#emc | Logs for 2007-10-23

Back
[00:06:31] <tomp> skunkworks: just madwifi
[00:13:00] <toast> THIS IS A HOLD UP
[00:13:34] <SWPadnos> for some reason, that reminds me of "Raising Arizona"
[00:14:01] <SWPadnos> "well son, which is it? iffin I freeze, I can't drop, and if I drop, then I'm gonna be in motion"
[00:15:01] <toast> lol
[00:22:12] <chr0n1c> woohoo i got my emc2 controller computer back up and running!
[00:22:27] <chr0n1c> *found a new power supply in a dumpster last night
[00:22:33] <chr0n1c> ;)
[00:25:15] <JymmmEMC> cool beans
[00:25:50] <skinnypuppy34> ha ... reminds me of a friend that calls dumpster diving a greenstore visit. I'm running a powersupply from a greenstore ;o)
[00:26:50] <jlmjvm> i finally got my motors mounted
[00:27:04] <toast> that's not all you mounted.
[00:27:53] <skinnypuppy34> Nice jlmjvm, I'm working out how my brackets/motors/belts are going to fit in mastercam.
[00:28:20] <skinnypuppy34> jlmjvm: you've probably said before but what kind of mill are you doing?
[00:28:47] <jlmjvm> bridgeport,older boss 2 or 3
[00:31:23] <skinnypuppy34> Cool, we an old boss with a busted controller at the school.
[00:38:21] <chr0n1c> yes.. dumpster diving is definatelyy saving the enviroment a little everytime i find something and take it hom to re-use
[00:38:39] <chr0n1c> lol... greenstore shopping. i like it
[00:39:17] <chr0n1c> the amount of wasted packing materials alone i find... hundreds of dollars in every dumpster
[00:41:00] <fenn> packing materials?
[00:41:08] <fenn> what do you do with that
[00:41:42] <chr0n1c> bubble wrap... shipping tubes.. any size box you could imagine.. i re-use some of it for shipping crazy whatever stuff...
[00:42:09] <fenn> go check out the grocery store
[00:42:29] <fenn> or a plastics distributor
[00:43:02] <fenn> i used to get carloads of books from the used bookstores to send to prisoners
[00:43:39] <chr0n1c> sweet... when i find programming books i grab them
[00:43:54] <chr0n1c> and there is some killer vinyl records out there in dumpsters
[00:44:22] <chr0n1c> c++, java, visual basic, html...
[00:44:37] <chr0n1c> books with tons of good knowledge just getting tossed out.. so sad
[00:44:41] <chr0n1c> lol
[00:44:44] <fenn> who cares
[00:44:51] <fenn> its all available for free on the internet
[00:45:31] <chr0n1c> well, i like learning.. and i don''t wanna spend 50 bucks on those books... and i could read one stranded with no power in a nuclear blast...
[00:45:44] <fenn> yeah lots of good that would do you :)
[00:45:48] <chr0n1c> thanks
[00:46:02] <chr0n1c> oh.. read that wrong
[00:46:06] <chr0n1c> anyways
[00:46:18] <chr0n1c> i just updated my emc2...
[00:46:22] <fenn> you could read a book about lisp simply "for the sheer enlightenment once you finally get it"
[00:46:25] <chr0n1c> is there any new toys in it?
[00:46:45] <fenn> um.. 2.2 is around the corner
[00:46:54] <fenn> i dont know how old your previous install is
[00:47:08] <chr0n1c> ahh i read a bit on the dev list about some bugs that are getting worked out.
[00:47:25] <chr0n1c> i forget what version i was using
[00:47:43] <chr0n1c> i think it was still whatever version is on the live cd unless it was updated recently
[00:47:53] <chr0n1c> *the live cd
[00:49:46] <jepler> chr0n1c: the latest released versions are bugfixes -- if you had a recent Live CD and upgraded, you probably got emc 2.1.7 instead of emc 2.1.6. There's a list of changes in each bugfix release: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Released
[00:50:06] <chr0n1c> ty, jepler.
[01:21:52] <jmkasunich> curse you Microsoft!
[01:21:55] <jmkasunich> and curse Epson too
[01:22:11] <skunkworks> * skunkworks shouts 'plug and play'
[01:22:47] <jmkasunich> would you believe that my shiny new printer came with a firewire cable but not a USB cable?
[01:22:51] <SWPadnos> it works better in engrish: "prug and pray"
[01:23:34] <jmkasunich> who the heck uses firewire anyway?
[01:23:38] <SWPadnos> I do
[01:23:44] <SWPadnos> but not for printers
[01:23:51] <SWPadnos> or motor control
[01:23:56] <chr0n1c> SPEAK ENGRISH!
[01:24:04] <jmkasunich> I made the mistake of trying, instead of immediately going to the computer store for a usb cable
[01:24:04] <chr0n1c> OMG ENGRISH... lol
[01:24:08] <jmkasunich> now the damn store is closed
[01:24:22] <SWPadnos> do you have no other USB cables around?
[01:24:31] <skunkworks> I have some usb cables in the kitchen
[01:24:29] <chr0n1c> i thought printers were parallel?
[01:24:34] <jmkasunich> not that I can find
[01:24:41] <chr0n1c> i have an extra usb cable ehre too
[01:24:43] <chr0n1c> here*
[01:24:48] <SWPadnos> then again, I have 100 cables, but they never seem to have the right A-B / A-miniB / whatever connections I need
[01:24:53] <jmkasunich> chr0n1c: they were before M$ came out with USB
[01:25:04] <chr0n1c> i do like usb...
[01:25:04] <SWPadnos> that was Intel's fault, I think
[01:25:13] <SWPadnos> well, shared blame anyway
[01:25:15] <fenn> i first saw it on apple computers..
