#emc | Logs for 2005-10-26

[00:00:14] <les_w> accelera not optima
[00:00:29] <les_w> http://www.galilmc.com/products/accelera/dmc18x6.html
[00:00:31] <icee> kz: 20 revs per UNIT (inch) * 400 half steps/rev = 8000
[00:00:41] <fenn> les_w: what are the symptoms of your problem?
[00:01:39] <icee> vw: is this your linear actuator? http://wrlabs.shacknet.nu/~vw/CNC/MechAxis1-002.jpg
[00:01:42] <anonimasu> les_w: but that's just the boards right?
[00:02:18] <anonimasu> $1895 for just the 3 axis cards..
[00:02:57] <les_w> fenn, we fixed the spurious acel spikes a while ago, We did not fix the breakdown of the blending algo at the triangular velocity profile threshold
[00:03:03] <anonimasu> hm, I guess you could throw g code with it from emc..
[00:03:19] <les_w> note the servo rates...
[00:03:27] <icee> spurious acceleration spikes?
[00:03:33] <anonimasu> nice
[00:03:47] <icee> * icee has had interesting problems trying to set accelerations low in emc2
[00:03:57] <les_w> yeah icee, we had a serious mathe error. We fixed in this spring.
[00:03:57] <Jymmm> les_w You KNOW you want to use steppers with that controller card
[00:04:11] <les_w> jymmm: haw
[00:04:13] <icee> hm. my checkout is much more recent than that
[00:04:44] <anonimasu> les_w: if you used emc to throw gcode to it it'd do any speed you would like easily..
[00:04:57] <les_w> icee, there are other serious problems with the trajectory planner that we did not fix.
[00:05:12] <anonimasu> but how does it solve the trajectory planning issues?
[00:05:13] <les_w> It is only an issue with high speed machining
[00:05:26] <icee> well, for very low speeds it was interesting when i was experimenting
[00:05:55] <les_w> I think emc is ok for anything up to a bridgeport or something
[00:05:57] <icee> if the axis accelerations are low in comparison to the feed rate, the default acceleration in the trajectory planner needs to be much slower
[00:06:04] <icee> or you get following errors
[00:06:25] <les_w> but for high speed modern machining...forget it.
[00:06:47] <icee> well, hopefully it'll continue to improve.
[00:07:02] <les_w> it would be nice
[00:07:10] <icee> Because I'm planning to get 5-10kHz updates with these big motors
[00:07:13] <les_w> but this is a daily bread thing for me
[00:07:17] <anonimasu> I wish somone understood the planning stuff..
[00:07:30] <les_w> you understand
[00:07:32] <icee> if no one understands it, maybe it needs to get replaced.
[00:07:49] <anonimasu> les_w: I thought it was more of a sparetime issue ;)
[00:07:50] <les_w> I do understand the planning maths...but cannot follow the code.
[00:08:23] <les_w> ha...spare time...what's that?
[00:08:35] <icee> I don't think the error handling is ideal either
[00:08:45] <icee> falling a little behind on a rapid traverse shouldn't necessarily cause things to blow up
[00:08:55] <les_w> right
[00:09:10] <icee> as long as the two axes are moving at the same rate with respect to each other, and you're reasonably close to the line you're supposed to be on..
[00:10:14] <icee> all should be OK.
[00:10:22] <les_w> well you should be familiar with my situation...exponential business growth...turning down orders (bad) need major productivity improvements like right now
[00:10:40] <icee> les: as much as that sucks, there are worse problems to have ;)
[00:10:45] <icee> too much business > too little business
[00:10:50] <les_w> haha yeah
[00:14:02] <les_w> knowing the math and not being able to follow the damn code sure is exasperating
[00:14:16] <icee> well, if/when i get a big mill up
[00:14:21] <icee> i may look at things some.
[00:14:33] <icee> depending on how much it sucks for my apps.
[00:14:38] <les_w> heh
[00:15:16] <icee> i've got motors in hand, so it's about time for me to work on design entry for the controller
[00:16:17] <les_w> if it's a bridgeport or something , you will be fine. I would not be fine for me, because this is not 1980.
[00:16:24] <les_w> it
[00:17:52] <CIA-6> 03rayhenry * 10emc2/configs/mazak_rf.hal: added spindle from emc
[00:18:09] <k4ts> ta tranquillo che mo il viaggetto me lo faccio
[00:18:11] <k4ts> davvero
[00:18:22] <les_w> hi anna.
[00:18:29] <k4ts> les
[00:18:31] <dmess> yeeee haaa... dads on his way home....
[00:18:32] <k4ts> sorry
[00:18:40] <CIA-6> 03rayhenry * 10emc2/configs/mazak_rf.ini: twiddled orient parameter
[00:18:41] <anonimasu> dmess: nice
[00:18:53] <dmess> VERY...
[00:19:40] <Jacky^> umpf ..
[00:19:51] <les_w> jacky: haha
[00:20:04] <Jacky^> * Jacky^ sigh !
[00:20:11] <Jacky^> :/
[00:20:23] <k4ts> J
[00:20:36] <les_w> what one eye?
[00:20:41] <Jacky^> shes hard to understand
[00:21:25] <les_w> ah...I learned italian from the colombo spindle manual a little
[00:21:45] <Jacky^> les_w: lol
[00:21:47] <Jacky^> :)
[00:22:41] <Jacky^> k4ts: what you want ?
[00:22:48] <Jacky^> talk, speak, say
[00:23:25] <Jacky^> uhmm
[00:23:26] <dmess> im tickled.... ; (
[00:23:49] <Jacky^> 00:35 -!- k4ts was kicked from #emc.it by emclogg3r [emclogg3r
[00:23:58] <Jacky^> again ?
[00:24:25] <Jacky^> dman
[00:24:32] <Jacky^> damn*
[00:24:51] <k4ts> cosa dici?
[00:25:00] <Jacky^> k4ts: youre too much nervous this evening ..
[00:25:12] <Jacky^> please..
[00:28:36] <vw> Just putzing around here, but given that distance traveled seems to vary with F value, I'd say it was likely that I am loosing steps in there...
[00:29:02] <vw> As in with F20 I travel 2" but with F200 I travel 1.
[00:29:27] <les_w> sounds like it
[00:29:32] <anonimasu> limit your accel :9
[00:29:59] <vw> Wilco, was just putzing :-)
[00:30:10] <anonimasu> hehe
[00:30:21] <anonimasu> I should go to bed and ponder about this stuff tomorrow
[00:30:24] <vw> Where was the 8000 scale factor supposed to go ?
[00:30:34] <vw> Not clear under calibration tab.
[00:30:35] <anonimasu> input_scale
[00:30:37] <anonimasu> ;)
[00:30:41] <anonimasu> in the ini
[00:30:45] <vw> I Gain?
[00:30:57] <anonimasu> input_scale
[00:30:59] <anonimasu> :)
[00:31:11] <anonimasu> it's there ;)
[00:31:17] <vw> ah ok lemme find that :), I'll be a do some work on that before I am done :)
[00:31:18] <Jacky^> ahh
[00:31:21] <Jacky^> good
[00:31:25] <vw> locate ini | grep emc
[00:31:45] <anonimasu> ah
[00:32:00] <anonimasu> heh :)
[00:32:08] <anonimasu> find ./* |grep emc.ini
[00:32:29] <vw> /usr/local/emc/emc.ini
[00:32:30] <vw> /usr/local/emc/generic.ini
[00:32:30] <vw> /usr/local/emc/mill_inch_freq.ini
[00:32:30] <vw> /usr/local/emc/mill_mm_freq.ini
[00:32:31] <anonimasu> or find ./* -name emc.ini
[00:32:32] <vw> a few
[00:32:43] <anonimasu> yep
[00:32:46] <anonimasu> what do you start emc with?
[00:33:08] <fenn> depends on the distro i think, it's in emc.run
[00:33:13] <anonimasu> yep
[00:33:20] <vw> K-Menu -> emc ->emc.generic
[00:33:38] <anonimasu> edit generic.ini :9
[00:33:39] <anonimasu> :)
[00:33:53] <anonimasu> night all
[00:34:15] <les_w> night anders
[00:35:31] <vw> Night and thanks!
[00:42:13] <dmess> i hatched a dream of using emc as a paragliding simulator platform...any thoughts??
[00:42:57] <vw> Hummm, the X display value is off by +.0005 when it should be 0.0000 and minus that much when it should be 1.0000, Strange...
[00:43:16] <vw> BUT if it is really that close I'll be happy :).
[00:44:33] <dmess> my issue is jog AND programmed commands need to apply ALL the time... a case 4 alex. real time interactive programming
[00:47:20] <les_w> also, you need a big electrical shock in the seat if you crash or something
[00:47:52] <les_w> but geezz, can't crash a paraglider?
[00:48:26] <dmess> wanna bet...
[00:48:38] <dmess> if ya bounce ya die
[00:49:05] <les_w> I was looking at galil accelera. Seems there is no commercial cnc control software to go with it...hmm
[00:49:11] <les_w> bounce?
[00:49:24] <les_w> as in not flaring?
