#emc | Logs for 2009-07-10

[03:35:14] <mozmck> when and where does iocontrol.0.user-enable-out get set to true?
[03:46:14] <SWPadnos> when the user clicks the "machine on" button or hits F2
[04:06:34] <mozmck> hi SWPadnos!
[04:06:46] <SWPadnos> hi
[04:06:48] <mozmck> it's not doing that here.
[04:08:11] <SWPadnos> not at all? it should be temporary
[04:08:24] <SWPadnos> unless I
[04:08:27] <mozmck> I'm watching it in the HAL configuration window and I never see it change.
[04:08:32] <SWPadnos> unless I'm thinking of another pin
[04:08:37] <SWPadnos> that's too slow
[04:08:39] <SWPadnos> try halscope
[04:10:11] <mozmck> don't see anything there either. I haven't used halscope much yet though so I might not have it set up right.
[04:10:58] <SWPadnos> well, I could be wrong, but I think that's what the pin is supposed to do
[04:11:18] <SWPadnos> I'm a bit tired to think about it now though, since it's past my bedtime :)
[04:11:55] <mozmck> that's what I was thinking too. the charge-pump.enable is connected to that signal in the stepconf generated config.
[04:12:08] <mozmck> but the charge pump never turns on.
[04:13:08] <mozmck> so I went looking at the signals. I did find one that turns on with machine on - motion.motion-enabled
[04:13:52] <mozmck> I wonder if there's any reason I can't use that signal to enable the charge-pump...
[04:14:18] <mozmck> I think I better hit the hay myself.
[04:14:32] <toastyde1th> what did the hay ever do to you
[04:15:11] <mozmck> :) made me a fairly decent mattress!
[04:15:33] <toastyde1th> haha
[04:29:44] <eric_unterhausen> why does firefox have such a problem with web pages that use php?
[04:35:32] <Jymmm> SWPadnos: It's always past your bedtime
[04:38:46] <eric_unterhausen> I think comcast is shipping my packets via pack mule
[04:39:00] <Jymmm> You know comcast filters, right?
[04:39:13] <eric_unterhausen> filters what though?
[04:39:25] <Jymmm> you'll have to ask them
[04:39:41] <eric_unterhausen> I thought it was mostly file sharing type things
[04:40:20] <eric_unterhausen> I can't even get a www.google.com to load
[04:43:28] <eric_unterhausen> I wish they had some kind of support
[06:16:35] <roh> mozmck did you try hitting F1 while the estop on the hw is pressed?
[06:17:11] <roh> i need to do that once after powering up to get the link led on (chargepump enabled), after that one releases the hw-estop and powers on via F2
[06:40:41] <nots_> nots_ is now known as nots
[11:08:57] <BigJohnT> http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/func,view/catid,9/id,497/lang,en/#507
[11:09:46] <BigJohnT> rtapi: no version for "nano2count" found: kernel tainted.
[11:24:52] <BigJohnT> anyone know what this means?
[12:24:34] <jepler> bigjohnt: it's a harmless message that means not all 'i's were dotted or not all 't's were crossed in building rtai. It's irrelevant to the poster's problem, which probably relates to the inifile he hasn't shown anyone.
[12:43:20] <skunkworks_> jepler: All of the servos seem to work (encoders read and they spin anyways)
[12:44:25] <skunkworks_> (hooking them into the printer port.) looks like I can read them up to 1500rpm or so. :)
[13:28:26] <alex_joni> quick question.. if I cat'ed from /dev/ttyS0 a bunch of data
[13:28:36] <alex_joni> but the port was set to 9600 bps, instead of 19200
[13:28:48] <alex_joni> would anyone think it's possible to reconvert the data captured?
[13:28:56] <SWPadnos> nope
[13:29:40] <SWPadnos> I don't even think it would work if you had the port set to 19200 instead of 9600 (ie, oversampling instead of undersampling)
[13:30:23] <alex_joni> maybe ssending it out through the port ;)
[13:30:35] <alex_joni> at the wrong baudrate, then trying to receive it at the right rate
[13:30:36] <archivist> RS232 is never forgiving
[13:30:36] <SWPadnos> maaaaybe :)
[13:30:48] <SWPadnos> character framing would also be relatively random
[13:39:41] <jepler> skunkworks_: cradek and I finished the motor mounts wednesday, but last night I watched DVDs instead of putting the system together
[13:51:12] <jepler> do we know of anybody who's tried "dual loop" PID in emc? I was imagining a 1kHz position pid (velocity out) and a 10kHz-20kHz velocity pid (pwm out).
[13:51:32] <jepler> I realized that for a velocity-input PID, FF0 makes sense
[13:53:26] <cradek> I don't know of that being done
[13:53:56] <cradek> you need velocity feedback for a velocity pid... not sure where you'd get that.