[01:25:20] <SWPadnos> no, that was ADB
[01:25:21] <fenn> firewire too
[01:25:24] <jmkasunich> SWPadnos: I have _one_ cable, that goes to the webcam - it has the tiny little device-end connector, not the squarish one
[01:25:27] <SWPadnos> that's what USB was invented to compete with
[01:25:48] <SWPadnos> hmmm
[01:25:48] <chr0n1c> i have my midi keyboard, a midi drum trigger pad, a printer, a wifi adapter, a usb speaker set all hooked up
[01:26:01] <chr0n1c> with serial or paralell that would take a lot of room on a box for connectors
[01:26:12] <jmkasunich> shove one of your extra cables thru the intraweb willya?
[01:26:22] <jmkasunich> sorry, intarweb
[01:26:25] <chr0n1c> **zap to jmk**
[01:26:35] <SWPadnos> you can probably go to Home Depot and get a USB cable
[01:26:41] <SWPadnos> they should still be open
[01:26:47] <jmkasunich> hmm, worth a try
[01:26:50] <fenn> not home depot.. a drug store maybe
[01:26:49] <jmkasunich> bbl
[01:26:50] <chr0n1c> jmkasunich: go dumpster diving... now is the perfect time if the store is closed
[01:26:55] <skunkworks> do you have walmart?
[01:26:59] <chr0n1c> i bet there is a usb cable in there
[01:27:18] <fenn> chr0n1c: not since they stopped shipping cables with printers
[01:27:30] <chr0n1c> i found one last night!
[01:27:32] <chr0n1c> ;)
[01:27:47] <chr0n1c> no kidding... and the power supply running this box
[01:28:00] <chr0n1c> a 400 watter, brand new
[01:28:17] <chr0n1c> they didn't even step on it to break the fan...(punks)
[01:32:45] <jepler> jmkasunich: which printer did you choose?
[01:32:51] <jepler> I hope you get to try it out some day soon
[01:34:01] <chr0n1c> my HP PSC-1350 XL all in one has started giving me an error after each page i have to reset it?
[01:34:14] <CIA-22> 03jepler 07TRUNK * 10emc2/docs/src/ (Submakefile lyxtree.py): produce xml files to temporary directory instead of cluttering docs/html/
[01:34:19] <chr0n1c> i am going to have to buy a new printer soon
[01:47:14] <jmkasunich> home desperate comes thru!
[01:47:31] <jmkasunich> jepler: Epson R1800
[01:48:14] <jepler> after you print on it you'll take a photograph for me to see the quality, right?
[01:48:29] <jmkasunich> of course
[01:55:46] <stustev> good evening
[01:55:56] <fenn> howdy
[01:56:16] <stustev> what's going on this evening?
[01:56:39] <fenn> i'm writing a grant proposal thingy
[01:56:53] <fenn> for development of low cost hexapods
[01:57:22] <stustev> for any industry in particular?
[01:57:37] <fenn> for hobbyists and small businesses in developing countries
[01:57:50] <fenn> and (eventually) consumers
[01:58:10] <fenn> once we get to the dreamland fabber in every basement open-source utopia
[01:58:38] <stustev> we're almost there
[01:58:50] <SWPadnos> and I'm finishing up (I think) the control software for a large ECM power supply
[01:59:09] <stustev> what is ECM?
[01:59:08] <SWPadnos> using HAL, the mesa 5i22, and other various and sundry things
[01:59:13] <SWPadnos> electrochemical machining
[01:59:38] <SWPadnos> your next investment ;)
[01:59:39] <stustev> does ECM use a mask?
[01:59:46] <SWPadnos> I have no idea
[02:00:06] <fenn> i think it uses shaped electrodes
[02:00:08] <SWPadnos> I'm just doing the pulse waveform controller part of it
[02:00:46] <stustev> shaped electrodes would be EDM - Electro Discharge Machining
[02:00:56] <fenn> no, it dissolves the metal instead of blasting bits off
[02:01:03] <SWPadnos> I think it's similar to EDM in that you have a shaped electrode, but it uses chemical etching rather than electrical sparking for the erosion
[02:02:07] <SWPadnos> they should be able to erode plenty of metal with their machine, the power supply is 36V 20000A
[02:02:23] <stustev> doesn't sound like there is much difference. What is the chemical?
[02:02:30] <SWPadnos> dunno
[02:02:36] <chr0n1c> battery acid!
[02:02:46] <chr0n1c> i jest i jest
[02:02:54] <SWPadnos> http://www.eod.gvsu.edu/eod/manufact/manufact-281.html
[02:03:00] <SWPadnos> http://electrochem.cwru.edu/ed/encycl/art-m03-machining.htm
[02:03:16] <SWPadnos> just a couple of google hits there, they may not be great
[02:03:30] <stustev> I will look at them
[02:04:56] <SWPadnos> interesting - in ECM, there is no tool wear
[02:05:30] <SWPadnos> no thermal or mechanical stress on the part either
[02:05:34] <SWPadnos> http://en.wikipedia.org/wiki/Electro_chemical_machining
[02:06:36] <fenn> you only need a small power plant to run the thing
[02:07:44] <SWPadnos> it's just 720 kW
[02:07:48] <SWPadnos> I don't see the problem ;)
[02:07:51] <jmkasunich> thats only 720KVA
[02:07:59] <SWPadnos> DC -kW
[02:08:05] <jmkasunich> pthbbt
[02:08:08] <SWPadnos> heh
[02:08:13] <fenn> its only 1000HP continuous
[02:08:20] <SWPadnos> roughly
[02:08:28] <fenn> any good train engine should do it
[02:08:50] <SWPadnos> lessee - on a normal mill, that would be roughly what, 4000 cubic inches/minutes of Aluminum?
[02:09:06] <SWPadnos> (Al is 0.25CI/Hp, isn't it?)