[00:50:05] <dmess> i have 2 broken wrists, a busted up vert and a scare on my foot
[00:50:19] <les_w> I remember
[00:50:25] <dmess> no.. thermal catassstrophy
[00:50:50] <dmess> i had no wing to flare... i was skiddin'
[00:51:15] <les_w> I have bounced once or twice. <ahem> full power. yoke forward.
[00:51:40] <dmess> ouch... HARD on the lg
[00:51:50] <les_w> yeah
[00:52:03] <les_w> I have to remember...your lg has shoes
[00:52:15] <dmess> les could your emc machine cut AERMET 100
[00:52:38] <dmess> yup... MESSIER landing gear
[00:52:42] <les_w> what is it (I should know)
[00:53:04] <dmess> steel... THE RIGHT STUFF
[00:53:19] <les_w> let me google it
[00:53:26] <dmess> from Carpenter i believe
[00:53:59] <Jacky^> carpenter ? :P
[00:54:08] <dmess> it needs to pass thru NDT after post h/t machining
[00:55:02] <les_w> ok...a hypoeutectoid crome/moly/cobalt alloy
[00:55:07] <les_w> looks tough!
[00:55:16] <Jacky^> !google carpenter
[00:55:22] <les_w> k
[00:55:25] <Jacky^> mmm
[00:55:32] <Jacky^> :)
[00:55:41] <dmess> it blows up rehardening and retempered areas..its tempering temp is 900 degrees...
[00:56:07] <Jacky^> 01:13 < emc_finder> Jacky^: http://www.cartech.com/ (Carpenter Technology | Manufacturer
[00:56:10] <Jacky^> of Stainless Steel, Industrial ...) - 13k
[00:56:11] <Jacky^> :)
[00:56:16] <dmess> its GOTTA be tool geometry
[00:56:50] <les_w> I'll bet that stuff work hardens a lot
[00:56:58] <les_w> looks nasty to machine
[00:57:37] <dmess> why do you say that...
[00:57:54] <les_w> the chromium content
[00:57:58] <dmess> it IS GROSS AND DISGUSTING
[00:58:09] <Jacky^> O_O
[00:58:10] <les_w> yeah?
[00:58:26] <dmess> oh yeah
[00:58:39] <dmess> and alot of material abuse
[00:58:43] <Jacky^> dmess: sure ?
[00:58:57] <dmess> of what??
[00:59:02] <Jacky^> tried ?
[00:59:22] <dmess> need to know basis...
[00:59:26] <Jacky^> oh.. ok
[00:59:32] <dmess> but yes
[01:00:17] <dmess> lg for many of today and tomorrow's fighters
[01:01:04] <dmess> i found a mountain bike frame out of this stuff and its 3 lbs.... 1.5 kg's
[01:01:06] <Jacky^> les_w: how are you ?
[01:01:11] <Jacky^> :)
[01:01:42] <dmess> .020" wall thickness for the tubing
[01:02:05] <les_w> amazing how those fighters get whacked doen at landing...and I have to eaaassse it in due to the nose wheel being just mounted to the sheet metal firewall
[01:02:12] <dmess> RETARDED... to try and ride it without breaking it
[01:03:01] <Jacky^> :P
[01:03:08] <dmess> our systems are hydaulic lg's
[01:03:44] <Jacky^> dmess: you was vw ?
[01:03:56] <Jacky^> ho no ..
[01:04:03] <Jacky^> k'zan
[01:04:05] <Jacky^> maybe
[01:04:08] <Jacky^> :)
[01:04:11] <dmess> they make nice Dune buggies
[01:04:50] <Jacky^> les_w: let me know when we start to search the motors I need :P
[01:04:53] <dmess> Messier-Dowty
[01:04:57] <Jacky^> can*
[01:06:26] <Jacky^> looking at the taxes i Payed for geckodrivers, id prefer to get the motors in my cuntry
[01:06:37] <Jacky^> at least, in EU
[01:06:59] <Jacky^> what you think about ?
[01:07:55] <Jacky^> escluding ebay ..
[01:15:40] <Jacky^> mmh
[01:15:49] <Jacky^> cell
[01:16:14] <Jacky^> driiinnn
[01:16:53] <Jacky^> oh oh
[01:19:13] <les_w> was on skype
[01:19:33] <les_w> jacky, let's review your router
[01:20:11] <les_w> 1x1 meter xy
[01:20:13] <les_w> z?
[01:20:21] <Jacky^> les_w: yeah
[01:20:44] <Jacky^> z ? i think 10 cm its enough
[01:20:54] <les_w> ok.
[01:21:06] <Jacky^> maybe less
[01:21:07] <les_w> twin ballscrewx....timing belt?
[01:21:19] <Jacky^> oh ..
[01:21:35] <Jacky^> ballscrew are 20 mm 5 mm lead
[01:21:42] <Jacky^> 20mm diameter
[01:21:54] <les_w> bought them?
[01:22:05] <Jacky^> not yet
[01:22:18] <Jacky^> we jaust plain them
[01:22:22] <Jacky^> just*
[01:22:43] <les_w> ok. 5mm is ok...for wood only 10 mm pitch would be a little better
[01:22:46] <les_w> faster
[01:23:01] <Jacky^> ok
[01:23:24] <les_w> want to analyze with 5 or 10?
[01:23:46] <Jacky^> mmhh
[01:24:07] <Jacky^> what the advantage of 5 respect to 10 ?
[01:24:14] <Jacky^> better precision ?
[01:24:42] <Jacky^> and lowest speed, i suppose..
[01:24:44] <les_w> with 10, ballscrew will be half speed. easier on the screw...but needs more motor torque
[01:25:09] <les_w> max speed with 10...about
[01:25:11] <Jacky^> oh, understood
[01:25:28] <les_w> 2 meters/min
[01:25:31] <Jacky^> so, the cohice depend on the motor type also ..
[01:25:36] <les_w> right
[01:26:05] <les_w> if cutting wood and metal, 5 mm is best
[01:26:15] <Jacky^> what you suggest ?
[01:26:17] <les_w> wood or foam only, 10 mm is best
[01:26:22] <les_w> hmmm
[01:26:28] <les_w> 10
[01:26:39] <Jacky^> ok
[01:26:49] <les_w> ok.
[01:26:55] <les_w> let's see.
[01:27:18] <les_w> moving gantry...I will say 50 kG with the spindle
[01:27:30] <Jacky^> yeah, right
[01:27:51] <les_w> max cutting force: hmmm 150n or so
[01:28:12] <Jacky^> yeah, i think so
[01:28:24] <les_w> 15 or 20 mm rails, right?
[01:28:33] <Jacky^> 20 mm
[01:28:37] <les_w> k
[01:28:49] <Jacky^> large flage types
[01:29:03] <Jacky^> flange*
[01:29:16] <Jacky^> bearings
[01:29:22] <les_w> all right I have enough information
[01:29:39] <les_w> I will build up a motioneering model
[01:29:49] <Jacky^> :)
[01:29:52] <Jacky^> okay
[01:30:01] <les_w> might need until tommorrow
[01:30:08] <Jacky^> sure
[01:30:14] <Jacky^> ta
[01:30:31] <les_w> ok I will vanish and get started
[01:30:33] <les_w> bbiaw
[01:30:39] <Jacky^> ;)
[01:31:32] <les_w> les_w is now known as les_w_away
[01:31:37] <fenn> nice, free engineering consulting
[01:31:47] <Jacky^> :-)
[01:31:52] <les_w_away> only after hours...
[01:31:55] <Jacky^> open source
[01:32:00] <les_w_away> yeah.
[01:32:00] <Jacky^> heheh
[01:32:50] <Jacky^> that the power of emc
[01:35:11] <Jacky^> not for all
[01:35:20] <Jacky^> :/
[01:40:36] <Jacky^> hi ccjoe
[01:48:45] <les_w_away> les_w_away is now known as les_w
[01:49:02] <les_w> mostly done jacky
[01:49:19] <les_w> you need
[01:50:08] <les_w> i nm cont torque
[01:50:14] <les_w> .231 kW
[01:50:57] <les_w> 3.2 nm peak torque
[01:52:07] <les_w> .000134 kg-m^2 armature polar moment of inertia
[01:52:10] <les_w> for x
[01:52:37] <fenn> how many nm cont torque?
[01:52:53] <fenn> sqrt(-1)? :)
[01:53:14] <les_w> 1 nm continuous
[01:53:20] <les_w> that is a good bit
[01:53:32] <les_w> yeah...i
[01:54:09] <fenn> what's the relevance of peak vs continuous torque in terms of acceleration?
[01:54:21] <fenn> amount of part detail?
[01:55:11] <les_w> well, peak is to do sharp corners and stuff. Continuous is to over come friction and cutting forces
[01:55:43] <fenn> ah, yes, cutting forces
[01:56:07] <les_w> I plugged in 150n...it's a wood router...very light cutting forces
[01:56:39] <les_w> that's 30 pounds...plenty.
[01:57:59] <fenn> how smart is motioneering? can you do funky kinematics like a hexapod?
