#emc | Logs for 2008-11-29

[08:13:56] <chainsaw_shave> I have a question about setting up the HAL for a 3 axis stepper type machine. My drivers take 3 inputs STEP, DIR, and ENABLE. I have a config that pegs the enables high and that works but it means the motors are drawing current all the time. Is there a way to tie the enable pins to some signal that is only active when a given axis is actually moving?
[13:09:53] <alex_joni> alex_joni is now known as NSA-1
[13:10:18] <NSA-1> NSA-1 is now known as alex_joni
[14:45:48] <Paragon> Hello All.....
[14:47:23] <archivist> ello
[14:48:41] <Paragon> I need some advice with regards to a 3 phase motor that I wish to connect to 240v coverted supply. The main motor on my mill is 240 / 440 6 lead so no worries there but the cross feed power appears to be rated at 440 only and only has 3 leads. What do i need to do to be able to run this motor from my 240v rotary converter?
[14:56:03] <archivist> I can think of a few methods, from cheap brave to costs
[14:57:32] <archivist> cheap..is get inside it see if it can be modded to 6 wire
[14:58:20] <Paragon> archivist: That was what I thinking but thought I'd get confermation here.
[14:59:40] <Paragon> I guess the other is to hook up a transformer to convert 240v in to 440v ???
[15:00:23] <archivist> is it one of those hang off the end of the table types, can the motor be changed, is it really 3ph
[15:01:17] <Paragon> Yes it hangs off the end but I dont think its a standard flange so mounting another motor will take some doing.
[15:02:06] <Paragon> I also have a suds pump which I am guesing may also be 440v but It's tucked away so I have not been able to confirm it as yet!
[15:02:12] <archivist> change to a servo ready for cnc :)
[15:02:49] <Paragon> :-)
[15:03:19] <Paragon> Catch yo soon.... heading back to the shop....
[15:03:36] <archivist> if you need an adapter...you have a milling machine....
[15:04:14] <Paragon> True. Just need to rewire the main motor :-)
[15:04:23] <Paragon> CU soon archivist
[20:04:51] <archivist_emc> jepler has gdepth been built on recent 8.04 I get a bunch of errors http://pastebin.ca/1270478
[20:26:28] <alex_joni> archivist_emc: I remember 8.04 having python2.5 on it..
[20:30:43] <archivist> hmm looks like its set up with 2.5
[20:31:42] <archivist> at least looking in synaptic
[20:31:46] <alex_joni> right
[20:31:57] <alex_joni> but the makefile for gdepth might be hardcoded for 2.4
[20:33:17] <archivist> that was it!
[20:35:04] <archivist> now to debug gdepth.py, the error message i get does not help me a lot
[20:35:35] <alex_joni> lemme see
[20:36:15] <archivist_emc> 0.ngc:0: Bad character 'g' used
[20:36:42] <archivist_emc> using a file from truetype tracer
[20:42:08] <JymmmEMC> man this hurts
[21:23:26] <archivist> cant find the error message in the source
[21:24:16] <alex_joni> gdepth-0.2.tar.gz?
[21:24:30] <archivist> except in streamer_usr.c and that has a lower case B
[21:24:39] <archivist> yes alllllllllllllllex
[21:24:44] <alex_joni> * alex_joni looks
[21:25:25] <archivist> ppppppppppppppppppp[p[iippppp''''''''''@@@@@'
[21:26:05] <alex_joni> sounds like a borken keyboard controller
[21:26:37] <archivist> just dug out two keys to remove crud
[21:26:42] <alex_joni> ah :)
[21:26:58] <alex_joni> well.. it imports 'gcode' which is the emc2 python interface to access g-code
[21:27:07] <alex_joni> and rs274.interpret
[21:27:23] <archivist> get full of brass filings needs a vacuum really
[21:27:24] <alex_joni> do you get a python trace?
[21:27:49] <archivist> did you look at the pastebin
[21:28:00] <alex_joni> didn't see one..