[02:09:15] <SWPadnos> err - CI/min per HP
[02:09:18] <jmkasunich> something like that
[02:10:03] <fenn> and this is my 2.5cfm aluminum compressor...
[02:10:18] <fenn> uncompressor
[02:12:54] <Ziegler> ecm cool process
[02:18:36] <stustev> the 2005 article on ECM stated the power supply is the bottleneck of the development of pulsed ECM.
[02:19:05] <stustev> all train locomotives are DC
[02:19:35] <SWPadnos> the company I'm doing this for regularly does multi-megawatt power supplies. I'm not sure that's going to continue to be the problem :)
[02:19:51] <SWPadnos> I think the largest I worked on was ~11 MW
[02:22:17] <stustev> I have been studying python.
[02:22:47] <stustev> today I was able to modify the /bin/axis file to get the behavior I want in the AXIS interface.
[02:23:01] <SWPadnos> so it appears. I think your changes are correct, but I'm not the python guru here
[02:23:08] <CIA-22> 03jepler 07TRUNK * 10emc2/src/hal/drivers/pluto_step_firmware/stepgen.v: fix description of the software
[02:23:09] <stustev> my efforts are very simple but I am making some progress
[02:23:18] <SWPadnos> that's the way to start
[02:23:55] <cradek> hi stuart!
[02:24:05] <stustev> hello Chris
[02:24:12] <cradek> be careful or we'll start assigning bugs to you
[02:24:29] <stustev> I will be VERY careful
[02:24:36] <cradek> haha
[02:24:48] <cradek> is your machine moving yet?
[02:25:06] <stustev> not even close
[02:25:28] <stustev> I have just been working on my desktop
[02:25:33] <cradek> ah
[02:25:56] <stustev> I will keep you posted on the progress as we wire it.
[02:26:11] <cradek> I'm itching to see it go
[02:26:29] <stustev> I should take some pictures. I can hardly wait to see it move again.
[02:27:17] <fenn> you are wiring up the vf-11 under emc? or am i getting mixed up
[02:27:56] <stustev> we are wiring a Giddings and Lewis boring mill. I will post some pictures.
[02:28:47] <cradek> hey did you find any leads on a transformer for me? I think I found one that will work at HGR, but I'd rather have an excuse to visit you than pay for shipping from there
[02:29:17] <stustev> the Haas five axis is a VR-11. The VF series are three axis mills.
[02:29:38] <tomp> SWPadnos: what sort of conductors do you use? we find lots of reflection & have no suitable cables for the range we want ( we make do with best midrange ) http://diglib.uni-magdeburg.de/Dissertationen/2007/fraluhn.pdf
[02:30:01] <stustev> I haven't found anything. I have asked around.
[02:30:35] <cradek> thanks for trying, it's an unusual transformer I think
[02:31:00] <stustev> I think so too. No one I talked to had ever seen one with the specs you want.
[02:31:26] <cradek> I think wanting to run such a large machine on single phase is uncommon
[02:31:38] <stustev> Is it a normal transformer wired different?
[02:31:59] <stustev> Or is it wound for the specs you want?
[02:32:01] <fenn> they all look like beige boxes to me :\
[02:32:14] <stustev> I can find a lot of boxes.
[02:32:16] <cradek> the one I found (actually jmk found it) is rated slightly less, but will be ok
[02:32:27] <fenn> haas mills, not transformers
[02:32:32] <stustev> Where is HGR?
[02:32:38] <tomp> SWPadnos: sorry, you might not be pulsed in ECM (never mind)
[02:32:39] <cradek> cleveland unfortunately
[02:32:43] <cradek> I really wish it was closer
[02:33:10] <stustev> no kidding. Cleveland is 12 hours from Lincoln?
[02:33:11] <SWPadnos> tomp, I don't do ECM, I just make control systems for the power supplies ;)
[02:39:55] <cradek> google says 13 hrs
[02:40:23] <stustev> maybe I drive a little faster :)
[02:40:47] <stustev> with the powerful car I drive
[02:40:50] <cradek> not through iowa, if you're smart...
[02:40:54] <cradek> haha
[02:41:13] <cradek> "brace yourself!"
[02:41:19] <stustev> Ha Ha
[02:41:27] <fenn> brace yourself for 60
[02:41:37] <fenn> where's my jetpack..
[02:42:03] <stustev> fenn, I drive a 1984 Mercury Lynx diesel.
[02:42:32] <stustev> The power is unbelievable. Unbelievably weak.
[02:42:50] <fenn> well.. it does the job right?
[02:42:58] <stustev> 0 to 60 is measured in minutes or miles.
[02:43:15] <cradek> for a 23 year old car it looks ... well, pretty much how you'd expect
[02:43:27] <fenn> you could convert it to vegetable oil if it would make you feel better
[02:43:38] <stustev> yes it gets me from here to there. I get 40 to 45 mpg in city driving.
[02:45:11] <stustev> I have checked out the waste vegetable oil conversion. All, and I mean all, the sources of WVO in Wichita have contracts for their waste oil.
[02:45:34] <jepler> 'night all
[02:45:48] <cradek> I bet the era of free waste oil came and went pretty fast
[02:45:55] <cradek> g'night jepler
[02:45:53] <skinnypuppy34> There's lots of stuff in waste oil that will cause top rings to stick and loose compression, pump problems too if it is acidic
[02:46:25] <stustev> There is virtually no waste oil dumped. It is all recycled. Yes the era was very short.