[01:58:35] <les_w> naw just cartesian...but for that...it's awfully handy
[01:59:24] <les_w> takes care of inertia matching and all that
[01:59:58] <les_w> for a hexapod you need to do a few jacobean transforms...
[02:00:55] <fenn> so it doesn't do robot arms or anything either huh
[02:01:53] <les_w> no, those are in general systems with poles and zeros
[02:02:07] <les_w> might be some software that can handle it though
[02:05:39] <fenn> so... what is the parameter you're optimizing for?
[02:06:04] <les_w> for jacky?
[02:06:06] <fenn> yeah
[02:06:13] <fenn> accel?
[02:06:18] <les_w> 10 m/min speed...
[02:08:02] <les_w> min corner radius at 10 meters/ min= (10/60)^2/5 m/s
[02:08:13] <les_w> given
[02:08:24] <les_w> 50 kg gantry mass
[02:08:53] <les_w> and a total of 2.6 meters of 20 mm ballscrew and it's contribution to inertia
[02:09:11] <les_w> .9 screw/slide efficiency
[02:09:19] <les_w> a few other things
[02:09:40] <fenn> is there like, a talking paperclip or anything, or you just have to know this stuff ahead of time?
[02:10:18] <les_w> well, engineering judgement I guess.
[02:10:28] <les_w> Like the 50 kg gantry
[02:10:37] <LawrenceG> hey les.... just installing bdi4.30 iso on a laptop and see you made it on the splash screens.. good work
[02:10:49] <les_w> that is 10 tomes lighter than my machine
[02:11:02] <les_w> oh heh thanks
[02:12:25] <les_w> but a 100 lb gantry weight would be very dooable for a light woodworking only machine
[02:12:33] <les_w> doable
[02:13:02] <les_w> fairly conservative
[02:13:09] <LawrenceG> I am working on a drive for those surplus ac servos... having problems getting a programmer functioning for burning the dspic chips
[02:13:42] <les_w> many are interested in your work!
[02:14:03] <LawrenceG> yea... a lot of motors got sold very quickly
[02:14:10] <les_w> I saw
[02:14:13] <les_w> what a run
[02:14:53] <LawrenceG> I'm just a little worried about all the hobby guys building ac drives
[02:15:49] <LawrenceG> icee, got his motors and has been confirming the wiring... looks very good so far
[02:16:27] <LawrenceG> with a single phase driven at 1 amp, he says he cant turn the motor shaft
[02:16:31] <les_w> well open source the plans...I don't think there is any liability
[02:16:57] <les_w> perhaps a short danger will robinson statement
[02:18:02] <LawrenceG> he he... a few blackened hulk pictures of the devel board maybe
[02:18:11] <les_w> ha yeah
[02:19:04] <vw> (My First CNC program :-) - Just a comment right?
[02:19:52] <vw> I ask because, I added that and now the G01 X0 F6 (the F6) seems to be ignored or set to something really slow.
[02:20:37] <les_w> ??
[02:20:40] <les_w> weird
[02:21:35] <vw> Thought M0 was program end, but apparently either % or M2 works.
[02:22:22] <les_w> I generally use %...saves g92 offsets etc
[02:23:38] <vw> Hummm, with % at the end I get a "Bad character used" error on it.
[02:23:57] <vw> Also now seems that feed rate is stuck at slow...
[02:24:07] <les_w> emc2?
[02:24:38] <vw> Whatever comes with bdi 4.30, the help->about (or anything else there) gives no version.
[02:25:19] <les_w> I think that is an emc1 ish branch....not sure
[02:25:41] <les_w> I ought to know if I am on the splash screen huh?
[02:26:19] <vw> Maybe it's brains are scrambled, I added G04 P1 codes after the moves (G01), dunno if that affected it, lemme remove a few.
[02:26:33] <les_w> k
[02:26:36] <vw> Is there a comment character?
[02:26:51] <les_w> just ( )
[02:27:44] <vw> Nope makes no diff, the F6 still ignored
[02:27:53] <vw> restarting it
[02:27:58] <les_w> k
[02:28:14] <vw> Maybe stepper is too hot, dunno, leaving it in a hold state, the thing gets toasty.
[02:29:27] <les_w> I am usung servo. Things ae pretty solid with it...some funky delayed massages and a screwy trajectory planner come to mind
[02:29:28] <vw> Ok, restarting it took care of the problem!
[02:29:52] <les_w> good...but bad...you know?
[02:31:25] <les_w> what did I say? delayed massages? heh. really. delayed MESSAGES
[02:35:12] <vw> Well, this is just too damn cool !!!!!!!!!!!!
[02:37:04] <les_w> ?
[02:43:56] <vw> Works :-)
[02:44:34] <vw> Wrote a quicky little program and the one axis just does what it was told. Now to build more hardware !!!!!!!!!! :-)
[02:44:34] <les_w> uh...working is cool.
[02:44:40] <vw> Yep, sure is.
[02:45:00] <vw> * vw does Happy Dance(TM)
[02:46:10] <vw> undoubtably have some configuration issues to deal with, the driver is not real fast and actual step rate changes with feed rate setting. Going to think on that for a bit.
[02:46:20] <les_w> I know the feeling. I loaded emc, phased the servos and switches,started it up, and carved a nice plaque. First try. Felt good.
[02:47:10] <vw> I can imagine. I need to kluge together 2 more axis :-) then I can do something other than watch the paperclip move back and forth :).
[02:47:25] <les_w> haha
[02:47:54] <vw> I am sure everyone will get a good laugh out of it when I get it done :-).
[02:48:23] <les_w> pictures. We need pictures.
[02:48:27] <vw> Home brewed sin :-). Deciding factor - what I can afford :-).
[02:48:46] <vw> Lemme see, got some pix of the axis I built, check out pix in:
[02:49:04] <vw> http://wrlabs.shacknet.nu/~vw/CNC/
[02:49:11] <les_w> looking
[02:49:25] <vw> :-)
[02:49:32] <vw> * vw goes and hides
[02:51:23] <les_w> ah, duct tape and hose clamps! very good! :)
[02:51:58] <vw> WOrks, not sure it will be accurate to 0.0001 but hey :-), gotta start someplace :-).
[02:52:24] <vw> Duct tape and hot glue make the world go around :)
[02:52:49] <les_w> well, I built a hang glider once where the wing was held in place by duct tape. I'm not dead.
[02:53:24] <vw> The Air Force refers to duct tape as "500 MPH tape", so I guess that is good :).
[02:54:38] <les_w> yeah, well, i'm a pilot, and more small planes I have flown have had "structural" duct tape on them than I care to remember
[02:55:06] <vw> :-), I'm just a sim pilot with a preference for helos :).
[02:55:16] <vw> * vw walks away from 'em all :)
[02:55:24] <les_w> I dosima lot...cannot fly helos
[02:55:53] <les_w> actually fs 9 is petty accurate
[02:56:00] <les_w> pretty
[02:56:00] <vw> Takes a LOT of practice and some good instruction (and controls) help. Check out hovercontrol.com
[02:56:21] <les_w> k
[02:56:40] <vw> fs9 sucks, but it serves, X-Plane is considered to be better, but there are SO MANY conflicting opinions.
[02:57:12] <vw> * vw doesn't remember the last time she flew one of those stuck winged things :-).
[02:59:00] <les_w> oh, with the real planes I fly fs9 has fair flight models. Spin is not correct, and forward slips don't work right though
[02:59:26] <vw> fs9 is OK really, best on the eye candy except where it counts :).
[02:59:58] <vw> F/W is pretty good, but you just don't get that fluid feeling you get with X-Plane.
[03:00:06] <les_w> I use it to practice navigation mostly...in case the gps goes out in realairplanes
[03:00:17] <vw> LOL, excellent for that.
[03:00:18] <les_w> I have xplane too
[03:01:07] <vw> Had XP running under linux and was real happy with that, then it started segfaulting on me and I haven't gotten back to see what that was all about.
[03:01:30] <les_w> as a student, the first time I left the airport pattern solo OI got lost. It was not a good feeling.
[03:01:36] <vw> I really need to go get some stick time :)
[03:01:55] <vw> Yeah, I can well imagine <shudder>
[03:02:22] <les_w> fun...expensive...basically no practical use...but what the heck?
[03:02:34] <vw> When I was flying for real I took IFR seriously...
[03:02:57] <vw> :-), well it does have it's point, but true for the most part.
[03:03:21] <vw> Them roads were damn important to know where I was :)
[03:03:25] <les_w> I only have vfr, complex , and high performance ratings.
[03:03:42] <les_w> Have I flown into ifr?
[03:03:44] <les_w> um
[03:03:46] <les_w> yes.
[03:04:00] <vw> Completely ahead of me, I quit doing that just after I solo'd the C150. Never got back to it :-(.
[03:04:23] <vw> IMC conditions suck, you were lucky.
[03:04:29] <les_w> I learned in a piper warrior.
[03:04:37] <vw> Nice wings !
[03:04:54] <les_w> c150 spins better though
[03:05:12] <vw> And you can see where with the low wings, you can't really.
