#emc | Logs for 2008-06-12

[00:46:53] <cradek> I'm more interested in running mine slower than normal
[00:47:07] <cradek> so I can stay in high gear a little longer
[01:00:22] <skunkworks> spindle?
[01:00:36] <cradek> yeah
[01:00:59] <skunkworks> so.. when is the conversion?
[01:01:01] <skunkworks> ;)
[01:01:24] <cradek> now that I have the probe mount done, I'm getting more antsy about it
[01:01:57] <cradek> it turned out very nicely. I was able to center it right up.
[01:02:39] <cradek> I used an 8-32 allthread "drawbar" which is flexible/stretchy enough to allow the adjustment screws to work
[01:03:59] <skunkworks> neat
[01:04:04] <cradek> the world is trying to kill us again... dangit
[01:04:10] <cradek> hope it fails again
[01:04:24] <skunkworks> ? weather?
[01:04:28] <cradek> yeah
[01:04:35] <skunkworks> yah - same here
[01:04:40] <cradek> tornados tearing stuff up every couple days lately
[01:05:48] <skunkworks> no tornados - just lots of flooding
[01:06:00] <skunkworks> anything close?
[01:06:30] <cradek> no, the closest houses successfully destroyed are 50 miles away so far
[01:06:38] <cradek> but it keeps trying
[01:06:55] <skunkworks> 50 miles seems close
[01:07:17] <cradek> yeah. that storm went right over us - we watched it
[01:07:34] <cradek> "look, all the clouds are rotating and smashing into one another"
[01:07:42] <cradek> fortunately (for us) it moved on in time
[01:11:34] <cradek> bbl
[01:16:51] <skunkworks> we had a really cool full rainbow.
[01:17:28] <skunkworks> double
[01:39:58] <jmkasunich> cradek: yuck
[02:17:26] <jepler> I'm still here too
[02:23:28] <SWPadnos> jepler, yuck
[02:27:28] <PMinMO> hi folks, new Hal but I still have issues. Couple of wrong seek directions for home and reset disables motors, but machine on and off doesn'thttp://www.pastebin.ca/1045650
[02:28:28] <PMinMO> EMC is really smooth with motors
[02:34:03] <jepler> PMinMO: to change the direction of homing, either invert the pin or negate the home search velocity. but if your switch is in the middle of travel, so that it's "open -- closed -- open", then you simply have to start homing on a known side of the switch. (that kind of configuration is discouraged)
[02:34:51] <PMinMO> home is end of travel, NC switches for logic low inputs
[02:34:54] <jepler> PMinMO: you may want to choose the "Amplifier Enable" signal instead of the "ESTOP out" signal to turn your amplifiers on and off. amplifier enable should be false any time estop out is true
[02:35:09] <jepler> that should make the amps turn off in "machine off" mode
[02:36:41] <jmkasunich> right - you are using axis.n.amp-enable-out to enable the stepgens, you should use the same signal to enable the actual drives
[02:37:01] <PMinMO> this is from stepconf
[02:37:18] <PMinMO> So it seems closer than with what I started
[02:37:49] <PMinMO> So change net estop-out => parport.0.pin-14-out
[02:37:55] <jmkasunich> stepconf gave you a starting point, now you can tweak it till it des what you want
[02:38:29] <PMinMO> to net Amplifier Enable => Paraport.0.pin-14-out
[02:39:06] <jmkasunich> "net xenable => parport.0.pin-14-out"
[02:39:52] <jmkasunich> note: hal pin names are case sensitive, using mixed case is allowed, but can be confusing
[02:39:57] <jmkasunich> spaces are NOT allowed
[02:40:38] <jmkasunich> so "net Amplifer Enable => etc" will create a net called Amplifier and try to connect it to a pin called Enable (which doesn't exist)
[02:41:02] <jmkasunich> also, "parport", not "Paraport"
[02:41:53] <PMinMO> line 35 vs line 52 which one are you referring too?
[02:42:13] <jmkasunich> I'm mostly referring to what you typed here in the IRC window
[02:42:59] <jmkasunich> in the hal file, you have a couple choices
[02:43:09] <PMinMO> line 35 is net estop-out => parport.0.pin-14-out
[02:43:29] <PMinMO> line 52 is net xenable axis.0.amp-enable-out => stepgen.0.enable
[02:43:45] <jmkasunich> you could delete line 35 and change line 52 to "net xenable axis.0.amp-enable-out => stepgen.0.enable parport.0.pin-14-out"
[02:43:57] <jmkasunich> yes, I know, I'm looking right at it
[02:44:42] <jmkasunich> your other choice is to leave line 52 alone, and change line 35 to "net xenable parport.0.pin-14-out"
[02:45:05] <jmkasunich> both approaches do the same thing
[02:45:40] <jmkasunich> they create a signal called "xenable", and connect it to three things: axis.0.amp-enable-out, stepgen.0.enable, and parport.0.pin-14-out
[02:45:50] <SWPadnos> or you can choose "Amplifier Enable" for pin 14 in stepconf
[02:46:01] <jmkasunich> no you can't
[02:46:10] <SWPadnos> it shows up in my version of stepconf
[02:46:10] <jmkasunich> no spaces in pin names
[02:46:18] <jmkasunich> oh, in stepconf
[02:46:21] <SWPadnos> that's a text display string
[02:46:30] <jmkasunich> fsck that - he has a hal file that mosty works
[02:46:47] <SWPadnos> it won't be editable by stepconf if it's modified
[02:46:56] <jmkasunich> no comment
[02:47:03] <PMinMO> :-)
[02:47:14] <SWPadnos> you can add the `net xenable parport.0.pin-14-out` to custom.hal though
[02:47:24] <jmkasunich> giving fish
[02:47:44] <SWPadnos> heh
[02:47:50] <PMinMO> for the time being ill change line 35 to net xenable => parport.0.pin-14-out
[02:47:52] <SWPadnos> stepconf is a barrel of fish
[02:48:28] <JymmmEMC> * JymmmEMC shoots the barrel
[02:48:47] <PMinMO> stepconfig was a huge help, I still haven't connected the hal "lingo"
[02:49:23] <jmkasunich> I understand - my comment was regarding the old proverb about "give a man a fish and he eats for a day, teach him to fish and he eats forever"
[02:49:35] <SWPadnos> I got that :)
[02:49:39] <jmkasunich> stepconf gets people going, but it also keeps them from learning
[02:49:46] <SWPadnos> stepconf gives out a lot of fish ;)
[02:49:56] <JymmmEMC> jmkasunich: Whaaaaaaaaaaat?! You don't like the way I fish? =)
[02:50:10] <SWPadnos> well I don't
[02:50:24] <SWPadnos> the fish tastes gunpowdery when you shoot it
[02:50:26] <JymmmEMC> * JymmmEMC tosses SWPadnos in the barrel
[02:50:43] <JymmmEMC> Fine....