[02:47:02] <fenn> well thats nice. i know in texas they're required to dump it, otherwise it's "illegal re-use of hazardous waste in a consumer product" or some nonsense
[02:47:10] <skinnypuppy34> jeez
[02:47:20] <SWPadnos> "Illegal cutting into oil company profits"
[02:47:40] <stustev> follow the money to the root of the reason
[02:47:58] <SWPadnos> money is the axis of evil :)
[02:48:08] <fenn> "Occasionally someone raised the possibility that the basic cause of poverty was that poor people had no money, but such suggestions were almost unanimously rejected as hopelessly naive and simplistic."
[02:48:45] <fenn> james albus is a pretty cool guy
[02:48:57] <stustev> I would think someone has figured out a way around the dumping. If there is money to be made someone is doing it.
[02:49:10] <fenn> he invented the robo-crane
[02:49:55] <fenn> stustev: that's pretty fatalist thinking, and wont get you anywhere
[02:50:20] <stustev> It may be fatalist but it is also realist.
[02:50:29] <fenn> all fatalists think they are realists
[02:50:41] <SWPadnos> let's rephrase that into pseudo-modernistic financial double-speak:
[02:50:43] <fenn> its no fun to say you're living in a terrible hopeless fantasy
[02:50:50] <stustev> but not all realists are fatalists
[02:51:25] <SWPadnos> "If there's a path to wealth, then surely some entrpreneur has codified the idea into a patentable business process, and is pursuing venture capital to realize their invention"
[02:51:36] <stustev> I have entered the Calvin and Hobbes zone.
[02:51:50] <SWPadnos> heh
[02:51:55] <fenn> yes, pursuing venture capital, or sitting on a patent, but probably not actually doing anything
[02:52:06] <SWPadnos> this is the transmogrifier. please step aside quickly to let others pass through
[02:52:43] <stustev> the days are just packed
[02:52:47] <fenn> hobbes is a weakly god-like artificial intelligence who granted calvin his wish to be a kid forever
[02:53:12] <stustev> where is my hobbes?
[02:55:08] <stustev> good night, gentlemen
[02:55:30] <fenn> "Tragically, utopian dreams went out of style just at the time in history when science and technology had reached a level where the elimination of physical poverty had become a real possibility."
[02:57:25] <fenn> maybe its just that once you see how they make sausage you arent hungry anymore
[03:00:03] <skinnypuppy34> Any of you guys ever see a laptop that will not turn on when the humidity is up? Place it atop a monitor or amp with a little rising heat and it works fine after about 30 minutes of heating. I'm stumped
[03:01:36] <fenn> i had one that turned off if it was cold in the room and you moved the mouse too much
[03:01:57] <skinnypuppy34> Works fine for weeks till it rains....
[03:02:30] <skinnypuppy34> Damn shame laptop is just outta waranty....
[03:02:32] <cradek> I've heard some strange hardware problems but that's not one I'm familiar with
[03:03:43] <Rabbitbunny> i had that happen once, lemme think what it was
[03:04:09] <skinnypuppy34> Same here... I've thought of cracking it open and reseating everything possible. Battery or wall same deal.
[03:05:25] <tomp> your solution was to heat the laptop, i didnt see humidity as the cause, but suspected heat make a better connection when something expanded ( what i pictured happening )
[03:06:44] <fenn> might want to re-solder anything that looks suspicious
[03:07:12] <skinnypuppy34> Yeah I guess there isn't any solution other than dig in.
[03:07:34] <tomp> shoot anythig suspicious ( the 'western' method haha a pun )
[03:07:41] <Rabbitbunny> lol
[03:08:05] <skinnypuppy34> pray for no extra screws
[03:08:23] <skinnypuppy34> when finished that is
[03:08:25] <tomp> if it >is< a bad connection, then digging in will be .... interesting
[03:09:37] <tomp> http://www.jir.com/ the journal of irreproducable results
[04:01:05] <fenn> toast!
[04:02:13] <fenn> what's your opinion of hexapods?
[04:05:19] <toast> NEAT
[04:05:21] <toast> not really sure of the accuracy
[04:05:52] <toast> they seem to be pretty experimental, it would be neat to see what kind of things would make them mainstream
[04:06:26] <fenn> why not sure of the accuracy?
[04:06:54] <fenn> just you dont know or have legitimate concerns?
[04:07:05] <toast> there are issues with measuring and correcting six points of control
[04:07:10] <toast> at the tool tip
[04:07:34] <fenn> issues beyond what you have with 3 axes?
[04:07:39] <toast> yeah
[04:08:16] <toast> i still think that they are cool, though
[04:08:23] <fenn> issues such as...
[04:08:43] <toast> i didn't read the paper, just made note of it
[04:08:46] <toast> sorry
[04:08:52] <fenn> hrm. ok
[04:10:01] <fenn> well its probably not all crap if NIST used one to machine master calibration fixtures for CMM's
[04:10:32] <toast> you mean the ball grids?
[04:10:36] <fenn> yes
[04:10:45] <toast> those don't have to be accurate, just the balls
[04:10:59] <toast> and i was pretty sure the balls were lapped?
[04:11:58] <toast> there's a reversal process that removes any error in the grid, but the balls have to be true
[04:12:24] <fenn> perhaps we arent talking about the same thing
[04:12:27] <toast> mabye?
[04:12:47] <toast> the cmm test matrix i'm familiar is a 4x4 square or a cross
[04:13:02] <toast> with lapped balls fixed at the intersection of the grid
[04:13:59] <fenn> this is what i thought they used the hexapod for: http://www.mel.nist.gov/photos/photo/ph046.html
[04:14:10] <toast> yar, those balls are lapped
[04:14:19] <toast> that grid can be out like, any amount they want
[04:14:37] <toast> they rotate the grid around and re-test the distance between the balls
[04:14:52] <fenn> why do they make it out of a solid chunk of aluminum then?
