#emc | Logs for 2008-08-15

[08:30:16] <christel> [Global Notice] Hi all, We appear to be having some European connectivity issues this morning, I'm about to do a tiny spot of re-hubbing, it should only hurt for a second! Have a great day and thank you for using freenode.
[08:39:34] <pjm> good morning
[08:59:49] <archivist_ub> spares buttons ? http://cgi.ebay.co.uk/Anilam-Crusader-II-CNC-control_W0QQitemZ230281152197QQcmdZViewItem?hash=item230281152197&_trkparms=72%3A1121|39%3A1|66%3A2|65%3A12|240%3A1318&_trksid=p3286.c0.m14&timeout=1218791090071
[09:04:26] <robin_sz> now if only i had a westinghouse Producer2 control spares ...
[09:05:37] <pjm> i need a CNC pendant control now!
[09:09:34] <mumu> I installed emc2 on ubuntu 8.04. Then i restarted the pc ant try to run "stepper-hyza", but i receive an error "can't read '/usr/realtime-2.6.24-19-generic/modules/rtai_hal.ko': No such file or directory". It seem it was not installed properly. I followed the instructions on the web site saying i should only run the installation script.
[09:15:09] <mumu> this is the error report i receive: http://pastebin.ca/1172708
[09:15:17] <mumu> any idea what could be wrong
[09:16:21] <archivist_ub> did you install from the live cd
[09:17:07] <mumu> no, i run the installation script
[09:17:24] <archivist_ub> to standard ubuntu
[09:17:27] <archivist_ub> ?
[09:17:35] <mumu> yes ubuntu 8.04
[09:17:51] <archivist_ub> EMC requires a realtime kernels to run
[09:18:11] <archivist_ub> these come on the live cd
[09:19:03] <mumu> i know, but installation script should probably also install realtime kernels
[09:20:17] <archivist_ub> 99 % use the live cd's I think loading realtime kernels is up to you if you go your route
[09:22:35] <mumu> this is bnot my route, there are the instruction i got from emc website: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_8_04_using_precompiled_EMC2_packages
[10:58:46] <pjm> btw, i need to cast a lead counterweight, anyone got any lead casting tips, apart from 'dont get burnt'
[10:58:58] <pjm> its for the Z axis counter balance weight on my CNC
[10:59:31] <archivist_ub> pjm, dont breath the fumes
[10:59:40] <pjm> ah yes that one too ;-)
[10:59:45] <archivist_ub> we do it here for clock weights
[10:59:50] <pjm> ahh ok
[10:59:53] <pjm> interesting
[11:00:16] <pjm> i thought it might be the answer since I can weigh it all accuratly then make a weight to match the Z axis weight
[11:00:25] <pjm> i was gonna make the mould out of wood
[11:00:31] <pjm> not sure if that is practical tho
[11:00:54] <pjm> but the plan would be to cast a bit of 10mm studding through it which I can screw a eye onto
[11:01:03] <archivist_ub> keep a little weight un balanced to keep the backlash hidden
[11:01:59] <archivist_ub> I should balance this one a bit
[11:02:20] <pjm> i have 2 sets of leadscrew nuts on the Z
[11:02:36] <pjm> that preload the leadscrew so there isnt any backlash
[11:03:08] <pjm> but re the mould, u reckon a wooden box would be ok/
[11:03:09] <pjm> ?
[11:03:16] <archivist_ub> my z is currently a 1mm pitch
[11:03:17] <pjm> the idea is to break it apart once its set
[11:03:44] <archivist_ub> sand/plaster
[11:04:07] <pjm> ah yes, that involves more muck tho
[11:04:11] <pjm> i wanted to try and be simple
[11:04:20] <archivist_ub> reasonably runny when molten
[11:04:32] <archivist_ub> and wood burns!