[02:50:55] <JymmmEMC> * JymmmEMC lites the M80 and tosses it in the barrel
[02:51:05] <JymmmEMC> happy now?
[02:51:07] <PMinMO> next question
[02:51:12] <JymmmEMC> SWPadnos: btw.... RUN!!!!!!!!!!!!!!!
[02:51:17] <PMinMO> inverting input signals
[02:51:22] <SWPadnos> nah, got 7 seconds before it blows
[02:51:34] <SWPadnos> blah-blah.in-not
[02:51:35] <JymmmEMC> negate them?
[02:51:38] <PMinMO> yes
[02:51:43] <SWPadnos> ^^
[02:51:50] <PMinMO> syntax
[02:52:04] <SWPadnos> add `-not` to the input pin name
[02:52:29] <jmkasunich> in this case, the -not is already there
[02:52:34] <PMinMO> ok and outputs
[02:52:38] <PMinMO> ?
[02:52:39] <jmkasunich> (lines 38-40 in the hal file)
[02:52:48] <SWPadnos> set the parameter blahblah.invert to 1
[02:53:06] <SWPadnos> like setp parport.0.pin-14-put-invert 1
[02:53:12] <SWPadnos> uh like setp parport.0.pin-14-out-invert 1
[02:54:25] <PMinMO> so it takes an extra line to define thae hadware pin when its inverted
[02:54:42] <jepler> luckily, even the freeware version of emc doesn't have a limit on the number of .hal lines permitted
[02:55:33] <renesis> wtf at freeware version
[02:55:43] <SWPadnos> uh-oh - I think jepler has been reading non-free software descriptions lately
[02:55:46] <PMinMO> is there a hal quick reference guide
[02:55:47] <renesis> * renesis shakes the shit outta jepler
[02:56:31] <SWPadnos> if you have an installed EMC2, there's an integrator manual in the CNC menu, which has a HAL section or two
[02:56:42] <SWPadnos> I'm not sure if it's called the integrator manual any more though
[02:56:50] <jepler> 'night all
[02:56:57] <jepler> sorry if the joke wasn't funny, renesis
[02:56:58] <SWPadnos> see ya
[02:58:01] <PMinMO> sorry I'm full of questions, but I'd sure like to get EMC running
[02:58:05] <jmkasunich> http://linuxcnc.org/docview/html/
[02:58:19] <jmkasunich> there is a hal section, including a tutorial
[02:59:48] <PMinMO> tnx
[03:00:09] <PMinMO> anybody here been to prior cnc workshops ?
[03:00:16] <SWPadnos> yes
[03:00:26] <cradek> several
[03:00:32] <jmkasunich> yes
[03:00:49] <PMinMO> I'm debating on the full week vs 2 days
[03:01:13] <jmkasunich> all three of us have been to two workshops before, and are doing all week this year
[03:01:40] <PMinMO> but I suspect you are doing code development?
[03:01:48] <PMinMO> or emc colaboration?
[03:01:49] <jmkasunich> we're doing all kinds of stuff
[03:02:14] <jmkasunich> we'll be able to help people do hands on work
[03:02:43] <jmkasunich> last year we even went to Peoria (about an hours ride) to help a guy troubleshoot a machine that was too big to bring to the workshop
[03:02:49] <cradek> wow there will be a lot of new emc-related folks there this year. that's great.
[03:04:10] <PMinMO> I had planned on Friday/Saturday but seeing the workshop schedule made me rethink
[03:04:18] <SWPadnos> I wouldn't mind finding that Mongolian Grill again :)
[03:04:29] <jmkasunich> it gets really busy fri/sat
[03:04:40] <JymmmEMC> * JymmmEMC LOL @ SWPadnos
[03:04:57] <JymmmEMC> SWPadnos <---- That boy can serioualy EAT!!!
[03:05:08] <SWPadnos> and it was all you can eat, I think
[03:05:32] <JymmmEMC> SWPadnos: You sure yo dind't put them out of business the last time you were there?
[03:05:56] <SWPadnos> I don't know - haven't been back to the scene of the crime since then ;)
[03:06:13] <JymmmEMC> lol, ok fair enough
[03:06:25] <PMinMO> so how many people were at last years?
[03:07:00] <SWPadnos> dunno the total, probably in the 150-200 range
[03:07:11] <PMinMO> for the whole week?
[03:07:19] <SWPadnos> total over the week, not necessarily all at the same time
[03:07:48] <SWPadnos> (I base that on the fact that the prior year was ~135, and I think there were more people last year than the year before)
[03:08:14] <PMinMO> any specific recommendation on place to stay?