[03:05:16] <vw> Ugh
[03:05:35] <les_w> do't like spins?
[03:05:39] <les_w> don't
[03:05:51] <vw> Not particularly...
[03:06:06] <vw> But good to know how to handle them.
[03:06:40] <les_w> I had first spin training in a c172. It just about will not spin in utulity cat
[03:07:00] <vw> So I have heard, but that is a good thing :)
[03:07:07] <les_w> yeah.
[03:07:59] <les_w> needs 40 degrees pitch up...and a blast of power at the break. Still, turns into spiral dive after a couple turns
[03:08:14] <les_w> but
[03:08:37] <K`zan> As long as it doesn't go flat on ya, life is good :)
[03:08:43] <les_w> put a couple lardasses in the back....spins just FINE
[03:08:52] <les_w> real flat!\
[03:08:53] <K`zan> LOL, I'll remember that!
[03:09:02] <les_w> heh
[03:09:55] <K`zan> Switched back to my main machine. That 1100 is a bit on the slow side...
[03:10:04] <les_w> but, as we know, spins on c172 are only aproved for utility...2 up front and half tanks.
[03:10:20] <K`zan> Heh, bureaucraps...
[03:10:28] <les_w> haha
[03:13:14] <les_w> I was thinking about xplane...seems too smooth-too easy. But then I always though fs9 was harder than the real thing
[03:15:04] <K`zan> Chow call, bbiab
[03:15:24] <les_w> k bye
[03:18:58] <les_w> hi john
[03:19:41] <jmkasunich> hi
[03:20:04] <les_w> did a complex contoured sign today
[03:20:16] <les_w> still have TP issues
[03:20:50] <les_w> it's better than it was though
[03:26:13] <Jacky^> ouch
[03:26:20] <Jacky^> 03:44 < Jacky^> ouch
[03:26:25] <Jacky^> G night !
[03:26:38] <Jacky^> Jacky^ is now known as Jacky^afk
[03:29:00] <les_w> I'll talk about your motioneering run tommorow.
[03:29:56] <les_w> les_w is now known as les_w_away
[03:41:30] <CIA-6> 03rayhenry * 10emc2/configs/mazak_rf.hal: added a few sigs
[04:04:58] <Jymmm> * Jymmm notes... NEVER EVER buy/use water (lytex) based spray paint - really bad stuff and will ruin whatever you use it on.
[04:58:06] <LawrenceG> any bdi-4.30 experts here? the install missed my network card... what is the network config utility called?
[05:08:04] <fenn> sysconf-network?
[05:10:25] <fenn> sorry, system-config-network
[05:10:39] <fenn> dunno if that's actually on bdi
[05:10:57] <LawrenceG> * LawrenceG looking
[05:12:17] <LawrenceG> no luck ... no netconfig, netcfg etc...
[05:13:47] <LawrenceG> the eth0 device is there but doesnt start... pick up the dhcp config or anything
[05:13:55] <fenn> eth0 is there?
[05:14:14] <LawrenceG> yes but not active
[05:14:24] <fenn> can't you just ifconfig eth0 up
[05:14:42] <fenn> "ifconfig eth0 up xxx.xxx.xxx.xxx"
[05:17:16] <LawrenceG> that takes it up so it shows up in ifconfig, but any pings etc show network unreachable
[05:22:39] <LawrenceG> ubuntu uses network-admin
[05:22:56] <LawrenceG> is there a similar debian app?
[05:26:18] <LawrenceG> cradek, you around tonight as guru of bdi?
[05:42:27] <LawrenceG> hmmm... setup /etc/network/interfaces with eth0 info... still no luck
[05:50:19] <LawrenceG> ok success.. also needed /etc/resolv.conf set up with name servers
[06:13:36] <CIA-6> 03paul_c * 10emc2-auto/wiki/ (11 files in 8 dirs): "Auto update wiki from a cron job. Wed Oct 26 05:30:01 BST 2005 "
[07:11:07] <fenn> i hope i never have to use a real PLC
[07:18:50] <K`zan> BTDT AB and GE Fanuc, Fun Fun Fun :-)
[07:53:59] <K`zan> Night folks and thanks!
[07:57:36] <Jymmm> fenn talk to Johm about them =)
[08:39:27] <alex_joni> good morning
[12:47:46] <cncuser> good morning :)
[12:48:29] <alex_joni> good morning
[12:48:36] <cncuser> hi alex_joni joni :)
[12:48:39] <alex_joni> obwohl.. ein wenig sp�t daf�r
[12:49:11] <cncuser> release is finished, only the shutdown hangs (busybox /sbin/poweroff)
[12:49:35] <cncuser> alex_joni: no, i wake up 30 minutes ago, and have just finished my kakao
[12:49:46] <cncuser> s/wa/wo/
[12:50:03] <alex_joni> hmm.. any idea why?
[12:50:07] <cncuser> adeos
[12:50:09] <alex_joni> it hangs.. I mean
[12:50:12] <alex_joni> rmmod adeos?
[12:50:17] <cncuser> isnt loaded ;)
[12:50:24] <alex_joni> compiled into the kernel?
[12:50:34] <cncuser> no, got no clou, reboot works so it must be some tiny little fuzzy thingy
[12:50:50] <cncuser> i blame a init script
[12:50:56] <cncuser> i hate bash
[12:50:57] <alex_joni> yeah.. do that
[12:51:03] <alex_joni> blame it on the scripts
[12:51:04] <alex_joni> :D
[12:51:07] <cncuser> :)
[12:51:07] <alex_joni> rofl
[13:03:57] <alex_joni> blame it on the weather :)
[13:06:45] <cncuser> i think its busybox or the initscript
[13:07:28] <alex_joni> I think it's the busy environment :)
[13:07:44] <cncuser> hahaha
[13:08:39] <cncuser> alex_joni: do you have space to host the dev packages (unleashed right now) ?
[13:08:58] <alex_joni> * alex_joni has plenty of space
[13:09:12] <cncuser> cool
[13:09:14] <alex_joni> not more than a few gigs, but plenty
[13:09:18] <alex_joni> I think
[13:09:34] <alex_joni> included a cvs version?
[13:09:38] <alex_joni> ready to build?
[13:09:44] <alex_joni> send me the link :)
[13:10:29] <cncuser> alex_joni: no, just setup some packages which you can add to puppy unleashed. to build a puppysit
[13:10:41] <cncuser> not online now. i cant upload from here.
[13:11:35] <cncuser> also i have to fix the shutdown bug. my pc is outside the room, i hate to standup when doing a shutdown and it hangs. so do many i assume
[13:11:53] <alex_joni> hang on.. about the shutdown
[13:12:03] <alex_joni> you mean it doesn't power off? or the shutdown hangs?
[13:12:13] <alex_joni> because power off is normal (as ACPI is missing)
[13:12:20] <alex_joni> or I hope that it's missing
[13:12:55] <cncuser> poweroff should work
[13:12:59] <cncuser> also without acpi
[13:13:12] <cncuser> its a old cpu instruction
[13:13:21] <cncuser> or not ?
[13:13:23] <cncuser> *hmmm*
[13:13:33] <alex_joni> nope.. it doesn't :)
[13:13:39] <alex_joni> it's part of ACPI and APM
[13:13:47] <alex_joni> both are bad for RT
[13:16:05] <cncuser> http://www.patmedia.net/marklevinson/cool/cool_illusion.html
[13:18:10] <alex_joni> nice
[13:26:05] <cncuser> the dot disapear because of the double focus. the colour, im not shure
[13:33:05] <cncuser> ok, it works, need to build one more iso, the package everything up, put it on a usbstick and search for a internetconnection to upload :)
[13:33:16] <cncuser> shutdown does not work
[13:33:25] <cncuser> it ha?lts, but without errormessages :)
[13:33:28] <cncuser> better then nothing
[13:34:26] <alex_joni> yup
[13:34:33] <alex_joni> it's the same on any BDI
[13:35:11] <cncuser> well, dont know, havent had the nerve to install it via qemu lately
[13:35:28] <alex_joni> heh
[13:35:30] <alex_joni> ok
[13:35:40] <alex_joni> you can mail me the iso if you want :)
[13:35:45] <alex_joni> if it's easier
[13:35:45] <cncuser> my last bdi install took 2 hours i think, as it disables dma by default in the install
[13:35:54] <alex_joni> yup
[13:36:00] <alex_joni> but you can hdparm it back up
[13:36:01] <cncuser> thats weird
[13:36:03] <cncuser> no
[13:36:08] <cncuser> last time i tried i couldnt
[13:36:13] <alex_joni> hmm.. dunno
[13:36:14] <cncuser> operation not permited :)
[13:36:22] <alex_joni> you're not allowed :D
[13:36:26] <cncuser> haha
[13:36:31] <cncuser> they know me
[13:36:35] <alex_joni> yup
[13:40:57] <les_w_away> good morning
[13:42:14] <alex_joni> morning les
[13:42:38] <les_w_away> les_w_away is now known as les_w
[16:29:48] <rayh> rayh is now known as rayh-away
[16:33:15] <les_w> hi jacky
[16:33:40] <les_w> I have some calculations for you
[17:09:54] <Jacky^> hey les_w :)
[17:09:57] <anonimasu> vw: *yawn
[17:09:59] <anonimasu> er
[17:10:01] <Jacky^> morning
[17:10:11] <Jacky^> was away ..