[11:04:46] <pjm> yeah i guess that could be a problem
[11:04:54] <pjm> the surface finish isnt too important
[11:04:59] <pjm> just need to get the weight right
[11:05:31] <pjm> i'm using that stepper motor that u pointed out on here
[11:05:45] <pjm> for my Z and its excellent, even with no counterweight it can move the Z perfectly
[11:09:02] <archivist_ub> we also have used metal moulds
[11:09:30] <archivist_ub> doesnt stick to rusty steel
[11:11:20] <archivist_ub> but sheet plaster board and plaster to seal with an external way of holding back the pressure during the pour (in sand in bucket)
[11:11:54] <archivist_ub> or cast iron lump from scrap yard :)
[12:11:50] <Guest290> Guest290 is now known as skunkworks_
[12:38:46] <mumu> i would like to controll emc from outside (computer vision), so i have to input commants to emc from another program somehow. Is this possible?
[12:46:16] <jepler> mumu: yes. One way you can control emc is through python using the 'emc' module. 'pydoc emc' for a summary of the commands and status available. this blog entry has an example program that remote controls emc in a very simple way: http://axis.unpy.net/01167419757 -- the program "mdi" installed by emc is another example.
[13:14:16] <pjm> archivist_ub ok thanks for the tips on lead, I'll look into it. I only need to cast this one counter weight then its all done
[13:14:50] <archivist_ub> pjm, what is the actual weight needed
[13:14:59] <pjm> 24Kg
[13:15:08] <pjm> that is the head weight
[13:15:31] <pjm> i have a weight on it at the moment that is about 13Kg
[13:16:34] <archivist_ub> you could do a built-up one so you dont need to cast all in one go
[13:16:48] <pjm> but the weight has to be max 70X70mm and i have about 400mm in length to play with
[13:17:02] <archivist_ub> or make a steal box and fill with lead shot
[13:17:16] <pjm> yeah thats a good idea actually
[13:17:45] <pjm> i was looking at the haas minimill and they use gas springs so i spose that could be another angle if I can find sometihng suitable
[13:18:16] <archivist_ub> car tailgate gas springs
[13:18:28] <archivist_ub> limited travel though
[13:19:09] <pjm> ah yeah that is a good idea actually
[13:19:48] <archivist_ub> boss intending to sell the building here so likely some clock weights lying around
[13:21:07] <pjm> i have a few bloody great lumps of lead that were counterbalances in my radome which has a az/el/roll steerable dish in it
[13:21:21] <pjm> and there is about 100Kg of weights but they are all stupid shapes
[13:33:19] <cradek> a piece of iron pipe with a cap on one end? fill with shot until it's the right weight
[13:34:01] <cradek> if it turns out you need a little more density, you can just heat the pipe with a torch and it'll melt (or at least I think it would be possible)
[13:34:47] <archivist_ub> melting IN the final container is a good idea
[13:43:31] <cradek> I bet 2" pipe will fit nicely in a 70mm square hole
[14:35:39] <mumu> i installed the emc and rtai kernel, and now ubuntu doesnt detect my sound card any more. Is there some kind of limitation for soundcard in rtai?
[14:36:14] <cradek> possibly a firmware or acpi problem
[14:36:16] <jepler> mumu: you may need to also install the package linux-ubuntu-modules-2.6.24-16-rtai
[14:43:52] <mumu> thank you, the installation of rtai-modules solved my sound card problem
[15:00:23] <micges> hello all
[15:13:17] <pjmemc> archivist_ub, u had one of these 220N motors from arc?
[15:13:28] <archivist_ub> yes
[15:13:45] <archivist_ub> and it worketh my z
[15:13:55] <pjmemc> u run it up yet? i think the one I have runs a little warm
[15:14:05] <pjmemc> so i probably need to reduce its hold current at a guess
[15:14:23] <pjmemc> yep i'm using the same motor on Z
[15:14:23] <archivist_ub> they er dont run cold
[15:14:30] <pjmemc> ahh
[15:14:37] <pjmemc> ok so praps nothing to worry about
[15:15:26] <archivist_ub> no steppers do get warm / hot I have mine at 2.5 A
[15:15:43] <archivist_ub> nope 2.4 A
[15:15:49] <pjmemc> and once its been on a while its warm to the touch
[15:15:56] <archivist_ub> * archivist_ub reading switches
[15:16:07] <pjmemc> i'm probably worrying about nothing
[15:16:25] <archivist_ub> at 2.7A it got to about 50 deg C while working
[15:17:56] <archivist_ub> if you can keep your hand on it, its probably not broken
[15:18:10] <pjmemc> ah yes its not that hot!!
