#emc | Logs for 2008-05-05

[01:12:17] <BigJohnT> http://www.youtube.com/watch?v=EC1_TTlsOrM
[01:12:32] <BigJohnT> Gamma-X that u?
[01:13:07] <BigJohnT> gotta go
[01:16:01] <Gamma-X> hahah
[01:16:02] <Gamma-X> yes
[01:19:05] <skunkworks> How is the build coming?
[01:19:50] <Gamma-X> not so good
[01:19:55] <Gamma-X> cant get the pid
[01:19:58] <Gamma-X> gave up a lil
[01:20:07] <Gamma-X> moved out also...
[01:41:41] <Gamma-X> i need auto pid or sumtin
[02:26:59] <stustev> good evengin
[02:27:03] <stustev> evening
[02:27:48] <SWPadnos> good evening, even
[02:28:10] <stustev> what's going on this evengin
[02:28:16] <stustev> again already
[02:28:21] <SWPadnos> heh
[02:28:34] <SWPadnos> I'm just doing some emailing, and deciding whether I should sleep or watch a movie :)
[02:28:59] <stustev> I was just playing minesweeper
[02:29:00] <eric_U> SWPadnos, ever work with pic?
[02:29:06] <stustev> no
[02:29:09] <SWPadnos> yes, in the distant evil past
[02:29:11] <stustev> brb
[02:29:37] <SWPadnos> I don't like them much, so I switched to AVR when they were released
[02:30:25] <eric_U> I'm a big fan of the msp430 myself
[02:30:33] <eric_U> but I have about 20 of these: http://www.rcgroups.com/forums/attachment.php?attachmentid=1643124
[02:30:35] <SWPadnos> haven't used that one
[02:31:06] <tomp> SWPadnos: i thought i might hook pos limit and neg limit switches in series with enA and enB on an L298, disabling one half of teh bridge. would that work? ( some pull downs liely )
[02:31:21] <SWPadnos> if you have a DC motor it might work
[02:31:31] <tomp> ng for stepper?
[02:31:33] <SWPadnos> nope
[02:31:55] <SWPadnos> you need both phases to move a stepper, the order in which you energize them determines the direction
[02:32:10] <tomp> i know its not the correct way but could any steps occur after 1/2 bridge was defeated?
[02:32:30] <eric_U> should successfully disable it
[02:33:12] <SWPadnos> I don't know the L298 specifically, but if you turn off phase A, you will still get alternating current to phase B, so the motor will sit and toggle +/- 2 steps around the point at which you disabled phase A (or something like that)
[02:33:38] <eric_U> I assumed he was going to run it bipolar
[02:33:53] <tomp> ah, so a somewhat safe limiting of travel? (even if it oscillates)
[02:33:53] <SWPadnos> even bipolar, the idea is to allow motion in one direction but not the other
[02:34:09] <SWPadnos> AFAIK disabling one phase of a stepper disables the stepper more or less totally
[02:34:21] <SWPadnos> sure, but you might as well disable both
[02:34:25] <SWPadnos> for either limit
[02:34:57] <tomp> good point, could do
[02:35:18] <SWPadnos> at which point, you may as well shut off DC power to all the motors ... :)
[02:35:39] <SWPadnos> (or something like that, apparently switching DC to the drives is a bad thing due to inrush currents or something)
[02:36:31] <tomp> by putting the limits to the 'amp', a parport could hand 5 stepper axis and 5 homes (freed of the limits inputs yet stopping travel )
[02:37:02] <SWPadnos> you'd want some logic to tell EMC to stop though
[02:37:11] <SWPadnos> so at least one input should be used for that
[02:38:00] <tomp> may have 1 (ext estop) gotta count :)
[02:38:04] <SWPadnos> but I think you get 5 inputs and 12 outputs, or 13 inputs and 4 outputs, there isn't a 5+12 option :)
[02:38:16] <tomp> yeh only 5 in
[02:38:19] <SWPadnos> err - 10 + 7 I meant
[02:38:41] <tomp> ooh? mongo go count
[02:38:49] <SWPadnos> mongo should read HAL manual ;)
[02:38:49] <Gamma-X> SWPadnos, hey!
[02:38:59] <SWPadnos> hi Gamma-X, cut off any fingers yet? :)
[02:39:11] <Gamma-X> lol didnt get it!
[02:39:14] <Gamma-X> cant get the pid
[02:39:16] <Gamma-X> and i moved out
[02:39:17] <tomp> thats where i get these ideas
[02:39:18] <Gamma-X> so no luck
[02:39:34] <SWPadnos> bummer
[02:39:39] <SWPadnos> or great! :)
[02:39:58] <Gamma-X> bummeer
[02:40:16] <Gamma-X> when u wanna come? lol
[02:40:51] <SWPadnos> err, well, I'm a bit busy at the moment (as usual)
[02:41:34] <Gamma-X> fudge lol
[02:41:57] <SWPadnos> unless you want to pay $750/day + expenses ;)
[02:42:59] <Gamma-X> ill pay 300 a day... plus my couch? lol
[02:43:03] <Gamma-X> i aint got that kidna money lol
[02:43:05] <SWPadnos> BZZZZT!