[03:08:28] <SWPadnos> two classes of hotel/motel in Galesburg
[03:08:55] <SWPadnos> fleabag motels and $90/night chain hotels (like Fairfield Inn and Country Suites)
[03:09:11] <PMinMO> I'm looking for cheap, but not roach infested
[03:09:19] <SWPadnos> well, not really fleabag, but not all that nice :)
[03:09:20] <JymmmEMC> * JymmmEMC hands out flea collars to all thw workshop attenties
[03:10:03] <SWPadnos> hmmm - JymmmEMC, could you ship my PZ3 and the ferrules to Cardinal Engineering?
[03:10:26] <JymmmEMC> SWPadnos: Provide proof that you completed a project already.
[03:10:31] <SWPadnos> to me, C/O Roland Freistadt
[03:10:40] <SWPadnos> don't make me zap you with the 20000A power supply
[03:11:04] <JymmmEMC> Recent project, like within the last 90 days
[03:11:05] <jmkasunich> PMinMO: the "fleabag" hotels aren't that bad
[03:11:15] <jmkasunich> I guess it depends on what you are used to
[03:11:25] <jmkasunich> SWPadnos likes to go first class
[03:11:28] <SWPadnos> heh
[03:11:35] <SWPadnos> or at least business classs
[03:11:37] <SWPadnos> sss
[03:11:38] <jmkasunich> I'll be in one of the fleabag places
[03:11:43] <JymmmEMC> I should post pics of his last hotel
[03:12:00] <JymmmEMC> arboretum, flowing pond/stream
[03:12:02] <SWPadnos> mailing address is here: http://www.cnc-workshop.com/costs.htm
[03:12:06] <SWPadnos> heh
[03:12:19] <SWPadnos> you should see where we're staying in Paris
[03:12:19] <JymmmEMC> SWPadnos: I'm still waiting for proof
[03:12:39] <SWPadnos> I'd have to kill you, so there's a race condition as to whether you can get the stuff in the mail before you're eliminated
[03:13:04] <JymmmEMC> SWPadnos: Fine, I'll take em to my grave... and be cremated
[03:13:05] <PMinMO> is the Vista a real dive?
[03:13:37] <jmkasunich> I think its probably the diviest of the three fleabags, but I have no firsthand experience
[03:13:49] <jmkasunich> I've stayed at the Econo Inn and America Best Value
[03:13:51] <SWPadnos> rumor has it that it's about the same as the Econo Inn and the Americas Best Value Inn, but it's nearly on top of the railroad tracks
[03:13:58] <jmkasunich> staying at the ABV this year
[03:14:06] <JymmmEMC> google satellite photos might help
[03:14:19] <SWPadnos> then again, I think ABVI also does double-duty as a coal stop or something :)
[03:15:28] <jtr> wasn't there a quiet side and a not-so-quiet side to one of them?
[03:15:54] <SWPadnos> I actually didn't mind the trains at the ABVI last year - I was on the parking lot side
[03:19:01] <jtr> Thought I remembered hearing about that. That was for PMinMO's benefit anyway - I'll be working in El Paso that week.
[03:22:36] <PMinMO> whats the difference between stand alone HAL and EMC HAL
[03:22:46] <PMinMO> I see two different classes
[03:24:47] <SWPadnos> HAL can be used standalone, but at this point is only distributed with EMC2
[03:25:28] <SWPadnos> I have an application that uses a "bare" HAL to control a large power supply
[03:27:17] <PMinMO> ok tnx
[03:28:29] <PMinMO> got to go, tnx fo rthe help tonight
[03:28:52] <SWPadnos> sure
[05:03:12] <gezr> well, the bike started, after roughly 14 years sitting in pieces
[05:13:54] <Zaarq> gezr: What kind of bike?
[06:39:28] <micges> good morning
[12:41:55] <skunkworks_> wow - it is raining here still.
[12:42:45] <archivist> you moved to Wales?
[13:06:36] <skunkworks_> heh - no
[13:07:18] <archivist> it allways rains in Wales when I visit
[13:07:47] <SWPadnos> I think a lot of people would leave off the "when I visit" from that sentence
[13:10:28] <archivist> I went a couple of weeks ago, annual trip to THE secondhand book shop town Hay on Wye
[13:11:38] <SWPadnos> which part of the country is that in?
[13:12:13] <archivist> south Wales near the english border
[13:12:17] <skunkworks_> neat.
[13:12:17] <skunkworks_> http://www.youtube.com/watch?v=quN37YskoaM&eurl=http://blog.makezine.com/
[13:12:32] <SWPadnos> ok - I've only been to the northern part (I think)
[13:12:50] <SWPadnos> we switched trains in Betwys-y-coed, went to some national park
[13:13:15] <archivist> SWPadnos, well worth a vist if you like books
[13:13:29] <SWPadnos> well, I do like books :)
[13:13:31] <archivist> ah thats further north
[13:13:48] <archivist> books are a bad habit
[13:14:13] <archivist> I would live there if I had the money
[13:14:42] <SWPadnos> there's an advantage to stores like Barnes & Noble (Or Waterstones there?)
[13:15:00] <SWPadnos> you can grab a coffee and read, and not feel bad about not buying the book :)
[13:15:19] <archivist> that new expensive books though
[13:15:36] <SWPadnos> that's why you buy the coffee instead ;)
[13:16:07] <archivist> I do sometimes go to Waterstones or Blackwells (Oxford)
[13:16:11] <skunkworks_> http://www.wkbt.com/
[13:17:50] <SWPadnos> yeah - weird weather this past week
[13:17:55] <SWPadnos> 90+ for a few days
[13:18:11] <SWPadnos> light frost in some places last night
[13:18:33] <SWPadnos> loads of rain, some tornadoes, big thunderstorms ...
[13:30:01] <skunkworks_> I wonder how low rolands shop is
[13:35:58] <Sweeper> woohoo, first test of my stepper driver and it worked :D
[13:39:21] <SWPadnos> skunkworks_, they've only gotten about 0.57" of rain in the last week. it should be OK unless something esle comes in
[13:53:02] <skunkworks_> That is good :)
[14:00:19] <renesis> urmom is 0.57" of rain!