[15:18:23] <pjmemc> but thanks for the info, I'll recheck my current settings
[15:18:34] <pjmemc> and what volts do u supply your driver with?
[15:18:51] <archivist_ub> nor sure if they spec temperature but coiled devices can often handle hot
[15:19:02] <archivist_ub> 30V on mine
[15:19:17] <pjmemc> yeah i looked in the data sheet for operating temp and no mention of it
[15:20:17] <ra3vat> ra3vat is now known as dimas_
[15:20:17] <archivist_ub> I want an 18tpi tip to mill a thread, cant find any that size on epay
[15:20:33] <archivist_ub> I hate paying full price
[15:21:46] <pjmemc> i did have some thread cutting inserts i'll find them and look
[15:22:05] <pjmemc> now i have finished my machine i need to seriously get shot of all the crap i've hoarded
[15:22:29] <archivist_ub> now the crap comes in useful
[15:23:08] <archivist_ub> or ready for the next machine
[15:36:59] <mumu> i am trying to command emc using python. I wrote a simple code ( http://pastebin.ca/1173060 ), but nothing happens and emc.stat().position always returns 0. What am i doing wrong?
[15:37:28] <cradek> s.poll()
[15:41:39] <mumu> what is with s.poll(), btw, where can i find all the command that are available?
[15:41:49] <jepler> "pydoc emc"
[15:42:18] <jepler> s.poll causes the status information to be updated. otherwise, the prior value is returned. before the first poll, the values are undefined.
[15:45:58] <pjmemc> i see quite a few ppl test their machines with the 3d face g-code, anyone know where to get this from?
[15:46:07] <pjmemc> i've had a quick google and it wasnt obvious
[15:47:37] <mumu> jepler: s.poll() returns "emc.error: emcStatusBuffer invalid"
[15:50:02] <jepler> mumu: probably you need to set emc.nmlfile before constructing an emc.command or emc.stat object, like jdi.py and mdi.py both do
[15:50:24] <jepler> jdi.py shows getting the path from an inifile, while mdi.py shows taking the nmlfile as a commandline argument
[15:50:37] <jepler> another reason you might get this error is if you're not actually running emc when you execute your program..
[15:51:12] <mumu> emc is running
[15:51:24] <jepler> then it's probably the first thing I mentioned: failure to set emc.nmlfile
[15:56:36] <mumu> jepler: where did you upload this mdi.py example, i can only find jdi.py
[15:56:50] <jepler> mdi.py is installed with emc, as /usr/bin/mdi if you installed from package.
[16:06:33] <pjmemc> archivist_ub, just one more stepper current question, do u have a seperate holding current setup and if so what is it?
[16:06:54] <archivist_ub> no
[16:07:17] <pjmemc> ok that was easy then ;-))
[16:07:35] <archivist_ub> have done that in the past but not for this sort of job
[17:19:08] <archivist_ub> hmm /me would like the mazak pointed to on the mailing list http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=360078008407
[18:17:49] <cradek> ha, good first project indeed
[18:20:52] <archivist_ub> that would be an excellent toy to learn on!
[18:22:42] <skunkworks_> hmm - don't think that would fit in the garage..
[18:23:12] <archivist_ub> adjust garage, its worth it
[18:24:04] <skunkworks_> actually - it would probably fit in the garage.. just nothing else..
[19:20:47] <mib_ggdfat> mib_ggdfat is now known as maddash
[19:21:02] <maddash> ooh yay.
[19:35:17] <maddash> hm, why so quiet?
[19:45:02] <jepler> maddash: I dunno, it's been a little less busy lately
[19:45:37] <archivist_ub> makin bits!
[19:48:12] <jepler> could be
[19:48:29] <jepler> after all, since 2.2.6 is bug free and the html/pdf documentation is now complete there's no reason to hang out on IRC anymore
[19:48:59] <maddash> :O 2.2.6 is bug free?!