[02:43:16] <SWPadnos> I guess I'll keep doing "real work" then :)
[02:43:29] <Gamma-X> lol
[02:43:31] <SWPadnos> but maybe I should watch a movie before the week begins
[02:43:52] <SWPadnos> yes, I think I will. bb-tomorrow :)
[02:44:23] <Gamma-X> true
[02:44:35] <Gamma-X> swp if ur ever in the area somehow let me know lol please!
[02:44:40] <SWPadnos> will do
[11:03:50] <BigJohnT> http://www.youtube.com/watch?v=qlXCdfPsQUU
[11:04:01] <BigJohnT> http://www.youtube.com/watch?v=EC1_TTlsOrM
[11:04:52] <alex_joni> hi BigJohnT ..
[11:05:01] <BigJohnT> morning
[11:05:43] <alex_joni> finally a longer cut ;)
[11:05:46] <alex_joni> looks great
[11:05:51] <BigJohnT> lol
[11:05:56] <BigJohnT> thanks
[11:06:09] <BigJohnT> I was cutting the slats that support the material
[11:06:37] <BigJohnT> so I used saw horses to temp hold the material up
[11:11:51] <BigJohnT> that cut was with o word calling a file to touch off the sheet at the beginning of each cut and with an o word loop for the small cuts
[11:15:07] <alex_joni> cool :)
[11:15:26] <alex_joni> BigJohnT: remember to put them not really parallel to X
[11:15:41] <alex_joni> (the slats I mean)
[11:16:36] <BigJohnT> their parallel to Y actually
[11:17:27] <BigJohnT> for long cuts on the Y axis I can move the sheet over to miss a slat
[11:33:22] <alex_joni> BigJohnT: Y was what I was meaning
[11:33:32] <alex_joni> can't you put them slightly misalligned?
[11:33:38] <alex_joni> 1" or so..
[11:33:47] <alex_joni> helps a lot with their lifespan
[11:34:44] <BigJohnT> not now, but I understand what you mean
[11:35:49] <BigJohnT> my plan was to be careful in positioning the material and planning my cuts to not cut right on top of a slat
[11:35:59] <BigJohnT> I can remove one if it is in the way
[11:39:51] <alex_joni> BigJohnT: another way is to make them slightly lonker
[11:39:54] <alex_joni> longer even
[11:40:30] <BigJohnT> to make them curve...
[11:40:55] <alex_joni> then bend them so that they align to the other slot
[11:40:57] <alex_joni> right
[12:41:54] <jepler> <svenk91> on crt a tv and monitor are different because one uses elektricity to guide the electron beam and the other magnetic stuff or something
[12:47:32] <skunkworks_> morning jeper
[12:47:37] <skunkworks_> jeper?
[12:47:41] <skunkworks_> jepler
[12:51:53] <jepler> yeah, jeper. rhymes with "leper".
[13:07:00] <cradek> jepler: there are both electromagnetic and electrostatic deflection CRTs, but TVs (and later, monitors) have been electromagnetic since about WW2. o'scopes are probably the only remaining electrostatic CRTs.
[13:07:16] <cradek> and yes, I know you probably didn't care
[13:07:21] <skunkworks_> heh
[13:08:01] <skunkworks_> how did they do electrostatic?
[13:08:07] <skunkworks_> plates?
[13:08:20] <skunkworks_> instead of coils?
[13:08:47] <skunkworks_> * skunkworks_ could google..
[13:08:58] <lerman> Are there still oscopes that aren't digial?
[13:09:07] <lerman> digital
[13:09:23] <lerman> Yes, parallel plates.
[13:09:34] <skunkworks_> Hi lerman
[13:09:46] <lerman> Hi skunkworks.
[13:10:42] <lerman> Two pairs of parallel plates. A horizontal set and a vertical set. The disadvantage of electrostatic is that they have a relatively small deflection angle.
[13:11:21] <cradek> lerman: sure, mine
[13:11:22] <lerman> The advantage is the high frequency response. A TV is raster scanned so the frequency is (more or less) constant.
[13:11:58] <lerman> cradek: do you have a new one that's a crt?
[13:12:26] <cradek> mine's a tek 466
[13:12:31] <lerman> Or is it like my first one -- an old Tektronix dual beam that had a zillion tubes and had a separate powersupply.
[13:12:46] <SWPadnos> my 7-year old digital scope still has a CRT :)
[13:12:57] <SWPadnos> my 2-year-old one has an LCD though
[13:13:05] <cradek> I had a tek tube beast too, but it's gone now
[13:13:25] <lerman> My teks (I had two) are both gone, too.