[14:00:25] <renesis> er, i mean, hi
[14:04:21] <Sweeper> hmm. even these 5w resistors are getting pretty hot D:
[14:04:39] <archivist> 5W is hot
[14:04:39] <Sweeper> and there's two of them!
[14:05:06] <Sweeper> should they be hot? it's just one little stepper :v
[14:05:56] <Sweeper> like, can't-keep-my-finger-on-for-more-than-5-seconds hot
[14:09:15] <renesis> depends on the device
[14:09:38] <renesis> heh, 5W on a small transistor = hot, dying
[14:10:00] <Sweeper> oh, these are RATED for 5w
[14:10:04] <renesis> 5W on a 100W resistor, prob rises like 5C if that
[14:10:12] <renesis> at what temp?
[14:10:20] <renesis> 5W at dying hot?
[14:10:28] <renesis> youre probably killing them =(
[14:10:30] <Sweeper> maybe 50-70C
[14:10:53] <renesis> power rating derates with temp
[14:11:13] <SWPadnos> if you can hold your finger on it for 5 seconds, it's not hot
[14:11:30] <SWPadnos> unless there's an accompanying sizzling sound
[14:11:35] <archivist> hehe
[14:11:49] <Sweeper> no sizzle!
[14:11:57] <renesis> yeah thats an odd time to test that
[14:11:57] <archivist> no worries
[14:12:09] <pjm> afternoon all
[14:12:30] <Sweeper> renesis: well, I was running the motor without any resistors, and it was getting a bit too warm after only a few minutes run time
[14:12:52] <Sweeper> so I rummaged in the parts box for big resistors
[14:13:01] <archivist> hot motors is not a real problem
[14:13:02] <SWPadnos> the motor can probably run to 150C
[14:13:08] <renesis> mexico does amazing things with sand in the 10% range
[14:13:11] <SWPadnos> and you really wouldn't want to touch that
[14:13:13] <renesis> heheh
[14:13:14] <Sweeper> orly?
[14:13:22] <SWPadnos> no, Orly is near Paris
[14:13:39] <renesis> orly?
[14:13:50] <archivist> oh really
[14:14:08] <Sweeper> hmm
[14:14:23] <Sweeper> I get a lot of missed steps at this speed with no resistors tho...
[14:14:28] <renesis> sweeper: do you has a fat diode?
[14:14:56] <renesis> if your driving a motor might as well put a fat rev protection diode across transistor
[14:15:16] <Sweeper> I've got one on each leg
[14:15:22] <Sweeper> err, coil
[15:02:31] <BigJohnT> OH BOY BOB Mills are out...
[15:03:27] <als> BOB Mills?
[15:04:08] <archivist> * archivist wonders what BigJohnT has broken today
[15:04:10] <BigJohnT> from Bobcat
[15:04:18] <BigJohnT> nothing knock on wood
[15:04:32] <BigJohnT> bobcad
[15:04:32] <BigJohnT> hmmm no link
[15:05:14] <als> the skid steer people?
[15:06:02] <BigJohnT> I see you have not downloaded BobCad before LOL
[15:06:18] <BigJohnT> http://www.bobcad.com/allproducts.html
[15:06:26] <archivist> we have not
[15:06:39] <BigJohnT> I got an email about the machine but I can't find a link to it
[15:06:57] <BigJohnT> If you ever download the trial version you will get a zillion phone calls from them
[15:08:10] <archivist> * archivist has no phone at home har har
[15:08:50] <BigJohnT> smart move archivist
[15:09:23] <archivist> also means no intarweb at home :((
[15:09:58] <BigJohnT> http://pastebin.ca/1046045
[15:10:12] <BigJohnT> that is the email I got less the pictures
[15:11:56] <skunkworks_> I get that from cnczone.. Once you try bobcam - they are relentless to make the sale
[15:12:07] <als> nema 42 steppers
[15:12:28] <archivist> BigJohnT, out machines are "ridged"
[15:12:35] <archivist> our
[15:16:10] <BigJohnT> trying to get a pic to image bin
[15:19:55] <cradek> archivist: oh I finally figured out what word that was supposed to be. All I could come up with was "rigged".
[15:20:37] <cradek> and this sentence just makes me shiver: "If your just starting out or currently have cnc machines, the value of our turn key packages are undeniable."
[15:20:48] <archivist> cradek yes, good typo for a sales email though
[15:21:14] <cradek> I hope the machine is made with more care than the spam
[15:21:30] <archivist> makes you really not want their stuff
[15:21:40] <cradek> yeah.
[15:21:45] <als> are they making it?
[15:22:04] <cradek> hard to say
[15:22:33] <archivist> missing from the all products tab on website
[15:22:47] <als> seems like they are in the software bis
[15:24:47] <BigJohnT> I think he meant rigid
[15:25:14] <cradek> I agree - but it took me a while.
[15:26:30] <als> that or they don't break the sharp edges
[15:29:29] <BigJohnT> http://imagebin.ca/view/IzYOyI.html
[15:29:42] <BigJohnT> took me a while to get
[15:30:09] <BigJohnT> * BigJohnT gotta go be back later
[15:59:44] <cradek> haha http://www.smithy.com/uploads/1240_new.jpg
[16:00:16] <SWPadnos> now where have I seen that before?
[16:00:40] <cradek> just like the smithy image, it has an estop button photoshopped on
[16:00:49] <cradek> but, it doesn't have the smithy logo photoshopped on.
[16:02:26] <SWPadnos> maybe it's a fuzzy estop button
[16:02:39] <SWPadnos> :soft-touch"
[16:02:39] <cradek> yeah, could be.