[21:28:35] <archivist> sorry no there is no backtrace for that error
[21:28:45] <alex_joni> can you put a log on pastebin?
[21:29:19] <alex_joni> did you try spiral.ngc ?
[21:29:57] <alex_joni> although I notice that jepler mentions that 2.2.x doesn't work with gdepth
[21:30:36] <archivist> that would be it then , need to find out why
[21:32:44] <archivist_emc> spiral.ngc:0: Bad character 'g' used
[21:33:13] <anonimasn> wow
[21:33:18] <archivist> something basic
[21:42:09] <alex_joni> maybe gcode.parse() has changes syntax?
[21:42:23] <alex_joni> changed* even
[21:42:36] <alex_joni> gdepth.py: code, seq = gcode.parse(args[0], Canon(), 'g20 g17 g64', '')
[21:44:11] <jepler> yeah it bitrotted and I never figured out why
[21:45:05] <jepler> it's probably something that will be obvious in retrospect
[21:50:40] <alex_joni> I bet it's the change to uvw
[21:50:54] <alex_joni> def straight_traverse(self, x, y, z, a, b, c): (from gdepth.py)
[21:51:08] <alex_joni> I think that needs to be (self, x,y,z,a,b,c,u,v,w):
[21:51:57] <alex_joni> * alex_joni is just speculating though
[21:52:10] <alex_joni> archivist: think you can try that out?
[21:52:32] <archivist> still errors
[21:52:52] <alex_joni> you need to do that for all similar functions
[21:53:27] <archivist> that would be a serious edit
[21:53:37] <alex_joni> hmm.. can't see others
[21:53:44] <archivist> possibly
[21:55:27] <alex_joni> tool stuff changed names
[21:57:32] <archivist> readme says to comment out tool commands
[21:58:14] <archivist> if calls changed names I would expect a better crash
[21:58:52] <jepler> try adding this in class canon: def get_axis_mask(self): return 7
[22:00:58] <archivist_emc> no change, before/after any particular def
[22:02:35] <jepler> * jepler shrugs
[22:02:38] <jepler> somebody should fix it
[22:03:23] <alex_joni> any idea how to make it more verbose?
[22:03:55] <archivist> poor me doesnt know which it is broke
[22:04:15] <alex_joni> * alex_joni starts his MV
[22:04:18] <alex_joni> VM
[22:04:52] <archivist> havnt seen and MV's for a while
[22:05:22] <jepler> don't have any bright ideas, sorry
[22:07:57] <alex_joni> jepler: tool.max = inf isn't right.. right?
[22:09:00] <jepler> alex_joni: beats me. I wrote that program several years ago and have barely used it
[22:09:19] <jepler> either -inf or +inf is "normal", because it represents the part of the tool's enclosing rectangle where no cutting takes place
[22:09:33] <alex_joni> ah, ok..
[22:11:03] <archivist> is there any way of single stepping python
[22:11:52] <alex_joni> you can't single step this
[22:12:00] <alex_joni> it goes into gcode.parse()
[22:12:13] <alex_joni> which is in src/emc/rs274ngc/gcodemodule.cc
[22:23:07] <archivist> I didnt find the error message in there though, I find similar message in streamer_usr.c else I cannot find where the error could be raised from
[22:23:24] <alex_joni> it's from the interpreter
[22:23:39] <alex_joni> the interpreter only sends the error number (228)
[22:23:55] <alex_joni> and then it gets converted to text: gcode.strerror(code))
[22:24:02] <alex_joni> line 242
[22:24:21] <alex_joni> raise SystemExit, "%s:%d: %s" % ...
[22:24:45] <alex_joni> first %s is filename, %d line number, %s error message
[22:27:34] <archivist> so where is the list of error messages as I grepped tunk and cant find
[22:27:47] <alex_joni> emc/rs274ngc/
[22:27:55] <alex_joni> rs274_errors or soemthing like that
[22:28:38] <alex_joni> 228: "Place holder for NCE_VARIABLE: shoult not occur" :-)
[22:30:50] <archivist> 13 is bad character
[22:31:35] <alex_joni> yeah, something doesn't match
[22:32:16] <archivist> also odd is the g is mid error message how has that happened
[22:32:29] <alex_joni> that gets preformatted somehow
[22:32:47] <alex_joni> change Canon(), 'g20 g17..' to something else
[22:32:58] <alex_joni> like 'a20' and you'll get Bad character 'a' ..