[13:13:31] <cradek> I'll use this one until it dies, which sure might be a while
[13:13:59] <lerman> I have a digital one now. -- I forget the number.
[13:14:44] <cradek> and let me plug the early television museum again: http://www.earlytelevision.org/museum_information.html
[13:14:46] <jepler> I like my digital one, since my scope before that wasn't a storage scope
[13:14:47] <lerman> When it comes to user interfaces, I find tek to be excellent. HP is also pretyy good.
[13:14:59] <jepler> but it's not new or featureful
[13:15:02] <jepler> (it's a HP)
[13:15:33] <lerman> On another subject, what's the best way to get to Galesburg?
[13:15:44] <SWPadnos> funny. I haven't really liked the tek interface, but I like the Agilent interface
[13:16:02] <SWPadnos> I did try the latest Tek at ESC though, and it was pretty nice
[13:16:07] <jepler> lerman: where are you coming from?
[13:16:10] <SWPadnos> driving
[13:16:18] <lerman> Connecticut.
[13:16:22] <SWPadnos> carpool ;)
[13:16:34] <lerman> SWPadnos: where are you coming from?
[13:16:48] <SWPadnos> it's a 2-day drive from here (Vermont), but since I bring a lot of crap, I can't use public transport
[13:17:55] <lerman> Unfortunately, I can't stay for the weekend -- we are having a family party.
[13:18:12] <lerman> SWPadnos: when are you going?
[13:18:13] <jepler> I don't think there's a major airport closer than chicago, and it's still a few hours drive from there
[13:18:17] <SWPadnos> if you're not bringing a lot of junk, the closest reasonable-sized airport is Peoria
[13:18:29] <SWPadnos> <1 hour away IIR
[13:18:32] <SWPadnos> IIRC
[13:18:51] <SWPadnos> or, if you have your own plane, you can fly to the airport 1 mile down the road
[13:19:16] <jepler> google calls it 3 hours from ORD to Galesburg .. and yeah, peoria is about 1 hour away but I don't know about its airport
[13:19:22] <lerman> Good idea, I'll buy a plane and take flying lessons.
[13:19:48] <SWPadnos> it's "reasonable-sized" ;)
[13:20:08] <SWPadnos> lerman, if you have a friend with a small plane, that works too :)
[13:20:26] <lerman> My friend with the small plane crashed it.
[13:20:30] <SWPadnos> bummer
[13:20:37] <lerman> But he did walk away.
[13:20:45] <skunkworks_> lucky
[13:21:07] <SWPadnos> skilled ;)
[13:21:08] <lerman> Lucky he crashed or lucky he walked away? :-)
[13:21:20] <skunkworks_> heh
[13:21:41] <lerman> Skilled he crashed or skilled he walked away? :-)
[13:21:44] <jepler> orbitz will give me a flight from Chicago ORD to Peoria via Intianapolis and Minneapolis for only $345 plus fees. Total duration: 9 hr 15min
[13:22:03] <SWPadnos> hmmm. I have a friend whose father usually has 5 planes or so, maybe I should convince my wife to finish her pilot training
[13:22:05] <cradek> ha
[13:22:08] <SWPadnos> heh
[13:22:10] <cradek> what a deal
[13:22:21] <jepler> the non-stop flight is more, starting at $553+
[13:22:40] <lerman> You can pick me up at Danbury airport. :-)
[13:23:01] <SWPadnos> jepler, is that POA?
[13:23:42] <jepler> it says PIA
[13:23:46] <SWPadnos> ah
[13:23:54] <SWPadnos> maybe that's why the search is taking so long
[13:24:05] <jepler> bbl
[13:24:08] <SWPadnos> hmmm. PIA is Porto Alegre :)
[13:24:23] <jepler> you mean, POA is
[13:24:31] <SWPadnos> yes, that's what I meant
[13:24:40] <SWPadnos> ~caffeine
[13:27:13] <SWPadnos> hmmm. $795 from here, one stop via chicago
[13:27:41] <SWPadnos> or less if I want to go to Pittsburgh and Washington DC
[13:28:18] <skunkworks_> * skunkworks_ is glad he is reletivly close.
[13:28:23] <SWPadnos> heh
[13:29:02] <skunkworks_> relatively
[13:29:21] <lerman> Hey I can go for $500 round trip. With a layover of 11 hrs 33 mins in Atlanta on the way back. Thank you Orbitz.
[13:30:24] <lerman> Hey here's another flight 11 hr 45 min layover on the way there and 11 hr 33 min on the way back.
[13:30:53] <lerman> Driving is looking better all the time. Maybe I should check out taking a train.