[17:10:47] <Jacky^> bad evening yesterday.. but ive all info saved in a log ;)
[17:10:54] <Jacky^> whats new ?
[17:11:26] <Jacky^> hi anonimasu
[17:11:53] <anonimasu> not much
[17:11:57] <anonimasu> just got home from work
[17:12:07] <Jacky^> mmh ..
[17:12:12] <anonimasu> I am tired and I need sugar and caffine badly
[17:12:21] <Jacky^> relax a bit then
[17:13:31] <anonimasu> I feel like coding
[17:13:40] <Jacky^> :)
[17:14:06] <Jacky^> i was tryng a small dc motor with encoder
[17:14:30] <Jacky^> but the encoder is not ttl, so can't work on geckos
[17:14:52] <anonimasu> just change it to tool..
[17:14:54] <anonimasu> ttl..
[17:15:11] <Jacky^> measured voltage is about 7 V
[17:15:44] <Jacky^> i'm not familiar with encodrs.. ill try with some other
[17:15:47] <anonimasu> what do you feed your encoders with?
[17:15:53] <Jacky^> where i'm sure its TTl
[17:15:54] <anonimasu> you've got +5v off the gecko..
[17:16:01] <anonimasu> that you should be using for the encoders..
[17:16:24] <Jacky^> i know, +5V 50 ma
[17:16:41] <Jacky^> i think this encoder cant work
[17:17:00] <Jacky^> have a lot of terminals too
[17:17:20] <Jacky^> tryng to find the 2 phases with the scope
[17:18:19] <Jacky^> ive another encoder from an hp printer
[17:18:41] <Jacky^> a small disc with optoled, im sure that is ttl
[17:19:29] <anonimasu> you need to feed it 5v also..
[17:19:33] <anonimasu> for the led..
[17:19:39] <Jacky^> yeah
[17:19:42] <anonimasu> :)
[17:19:46] <anonimasu> just checking ;)
[17:19:55] <Jacky^> but it coud be work under 50 mA
[17:20:07] <Jacky^> already tried, it work
[17:20:15] <anonimasu> get datasheets..
[17:20:22] <Jacky^> now going to test it on the gecko
[17:20:32] <Jacky^> no datasheet online
[17:20:39] <Jacky^> its an agilent component
[17:20:45] <Jacky^> from hp
[17:20:58] <Jacky^> but i found the leads with the scope ;)
[17:21:04] <Jacky^> not difficul
[17:21:37] <anonimasu> what+
[17:21:40] <anonimasu> ?
[17:21:48] <Jacky^> hp seem to use 'not standard' components on theyr printers
[17:22:06] <Jacky^> + is 5V
[17:23:02] <Jacky^> it was a very small inkjet printer, i guess very fast
[17:23:34] <Jacky^> 2 small dc motors, one with ancoder disc the other on the head with a linear bar encoder
[17:23:50] <Jacky^> first printer ive seen to use dc motors
[17:24:43] <Jacky^> ive a friend that have a PC shop, so when i have old printers he call me to take them for free ;)
[17:24:51] <Jacky^> he*
[17:26:30] <Jacky^> im looking for heatskins on drivers, socket370 model should be ok for what i read on the web
[17:26:50] <anonimasu> I've bolted mine to a 1cm alu late..
[17:26:53] <anonimasu> plate..
[17:27:25] <Jacky^> good ..
[17:27:38] <Jacky^> ive some piece of alu, but its too small for it
[17:27:46] <anonimasu> I am going to re make it.
[17:27:51] <anonimasu> it was a scrap piece I had lying around
[17:28:05] <anonimasu> going to make a real heatsink when I get the spindle back togther
[17:28:12] <Jacky^> have fan up too ?
[17:29:13] <Jacky^> I found an interesting diagram online
[17:29:25] <Jacky^> i think maded by steve
[17:29:32] <Jacky^> er
[17:29:35] <Jacky^> i'm sure
[17:29:37] <Jacky^> :)
[17:29:52] <Jacky^> looking at that
[17:30:04] <anonimasu> no
[17:31:18] <anonimasu> brb going outside to the ship
[17:31:21] <anonimasu> shop
[17:31:29] <Jacky^> ok, later
[17:31:40] <anonimasu> I'll be back in maybe a hour..
[19:12:23] <etla> ?
[19:12:32] <alex_joni> !
[19:12:45] <etla> hi
[19:12:54] <alex_joni> hi
[19:13:01] <etla> do you know anything about the parallell port motion control / pico-systems
[19:13:08] <alex_joni> yes
[19:13:20] <etla> do you have it ??
[19:13:24] <alex_joni> no
[19:13:29] <Jymmm> YES! YES! YES! OH GAWD YES!
[19:13:39] <alex_joni> Jymmm: hi
[19:13:52] <etla> I'm wondering if the EPP approach would be the way to go
[19:13:59] <etla> with a home developed system
[19:14:03] <alex_joni> * alex_joni will wonder with you
[19:14:16] <alex_joni> Jymmm: care to join us wondering?
[19:14:27] <Jymmm> * Jymmm will fart instead of wonder... more productive this way.
[19:14:33] <alex_joni> oh.. let's wonder together.. that would be nice
[19:14:36] <etla> would be cheap, all computers have parallell ports, no need for special servo pci cards
[19:14:52] <alex_joni> imagine PC's without PCI, only parports
[19:14:55] <alex_joni> dreamy
[19:15:00] <Jymmm> ewwwwwwwwwwwwwwwwwwww
[19:15:19] <alex_joni> you could have a soundblaster on parport (like in the old days)
[19:15:35] <Jymmm> eth on paraport, but never heard of SB
[19:15:36] <alex_joni> actually it was more like a resistor-net
[19:15:51] <alex_joni> R2R if it rings a bell
[19:15:59] <Jymmm> * Jymmm still has a working ethernet paraport adapter.
[19:16:16] <Jymmm> can even boot from dos to the net with it
[19:16:19] <alex_joni> * alex_joni would like to try paraporting
[19:16:36] <Jymmm> alex_joni: Actually, it wasn't bad at all.
[19:16:40] <alex_joni> yup
[19:16:49] <alex_joni> see.. now imagine a PC with 10 parports
[19:17:02] <alex_joni> all different colours
[19:17:12] <Jymmm> I2C instead thank you very much
[19:17:23] <alex_joni> purple, blue, yellow, pink, gray, red,...
[19:17:36] <alex_joni> oh.. and different sizes of course
[19:19:53] <Jymmm> * Jymmm shoves alex_joni back into reality
[19:39:44] <alex_joni> * alex_joni doesn't fit in there
[19:40:35] <les_w> ?
[19:40:39] <etla> hi
[19:40:47] <les_w> hi etla
[19:41:29] <etla> thinking about electronics for new servo setup...
[19:41:46] <les_w> amps?
[19:41:58] <etla> yes, at first just amps
[19:42:10] <etla> but then also maybe separate encoder counters
[19:42:11] <les_w> full servo or step/servo?
[19:42:14] <etla> and the pc interface
[19:42:24] <etla> I think I want to move away from step/dir
[19:42:30] <etla> had enough of that already
[19:42:36] <les_w> good idea
[19:42:51] <les_w> big or little machine?
[19:43:16] <etla> right now we have a stepper controlled mill with dovetails+acme screws
[19:43:22] <etla> max travel around 250mm
[19:43:30] <etla> first get the servo system working on that
[19:43:45] <etla> then if we find cheap linears+ballscrews maybe build/retrofit something new
[19:44:14] <les_w> I see
[19:44:37] <Jymmm> les_w Do you have a DVD drive in your computer?
[19:44:54] <les_w> yeah!
[19:45:35] <Jymmm> was that to me?
[19:45:46] <les_w> yup
[19:45:55] <Jymmm> les_w ok, cool.
[19:46:01] <les_w> guess I am not registered again
[19:46:22] <Jymmm> les_w: /ns identify <password>
[19:46:34] <les_w> k
[19:47:47] <les_w> done
[19:49:05] <les_w> etla: the 4 amp 35v or so ic servo amps might work for you
[19:49:12] <les_w> forgot the part number
[19:49:33] <etla> les_w I have the surpluscenter sanyo denki 400W ac servos on order
[19:49:45] <etla> 100VAC, 3-phase sinusoidal drive is what I want
[19:50:30] <les_w> oh...ok. Well of course icee and lawrence are working on it...
[19:50:52] <les_w> jymm getting my pm?
[19:51:00] <Jymmm> les_w no
[19:51:03] <etla> yep, I've talked to them about it
[19:51:04] <anonimasu> les_w: what happened to "les"
[19:51:12] <alex_joni> it grew
[19:51:20] <les_w> had to change it
[19:51:21] <anonimasu> :)
[19:51:30] <les_w> too many les
[19:51:52] <les_w> or something.