[04:15:05] <toast> so the distance between the balls doesn't change with shock
[04:15:11] <toast> welded aluminum shifts a lot
[04:15:45] <SWPadnos> NIST machined the titanium encasements for the Declaration of Independence (and other national charters) on a hexapod
[04:15:48] <fenn> but you're saying it doesnt matter what the distance is
[04:15:53] <toast> right
[04:16:00] <toast> it doesn't matter what the distance is
[04:16:12] <toast> but the qualification there is "As long as that distance doesn't change"
[04:16:42] <toast> between any two balls, that is
[04:16:51] <toast> they don't decide, hey, 40 mm between balls
[04:16:59] <toast> they put the balls on and wherever they land, they measure
[04:17:01] <toast> and then that's the map
[04:18:03] <fenn> my understanding was that the balls were machined out of the same piece as the rest of the block, which is why they needed super precise 5 axis milling
[04:18:21] <toast> if that is true, that is pretty impressive
[04:18:27] <fenn> of course i have no idea where i read that now
[04:18:31] <SWPadnos> http://www.mel.nist.gov/photos/photo/close.htm
[04:18:36] <toast> but i've never seen even a diamond turning machine that could make those balls
[04:18:41] <SWPadnos> http://www.mel.nist.gov/photos/photo/wide.htm
[04:18:48] <fenn> other ball plates just look like some steel balls set into an aluminum plate
[04:18:57] <toast> yep, but they're lapped balls
[04:19:00] <toast> very very accurate
[04:19:25] <toast> the cmm probes the balls and calculates the geometric center from those points
[04:19:28] <toast> then moves to a new ball, and does the same
[04:19:42] <toast> the distance between those two points is recorded
[04:19:45] <toast> then the whole thing is rotated
[04:19:52] <toast> do it again
[04:19:59] <toast> if it changes, you have an error in your cmm
[04:20:23] <toast> but you know, this could be some new test plate i've never heard of
[04:21:50] <toast> but like, the specs on those balls
[04:21:50] <toast> are ridiculous
[04:22:01] <toast> and if you are right, and that machine could mill those balls in place
[04:22:10] <toast> that would be something I heard about
[04:22:17] <toast> or i'm really living in a cloud somewhere
[04:23:32] <fenn> within 127 nanometers
[04:23:33] <tomp> if i read correctly, each hexapod has a volume in which it is achieves a certain accuracy, and that volume is unique to that machine. I think this is true of many machine where more than 1 joint defines an 'axis' ( worse at furthest reach, better in center of reach )
[04:24:21] <toast> fenn: link?
[04:24:33] <fenn> just some ball plate manufacturer http://www.precisionballs.com/CMM/Ball_Plate.htm
[04:24:51] <toast> oh, i thought you said the MACHINE was doing 127 nm
[04:24:55] <fenn> er, i wasnt talking about the hexapod
[04:24:59] <toast> and i was like, "that's pretty good for a milling machine"
[04:25:26] <fenn> there was some japanese study where they had within 10 um across the work envelope
[04:25:48] <toast> that's what... .0010 mm?
[04:25:51] <fenn> but it went up to 20-30 with lots of rotation
[04:26:08] <fenn> .01
[04:26:11] <toast> oh
[04:26:19] <fenn> not that good is it :\
[04:26:20] <toast> that's like, a standard milling machine
[04:26:42] <toast> i mean it's not bad by any stretch of the imagination
[04:26:45] <toast> not at all
[04:26:54] <fenn> that was after doing their automatic calibration procedure
[04:27:01] <fenn> http://mmc.me.kyoto-u.ac.jp/research/para/ishida/pa_ishida.html
[04:27:30] <toast> well doing it via calibration is still valid, accuracy is accuracy i suppose
[04:27:51] <toast> and they can certainly machine more shapes with greater flexibility than most 5 axis machines can
[04:28:09] <fenn> really? i dont know much about 5 axis mills really
[04:28:29] <toast> yeah, they're usually pretty limited in range in one way or another
[04:28:38] <toast> by how the manufacturer stacks the rotary axes
[04:28:50] <fenn> i figured the hexapod would be much more limited in angular motion (the simulation i've been playing with does about 50 degrees)
[04:29:06] <toast> 50 degrees is pretty good for a lot of moulding applications, like plastic injection
[04:29:18] <fenn> yeah anything more and you're doing candlesticks
[04:29:35] <toast> well any more and you're usually not doing contouring
[04:29:49] <toast> but like, tombstone milling on fixtures and pallet changers
[04:29:53] <fenn> its six axes though so you can mount an angle head on it to get more range in one of the rotational axes
[04:30:31] <fenn> a fixed angle, not an additional axis
[04:30:35] <toast> but they do that on 5 axis machines too
[04:30:50] <fenn> am i making sense?
[04:30:51] <toast> yes, you are
[04:30:59] <toast> very much so
[04:31:41] <toast> i guess there's no ADDITIONAL mechanical noise
[04:31:48] <toast> on a hexapod
[04:31:50] <toast> when you do contouring
[04:31:53] <toast> which is sometimes an issue
[04:32:07] <fenn> noise like in a signal processing sense?
[04:32:07] <toast> on conventional axes, unless you spend a ton of money on brushless drives or linear motors
[04:32:28] <toast> noise like ballscrews, worm gears, brushed servos, ec
[04:32:30] <toast> *etc
[04:32:33] <toast> physical noise
[04:32:35] <fenn> clank clank whirrr
[04:32:37] <toast> inside the machine
[04:32:43] <fenn> that messes things up eh
[04:32:46] <toast> yeah, all that crap causes a poor surface finish
[04:34:09] <toast> and i recall servohydraulics, while being really damn finnickey in their own right, are very quiet mechanically
[04:34:43] <toast> what would be really cool is if you made a hexapod lathe
[04:34:50] <toast> where the chuck was on the hexapod, and the tool was truly fixed.