[13:32:11] <skunkworks_> greyhound.. ;)
[13:34:19] <SWPadnos> Davenport, IA is about the same distance as Peoria, just in the opposite direction
[13:34:40] <SWPadnos> Bloomington IL is 2 hours away (past Peoria)
[13:35:12] <SWPadnos> also Springfield and Dhampaign are within 2-3 hours (looking at Google, I haven't driven those routes in many years)
[13:35:30] <SWPadnos> err - Champaign
[13:35:40] <SWPadnos> and then you're in the Chicago range
[13:35:44] <SWPadnos> (for drive time)
[13:36:06] <SWPadnos> I think the major train route goes to Chicago
[13:36:35] <SWPadnos> (from CT or MA anyway)
[13:39:20] <lerman> $136 round trip by Greyhound.
[13:39:34] <SWPadnos> that's much less expensive than driving
[13:39:34] <lerman> Whoops I must leave you now.
[13:40:00] <SWPadnos> but sleeping in a hotel is much nicer than on a bus ;)
[13:51:29] <skunkworks_> wow geyhound is cheap
[13:51:53] <SWPadnos> do they have fuel surcharges? :)
[13:54:57] <skunkworks_> I was just going by <lerman> $136 round trip by Greyhound.
[13:55:13] <SWPadnos> oh
[13:55:40] <SWPadnos> that's probably less than fuel cost for me to drive, let alone food and lodging on the way
[14:08:52] <jepler> you'd still need food on the way
[14:09:18] <SWPadnos> heh, but there aren't as many opportunities
[14:12:38] <cradek> I bet person-miles per gallon is pretty high on a greyhound
[14:13:00] <SWPadnos> much higher than in my Jeep, that's for sure
[14:14:19] <cradek> we'll easily get 50-60 depending on whose car we take
[14:14:56] <cradek> a greyhound could get 200?
[14:15:42] <SWPadnos> 184.4, apparently
[14:15:47] <SWPadnos> http://www.buses.org/node/737
[14:19:36] <cradek> maybe I should try taking one somewhere. I wonder how they are.
[14:19:58] <SWPadnos> not great, last time I took one
[14:20:13] <SWPadnos> for a short trip (12 hours or less), it probably wouldn't suck much
[14:21:04] <SWPadnos> my data is 20 years old though, and I know there are nice improvements to charter buses, like in-seat video systems and the like
[14:22:24] <cradek> more important than videos to me would be leg room and the ability to sleep comfortably
[14:23:08] <SWPadnos> oh. well it would be more comfortable than a car or many airplanes, but less space than a train (even in the US)
[14:23:41] <skunkworks_> hmm - train. Still want to try that.
[14:24:31] <SWPadnos> just make sure you sit in an area surrounded by adults only, with no empty seats
[14:24:50] <SWPadnos> empty seats == wailing babies some time later (ask me how I know this)
[14:24:59] <cradek> ugh
[14:25:09] <SWPadnos> 36 hours of ugh
[14:25:37] <SWPadnos> correction: only around 28 hours of ugh, but it was after 8 hours or so of relative peace
[14:29:26] <skunkworks_> that is when you start acting like you're really sick.
[14:30:03] <SWPadnos> and kill the baby?
[14:30:08] <SWPadnos> or the parents
[14:30:18] <skunkworks_> 'wow - I didn't know that kind of stuff would come out of my nose' <- things to say out loud to yourself.
[14:34:29] <skunkworks_> both?
[14:34:31] <skunkworks_> ;)
[14:35:27] <SWPadnos> actually, removing the parents would have been the correct choice in my particular case
[14:39:18] <skunkworks_> we are actually going to see talk of the nation - science friday this month.. it is going to be in millwaukee.
[14:41:11] <skunkworks_> Ira flatow is the bomb.. :)
[14:41:22] <SWPadnos> cool
[14:41:47] <SWPadnos> oh hmmm. I should get tickets to "A Prairie Home Companion" - they'll be at the fair this year
[14:41:57] <SWPadnos> I probably missed out on Elton John tickets :(
[14:42:25] <SWPadnos> but I may be able to hear him from the back porch, so it might not matter
[14:43:21] <skunkworks_> we went to that this spring. it was in winona. It was awesome.
[14:43:30] <skunkworks_> (PHC)
[14:43:45] <tom1> re: galesburg transport from ohare... amtrak makes you go to Naperville to get on train (google amtrak then add 40$ taxi from ohare each way).
[14:43:45] <tom1> Trailways Bus makes you go downtown Chicago (figure 5$ 'EL' train or 40$ taxis each way plus 38.50 bud fare)
[14:43:54] <SWPadnos> yeah. we went to see Garrisson Keillor here, but it wasn't for a PHC show
[14:45:54] <skunkworks_> bud fare? beer?
[14:45:59] <skunkworks_> ;)
[14:46:10] <SWPadnos> small plants
[14:47:51] <skunkworks_> it was fun before the phc show. Garrisson would walk thru the audience helping people to thier seats. (with his microphone)
[14:47:59] <SWPadnos> heh
[14:48:24] <SWPadnos> he didn't do that here. it was an outside show, and it ended up raining or sprinkling most of the time :)
[14:49:41] <skunkworks_> yeck - this was in a huge gymnasium at winona state.