[16:03:25] <SWPadnos> after all, you wouldn't want to worry about breaking a nail or hurting your hand when the machine is about to rip someone's arm off
[16:03:44] <SWPadnos> so a soft, fuzzy "personable" estop button could be a great selling point
[16:04:19] <archivist> * archivist thinks someone photoshopped a label off
[16:05:35] <cradek> archivist: no, I'm pretty sure smithy photoshopped it on
[16:06:18] <archivist> all the shadows look photoshop
[16:06:19] <cradek> huh, that image won't load for me now.
[16:06:37] <als> you can look at mine
[16:06:40] <cradek> yeah it's not a great image. it probably had ugly stuff in the background.
[16:06:56] <archivist> I had trouble earlier on the imagebin site
[16:07:03] <archivist> methinks load
[16:07:54] <cradek> does look like a decent machine though.
[16:08:03] <als> looks like the same machine BOB mills is selling?
[16:08:04] <SWPadnos> I think the Smithy label was added - it's a little too sharp and not quite in the same plane as the front of the machine
[16:08:17] <cradek> I agree
[16:08:31] <cradek> and of course the bizarre fuzzy button
[16:08:45] <archivist> same oils stain on both images
[16:09:29] <SWPadnos> well, it's $7k (machine) + $4k (control/drives) from Bobcad, what does Smithy charge for it :)
[16:10:11] <als> Head Swivels - + or - 90 Degrees
[16:11:05] <SWPadnos> only $10,290 from Smithy, including Synergy
[16:14:49] <cradek> is it only because I've never had a mill that does that, that I don't think I'd use it?
[16:15:38] <cradek> seems like all I see about it on bridgeports is "this or that knocked it out of tram, and now I have to mess with it again to put it back"
[16:16:35] <SWPadnos> for a manual machine, I can see it being useful
[16:16:55] <SWPadnos> it does give you some semblance of 5-axis operation
[16:17:03] <cradek> do you use it?
[16:17:07] <als> odd jobs and strange setups
[16:17:09] <SWPadnos> no
[16:17:17] <SWPadnos> then again, I hardly use the mill so ...
[16:20:06] <als> i don't see how they swivel it though
[16:25:08] <als> ah maybe behind the fuzzy estop button looks like a boss on the casting in front of the oiler
[16:43:29] <BigJohnT> so who actually makes it?
[16:44:08] <skunkworks_> china? ;)
[16:46:55] <BigJohnT> might make a good dust magnet
[16:47:40] <skunkworks_> * skunkworks_ wouldn't mind having a machine that size
[16:48:05] <skunkworks_> it would fit in the corner of the garage real nice
[16:48:09] <BigJohnT> with steppers?
[16:49:21] <skunkworks_> me? I would really rather have servos. (but that is just me.
[16:49:22] <skunkworks_> )O
[16:50:11] <BigJohnT> thats what I thought with a machine that size
[16:50:24] <BigJohnT> I would not want R8 on a CNC mill
[16:51:16] <als> build a retention knob setup
[16:51:47] <cradek> with tormach stuff R8 might be ok on a small machine
[16:52:02] <BigJohnT> I'm spoiled by Rapid-Switch
[16:52:29] <cradek> R8 is good for hobbyists because import tooling is dirt cheap
[16:52:35] <als> ya
[16:52:48] <BigJohnT> I can use any tooling too
[16:53:04] <BigJohnT> but I can take it out and put it back in a second and the Z stays the same
[16:53:36] <als> R8 will too
[16:54:14] <cradek> I like my QC30 better than R8, but I can sure understand why you'd choose to make a machine this size R8
[16:55:45] <als> I have kwik switch 200 and its painful to buy
[16:55:53] <BigJohnT> R8 is a cheaper way to go
[16:56:04] <BigJohnT> als why is it painfull?
[16:56:05] <cradek> yep I think that's much worse than QC
[16:56:15] <cradek> rare and expensive
[16:56:30] <BigJohnT> not really flea bay has them all the time
[16:56:39] <BigJohnT> and you can get new from many places
[16:56:39] <als> can't find it and every pcs is 50 bucks
[16:56:53] <cradek> oh only 50? :-)
[16:57:17] <BigJohnT> I just bought 6 collet holders from flea bay for $49 each delivered
[16:57:23] <BigJohnT> all brand new!
[16:57:36] <als> that's for the ordinary
[16:57:52] <BigJohnT> ordinary what?
[16:58:19] <als> end mill holders and such
[16:58:21] <cradek> the killers are stuff like boring heads, jacobs tapers
[16:58:34] <als> tap holders
[16:58:38] <cradek> and renishaw arbors, haha
[16:59:03] <cradek> shell mills...
[16:59:41] <BigJohnT> I have everything but a renishaw and didn't pay much for any of it
[16:59:58] <BigJohnT> all rapid/kwik switch 200
[17:00:28] <als> yours has the ears on them?
[17:00:32] <BigJohnT> yes
[17:02:05] <BigJohnT> http://business.search.ebay.com/rapid-kwik-switch-200_Business-Industrial_W0QQcatrefZC12QQfromZR9QQfromZR9QQfsooZ1QQfsopZ1QQfstypeZ1QQftrtZ1QQftrvZ1QQsacatZ12576QQsatitleZQ28rapidQ2ckwikQ29Q20switchQ20200QQsspagenameZSTRKQ3aMEFSRCHQ3aSRCH
[17:03:10] <BigJohnT> http://cgi.ebay.com/Kwik-Switch-200-80235-3-8-Collet-Chucks-22-Available_W0QQitemZ220243377757QQihZ012QQcategoryZ41946QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[17:03:50] <cradek> huh, maybe that stuff is cheaper than QC
[17:04:17] <alex_joni> 'lo
[17:05:10] <als> that is for one pcs
[17:06:18] <BigJohnT> hi alex_joni
[17:06:28] <BigJohnT> $35 for one
[17:06:33] <alex_joni> hey BigJohnT
[17:06:39] <als> yep
[17:06:56] <BigJohnT> alex_joni: you back in the saddle?