[04:34:54] <fenn> i've thought about it
[04:35:00] <fenn> makes more sense the other way around
[04:35:11] <fenn> then you can keep the hexapod spindle for live tooling too
[04:35:22] <fenn> the spindle requirements are totally different
[04:35:25] <toast> yeah
[04:35:28] <toast> well, sort of
[04:36:07] <fenn> lathe requires a large through bore, bar changer
[04:36:21] <fenn> high resolution encoder
[04:36:27] <toast> you don't NEED those things
[04:36:47] <toast> there are a couple lathes with no through bore
[04:36:53] <fenn> no, but it makes more sense to put the spindle on the bed so later you can add them
[04:36:59] <toast> perhaps!
[04:37:34] <toast> i think it would be easier to just move the spindle
[04:37:34] <fenn> otherwise you will have to compromise your "milling" spindle to satisfy lathe requirements
[04:37:36] <toast> to a cutoff machine
[04:37:53] <toast> clamp down, and have it cut off
[04:37:56] <toast> then move the spindle back up to the tool.
[04:38:05] <toast> and it could be gang tooled.
[04:38:11] <fenn> hexapod can move very quickly so you can take advantage of that by using high speed tooling
[04:38:30] <fenn> on a lathe there's limits to how fast you can spin things
[04:38:43] <fenn> just because of safety and balancing
[04:38:48] <toast> i'm not really sure i agree with that
[04:39:22] <toast> 10k rpm lathes are just as common now as 10k rpm mills
[04:39:22] <fenn> well i wouldnt want to be near a big flywheel going 50krpm
[04:39:29] <toast> haha.
[04:39:40] <toast> point taken.
[04:40:02] <toast> apparently cincinatti did some research
[04:40:19] <toast> that, in aluminum, tool wear disappears at some ridiculous spindle speed
[04:40:22] <toast> using carbide
[04:40:26] <toast> any carbide
[04:40:29] <fenn> heh
[04:40:37] <toast> do you know what i am talking about
[04:40:44] <toast> like a 1" tool at 40k rpm?
[04:40:47] <fenn> yeah machining force goes negative once you reach the speed of sound in the material you're machining
[04:40:57] <toast> hmm
[04:40:58] <toast> i did not know that
[04:41:00] <fenn> they were shooting bullets across tooling right?
[04:41:04] <toast> no?
[04:41:11] <toast> oh, you are just kidding around
[04:41:13] <fenn> different study i guess
[04:41:21] <toast> yeah
[04:41:24] <toast> it was normal machining
[04:41:34] <toast> they found that at ridiculous speeds the aluminum is fully plastic
[04:43:31] <fenn> well 40k isnt unthinkable
[04:43:41] <toast> indeed, especially with an air spindle
[04:43:50] <fenn> TURBOTOOL
[04:43:54] <toast> BZZZ BZZZ BZZZZZZZZZ
[04:44:01] <fenn> squeeee
[04:44:18] <toast> i keep trying to find videos of freeform machining
[04:44:21] <toast> i can't find any
[04:44:30] <fenn> that's like a shaper sorta right?
[04:44:38] <Jymmm> hey, who has the dune buggy that jumped the flood canal
[04:44:42] <Jymmm> ????
[04:44:43] <toast> it's a lathe
[04:44:44] <fenn> skullworks
[04:44:47] <toast> that can do millwork
[04:44:53] <toast> by moving the tool around in the proper shape
[04:45:19] <toast> both the tool and the work have like 2/3 axes of freedom each
[04:45:21] <fenn> toast: like machining splines on a shaft? or like making a pocket in a plate
[04:45:30] <toast> like making a mould cavity or core
[04:45:54] <toast> but it can do anything a mill can do
[04:46:01] <fenn> does it spin?
[04:46:03] <Jymmm> who has the sandrail???
[04:46:02] <toast> yes
[04:46:23] <toast> fenn: both the tool and the spindle
[04:46:25] <toast> are axes of rotation
[04:46:30] <toast> so you can keep the tool normal to your surface
[04:46:35] <toast> or get it out of the way of a feature
[04:46:54] <toast> usually tool has a X and Z
[04:47:05] <toast> and the spindle has like, a Y axis and an additional rotary
[04:47:17] <toast> lots of motion
[04:48:58] <fenn> weird stuff like that goes by different names depending on who you talk to
[04:49:01] <toast> yeah
[04:49:21] <toast> well freeform machining refers to something very specific
[04:49:31] <toast> there is also slow tool servo, asymmetric turning, etc
[04:49:38] <toast> which all refer to specific capabilities
[04:49:55] <toast> tool normal turning, etc
[04:50:02] <toast> i can't think of any more
[04:50:16] <fenn> orbital boring
[04:50:25] <toast> that's a rad name
[04:50:58] <fenn> i may have made it up in the same delusional fit that generated the cmm plate fantasy
[04:51:18] <fenn> i'm terrible at keeping references
[04:52:04] <toast> hahaha
[04:52:09] <fenn> Similar to circular milling, orbital boring creates the hole incrementally through forward drilling and a circular motion, thus substantially lowering radial cutting forces and minimizing deflection while cutting dissimilar material.
[04:52:11] <toast> same, but it does sound like something that exists
[04:52:25] <fenn> that is totally not what i'm talking about
[04:52:28] <toast> hahahaha
[04:52:33] <toast> you're talking about like an elliptical bore?
[04:52:37] <fenn> yeah
[04:52:47] <toast> yeah orbital boring sounds cooler
[04:52:49] <fenn> or an incolute spline
[04:52:53] <toast> loooool
[04:52:53] <Jymmm> http://www.gibbstech.co.uk/mediacentre/quadski.php
[04:52:59] <toast> there's apparently some specific language
[04:53:03] <toast> to program those machines
[04:53:06] <toast> called c-code
[04:53:23] <toast> it uses very high order polynomials to describe curves
[04:53:35] <toast> and 3d surfaces
[04:53:35] <fenn> jymmm its like a hovercraft except not as cool
[04:54:27] <fenn> toast: high order polynomial == NURB
[04:54:32] <fenn> S
[04:54:38] <toast> are you sure?