[14:50:12] <SWPadnos> drier at least, though the outdoor bandstand was surrounded by nice views
[14:50:27] <skunkworks_> I didn't believe it was live. But it was.
[14:52:19] <SWPadnos> oh interesting. Feist will be playing there in July
[14:52:31] <tom1> yeah, greyhound is the cheap way, as little as 29$, and you can begin bus trip from the 'Cumberland Avenue" station, only 2 stops from Ohare on the 'EL'. http://www.yourcta.com/maps/rail/Rail.html ( look top left for airport symbol on blue line, then follow it 2 dots toward city)
[14:52:33] <SWPadnos> I was looking for photos of the green at Shelburne Museum, and that popped up :)
[14:52:39] <skunkworks_> Saw her in madison :) very good concert
[14:53:02] <skunkworks_> She is impressive. She does the background looping live on stage.
[14:53:13] <SWPadnos> interesting
[14:54:28] <SWPadnos> hmmm, looks like a good series this year: Feist, Mellissa Etheridge, John Fogerty, and Rober Cray
[14:54:33] <SWPadnos> Robert
[14:54:56] <skunkworks_> I think I like her mainly (other than she is really good) because all her songs don't have to do with how bad life is and how everything sucks. ;)
[14:55:19] <SWPadnos> with a couple others thrown in at other venues: Elton John and the Black Crowes
[14:55:27] <skunkworks_> Neat.
[14:55:50] <skunkworks_> I don't know what is playing around here.. but it is usually country music. :(
[14:55:51] <SWPadnos> I wonder who'll be at the fair this year
[14:57:18] <SWPadnos> hmmm. mostly people I don't know or don't like (other than PHC)
[15:09:45] <tom1> libsixdof0.0.2 a library for 6 degreee of freedom devices (sounds like input devices, haptic? )
[15:09:47] <tom1> http://freshmeat.net/projects/libsixdof/?branch_id=74219&release_id=277071
[15:31:06] <SWPadnos> tom1, looks like a wrapper around devices like the Spaceball/SpaceNavigator and applications that might want to use them
[15:31:25] <SWPadnos> or other devices, like the cyberpuck, wii controller, etc.
[16:24:38] <tomp2> BigJohnT: some notes for you on #cam about gcnccam
[16:24:57] <BigJohnT> I was just typing at you over there
[16:25:58] <BigJohnT> tomp2: thanks
[16:28:03] <jymm> SWPadnos: You're going to control your mill usiang a Wii controller?! COOL!!!
[16:34:56] <chrispol> Hey guys, got emc running peachy, the only issue is my z axis max travel is 60ipm when i have jog speed set to 300 for my x&y z tries to jog at 300??
[16:36:41] <skunkworks_> how did you setup emc? stepconf or editing the ini file directly?
[16:41:36] <skunkworks_> and if you did edit the ini directly - did you enter the max_velocity as in/sec and not in/min? should be 1 (in/sec) for 60in/min.
[16:42:31] <skunkworks_> for axis_2 (z axis)
[16:43:16] <chrispol> ini file ditrectly
[16:43:35] <chrispol> it's set to 1 max valocity
[16:44:09] <chrispol> stepgen max accel/max accel is set to 10
[16:47:50] <jepler> the GUI sets the top limit of "Jog Speed" to the fastest axis (300 in/min in this case)
[16:48:24] <jepler> if you set the slider to 300 and then jog an axis whose max speed is lower (e.g., 60 in/min) the max jog speed will be the axis limit
[16:48:37] <jepler> I set up a configuration similar to what you described using stepconf
[16:49:10] <jepler> I set the slider to 300, turned on "show velocity" in the View menu, then held the "Page Down" key to move "Z". During the movie, it showed "Vel: 60.0000"
[16:49:37] <jepler> what is it you're seeing that makes you think it's trying to move at 300 and not at 60 when you are jogging Z?
[16:49:40] <cradek> I can't tell if you're saying the gui's jog speed slider is different from how you expect, or if the velocity display is wrong, or if the machine motion is wrong, or something else. can you describe what you're actually seeing that's wrong? I see that everyone is guessing (differently) what the problem is.
[16:50:36] <cradek> also, putting your ini on pastebin (http://www.pastebin.ca) could let us see if something is wrong with it
[16:57:26] <chrispol> i get following errors when above 60
[16:57:56] <chrispol> the machine is currently cutting will check the velocity once it's done
[17:00:58] <gefink> 60 feederoverride?
[17:19:21] <skunkworks_> chrispol: what version of emc?