[17:07:04] <alex_joni> BigJohnT: shortly :)
[17:07:14] <alex_joni> 2 more days at home
[17:07:42] <BigJohnT> cool
[17:09:44] <alex_joni> then I'm gone for another 2 weeks
[17:23:27] <BigJohnT> http://imagebin.ca/view/s4342Ct.html
[17:25:20] <als> sweat :)
[17:27:08] <als> I have about half that
[17:27:41] <als> are those er collets
[17:28:26] <BigJohnT> yes 75's and I forget the small ones
[17:28:42] <BigJohnT> * BigJohnT pulls out his MSC catalog
[17:29:07] <BigJohnT> Opps no they are AF75 and AF38 collets
[17:30:03] <als> the small ones AF38
[17:30:53] <BigJohnT> yep they go down to 1/16"
[17:31:02] <BigJohnT> up to 3/8"
[17:31:22] <BigJohnT> http://imagebin.ca/view/ISdoImo.html the parts I'm making at the moment
[17:34:13] <BigJohnT> the finished part http://imagebin.ca/view/Ssh8vKuE.html
[17:35:14] <als> where did you find the AF75 collet holders?
[17:35:52] <BigJohnT> flea bay
[17:36:18] <als> you don't see many of those do ya
[17:36:39] <BigJohnT> usually more than the 38's
[17:36:48] <BigJohnT> but not this week
[17:37:17] <BigJohnT> they show up in spurts
[17:37:25] <als> I guess I have to be on the look out
[17:37:53] <BigJohnT> i have a search for it so its easy
[17:38:31] <BigJohnT> (rapid,kwik) switch 200
[17:38:41] <BigJohnT> finds them all
[17:39:18] <als> I I've bidding against ya
[17:40:08] <BigJohnT> i'll look out for that now
[17:41:51] <BigJohnT> i find the collet holders more useful than the end mill holders as not all end mills have weldon shanks
[17:43:02] <skunkworks_> http://www.youtube.com/watch?v=o-9J4KsXYwU
[17:43:04] <als> tolerance wise or the flat
[17:43:58] <BigJohnT> flat
[17:44:21] <als> grind one on
[17:45:20] <BigJohnT> too much trouble
[17:45:34] <als> sometimes the collet holders let the mill get pushed up
[17:46:29] <BigJohnT> replace dull mill or worn out collet
[17:46:42] <BigJohnT> end mill
[17:47:12] <BigJohnT> als do you have the holder for changing tools?
[17:47:31] <als> no
[17:47:48] <BigJohnT> that is the problem i bet
[17:48:34] <BigJohnT> let me finish lunch and i'll take a pic of mine real simple
[17:49:08] <BigJohnT> http://imagebin.ca/view/ke5bYs.html
[17:49:49] <als> Mostly I've had it happen on large hmc the problem is a hard shank and a hard collet
[17:50:16] <BigJohnT> hmc?
[17:50:37] <als> large machining center
[17:50:44] <BigJohnT> ok
[17:51:23] <als> the larger stuff uses a stop inside
[17:52:30] <BigJohnT> my 38's have a stop inside...
[17:53:10] <als> adjustable?
[17:53:44] <BigJohnT> yes
[17:55:55] <BigJohnT> http://imagebin.ca/view/5_KVrCa.html
[17:56:57] <BigJohnT> http://imagebin.ca/view/FRp9-eID.html
[18:00:55] <als> is that a wood machine enclosure?
[18:01:08] <BigJohnT> LOL yes
[18:01:20] <BigJohnT> chip control
[18:01:50] <BigJohnT> sheet metal on is on the drawing board and will be cut out with the EMC Plasma Cutter
[18:02:01] <als> niffty
[18:03:02] <BigJohnT> did the holder pics makes sense?
[18:04:15] <als> I don't have any AF75 at the moment so the spindle works ok
[18:04:52] <BigJohnT> ok
[18:05:28] <als> but yes I have been going to make one
[18:07:18] <BigJohnT> I don't think the taper is needed just a big hole on top and a smaller hole for the bottom with the two slots
[18:11:55] <als> BigJohnT, you going to fest?
[18:12:54] <BigJohnT> als I can't I have three machines to finish building by the end of the month
[18:13:07] <BigJohnT> it's going to be tight
[18:14:23] <als> bbl
[18:50:48] <pjm> btw, last night i was asking about dynamic breaking resistor values for use with VFD's and today I found this PDF: http://www.galco.com/techdoc/chgp/mvx9000/mvx9000_ts.pdf which gives a clue as to suitable values / wattages for a given HP motor
[19:06:03] <awallin_ubu> how do I checkout the version of emc2 that runs as sim, i.e. on a normal kernel (not RT)
[19:07:24] <SWPadnos> check out from CVS?
[19:07:33] <SWPadnos> or install as a package?
[19:07:57] <awallin_ubu> I guess it's just --enable-simulator
[19:08:02] <SWPadnos> yes
[19:08:12] <SWPadnos> same version, hence my confusion :)
[19:08:52] <awallin_ubu> ok, then I just need to add the emc2 cvs server to sources.list... if I can find the correct line in the wiki
[19:09:57] <awallin_ubu> hm, maybe that's just for the packages...
