmware. Looks good now so nex t week or so we will send a couple 8I20s to Andy Pugh
[15:29:51] <pcw_home> and see how bad making a driver looks (were trying to hide as much protocol complexity from the driver as possible)
[15:30:54] <Eric_K> Eric_K is now known as EricKeller
[16:21:39] <awallin_> I wonder if this gpu computing with OpenCL is worth getting into... or will there be an OpenMP or gcc version that does the parallellization magically just around the corner...
[16:22:27] <morficmobile> if the gpus still out calculate a dual core, why not
[16:29:27] <Jymmm> LOL http://www.citycarshare.org/FAQs.do#smokingandpets
[16:50:25] <IchGuckLive> hi from the wet Germany 80+liters today
[17:29:08] <awallin> is "comp" included with a standard emc2 install, or does it require build-dep or something similar?
[17:32:24] <awallin> wiki says "emc2-dev" and "build-essential" are required...
[18:00:57] <awallin> this one says the item is new: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=250687278583
[18:01:12] <awallin> but who makes that one? do they have a website? dimensions, drawings, specs??
[18:39:08] <Jymmm> awallin: http://www.boxford.co.uk/boxford/
[19:20:14] <awallin> Jymmm: looking...
[19:20:38] <Jymmm> best I could find for that tool changer
[19:22:01] <awallin> http://www.boxford.co.uk/boxford/docs/products/productimg/cnclathes/160bit1.jpg
[19:22:12] <awallin> that looks a bit smaller?
[19:30:39] <Jymmm> I really can't tell, doens't look the same to *me*
[19:44:34] <awallin> in a normal emc2 install, shouldn't the . ./scripts/emc-environment be run automatically?
[19:45:11] <cradek> do you mean an installed deb package?
[19:46:10] <awallin> trying to help a friend, probably installed with the automagic shell-script I would guess
[19:46:40] <awallin> he doesn't find "comp", and I suggested packages emc2-dev and build-essential, but that did not help
[19:46:45] <cradek> if a deb package is installed, appropriate things are in the system PATH and MANPATH so you don't need to run any special thing
[19:47:07] <cradek> yes, if you install emc2-dev you'll automatically get "comp" in your path
[19:47:08] <awallin> I suggested trying some other normal tools: halmeter, halscope, pyvcp
[19:47:22] <awallin> if the other tools are in the path it should be ok?
[19:48:10] <cradek> if you install the emc deb, halmeter and halscope are in the system path
[19:48:26] <cradek> if you install the emc-dev deb, comp is in the system path
[19:48:40] <cradek> (er I mean emc2 and emc2-dev)
[19:49:44] <Jymmm> Anyone have any suggestion on attaching a duct flange to the SIDE of a pipe?
[19:50:00] <cradek> use a ductwork T
[19:50:16] <JT-Work> learn how to be a tinsmith
[19:50:33] <Jymmm> The pipe in this case in a 12" cardbaord concrete form.
[19:50:50] <skunkworks> hot glue?
[19:50:55] <skunkworks> ;)
[19:51:02] <Jymmm> silly putty
[19:51:21] <Jymmm> it's the bevel that I can't figure out
[19:51:27] <Jymmm> err arc
[19:51:52] <JT-Work> you make a ring then cut the pipe to fit the larger pipe then bend the edge of the pipe over the ring then screw it on
[19:53:03] <Jymmm> JT-Work: flex duct (sorry)
[19:53:04] <JT-Work> Jymmm: I have the methods for laying that out that at home somewhere
[19:53:22] <Jymmm> JT-Work: I'll take what you got when you have a chance
[19:53:58] <JT-Work> basically you lay it out on flat paper then wrap that around the pipe to give you the cut line
[19:57:06] <Jymmm> I need something like this ----> |===(O where the 'O' is looking into the end of the concrete form, and the |==( is the duct that has been concaved
[20:08:42] <JT-Work> yep or just just a hole in the concrete tube the size of the small tube, stick it in, whoop out your magic marker and lay it out on the smaller pipe
[20:17:15] <Dave911> Jymmm that is called "coping" a joint....as in a coping saw.. etc which can be used to "cope" moulding etc. http://www.metalgeek.com/static/cope.pcgi
[20:41:18] <odiug> Hi! Does anybody here use EMC for driving a RepStrap?
[20:45:36] <skunkworks> odiug:
[20:45:39] <skunkworks> http://objects.reprap.org/wiki/EMCRepStrap
[20:46:19] <odiug> Yes, I was currently looking through the wiki pages.
[20:46:44] <odiug> I am a bit confused by the many different extruders.
[20:49:05] <Fox_M|afk> Fox_M|afk is now known as Fox_Muldr
[21:02:20] <packrat> sup dudez
[21:32:53] <Jymmm> Dave911: thanks!
[21:34:59] <Jymmm> Dave911: which tube is which?
[21:39:02] <Jymmm> ok I think I figured it out
[22:45:05] <ries_> ries_ is now known as ries
[23:04:04] <jthornton> to setup this router your computer must be connected to the internet...
[23:04:30] <jthornton> please install this CD on a computer that has an internet connection LOL
[23:08:58] <ries> lol
[23:13:07] <jthornton> now that internet has been restored time to clean a spot 72" x 75" and a path to that spot :)
[23:16:04] <ries> jthornton: you must have a big router!
[23:18:37] <JT-Hardinge> and it appears to work out here in the shop too HORRAY!!!
[23:22:47] <Dave911> Jymmm: Np Happy coping ;-)
[23:22:58] <Jymmm> lol
[23:23:33] <Dave911> Hey Jthornton: What kind of router needs to be connected to the internet??
[23:30:25] <JT-Hardinge> the sticker on the wrap for the CD lol
[23:30:40] <JT-Hardinge> hey anyone use DNMG432 inserts?
[23:54:35] <Fox_Muldr> Fox_Muldr is now known as Fox_M|afk