[17:31:24] <chrispol> 2.2.4
[17:31:51] <chrispol> ok so jogging set to 60 velocity = 60 jogging set to 100 velocity of 100 and i can definatly hear it moving faster
[17:32:53] <chrispol> http://pastebin.com/m553d042a
[17:35:33] <skunkworks_> STEPGEN_MAXACCEL = 10
[17:35:51] <skunkworks_> that needs to be 20% or so higher than MAX_ACCELERATION = 10
[17:36:11] <skunkworks_> that is most likly your problem
[17:36:58] <skunkworks_> all your axis need to have stepgen_maxaccel headroom
[17:37:27] <chrispol> ok will try
[17:38:01] <skunkworks_> so - make it 12 or so and 48 for the other 2 axis
[17:38:27] <skunkworks_> what version of emc are you running?
[17:38:30] <skunkworks_> sorry
[17:38:34] <skunkworks_> just read it.
[17:39:24] <chrispol> yep cool works :)
[17:50:32] <skunkworks_> Good :)
[17:52:54] <skunkworks_> in the normal ini file - there is text that says http://pastebin.ca/1008457
[17:54:21] <skunkworks_> actually - missed a line
[17:54:23] <skunkworks_> http://pastebin.ca/1008459
[19:20:04] <alex_joni> quiet today
[19:20:46] <skunkworks_> Hi alex
[19:21:05] <K`zan> Contemplating where to go from here....
[19:21:32] <alex_joni> K`zan: that is a big step :)
[19:21:37] <alex_joni> hey samco
[19:22:00] <K`zan> Yep, now I know everything (heh) that went wrong / didn't work out. Just what to do is becoming clearer
[19:23:17] <K`zan> These heavier steppers are a bit of a problem for the Y transversal as it causes the X table to tilt, bracket with a bearing under the "ways" should take care of that.
[19:23:49] <fenn> flexing under its own weight isn't a particularly good sign
[19:23:52] <K`zan> Z axis mount is too compact so I'll expand that by, at least, the size of the coupler to the lead screw.
[19:24:08] <K`zan> Yep, it is a pitiful little thing, but educational :).
[19:24:54] <K`zan> And it is something to do that will give me a PCB / engraving machine of some reasonable accuracy for the task.
[19:25:21] <fenn> how are you going to take care of backlash in the screw?
[19:25:22] <K`zan> And it just might allow me to (Long run) form the stepper mounts for the uMill.
[19:25:51] <K`zan> 2 nuts and spring if just a joining nut is too far out.
[19:26:03] <K`zan> coupling nut
[19:26:55] <K`zan> I don't expect a lot out of it but it is something that I can do with the tools I have :).
[19:27:25] <K`zan> And it is fun if occasionally frustrating. Not going to be doing anything for the aerospace industry with it, I am SURE :-).
[19:27:37] <fenn> you'd be surprised what's possible with a small set of tools
[19:27:48] <K`zan> Yeah, but I am impatient :).
[19:28:18] <K`zan> With a piece of flint I could probably do this if I was willing to make it a life work...
[19:28:43] <K`zan> My rational is that anything will inspire more / better.
[19:28:53] <fenn> as long as you finish each step
[19:29:01] <K`zan> Rather than sitting and waiting to acquire something better.
[19:29:06] <K`zan> Yep, true.
[19:30:53] <K`zan> I did get a better drive and steppers than the project calls for though.
[19:38:01] <K`zan> FOOD! bbl
[19:57:17] <alex_joni> yumm.. freshly baked croissants
[20:04:11] <tomp2> gimpels? email me one please
[20:04:12] <rayh> found a link to a winders based control. http://www.hoffhilk.net/cgi-bin/mnlist.cgi?hoffhilk21/16/1
[20:04:49] <tomp2> is that a disconnect switch or a hand crank :)
[20:05:52] <tomp2> looks like someone got pissed at a touch screen
[20:06:40] <alex_joni> don't think that's a touchscreen
[20:07:02] <alex_joni> well.. not in the mouse-pointy-thingy kind of way anyways
[20:07:05] <tomp2> :) but it got touched
[20:07:09] <tomp2> hard
[20:07:15] <alex_joni> tomp2: I can see that
[20:07:22] <alex_joni> maybe it's a punchscreen ?
[20:07:39] <tomp2> http://content.honeywell.com/sensing/datainstruments/wintriss/wintriss_smartpac.stm
[20:08:11] <alex_joni> scratchpad?
[20:08:18] <tomp2> heh\
[20:08:22] <alex_joni> maybe they confused it for a scratchscreen
[20:08:45] <jepler> rayh: any particular reason we're looking at this system?
[20:08:48] <tomp2> winders/wintriss
[20:09:08] <cradek> looks like someone stole buttons and switches off of it already. that's what I would have used it for too.
[20:09:19] <rayh> Just thought you'd like to see a modern style control cabnet.
[20:09:58] <alex_joni> define modern?
[20:09:58] <tomp2> nice app for hal, how hard it hits, dwell, press control
[20:10:02] <rayh> They've got a couple of nice welding robots in that list.