[19:10:12] <SWPadnos> yes, apt/synaptic are only used for package management
[19:10:25] <acemi> to get emc2 source, cvs -z5 -d:ext:anon@cvs.linuxcnc.org:/cvs co -dtrunk emc2
[19:10:27] <SWPadnos> you'll need to install emc2-dev (probably)
[19:10:44] <SWPadnos> and also from a command line do `apt-get build-dep emc2`
[19:11:24] <SWPadnos> I'm not sure if emc2-dev gets everything build-dep does - I'd be interested in knowing if you can build emc2 after only installing emc2-dev
[19:12:11] <awallin_ubu> well I already installed build-dep :)
[19:12:51] <SWPadnos> next time then :)
[19:15:10] <awallin_ubu> W: GPG error: http://www.linuxcnc.org hardy Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 3CB9FD148F374FEF
[19:15:31] <awallin_ubu> I got that when running 'sudp apt-get update
[19:17:37] <gefink> look in this sctipt how to get the key http://www.linuxcnc.org/hardy/emc2-install.sh
[19:23:31] <micges> SWPadnos: in fresh system it is imposible to build emc2 only after installing emc2-dev
[19:23:56] <micges> about 10 or more packets must be chosen manually
[19:24:00] <BigJohnT> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_6_06_or_8_04_from_source
[19:24:26] <SWPadnos> micges, have you tried apt-get build-dep emc2? I'm swuite sure that works
[19:24:28] <SWPadnos> quite
[19:25:22] <gefink> micges: i tried it with synaptic, and got all packages
[19:25:40] <micges> I have new pc to install emc2.2.5 from source and Ill try it tommorow
[19:26:02] <micges> probably I choose only emc2-dev
[19:26:05] <fenn> you also need build-essential
[19:26:53] <micges> ok so I need build-dep, build-essential
[19:27:12] <micges> eventually emc2-dev
[19:27:42] <awallin_ubu> lots and lots of packages related to documentation (tex-stuff) being installed right now
[19:27:47] <micges> ok note it and will see
[19:27:55] <SWPadnos> emc2-dev depends on build-essential and several other packages
[19:28:10] <alex_joni> you don't need emc2-dev
[19:28:12] <micges> ok
[19:28:24] <alex_joni> emc2-dev is needed only if you want to build things that go along with emc2
[19:28:24] <SWPadnos> build-dep should pull in everything needed to make the packages, which is a lot more than emc2-dev does
[19:28:30] <alex_joni> but not to recompile a new emc2
[19:28:38] <alex_joni> apt-get build-dep emc2 is what you need
[19:28:40] <SWPadnos> right - you can build comps with emc2-dev installed
[19:28:55] <SWPadnos> (for example)
[19:29:03] <alex_joni> right
[19:29:36] <awallin_ubu> is the doc build included by default? I remember it being lengthy and verbose...
[19:29:53] <SWPadnos> no
[19:29:58] <alex_joni> only pdf is by default in there
[19:30:00] <SWPadnos> you need to --enable-build-doc or something
[19:30:12] <alex_joni> html builds is something you need to install separately than build-dep
[19:30:12] <awallin_ubu> ok. I'll run make then.
[19:31:17] <awallin_ubu> some warnings about truncated integers
[19:32:55] <awallin_ubu> done, now let's see if it runs
[19:34:36] <awallin_ubu> yep, seems to work. I might look at pyvcp widgets a bit next week if I get really inspired
[19:35:53] <micges> SWPadnos: Is there any solution/fix/patch/change to get nan from position file in emc ?
[19:36:06] <SWPadnos> not so far
[19:37:39] <awallin_ubu> any ideas about what new stuff we can expect during fest?
[19:41:17] <cradek> on the wiki some of us have assembled a list of things we hope to work on, but that's an answer to a different question
[19:41:36] <cradek> in the past we've not been good at predicting the outcome
[19:42:29] <awallin_ubu> ok, I guess by following #emc (or #emc-devel) we can get some idea of what is going on
[19:44:00] <cradek> definitely
[19:44:34] <skunkworks_> * skunkworks_ is excited
[19:44:58] <cradek> * cradek goes and looks at the wiki
[19:47:22] <awallin_ubu> parameter access from HAL would be nice, or get rid of all parameters and make them pins
[19:47:50] <cradek> that's just lots of busywork though...
[19:48:03] <alex_joni> micges: what nan ?
[19:48:39] <skunkworks_> I for some reason is excited about the kins work that may or may not be done :)
[19:49:44] <awallin_ubu> skunkworks: clearing up the joints vs. axis confusion?
[19:49:50] <skunkworks_> yes
[19:49:50] <cradek> yeah I hope we can do something to help 5axis along. I think that's what jmk wants to do the most.
[19:50:23] <cradek> more than that. replacing teleop, implementing joint constraints
[19:50:26] <skunkworks_> constraints
[19:50:29] <skunkworks_> right
[19:51:02] <awallin_ubu> I thought we had joint constraints (as in vel acc limits for joints) ?
[19:51:27] <skunkworks_> not after you run it thru kins
[19:51:37] <cradek> nope we have axis constraints
[19:52:32] <awallin_ubu> so will the axis constraints move to [TRAJ] and then there will be new joint constraints for each joint?
[19:53:02] <cradek> it's not yet clear to me how it should be fixed
[19:53:08] <skunkworks_> I think there is a wiki on that also
[19:53:51] <awallin_ubu> doesn't that require calculating first and second derivatives of the forward kins to get vel/acc of each joint when the user commands a certain axis vel/joint
[19:54:53] <cradek> that does seem like one way
[19:55:06] <cradek> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?JointsVsAxes
[19:55:39] <micges> alex_joni: my machine is driven by PID, full feedback from encoders
[19:55:43] <alex_joni> awallin_ubu: another option is to simulate the whole move, and check how the joints move..
[19:55:44] <cradek> I think this plan of simulating it and then scaling back the cartesian vel/acc according to how badly it exceeded joint constraints is promising
[19:56:19] <micges> and when in some way comunication or sth is wrong I see nan in preview position
[19:56:27] <alex_joni> cradek: not sure how math intensive it is
[19:56:28] <cradek> that takes trouble off the kins writer. you will get discrete approximations of those derivatives and that seems fine.