[19:52:05] <Jymmm> somebody registered the nick 'les'
[19:52:07] <les_w> registration is not working too great for me here
[19:53:02] <Jymmm> les_w: /ns help register
[19:53:13] <les_w> thanks
[19:54:03] <Jymmm> les_w: /ns register <password> is the direct command
[19:54:05] <anonimasu> /msg les you fake!
[19:54:11] <anonimasu> *grins*
[19:55:00] <les_w> I did register but it seemed not to take
[19:55:02] <les_w> hmm
[19:55:14] <les_w> try again...
[19:55:28] <Jymmm> nope, it's still not registerd
[19:55:45] <Jymmm> try /msg nickserv register <password> (without the brackets)
[19:55:52] <les_w> k
[19:56:30] <les_w> ieee
[19:56:54] <Jymmm> ?
[19:57:08] <les_w> nope...no register
[19:57:41] <Jymmm> what happens?
[19:58:31] <les_w> got "nickserve register <password>
[19:58:36] <anonimasu> nickserv
[19:58:38] <les_w> but pms don't send
[19:58:38] <Jymmm> no e
[19:59:00] <Jymmm> les_w: /msg nickserv not nickserve
[19:59:06] <les_w> ah ok
[19:59:08] <Jymmm> 8 chars
[19:59:08] <les_w> duh
[19:59:14] <les_w> trying
[20:00:27] <les_w> ok regitered I think
[20:00:43] <Jymmm> yep it worked
[20:00:45] <les_w> registered
[20:00:51] <les_w> good. bleh
[20:01:28] <Jymmm> in some clients you can abbr /msg nickserv to just /ns
[20:01:38] <Jacky^> les_w: you need to identify too
[20:01:54] <Jacky^> anytime you enter
[20:01:59] <Jymmm> he's identified right now, but the next time he logs in he'll have to
[20:02:02] <Jacky^> /msg nickserv identify passwd
[20:02:20] <les_w> right understand jacky
[20:02:25] <Jacky^> put this line in the irc client to get auto-identify
[20:02:29] <les_w> ready for some motor data?
[20:02:49] <Jacky^> yeah
[20:03:20] <Jacky^> les_w: got some good info ?
[20:03:38] <les_w> ok max 1000 rpm,,,need:
[20:04:18] <les_w> .000593 Kg M^2 rotor inertia
[20:04:21] <les_w> and
[20:04:59] <les_w> 3.5 nm cont torque
[20:05:08] <les_w> that is a fairly big motor
[20:05:16] <les_w> y and z much smaller
[20:05:23] <les_w> all direct drive
[20:05:44] <Jacky^> oh, Z withiut pinion too then
[20:06:04] <les_w> with simple/fixed ballscrew supports (MUCH easier to make)
[20:06:23] <Jacky^> yeah
[20:06:29] <Jacky^> good :)
[20:06:29] <les_w> noe 3.5 nm is a lot of torque...but it allows easy direct drive
[20:06:41] <Jacky^> understood
[20:06:54] <les_w> that is x
[20:07:05] <les_w> let's do y
[20:07:19] <les_w> assume moving mass is....hmmm...15 kg
[20:08:14] <Jacky^> i think i will use aluminum for the gantry
[20:08:48] <Jacky^> 20 mm tickness its enough ?
[20:10:07] <les_w> for y, 1 nm m torque 10^-7 kg m^2 rotor inertia
[20:10:56] <les_w> 3 times smaller motor
[20:11:17] <les_w> for the gantry you can use aluminum...steel is better
[20:11:50] <Jacky^> i could use steel too, but what about the weight ?
[20:12:01] <les_w> steel rail bolted to aluminum = bimetallic strip
[20:12:18] <alex_joni> yeah.. bends nicely on temperature
[20:12:35] <les_w> jacky, the stiffness to weight ratio of aluminum and steel is virtually identical
[20:12:47] <Jacky^> ouch :(
[20:12:58] <Jacky^> thts sound new to me ..
[20:13:01] <alex_joni> les_w: if you use stock aluminum
[20:13:01] <Jacky^> okay
[20:13:13] <alex_joni> but you can have extruded profiles, lots more strength
[20:13:39] <les_w> aluminum is often used...but it will bow some with temperature
[20:13:48] <les_w> but I like steel because:
[20:13:53] <les_w> cheaper
[20:14:02] <les_w> same specific stiffness
[20:14:04] <alex_joni> easier to weld :)
[20:14:08] <les_w> easier to weld
[20:14:17] <les_w> no bimetallic strip effects
[20:14:23] <alex_joni> * alex_joni was first :)
[20:14:35] <les_w> you win!
[20:14:36] <les_w> haha
[20:14:56] <Jymmm> alex_joni just likes being on top
[20:15:20] <les_w> I will lose a typing contest every time
[20:15:25] <alex_joni> Jymmm: depends *g*
[20:15:55] <Jacky^> les_w: thats ok for y
[20:15:56] <Jymmm> lol
[20:16:07] <les_w> People often ask about the bimetallic strip thing...if the bowing will really be significant
[20:16:19] <alex_joni> depends on size & temp
[20:16:26] <les_w> I have not calculated it...but yeah I think it will be big
[20:16:42] <les_w> I might calculate it if I have time
[20:16:56] <fenn> just get hardcoat anodized aluminum rails :P
[20:17:02] <les_w> yeah
[20:17:09] <Jymmm> ROTFLMAO... <Jymmm> alex_joni just likes being on top <alex_joni> Jymmm: depends *g* <alex_joni> depends on size & temp
[20:17:24] <les_w> hahaha
[20:17:32] <alex_joni> that's what I meant :D
[20:17:53] <Jymmm> well, if it aint hot, not really worth it
[20:17:54] <Jacky^> I must go fro dinner :(
[20:18:01] <Jacky^> came back in 10 min
[20:18:10] <Jacky^> Jacky^ is now known as Jacky^afk
[20:19:17] <les_w> well the closer to the neutral axis the more dissimilar metal structures will bend
[20:19:33] <les_w> but just look what a thermostat does!
[20:20:16] <les_w> also, unless pinned, it would slide all over
[20:20:35] <les_w> even with very tight bolt clamp loads
[20:21:02] <les_w> btw I do pin AND bolt rails
[20:24:09] <Jymmm> dowel pins?
[20:24:27] <alex_joni> needles & pins
[20:27:19] <les_w> heh
[20:27:24] <les_w> yeah dowels
[20:27:47] <Jymmm> http://www.rockler.com/ecom7/product_details.cfm?cookietest=1&offerings_id=10631&sid=AFD70
[20:27:52] <les_w> wow found calcs for simple rectangular cross section bimetal strip:
[20:27:59] <Jacky^afk> Jacky^afk is now known as Jacky^
[20:28:07] <les_w> http://www.doitpoms.ac.uk/tlplib/thermal-expansion/printall.php
[20:29:02] <les_w> need it for arbitrary beam second moment of area
[20:29:57] <les_w> quick dinner jacky
[20:30:05] <Jacky^> yeah :)
[20:30:07] <les_w> ok we have x and y
[20:30:20] <les_w> now z
[20:30:50] <les_w> assume 10 kg with gravity...
[20:30:55] <les_w> running...
[20:33:18] <les_w> ok. about 0.5 nm with 10^-5 kg -m^2 inertia
[20:33:50] <les_w> so x...3.5. Y...1 Z....0.5 nm
[20:34:01] <les_w> continuous
[20:34:25] <Jacky^> well :)
[20:34:49] <les_w> those are the motors you need.
[20:34:59] <Jacky^> finally i can start to ask for quotes around here
[20:35:13] <Jacky^> les_w: thanks a lot :P
[20:35:33] <les_w> right. When you find a potential motor I will check the specs for you
[20:35:35] <les_w> yw
[20:36:13] <anonimasu> hm../win 11
[20:36:14] <Jacky^> i suppose the price of Z motor should be about 1/3 of X motor ..
[20:36:30] <les_w> this will give you a VERY fast 10 m/min 0.5g machine
[20:36:42] <Jacky^> cool :P
[20:36:44] <les_w> yes z much cheaper
[20:37:04] <les_w> rpm can be anything...but you will only be using 1000
[20:37:09] <les_w> so min 1000 rpm
[20:37:23] <Jacky^> yeah
[20:38:01] <les_w> we run your ballscrews slow...so no crirical speed problems and easier to make end bearings
[20:38:29] <Jacky^> nice
[20:38:47] <Jacky^> les_w: any suggest about ecnoders type ?
[20:39:02] <Jacky^> i think it should not be a great problem
[20:39:05] <les_w> us digital is good
[20:39:08] <les_w> low cost
[20:39:14] <Jacky^> assuming G340 have the step multiplier too ..