[20:10:17] <alex_joni> post-transister modern? ;-)
[20:12:40] <alex_joni> now this is a useful interface: http://mustang.hoffhilk.com/maxanet/hermanmiller3/DSC06191.jpg
[20:18:34] <jymm> alex_joni: Not made from straw and mud or from rough cut timber product
[20:19:04] <alex_joni> * alex_joni likes rough cut timber
[20:19:39] <jymm> alex_joni: Great! I'll send you all mine, you plane it smooth and send it back
[20:20:55] <alex_joni> lol.. http://www.theregister.co.uk/2008/05/01/dodgy_cheque/
[20:25:24] <cradek> ... while carrying his weed
[20:26:24] <assargadon> is emc_dev (with prerequesties) hold all needed for HAL components development?
[20:26:54] <cradek> emc2-dev
[20:26:58] <cradek> and yes I think so
[20:27:08] <cradek> This package includes files needed to build new realtime components and
[20:27:09] <cradek> alternate front-ends for emc2
[20:27:42] <assargadon> there are no "all" word, isn't it?
[20:27:55] <cradek> ?
[20:27:58] <gefink> assargadon: yes i tried with comp. All needed packages got auomatic installed
[20:28:06] <alex_joni> there is no "some" either
[20:28:16] <assargadon> :)
[20:28:39] <assargadon> "you never know with bees" (c) Vinni-the-Pooh
[20:29:43] <assargadon> As far as I understand, there are no possibility for combining "simple" HAL components to larger ones?
[20:30:21] <assargadon> is it correct?
[20:30:30] <cradek> HAL is all about building what you need out of simpler parts. But maybe you should give a concrete example, since it is not clear what you are asking.
[20:31:29] <assargadon> cradek, as the exmaple...hm...
[20:31:43] <alex_joni> assargadon: you can combine things at run-time, not at compile-time
[20:32:04] <assargadon> I going to describe pulse_counter component
[20:32:42] <gefink> assargadon: you can only build own complex realtime-objekts
[20:32:47] <assargadon> it should work as "normal" one, but inputs are "pulse" and "direction"
[20:33:12] <assargadon> inspead of pulseA and pulseB
[20:33:19] <assargadon> *instead
[20:33:31] <assargadon> it's native for every step counter
[20:33:42] <assargadon> to make mesurement in steps, isn't it?
[20:34:15] <assargadon> and, as far as we need float numbers
[20:34:55] <assargadon> some "sclale" component for converting this integer "steps" to float "coordinate"
[20:35:14] <assargadon> using the current scale, of course
[20:35:37] <assargadon> it's natural thing then
[20:36:21] <assargadon> to bild larger component, which consist of "pulse_counter" and "scale" subcomponents
[20:37:01] <assargadon> we call it "decomposition"
[20:37:03] <gefink> with the comp tool its easey to create such a counter
[20:37:39] <gefink> but its not possible to count shorter pulses
[20:37:43] <assargadon> yes. _It_ is simple
[20:38:18] <gefink> as 2 callingfrequence
[20:39:06] <gefink> au if you use 25us clocktime the Pulse must be longer as 25 usec
[20:39:33] <gefink> think thats verry low
[20:40:13] <assargadon> gefink, I do not understand your criticizm :)
[20:40:27] <assargadon> do you unhappy with decomposition ideology
[20:40:44] <assargadon> or with this encoder project?
[20:40:51] <gefink> its not the decomposition
[20:41:16] <gefink> its the problem to detecht clockpulses without hardware
[20:42:12] <assargadon> gefink, do you mean decomposition same to one we use with hardware can hardly be implemented with software
[20:42:22] <gefink> say if the maximum required inputrate is lower 20 khz all is ok
[20:42:30] <assargadon> becouse of problems of sinhronisation?
[20:42:43] <assargadon> *syncronisation
[20:42:57] <gefink> because of sampling
[20:43:53] <gefink> realtimecomponents are called periodical. to detect a inputpulse you need 2 samples
[20:44:06] <gefink> 1 sample need 25usec....
[20:44:33] <assargadon> gefink, do you talking especially about my pulse_encoder?
[20:44:42] <gefink> yes
[20:45:14] <assargadon> gefink, "step" is not real "pulse"
[20:45:16] <gefink> its a good idea to build own components if it helps to solve your problem
[20:45:21] <assargadon> it's change of input
[20:45:30] <assargadon> from 1 to 0 or from 0 to 1
[20:45:41] <gefink> howfast?
[20:45:48] <gefink> and how often
[20:46:44] <assargadon> gefing, rarelyenought, as far as I can detect blinkings by my eyes (there are leds for encoder on controller box)
[20:46:56] <assargadon> but it's not really a question
[20:47:15] <assargadon> becouse of if my ciunter is not applicabel
[20:47:22] <assargadon> *not applicable
[20:47:33] <assargadon> *encoder
[20:47:36] <assargadon> :)
[20:47:55] <assargadon> if my encoder is not applicable
[20:48:08] <assargadon> normal one is not applicabel, too - isn't it?