[19:49:58] <cradek> yay
[19:50:12] <maddash> wow, kudos
[19:50:23] <maddash> I'm trying to find the bug status page right now
[19:50:59] <jepler> * jepler wonders whether it's his sarcasm transmitter or receptor that is malfunctioning right now
[19:53:07] <maddash> oh, you were joking.
[19:53:08] <archivist_ub> * archivist_ub has a doc comment in store
[19:53:31] <maddash> jepler: it's your transmitter and my receptor that are malfunctioning
[19:54:48] <jepler> software is never bug-free, documentation is never complete, and even when it is people still don't find the part that answers their questions..
[19:58:36] <maddash> only one bug opened this month? looks like i'll have something to do this weekend...
[20:03:49] <archivist_ub> doc bug, go straight to g83 try and peck drill without an R in the code read doc to see what R should do!
[20:23:34] <maddash> on an unrelated note, I'll be applying for my first IT job in a while.
[20:26:08] <jepler> yuck, job interviews
[20:26:44] <cradek> yuck, IT jobs
[20:26:48] <cradek> oh wait
[20:27:09] <skunkworks_> heh
[20:27:11] <archivist_ub> boss says he is selling the building here, /me has to do some job hunting!
[20:28:03] <maddash> any tips? don't bring my own keyboard? hide all usb flashdrives? etc?
[20:28:33] <skunkworks_> change all administator passwords..
[20:28:35] <cradek> for the interview? I don't know, it's been so long
[20:29:11] <skunkworks_> (then you know who thinks they should have access) ;)
[20:30:08] <maddash> meh, i'm still psychoanalyzing the job recruiting and editing my resume to fit
[20:30:55] <cradek> I can only think of the obvious stuff: be sure to have no grammar or spelling errors on your resume, and use an appropriate font
[20:32:10] <alex_joni> not comic sans?
[20:32:16] <cradek> but that's ancient 20th century advice
[20:32:37] <alex_joni> hi guys
[20:32:40] <cradek> hi
[20:32:52] <alex_joni> 'nother really long day
[20:33:07] <alex_joni> but at least it's weekend now
[20:33:12] <alex_joni> (and soon over there)
[20:33:17] <jepler> maddash: http://typicalprogrammer.com/?p=13
[20:35:06] <maddash> jepler: thanks
[20:35:13] <jepler> warning: that list contains sarcasm
[20:35:21] <maddash> crap.
[20:37:48] <alex_joni> http://eneas.juve.ro/~juve/servere/
[20:52:43] <skunkworks_> alex_joni: kitchen?
[20:52:45] <skunkworks_> ;)
[20:55:11] <alex_joni> skunkworks_: close :P
[20:56:08] <alex_joni> skunkworks_: those are the servers I got last month
[20:56:59] <alex_joni> I'm thinking about selling them atm..
[21:15:53] <maddash> gnite, folks
[21:32:22] <alex_joni> good night all
[21:45:45] <archivist_ub> ew /me hits a limit on the A axis at 12991 degrees oorf in the other dir now
[21:48:12] <archivist_ub> oops was -12991 deg
[22:01:32] <pjm> did u build your own A axis?
[22:01:40] <pjm> its one of the things on my list to make
[22:03:42] <archivist_ub> its a vertex HV4 with a stepper
[22:03:48] <tomp> 12991~=13million detect units (13m * .001) are there .ini limits on rotaries?
[22:04:21] <archivist_ub> dunno but there was some natter about math limits the other day
[22:04:32] <tomp> heh
[22:04:54] <tomp> "err: outta fingers & toes"
[22:05:13] <archivist_ub> 32 fingers aint enough
[22:05:37] <jmkasunich> archivist_ub: exactly what are you seeing?
[22:05:40] <tomp> 4.3billion
[22:06:01] <archivist_ub> it just stopped as if it was a limit
[22:06:20] <archivist_ub> was hand milling a pocket
[22:06:23] <jmkasunich> you can't jog it or anything in that direction?