[20:39:20] <les_w> right
[20:39:25] <Jacky^> ok
[20:39:49] <etla> you need G340s ? I might have for sale in a while
[20:39:56] <les_w> ok...happy motor hunting...back to the shop for me
[20:40:11] <Jacky^> etla: alredy buyed :/
[20:40:17] <Jacky^> too late
[20:40:19] <etla> ok
[20:40:20] <Jacky^> for me
[20:52:23] <les_w> back for a sec...jacky...one thing I forgot...with no power z may drop from gravity. May need a simple brake actuated on estop
[20:52:51] <Jacky^> les_w: ok, :)
[20:53:33] <etla> counterweight ala elevators for z !
[20:54:05] <les_w> you can use that...but it adds mass
[20:54:10] <les_w> springs are good
[20:54:50] <Jacky^> les_w: i was looking here: http://www.infodelta.it/motori.htm
[20:54:54] <les_w> I included a bias torque in the motor calcs for z to couteract gravity while it's on
[20:55:44] <Jacky^> the prices are good.. but can't see any datasheet ..
[20:55:49] <Jacky^> cheaper motors
[20:57:15] <Jacky^> SM5 1,30 nM 400W shaft 11 mm 4,8 kg = E. 180
[20:57:38] <Jacky^> from 1500 to 3000 rpm
[20:59:50] <etla> youre converting normal AC motors to servos ?
[21:00:14] <Jacky^> im not
[21:00:36] <Jacky^> they are DC motors
[21:00:44] <etla> ok
[21:01:15] <Jacky^> they write cc (corrente continua) = DC
[21:01:24] <les_w> brb phone
[21:02:21] <Jacky^> looking at rs components too, but isnt cheaper :/
[21:04:02] <etla> dc servos are becoming hard to find I think
[21:04:11] <etla> eveyrithing is AC or brushless now
[21:05:19] <anonimasu> galil sells new ones..
[21:06:28] <etla> where ?
[21:07:01] <etla> ok, found it
[21:20:28] <les_w> ok off the phone
[21:20:55] <les_w> lots of dc servos around, but of course brushless will eventually dominate
[21:21:12] <les_w> still a price issue now
[21:21:42] <Jacky^> yeah
[21:21:53] <les_w> you need to pay particular attention to rotor polar moment of inertia
[21:22:06] <les_w> if too big, much more torque is needed
[21:22:20] <les_w> if too small servo instability can result
[21:22:29] <Jacky^> mmhh
[21:22:53] <Jacky^> whats the 'default value' ?
[21:23:04] <Jacky^> it depend on motor nm ?
[21:23:22] <les_w> I did a fairly good inertia matching on the run
[21:23:36] <fenn> you can have too little inertia?
[21:23:53] <les_w> torque/ moment of inertia= how fast the motor can spin up
[21:24:08] <les_w> fenn, yes.
[21:24:36] <fenn> the drive amp just cant react quickly enough?
[21:24:44] <les_w> ideally you want to match mechanical impedance of the motor and load
[21:24:56] <les_w> it is not horribly critical
[21:25:06] <les_w> but deserves some attention
[21:25:21] <les_w> too small can cause resonances
[21:25:23] <les_w> and
[21:25:46] <les_w> if the load inertia varies with position....servo instability
[21:26:01] <Jacky^> :/
[21:26:36] <fenn> "load inertia varies with position" like in a robot arm? or like cogging in any old motor?
[21:27:55] <les_w> for example jacky's x axis (reading back) is 6x10^-4 kg -m^2 equivalent moment of inertia
[21:29:08] <les_w> fenn: dual drive ganties have variable inertia. A part of the cutting load is inertial too.
[21:29:49] <fenn> they do?
[21:30:01] <fenn> how is that possible?
[21:30:09] <fenn> the gantry weighs the same no matter what
[21:30:34] <les_w> and at higher frequencies the impedance goes from springlike, to damper like, to inertial as it passes through a pole pair. Thes locations vary with position.
[21:31:31] <les_w> on dual drive x when y is on one end...that motor sees it's mass more than the other
[21:31:41] <les_w> in the middle they see the same
[21:31:44] <les_w> and so on
[21:31:57] <fenn> okay i see that
[21:32:01] <les_w> now jacky is ising one motor so no big problem there
[21:32:09] <les_w> using
[21:32:35] <les_w> si it's just a resonance issue
[21:32:40] <les_w> so
[21:32:43] <Jacky^> one motor, 2 twin screw with strap
[21:32:51] <les_w> right
[21:32:53] <etla> resonance of ballscrew ?
[21:32:56] <les_w> gt belt
[21:33:02] <Jacky^> yeah
[21:33:45] <les_w> etla: yes, ballscrew torsional resonance and other structural resonances that change as the gantry mass is moved along
[21:34:05] <les_w> for example the ends of travel are usually stiffer than the middle
[21:34:26] <fenn> can you run the servo amp through a lowpass filter somehow?
[21:35:01] <les_w> yes you certainly can...a low pas or band pass. It helps a good bit.
[21:35:10] <les_w> emc does not have that though.
[21:35:33] <etla> you filter velocity ?
[21:35:34] <les_w> the filter needs to be in the pid loop.
[21:35:34] <fenn> i was thinking inside the drive electronics
[21:35:49] <etla> then you will get following errors
[21:35:50] <les_w> right
[21:36:17] <etla> I think it's better to put the hard calculations in emc and have 'dumb' hardware
[21:37:01] <les_w> if it is a gecko type, yes you could filter it there. I don't know what mariss does with his drives. Almost always the d term needs a lp, because an ideal differentiator has infinite gain as frequency goes high
[21:37:40] <les_w> actually emc does have that...it is called SMOOTH_D or something.
[21:37:59] <les_w> it is a rectangular convolution...i.e. moving average
[21:38:25] <etla> I've found the G64 mode to be quite unsatisfactory with steppers and freqmod
[21:38:55] <les_w> well I have found it a bit unsatisfactory as well.
[21:39:14] <etla> it blends moves the same way regardless of angle between them
[21:39:34] <etla> not lookahead, just blending...
[21:39:58] <les_w> I am doing some tests to further characterise some problems
[21:40:09] <etla> do you have steppers or servos ?
[21:40:22] <les_w> well trapezoidal blending requires look ahead...but only one point.
[21:40:32] <les_w> servos etla
[21:40:57] <etla> yes it looks at the code alright but it does not change behavior if the bend is 1 degree or 180 degrees !
[21:41:17] <les_w> really
[21:41:18] <etla> servos: ok so the problems are in the planner and not freqmod or something stepper related
[21:41:39] <les_w> I suspect so since I have problems too
[21:41:42] <etla> try g0x100x0 x100 x0 x100 x0 x100 x0 etc. with acc set low
[21:42:03] <etla> it will go nowhere near 0 and 100
[21:42:18] <les_w> oh really
[21:42:20] <etla> with G64 that is, with G61 and G61.1 it's ok
[21:42:31] <les_w> I have NOT had any problem with that
[21:42:45] <etla> I think I have max vel=3 and max_acc=15 or so
[21:42:51] <etla> inch/s and inch/s^2
[21:43:02] <les_w> blends are supposed to be turned off for any angle more than 90
[21:43:15] <etla> don't think they are
[21:43:27] <les_w> hmm ok. making a note of that
[21:43:46] <etla> do you actually know your way around the blending code ?
[21:44:33] <les_w> I know the math very well. I think few (if any) can really figure out the code. It is too convoluted.
[21:44:53] <etla> ok.
[21:44:53] <les_w> we try, though.
[21:45:06] <etla> it fit's a polynomial to the endpoints ?
[21:45:14] <cradek> les_w: we?
[21:45:26] <les_w> haha
[21:46:09] <fenn> i tried, just not very hard
[21:46:41] <les_w> etla, a cubic polynomial is fitted to servo update ponts.
[21:47:08] <cradek> etla: is this emc1 or 2?
[21:47:11] <les_w> A second order fuction is fitted to trajectory points...parabolic
[21:47:11] <etla> so you need three points to fit to ?
[21:47:21] <etla> I'm using emc1, BDI4.30
[21:47:21] <les_w> 4 for cubic
[21:47:36] <etla> ah the constant also
[21:48:16] <etla> the problem is not so much the fitting I think
[21:48:33] <etla> but that there is no tolerance btw. what planner plans and actual g-code trajectory
[21:48:43] <les_w> one point behind and one point ahead for trapezoidal
[21:48:49] <les_w> total 3
[21:50:02] <etla> and it's all supposed to be in tp.c ?
[21:50:05] <les_w> BTW Chris...I got some real TP nasties doing a big contours sign yesterday.
[21:50:29] <les_w> It's better that it was, but some significant problems left.
[21:50:53] <les_w> etla: tp.c and tc.c
[21:51:50] <etla> the algorithm is quite well hidden...
[21:52:07] <les_w> it is to me ;)
[21:52:27] <les_w> I can identify some parts
[21:52:57] <les_w> for trapezoidal primer I think I still have a web page up
[21:53:00] <les_w> let me check
[21:53:11] <les_w> for got thr name of it
[21:56:06] <les_w> yeah still there
[21:56:07] <etla> you should put everything you know in the wiki !