[04:54:54] <fenn> no
[04:54:57] <toast> this was like, "this surface is 24x^24 + 2x^17 + etc etc
[04:55:02] <toast> this huge polynomial
[04:55:11] <fenn> why do they do that
[04:55:34] <fenn> nurbs is like x^5 i think
[04:55:36] <toast> because it gives you the optically perfect solution
[04:55:47] <toast> which is usually what those machines do
[04:55:57] <toast> although it can be used for molds and other stuff too
[04:56:01] <fenn> well, that sounds ok to me too
[04:56:17] <toast> nothing wrong for nurbs, that's the easier aerospace solution
[04:56:19] <toast> *with
[04:56:20] <fenn> if the machine can calculate the desired shape in real time you might as well tell it the real shape you want
[04:56:29] <toast> yeah
[04:56:39] <toast> but i think nurbs solves a different, but related problem
[04:56:43] <toast> not sure though.
[04:56:44] <fenn> when you say c-code it sounds like a joke to me :)
[04:56:50] <toast> lol =(
[04:57:19] <toast> i thought the guy was kidding
[04:57:23] <fenn> yeah, we got this cool new machining language, you can program any shape you can think of. it's called ansi-c-code
[04:57:25] <toast> then i heard like eight people say "c-code"
[04:57:40] <toast> and i was like, if this is a joke, there are sure a lot of people in on it
[04:57:52] <toast> then i saw it in a machine's manual
[04:58:03] <toast> or rather the literature
[04:58:27] <fenn> its not like there are 24 other letters to choose from
[04:58:36] <toast> "q-code"
[04:58:42] <toast> that sounds very 007.
[04:58:58] <fenn> n-code and d-code
[04:59:08] <toast> orrr
[04:59:09] <fenn> not commutative
[04:59:13] <toast> g^2-code
[04:59:16] <toast> CUTTING edge.
[04:59:23] <toast> or we can make it web-2.0:
[04:59:25] <fenn> g+code
[04:59:28] <toast> cuttingEDGE
[04:59:32] <toast> (beta)
[04:59:42] <toast> haahaha
[04:59:43] <toast> g+code
[04:59:50] <fenn> just below F-
[05:00:22] <toast> zing
[05:00:34] <toast> my co-worker is bringing up a lathe on.. some sort of ladder logic?
[05:00:41] <toast> he has never machined anything before
[05:00:48] <toast> and what he is talking about is starting to concern me
[05:01:01] <toast> but we shall see
[05:53:53] <toast> http://www.youtube.com/watch?v=-Z7zThqVBhU
[05:53:57] <toast> new radiohead
[05:56:36] <skinnypuppy34> I was just listening to some frontline assembly with an interesting backvid...
[05:56:49] <skinnypuppy34> http://youtube.com/watch?v=yQMwimh8QdI
[05:59:46] <toast> what album is this from
[06:01:15] <skinnypuppy34> tactical neural implant
[06:02:10] <toast> ty
[06:02:22] <toast> i recall not liking frontline assembly in the past
[06:02:27] <toast> because they were very synthpoppy
[06:03:53] <skinnypuppy34> A lot of the stuff after millenium sucked, rhys fulber quit playin with fla and focused on producing fear factory and the expansion of delerium
[06:04:31] <ds2> hey, is it an abuse of gage pins to clamp one in a collet to check for collet runout?
[06:06:38] <toast> not really, no
[06:06:47] <toast> i would not do that with a class X pin
[06:06:53] <toast> but i'd do it in a heartbeat with a zz
[06:07:02] <ds2> is ZZ 'shop grade'?
[06:07:07] <toast> yar
[06:07:28] <toast> X is "dead nuts"
[06:07:34] <toast> zz is "darn close"
[06:07:35] <ds2> that means my assortment of second hand surplused shop grade pins are fine for this use then ;)
[06:07:50] <toast> absolutely
[06:07:55] <toast> but you must be careful
[06:08:01] <toast> in that the pins
[06:08:02] <toast> are lobed
[06:08:04] <ds2> I am suspecting the dowel pin I am using for checking is bad in some way
[06:08:05] <toast> and not perfectly concentric
[06:08:10] <ds2> lobed? ack
[06:08:11] <toast> er, circular
[06:08:14] <toast> but it is very small
[06:08:18] <toast> and you should not notice it
[06:08:28] <toast> compared to the 3-4 tenths i assume your collets are going to be out
[06:08:28] <ds2> thought gage pins are ground with centers so they aren't lobed?
[06:08:35] <toast> gage pins are centerless
[06:08:50] <toast> that's how you get zz class so cheap
[06:08:52] <ds2> hmmm
[06:09:03] <toast> it's not a bad thing
[06:09:06] <toast> i just wanted you to be aware
[06:09:09] <ds2> know off hand, what's the accuracy of a dowel pin?
[06:09:13] <toast> you should still be able to get dead close
[06:09:17] <toast> lol, like a wooden dowel pin?
[06:09:25] <ds2> from what you said, the pin lobes must be sub 0.0001?
[06:09:34] <toast> should be, yes
[06:09:39] <ds2> no, the hardened steel pins used for locating stuff
[06:09:49] <toast> those should also be very accurate
[06:09:54] <toast> what are you reading on your collet
[06:09:56] <toast> that makes you think it is off
[06:10:33] <ds2> cuz I am measuring 0.003-0.004 TIR on a collet chuck mounted on a spindle that didn't move my 0.0005 indicator
[06:10:43] <toast> hmm
[06:10:53] <skinnypuppy34> Rollers for needle bearings aren't too shabby either, trans shops will have them by themselves
[06:11:26] <toast> those should be top notch
[06:11:26] <ds2> skinnypupp34: new ones or you refering to their discards piles?