[22:07:11] <archivist_ub> yup went to -12991.xxx and stopped would jog back though
[22:07:35] <jmkasunich> what are the limits in your ini file for that axis?
[22:07:55] <archivist_ub> no idea set it via stepconf
[22:08:04] <jmkasunich> look in the ini file ;-)
[22:08:32] <jmkasunich> section [AXIS_3]
[22:09:23] <jmkasunich> MIN_LIMIT and MAX_LIMIT
[22:10:28] <archivist_emc> -9999 and 9999
[22:10:37] <jmkasunich> well there you have it
[22:10:45] <archivist_emc> yup
[22:11:01] <jmkasunich> those are in machine coords, if you have any offsets or such that could account for the other 2000 degrees
[22:11:08] <anonimasu> hm, why not have a min_limit = disabled or something ?
[22:11:11] <archivist_emc> * archivist_emc pokes stepconf for low numbers :)
[22:11:13] <jmkasunich> what is the SCALE value for that axis
[22:11:30] <anonimasu> rotary tables with limits is kind of a paradox :p
[22:11:31] <archivist_emc> 100
[22:11:56] <jmkasunich> then you won't hit a real limit for some millions of degrees
[22:11:57] <archivist_emc> anonimasu: er yes thats what I thought
[22:12:30] <jmkasunich> stepconf's default limits are "very big" when measured in inches and applied to a linear axis
[22:12:36] <jmkasunich> even in mm they are big
[22:12:53] <jmkasunich> dunno if there is code in stepconf that would treat rotaries differently
[22:13:38] <tomp> how does +/-9999 relate to |12991| ?
[22:13:50] <archivist_emc> other axis have sensible limits for the travels I have
[22:14:20] <jmkasunich> tomp: I suspect that if archivist went the other direction, he'd hit the limit at 6999 or so on the DRO
[22:14:57] <skunkworks> switch to machine coords and see?
[22:15:00] <jmkasunich> there is probably a 3000 deg offset between his DRO numbers and the actual motor position - motor pos is zero on power up and never gets changed (except by moving), while DRO changes due to offsets, touch-off, etc
[22:15:07] <archivist_emc> hehe long time to hold a pinky on the kb to find the other limit :)
[22:15:19] <tomp> ok, so it's just the '0' posn offset, thx
[22:15:28] <jmkasunich> g0 30000 ;-)
[22:15:31] <jmkasunich> (mdi)
[22:15:40] <archivist_emc> I did touch off to save returning to 0
[22:16:19] <SWPLinux> stepconf sets the FERROR to 1 and MIN_FERROR to 0.25 when the axis number is 3 or the uits are mm
[22:16:21] <SWPLinux> units
[22:16:33] <SWPLinux> vs. 0.05 and 0.01 for axes 0-2
[22:19:17] <archivist_emc> bleh slow steppers....900 deg a minute
[22:27:42] <tomp> copley has actuators using linear motors ('Servotube') and they say it has 'Built In Position Sensor, no encoder required, industry standard sine/cos output'. I only know of Heidenhain sin/cos encoders, are there industry standard converters (sinus/ttl)?
[22:28:54] <skunkworks> sounds like a resolver style feedback.
[22:29:34] <skunkworks> I know jonE is making some resolver to quad converters.
[22:29:45] <tomp> didnt think of resolver, thx
[22:41:08] <tomp> wow, yaskawa, miranova, indutek all make sin/cos->ttl converters. and an ap note from TI on how to dsp your own http://www.eetindia.com/ARTICLES/2001MAY/2001MAY14_AMD_DSP_AN1.PDF?SOURCES=DOWNLOAD
[23:04:35] <JymmmEMC> SWPLinux:
[23:05:39] <JymmmEMC> Well, I'm getting better.... Instead 5 seconds AFTER he leaves, I'm getting him 5 seconds BEFORE he leaves
[23:10:17] <skunkworks> you're getting better
[23:28:25] <skunkworks> tom1: http://www.analog.com/static/imported-files/Data_Sheets/AD2S1200.pdf
[23:28:49] <skunkworks> tom1: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=AD2S1200WSTZ-ND