[13:54:15] <jepler> hostmot2 at least has the high precision velocity output based on the arrival time of individual encoder pulses
[13:54:36] <cradek> true - wonder how that would work.
[13:54:41] <micges> jepler: you mean position PID and lower velocity PID ?
[13:55:19] <jepler> velocity pid in base-thread
[13:55:27] <micges> we tried this here
[13:55:55] <micges> very low ferrors
[13:56:47] <jepler> so it was helpful?
[13:57:52] <micges> yes
[13:58:07] <jepler> micges: what base-period did you use for the velocity PID?
[13:58:15] <micges> this was on hm2
[13:58:20] <jepler> yes, I'll be using hm2 as well
[13:59:03] <cradek> so you turned on fp in the fast thread?
[13:59:30] <jepler> I suppose that would be a requirement
[14:00:07] <micges> I don't know but it was 2x 1khz
[14:00:32] <micges> maybe vel thread should be faster but motor wasn't connected to machine
[14:01:19] <micges> guy whose wrote that config is on vacation :)
[14:01:43] <micges> bbl
[14:24:37] <skunkworks_> jepler: cool! it would be interesting to see how tight you could tune it that way.
[14:25:00] <skunkworks_> velocity amps without them being velocity amps ;)
[14:25:33] <skunkworks_> SWPadnos: did you put the g540 config somewhere? they don't seem to have it on geckos site yet.
[14:25:44] <SWPadnos> hmmm
[14:25:55] <SWPadnos> I put it somewhere. let me see where :)
[14:27:03] <SWPadnos> ok, it's in the wiki uploads directory, but I don't know if any page links to it
[14:27:29] <SWPadnos> http://www.linuxcnc.org/wiki/uploads/G540.stepconf might work
[14:28:09] <cradek> I don't understand why you haven't put that in the distribution.
[14:30:42] <SWPadnos> hmmm. I'm not sure why
[14:31:15] <SWPadnos> I think only a couple of people have tried it out, so I might just be waiting for more user feedback
[14:31:27] <SWPadnos> which is hard to get when nobody knows about it ...
[15:00:40] <skunkworks_> SWPadnos: you must have set that up on one of the goal3 motherboars (15000us base thread) ;)
[15:01:38] <skunkworks_> heh - never mind. No coffe yet this morning
[15:24:03] <skunkworks_> I thought that was what you guys did at stuarts shop. (granted you used 2 different feedback devices - scales and encoders)
[15:24:10] <skunkworks_> (dual loop)
[15:29:51] <skunkworks_> I think peter made some comment about having to
[15:30:15] <skunkworks_> 'work' on the velocity output from the mesa card.
[15:30:25] <skunkworks_> * skunkworks_ waits for peter to show up.
[15:35:33] <cradek> skunkworks_: those were two pids with their outputs summed, not one inside another
[15:35:38] <cradek> they were both position loops
[15:36:08] <skunkworks_> ah
[15:38:13] <skunkworks_> is someone going to add a velocity input to the pid componant now...... ;) jepler?
[15:38:49] <jepler> that's mostly unrelated to this
[15:39:26] <skunkworks_> That hasn't stopped me before.\
[15:39:28] <jepler> incidentally, I'm not sure whether that other input should be called "velocity"
[15:40:01] <jepler> rather, it's the first derivative of feedback -- but feedback is not necessary a position; it could be a velocity, or even a temperature
[15:42:34] <cradek> yes
[15:51:46] <skunkworks_> jepler: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2009-06-19.txt
[15:51:57] <skunkworks_> 2009-06-19 14:16:35
[15:52:08] <skunkworks_> I don't know if it will effect you.
[15:52:31] <jepler> thanks
[15:52:41] <jepler> I think he's talking about the communication between FPGA and driver
[15:52:47] <jepler> the HAL connections would be the same in either case
[15:52:54] <skunkworks_> ah - ok
[19:02:26] <Ito-Brazil> Hello there, someone can help me to configure my steppers? Thanks !!!
[19:02:58] <cradek> Ito-Brazil: go ahead and ask your questions and someone may be able to help.
[19:03:53] <Ito-Brazil> OKe, i have no signal from LPT. I have a 12 Leds atached on it to test the computer
[19:04:25] <cradek> perhaps you have not specified the right parport address in stepconf.
[19:04:30] <Ito-Brazil> This computer was used to command a CNC with EMC, a friend gave it to me qhen got a P4 to do it
[19:06:06] <Ito-Brazil> first parport base Address (im using the 378 ECP on BIOS) what should i put on wizard?
[19:07:06] <cradek> 378, sounds like
[19:07:54] <Ito-Brazil> on wizard it is 0x378 , change it to ONLY 378 ?
[19:08:31] <cradek> what do your LEDs show?