[06:11:35] <skinnypuppy34> new yes
[06:12:04] <skinnypuppy34> I get some discard gears and shaft to anneal sometimes.
[06:12:28] <ds2> do trans shops understand if I ask for a 0.7500 pin? I know automotive places insist on a catalog number instead of specs
[06:13:17] <toast> probably not
[06:13:23] <skinnypuppy34> I don't know about that large, take a mic and ask for the needle type rollers.
[06:13:37] <ds2> hmmm
[06:13:42] <toast> ds2: i think you are going to be just fine
[06:13:44] <toast> with a pin
[06:13:46] <toast> from your box
[06:14:04] <ds2> toast: some of my pins are dirty/rusted as they are from industrial surplus
[06:14:09] <toast> o
[06:14:29] <toast> dunno man
[06:14:34] <toast> are your collets like
[06:14:41] <toast> Hardinge Super Precision collets
[06:14:53] <ds2> no, import ER32's and ER40's
[06:15:05] <skinnypuppy34> 5clathe type?
[06:15:08] <toast> are you locking the ER collets
[06:15:11] <toast> into the nut
[06:15:22] <toast> BEFORE you put them into the collet chuck
[06:15:25] <ds2> btw, another reason to doubt the pin is I get similar numbers on 2 different chucks
[06:15:43] <ds2> toast: yep, seat them into the ejector nut first, start them on a bit then insert pin
[06:15:49] <toast> whack man.
[06:21:52] <fenn> how bout you put whatever you plan on machining into the collet and see if it's concentric
[06:22:50] <fenn> or turn a .7500 rod
[06:22:59] <fenn> i mean you've got a lathe, whats the problem
[06:23:34] <fenn> ok so tenths is pushing it maybe
[06:23:39] <toast> yar
[06:23:49] <toast> most like, really expensive 5c collets i see only hold .0002ish
[06:23:52] <toast> "only"
[06:23:53] <fenn> it doesnt have to be the right diameter though, just straight and good surface finish
[06:24:25] <fenn> and if the lathe wont do it, why bother
[06:25:48] <toast> tru
[06:27:46] <toast> what is the lathe doing, by the way
[06:28:16] <toast> while we're on that subject
[07:07:44] <ds2> the lathe itself is being checked out mechanically
[07:08:38] <ds2> anyways, right now all I have time for is tought experiments and data waiting to be gathered when I get a free minute in teh shop
[07:09:30] <toast> lol
[07:09:35] <toast> SUCH IS LIFE, EH
[08:46:45] <alex_joni> * alex_joni will be back tomorrow evening
[12:36:39] <The_B> What free programs does people here use to generate g-code for engraving simple text?
[12:42:55] <jepler> there's this program written by our own cradek: http://www.timeguy.com/cradek/truetype
[12:42:59] <jepler> I don't know offhand of any others
[12:46:37] <The_B> cheers, I will try it. I also found a very nice post from Paul Fox on the mailing list, he points out http://autotrace.sourceforge.net/ to convert bitmaps to vectors, very handy
[13:45:08] <cradek> I've used autotrace for other stuff, but I think going through a bitmap representation is the wrong approach for fonts
[13:45:22] <cradek> (truetype and postscript fonts are vectors so they can scale nicely)
[13:56:23] <The_B> cradek, making some dust now, works like a charm, cheers
[13:58:20] <cradek> The_B: you mean truetype-tracer?
[14:07:54] <The_B> cradek, yes, one thing though, in your screen shot you show the output on two lines, i tried \n in the string and also a normal new line in the string, both produced a error, is it possible to start a new line?
[14:10:06] <cradek> hmm, I may have done that in autocad, I don't remember. Surprisingly I don't think starting a new line is supported right now.
[14:10:28] <cradek> but, you can use an offset in gcode to place a second line where you want it
[14:11:21] <cradek> support for multiple lines, different justifications, etc., would be nice
[14:12:08] <The_B> yeah, maybe i can throw together a gui for it, i've started learning qt
[14:13:55] <cradek> I'm not sure it needs/wants a gui - support for \n would be a nice incremental improvement though
[14:18:19] <The_B> maybe not
[14:19:33] <cradek> in the loop that renders the characters, you'd detect \n and move down by face->glyph->advance.y, and set offset=0 to move back to the left
[14:19:53] <cradek> I think it's just a few line change
[14:21:21] <cradek> right/center justification is a little harder, you'd have to measure the lines first (sum up the advances)
[14:26:53] <The_B> maybe a command line option for line spacing, the normal is 1.5 i think, but you would just multiply face->glyph->advance.y with whatever the user specifies
[16:34:22] <[1]a-l-p-h-a2> hey gang
[16:34:31] <[1]a-l-p-h-a2> I've got a giant sql file... I have a specific line number I want to start displaying at... what command would allow me to say start display at a given line number, but ignore anything above it?
[16:40:27] <jepler> tail +<num>
[16:40:41] <jepler> e.g., "seq 1 10 | tail +8" will show 8, 9, 10
[16:59:40] <[1]a-l-p-h-a2> [1]a-l-p-h-a2 is now known as a-l-p-h-a2
[17:02:35] <a-l-p-h-a2> jepler, awk worked better
[17:02:48] <a-l-p-h-a2> awk 'NR>linenumber{print $0}' sourcefile
[21:05:50] <CIA-22> 03jepler 07TRUNK * 10emc2/lib/python/rs274/interpret.py: fix for certain 360 degree arcs in preview plot