[19:57:22] <micges> and when I exit AXIS in this moment nan are write to [TRAJ] position_file
[19:57:39] <cradek> alex_joni: seems like you wouldn't need to simulate every single servo cycle.
[19:57:42] <fenn> you could also provide hooks for the kins writer to use their own derivative function if they have an analytical solution
[19:57:46] <alex_joni> cradek: right
[19:58:01] <alex_joni> fenn: if.. but even for kins you don't always have one
[19:59:20] <micges> problem is that emc is blindly load nan posistion from posiition file without checking
[19:59:59] <micges> and user have no way to know why nan is showing
[20:01:02] <alex_joni> micges: then I guess you should "fix" the loading
[20:01:18] <cradek> and more importantly, fix whatever caused it
[20:01:29] <alex_joni> yeah, that too
[20:01:47] <alex_joni> but I would happily apply a patch which checks the position file for sanity before loading it
[20:01:58] <alex_joni> (for example if it's out of bounds or something)
[20:02:08] <alex_joni> (or nan, or inf ;)
[20:02:24] <alex_joni> but I must say I've never seen nan on regular kins..
[20:03:30] <micges> alex_joni: when you have dynamically changed scale and you are developing interface to it you'll see many nan :)
[20:03:53] <alex_joni> micges: might be a reason it doesn't work by default than
[20:04:07] <alex_joni> or as they say so nicely "you broke it, you get to fix it..."
[20:06:01] <micges> Im doing it now ;) in meantime
[20:08:24] <SWPadnos> is it an error if a POSITION_FILE specified in the ini doesn't exist?
[20:08:39] <SWPadnos> ie, the ini has POSITION_FILE=fred.txt, but fred.txt doesn't exist
[20:09:00] <alex_joni> SWPadnos: I'd think so
[20:09:03] <cradek> doubt it. it is probably created.
[20:09:15] <SWPadnos> heh
[20:09:27] <cradek> haha, different guesses. maybe you will have to try it or read the source
[20:10:04] <SWPadnos> my thought was that if there's a nan when you shut down, the position_file should be deleted, but that just puts off the silent error "recovery" until the next startup
[20:10:20] <SWPadnos> unless that's an error, which would be a bad thing
[20:10:33] <micges> its created with zeros
[20:10:38] <alex_joni> + FILE *f = fopen(posfile, "r");
[20:10:38] <alex_joni> + if(!f) return 0;
[20:11:03] <alex_joni> if it's not found, nothing happens.. it will be written on shutdown though
[20:11:11] <alex_joni> SWPadnos: http://emergent.unpy.net/files/sandbox/emc2-remember-position2.txt
[20:11:13] <SWPadnos> ok
[20:11:53] <alex_joni> I'm surprised fscanf() reads a nan
[20:11:54] <SWPadnos> so deleting the file instead of writing out nan might be OK, though it's the same thing as writing zeroes to it, since that's how emc will start the next time if the file doesn't exist
[20:11:59] <SWPadnos> heh
[20:12:08] <chr0n1c> yo yo!!
[20:12:14] <alex_joni> +int r = fscanf(f, "%lf", &positions[i]);
[20:12:15] <alex_joni> +if(r != 1) { fclose(f); return -1; }
[20:12:35] <cradek> there's an error, so delete any evidence of it?
[20:12:42] <cradek> I think that's otherworldly
[20:12:55] <alex_joni> no, the code looks ok imo
[20:13:11] <chr0n1c> the code looks like chinese to me.. ;)
[20:13:13] <alex_joni> but I'm surprised the %lf reads a nan as a valid float
[20:13:15] <SWPadnos> no, there are two choices when there's an error, don't write it out, or don't read it in
[20:13:39] <cradek> or abort, giving a meaningful error message to the user
[20:13:58] <chr0n1c> ** i only say chinese because i don't read chinese (it's not because i am racist)
[20:14:10] <fenn> chr0n1c: they both begin with a C
[20:14:29] <SWPadnos> yes, of course - so when a nan appears, what happens? you abort, the user says "gee, I should plug my cable in", they shut down EMC, and now what?
[20:14:39] <SWPadnos> abort = go to estop mode
[20:14:59] <SWPadnos> you can't prevent the user from exiting the program, which is what an abort at emc shutdown would be
[20:15:23] <micges> meaningful error message to the user is the most important
[20:15:28] <alex_joni> good night all
[20:15:29] <chr0n1c> ohhh, i had a question last night when i was running a program.. where do i tell emc to stop and display messages during a tool change? i have see this option somewhere (it may have been in mach3)
[20:15:30] <cradek> when you detect bad state, you don't write it out (obviously), and you give an error message
[20:15:40] <cradek> you don't go around deleting files
[20:16:01] <cradek> chr0n1c: use hal_manualtoolchange. several of the sample configs use it
[20:16:03] <micges> in my machine nan are created when:
[20:16:23] <cradek> chr0n1c: it prompts and then waits for you to confirm the change
[20:16:24] <chr0n1c> ok, will look into now... thanks cradek
[20:16:31] <micges> heavy distortions
[20:16:32] <SWPadnos> ok, that's valid, though the position file will have known bad information in it (or very nearly so - it could be correct if you happened to move the machine to exactly where it was at startup)
[20:17:00] <cradek> oh hmm, I hadn't thought of that
[20:17:14] <cradek> I may be wrong
[20:17:15] <awallin_ubu> is there a good cpu benchmark program for multi-core machines?
[20:17:20] <SWPadnos> things get weird with persistent data ;)
[20:17:31] <micges> 2. bad config created by user that prevent him to use machine without deleting/editing files
[20:18:10] <chr0n1c> * chr0n1c loves qemu
[20:25:10] <micges> (all user config is created in Axis)
[20:25:22] <micges> no editing files by user
[21:33:34] <dmess> hi all
[21:35:38] <chr0n1c> yo!
[21:39:07] <dmess> sup