[19:09:01] <Ito-Brazil> 1,3,4 ON .... 10,11,12 ON
[19:09:13] <Ito-Brazil> But all leds was on on startup
[19:09:35] <cradek> ok, then what do you do?
[19:10:36] <Ito-Brazil> i wil save the wizard configuration on 0x378 then i post a test
[19:10:55] <Ito-Brazil> i did nothing, the computer turned off alone those LEDs
[19:16:27] <Ito-Brazil> configuration returned error, tryed to start emc, error too
[19:16:39] <Ito-Brazil> i have another installation, better than ubuntu
[19:17:11] <Ito-Brazil> i instaled on debian netinst and put only a 32kb interface in it, called aewm++
[19:17:31] <Ito-Brazil> runs like hell, much faster than ubuntu-emc
[19:18:47] <cradek> if you want help with an error, you must say what the error is
[19:18:51] <cradek> use pastebin
[19:19:00] <Ito-Brazil> on this install, the 3-pin turns ON/OFF qhen a click a butt on axis, but just that, no other LEDs change
[19:19:35] <cradek> depending on what button you click, that may be right
[19:19:54] <cradek> you need to describe your problem much more precisely to get good help
[19:20:05] <cradek> brb
[19:20:21] <Ito-Brazil> i´m updating ubuntu now and upgrading emc. I think it´s better i change to my own installation... I will be bak then, thanks!
[19:22:13] <skunkworks_> wow - you do a search for hexapod now and the results are a flood of that 6 leg robot. Yeck
[19:30:40] <fenn> skunkworks_: something i saw recently re: hexapods http://tvalley.nl/fckfiles/file/UTwente_-_Dannis_gehele_presentatie.pdf
[19:30:58] <fenn> (MEMS hexapod using electrostatic comb actuator)
[19:34:59] <skunkworks_> fenn: wow. neat
[19:46:48] <Jymmm> I heard that when Farah Faucett went to heaven, God asked her what her final wish would be. She said "I want all the little kids on Earth safe". So he killed MJ."
[19:57:47] <skunkworks_> did you hear that they are going to recycle him? He is mostly plastic anyways.
[19:58:35] <Jymmm> "Billy Mayes here for the MJ Paintbrush..."
[20:41:50] <skunkworks_> http://imagebin.ca/img/xCPTcJ.jpg
[20:42:06] <skunkworks_> Not ul approved
[20:42:35] <skunkworks_> http://imagebin.ca/img/hR9ceTb7.jpg
[20:42:42] <skunkworks_> did I mention how much I like hal?
[20:42:44] <skunkworks_> :)
[20:48:45] <cradek> love your wiring
[20:48:51] <cradek> what's the WIBU dongle?
[20:49:34] <skunkworks_> for some old rip software
[20:50:27] <archivist> * archivist sees no dodgy wiring
[22:06:26] <Ito-Brazil> hello there! There is some how to test my LPT?
[22:06:45] <Ito-Brazil> I mean something to see if my config is right
[22:14:35] <Ito-Brazil> no ideas?
[22:24:09] <anonimasu> do you have a meter?
[22:31:43] <Ito-Brazil> yes
[22:31:47] <Ito-Brazil> i have a mult meter
[22:36:55] <jmkasunich> someone has to use EMC2 to do this: http://www.youtube.com/watch?v=iX93vj2zclM
[22:44:36] <archivist> hmm ink jet...but bigger
[22:52:48] <anonimasu> Ito-Brazil: well then measure the pins
[23:00:00] <Jymmm> jmkasunich: Just curious, why do you say that?
[23:00:40] <Jymmm> jmkasunich: and do you mean the happy face or Monalisa?
[23:02:31] <jmkasunich> the happy face
[23:02:58] <Jymmm> jmkasunich: Oh, you mean someone SHOULD do that with emc?
[23:03:05] <jmkasunich> yes
[23:03:33] <Jymmm> Ah, I thought you meant it was done with emc. I thought someone already did but using two motors hanging on a wall
[23:05:08] <Jymmm> What would you guess the distance was from end of barrel to canvas?
[23:05:19] <jmkasunich> dunno, maybe 10-20 feet
[23:05:22] <Jymmm> and the diameter of each splatter
[23:05:45] <jmkasunich> however big a paintball splat is - I've never played paintball, so I dunno
[23:06:01] <Jymmm> jmkasunich: well depends on distance =)
[23:06:09] <Jymmm> But I meant on that video.... 4" ?
[23:07:12] <jmkasunich> looks like the paper is about 4-5 feet tall
[23:07:17] <Jymmm> Heh.... RGB painball
[23:07:23] <Jymmm> or CYKM???
[23:07:34] <Jymmm> err CYMK
[23:08:06] <Jymmm> Hmmm, couldn't "mix" colors as they overlay on top of each other.
[23:09:45] <archivist> L84Supper, should watch that too
[23:53:19] <L84Supper> a bit low res but fun