[20:49:09] <alex_joni> good night all
[20:49:21] <assargadon> good night alex
[20:49:31] <gefink> good night
[20:50:26] <gefink> assargadon: yes. im thinking about 2 counters and sum
[20:52:10] <assargadon> gefink, I do not understand
[20:53:01] <assargadon> firstly, is there ANY encoder which works faster than 20khz?
[20:53:04] <gefink> assargadon: you can use one counter for up and one for down and subtract one from the other
[20:53:29] <gefink> assargadon: unly the encoder
[20:53:55] <assargadon> gefink, yes I can use 2 counters, but how it helps?
[20:54:26] <gefink> assargadon: it dont help for speed, but solfes the step-dir problem
[20:55:35] <assargadon> gefink, what is the problem with step-dir? Do you mean "how to solve this problem with standart components without making new ones" ?
[20:57:13] <gefink> there is no problem. both work. standard like custom. Bat quadrature is 2time faster then step-dir
[20:57:34] <assargadon> gefink, ah. Now I get the point.
[20:57:45] <assargadon> Yes, I know it perfectly
[20:57:59] <assargadon> thet's even described in robot's manual
[20:58:22] <assargadon> unfortunately, there are no quadrature counter
[20:58:28] <assargadon> only pulse one
[20:58:35] <alex_joni> sure there is.. it's called encoder
[20:58:58] <gefink> encoder is a quadraturcounter
[21:00:02] <assargadon> gefink, this manual is wreitten on Russian
[21:00:19] <assargadon> so therms you hear from me is just my translation
[21:00:24] <gefink> tried "man encoder"
[21:01:13] <assargadon> yes, I know the way classical encoder works
[21:01:29] <assargadon> but I have no classical encoder
[21:01:50] <gefink> yes. so you cant use encoder
[21:02:51] <assargadon> yes, and I need some component to use. I call this component pulse_encoder
[21:02:58] <assargadon> insmod: error inserting '/usr/realtime-2.6.15-magma/modules/emc2/pulse_encoder.ko': -1 Unknown symbol in module
[21:03:00] <assargadon> <stdin>:3: exit value: 1
[21:03:02] <assargadon> <stdin>:3: insmod failed, returned -1
[21:03:25] <assargadon> hmmm....
[21:03:38] <jepler> look in "dmesg" to find out what the undefined symbol is
[21:04:11] <gefink> use halcmd to load the modules
[21:04:44] <assargadon> Unknown symbol extra_setup...
[21:04:46] <assargadon> hmmm
[21:04:53] <assargadon> thanks, jepler
[21:05:10] <assargadon> problem is I just copy the example from the manual :)
[21:05:42] <jepler> hm I have no idea why "option extra_setup yes" is in the 'constant' example
[21:05:44] <assargadon> gefink, instead of hulrin you mean?
[21:05:47] <jepler> you should remove that line
[21:06:01] <assargadon> jepler, yes, I think so too :)
[21:06:08] <assargadon> And I know why it there
[21:06:17] <jepler> in 'constant'?
[21:06:22] <assargadon> to prevent user from too easy results :)
[21:06:29] <gefink> halrun is ok
[21:07:09] <gefink> i thought you call insmod manual
[21:07:09] <jepler> (it is rightfully in out8)
[21:08:11] <assargadon> ok, it works. Greate.
[21:08:37] <jepler> I corrected the documentation. thank you for reporting the problem.
[21:10:13] <assargadon> you are welcome :)
[21:10:50] <assargadon> * assargadon feels himself Greate Betatester :)
[21:12:38] <assargadon> I trying to get the point of floats using for coordinate mesuring
[21:12:56] <assargadon> *measuring
[21:13:10] <assargadon> I feel floats as analogue signals
[21:17:16] <assargadon> why difference between fp an nofp functions needed?
[21:20:46] <gefink> assargadon: its for speed
[21:22:30] <tomp2> assargadon: it is not really a float to the system, it is an intger count of how many small distances, if the small distance is .001mm then it is displayed as a float. this is display versus internal counting.
[21:23:18] <gefink> therefore floatingpoint calculation is most don in the 1ms task
[21:24:33] <gefink> the raw s32 counter run in the 25us task
[21:24:48] <tomp2> some functions use floats, like sine, cosine, atan and thus need fp
[21:30:11] <K`zan> Sigh, space is really tight in there, I think I am going to see if that old 266MHz laptop will run EMC.
[21:31:28] <skunkworks_> double wammie.. 266mhz and laptop.. :)
[21:31:45] <K`zan> Sigh, yeag
[21:31:47] <K`zan> yeah
[21:33:46] <gefink> assargadon: knob2float is a block which can make the scaling and floatingpoint for you
[21:36:11] <micges> hello
[21:37:33] <gefink> micges: hello
[21:39:50] <gefink> by
[22:14:00] <tomp2> do i need the inductance of a brake coil to size the blocking diode?