[21:56:09] <les_w> http://www.lmwatts.com/traj/traptp.html
[21:56:13] <les_w> enjoy
[21:56:13] <dmess> hi
[22:02:22] <les_w> hi dmess
[22:26:22] <anonimasu> iab
[22:26:40] <alex_joni> wb
[22:26:49] <anonimasu> thanks
[22:26:52] <anonimasu> what's up?
[22:27:00] <alex_joni> fighting tool-prepare logic
[22:27:02] <alex_joni> and tool-load :)
[22:27:23] <anonimasu> ehe
[22:27:23] <anonimasu> nice
[22:27:28] <les_w> yeah
[22:28:39] <les_w> blah writing, writing, writing reports
[22:29:27] <les_w> oh well got out inthe shop a little. Surface ground a small machine base.
[22:29:31] <anonimasu> hm
[22:30:06] <les_w> sparked it off with .01 mm down feed
[22:31:55] <anonimasu> :)
[22:31:56] <anonimasu> nice
[22:35:30] <les_w> ok proofread report once more and bam...off to client
[22:42:01] <fenn> anyone know if cncuser ever finished his "new and improved" puppy distro?
[22:42:45] <alex_joni> fenn: hang on
[22:43:36] <alex_joni> doesn't look like it
[23:15:04] <lerman> les_w: I just reread your blending notes an dhave some questions.
[23:15:06] <lerman> (1)Can the same process be used to limit the jerk? That is, instead of using a trapezoidal velocity profile, use a trapezoidal acceleration profile? How hard would that be to do (assuming one could understand the present code)?
[23:15:07] <lerman> (2)How and where are circular interpolation done (to machine circles)?
[23:19:33] <anonimasu> hm
[23:19:36] <anonimasu> I need to call surpluscenter..
[23:20:50] <CIA-6> 03alex_joni * 10emc2/ (3 files in 2 dirs): work done for iocontroller to support tool-changing, right now there's a problem that Tx and M6 can't be on the same line (probably some weird message passing), but other than that it seems functional and ready to test
[23:24:54] <anonimasu> god damn it!"�#
[23:25:01] <anonimasu> fsck surpluscenter.
[23:25:21] <icee> anon: what's wrong?
[23:26:12] <anonimasu> They sold my servos.
[23:26:22] <icee> sold them?
[23:26:24] <anonimasu> yes.
[23:26:36] <anonimasu> they sent me a mail yesterday wanting to verify my credit card number.
[23:26:47] <anonimasu> and I called to verify it, and they've sold my servos.
[23:26:55] <anonimasu> I am so mad right now.
[23:27:09] <icee> :/
[23:27:28] <icee> which ones were you getting?
[23:27:40] <anonimasu> the 400w ones
[23:27:51] <anonimasu> I dropped them another mail asking them to remove my user from their database.
[23:34:45] <les_w> hi icee
[23:35:12] <icee> hey les.. how goes it?
[23:35:23] <icee> * icee is beginning design entry for the servo controller
[23:35:40] <les_w> anon: sold your motors??? #%@!
[23:35:55] <les_w> going ok icee...long day
[23:37:31] <anonimasu> les_w: I placed a order..
[23:37:40] <les_w> grinding this...worker painting that. Writing reports and talking on the phone like a girl (but being paid better)
[23:37:50] <anonimasu> les_w: they dropped me a mail about verifying my card.
[23:38:04] <les_w> and?
[23:38:17] <anonimasu> well I called them to verify my card and they've sold my motors..
[23:38:26] <les_w> shit
[23:38:29] <anonimasu> yeah..
[23:38:56] <anonimasu> makes me so mad..
[23:39:13] <anonimasu> "Is there anyting else you would like to buy?"
[23:39:35] <anonimasu> yeah, Got explosives?
[23:39:43] <lerman> les_w:[repeat of previous note] I just reread your blending notes an dhave some questions.
[23:39:44] <lerman> (1)Can the same process be used to limit the jerk? That is, instead of using a trapezoidal velocity profile, use a trapezoidal acceleration profile? How hard would that be to do (assuming one could understand the present code)?
[23:39:46] <lerman> (2)How and where are circular interpolation done (to machine circles)?
[23:40:10] <les_w> well bet you can get some from others who are hoarding them
[23:40:18] <anonimasu> ah, I dont give a shit..
[23:40:24] <icee> * icee wishes he had hoarded some
[23:40:32] <anonimasu> not right now..
[23:40:48] <icee> they still have a few 300W ones
[23:40:49] <anonimasu> I might aswell order some galil motors.. from galil in US..
[23:40:55] <les_w> lerman: yo can readily generate jerk limited profiles with a fifth order polynomial
[23:41:00] <les_w> and
[23:41:12] <les_w> forgot where the circle stuff is
[23:41:24] <anonimasu> les_w: I am not desperate after motors but it makes me mad that they've sold the stuff I placed a order for..
[23:41:49] <les_w> yeah
[23:42:06] <anonimasu> placing a order and giving them my CC
[23:42:12] <les_w> I would be bouncing off the walls
[23:42:17] <anonimasu> usually that meand your stuff is yours.
[23:42:47] <anonimasu> I can understand that they have a few day's expiration for verifying
[23:43:00] <anonimasu> but I got a mail late last night.. saying I had to verify my card number..
[23:43:16] <anonimasu> :/
[23:43:20] <icee> yah.. you ought to have at least a day or two
[23:43:24] <les_w> bleh
[23:44:29] <les_w> oh...Lerman: jerk limited profiles can have consequences since arbitrary shapes you might want to generate might not have such a profile!
[23:44:30] <anonimasu> maybe I should build my own motors..
[23:44:40] <les_w> ha yeah
[23:44:53] <anonimasu> out of 5c coins and cast iron shavings..
[23:45:05] <anonimasu> *grins*
[23:45:37] <les_w> I had a similar rapid run on those 105 sevo motors we sold
[23:45:50] <les_w> they sat there for the longest time then...
[23:45:55] <les_w> poof! gone.
[23:46:11] <anonimasu> yep
[23:46:24] <anonimasu> they said they were gone very fast..
[23:46:30] <les_w> All I was was a shipping department
[23:46:41] <les_w> I promised same day ship.
[23:46:53] <anonimasu> I countered with "I knew that would happen that's why I placed the order in time."
[23:47:43] <les_w> OOPS desk chair rolled on the cat's tail
[23:47:49] <lerman> les_w: How can that be? I could understand that I might have to go slower than I would like, but since the surface is continuous (in a mathematical sense), shouldn't the third derivative be have a limit?
[23:47:51] <anonimasu> :/
[23:47:53] <les_w> he's ok.
[23:47:59] <anonimasu> ah good
[23:48:59] <les_w> hmm lerman is making me think....
[23:49:56] <les_w> ok. imagine a customer wanted a precise s shape consisting of two tangent arcs
[23:50:46] <les_w> jerk at the join point is infinite...at any speed....so a jerk limited planner would have to make a little wiggle at that join point
[23:51:51] <icee> though if you're willing to go arbitrarily slow, you can end up with arbitrarily low error
[23:52:00] <les_w> right
[23:52:27] <lerman> les_w: looking at it another way, in any real system, the jerk IS limited. So, in such a system, there is a limit to what you can machine. But if you halve the the velocity, you halve the acceleration and the jerk.
[23:52:36] <les_w> because the infinite jerk region is infinitely small!
[23:53:23] <icee> lerman: sure, but that's made up for in the mechanical system now, not EMC
[23:53:50] <icee> I really need to read the trajectory planner source etc to be sure of what i'm saying
[23:54:01] <anonimasu> hm
[23:54:41] <anonimasu> hm, I saw at some higher end machines you could adjust it..
[23:54:49] <les_w> icee: please do read it. be my guest. Then tell us all how the heck it works!
[23:55:17] <lerman> icee: I don't worry about the facts. They might just confuse me. You shouldn't either :-) . But if you really feel like reading it, why don't you document it for the rest of it. -- Drat, les_w beat me to it. KL
[23:55:32] <icee> les: if none of the developers know how it works.. there really should be a concerted effort to deconstruct / document it or replace it
[23:55:39] <les_w> lerman: yes real physical systems are inherently jerk limited
[23:55:40] <icee> especially if it's known to bre broken
[23:55:59] <anonimasu> agreed
[23:56:00] <les_w> and I'm glad...I would hate for the universe to blow up and stuff.
[23:56:19] <anonimasu> les_w: ?!
[23:56:21] <les_w> icee: agreed
[23:56:22] <icee> is there any kind of release or project management process behind emc2?
[23:57:04] <icee> a running cvs branch is nice and all, but really there should be stable releases that are maintained for a known lifecycle, and then a trunk that converges to new releases periodically
[23:57:07] <les_w> don't know. Not involved a lot since I have to devote my time to production
[23:57:21] <icee> That kind of infrastructure is necessary if someone's going to do something like take on the trajectory planner
[23:57:28] <icee> where the trunk might have to be broken for a couple weeks
[23:58:18] <les_w> Well, a lot of that needs a serious re write...and a structure where the math guys can talk to the code guys
[23:59:28] <lerman> les_w: are you a math guy or a code guy or both? I have a degree in math (Cornell class of 1966), but have been a practicing C programmer for most of my career.