[22:33:16] <alex_joni> so definately something is borken in gcode-parse
[22:36:59] <archivist> changing to empty string in the canon() line still gets a g
[22:37:24] <archivist> but a got a as the error
[22:38:58] <JymmmEMC> * JymmmEMC taking a break... howdy!
[22:42:52] <DanielFalck> hiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii
[22:43:10] <DanielFalck> what are you taking a break from JymmmEMC
[22:43:15] <DanielFalck> ?
[22:43:49] <JymmmEMC> DanielFalck: Inventorying all the stuff that I'll be selling with the CNC machine =(
[22:44:14] <archivist> dont be silly, just fix it and use it
[22:46:50] <jmkasunich> he has been trying to do that for a couple years
[22:51:14] <JymmmEMC> archivist: It's time to move on
[22:51:33] <DanielFalck> to what? lasers?
[22:51:37] <JymmmEMC> I just don'r even have the time for it any more.
[22:52:12] <JymmmEMC> DanielFalck: Well, that's the intension. But pre-made, not home built.
[22:53:04] <JymmmEMC> DanielFalck: Considering one of these... http://www.ulsinc.com/english/laser_systems/product_line/VersaLaser.html
[22:56:08] <alex_joni> archivist: fsck if I know what's wrong :/
[22:56:35] <JymmmEMC> wb jmkasunich
[22:56:43] <jmkasunich> hi jymm
[22:57:04] <jmkasunich> I decided that since the kernel got updated, I might as well reboot
[22:57:14] <archivist> alex_joni, thanks for trying, do you thing it affects anything else
[22:57:20] <jmkasunich> instead of waiting a couple months like I usually do
[22:57:32] <archivist> * archivist been waiting
[22:58:02] <archivist> I dont want to reboot....
[22:58:04] <JymmmEMC> jmkasunich: Heh, it seems like ubuntu is turning more and more like M$ in reboots after updates it seems (not including kernel updates)
[22:58:11] <jmkasunich> now I'm restarting and updating the seven VMs that are the compile farm
[22:58:31] <jmkasunich> JymmmEMC: I haven't seen it request a reboot unless it was a kernel update
[22:58:32] <archivist> firefox is a bugger for wanting reboots of itself as well
[22:58:58] <JymmmEMC> jmkasunich: the last update I did (3d ago) wants an update
[22:59:00] <jmkasunich> "itself" you mean restart firefox, or restart the whole box
[22:59:16] <archivist> restart firefox
[22:59:34] <jmkasunich> JymmmEMC: you sure that didn't include a kernel update?
[22:59:39] <JymmmEMC> archivist Well, DUH!
[22:59:48] <JymmmEMC> jmkasunich: something about headers
[23:00:08] <jepler> archivist: well, you can try this, no guarantees (but it runs for me again, albeit on a 6.06 system and I didn't try the preview, just the png output) http://emergent.unpy.net/index.cgi-files/sandbox/gdepth-0.2.1.tar.gz
[23:00:52] <alex_joni> jepler: what did you change?
[23:00:55] <jepler> alex_joni: stuff
[23:01:04] <alex_joni> :?
[23:01:15] <JymmmEMC> jepler: "Good answer! Good Answer!" LOL
[23:01:18] <alex_joni> around Canon() ?
[23:01:19] <archivist> hehe
[23:02:06] <jepler> here's most of the change: http://pastebin.ca/1270622
[23:03:27] <alex_joni> probably would have missed the tool_offset(x,y,w) :/
[23:03:45] <alex_joni> I did notice set_origin_offsets though
[23:04:00] <jepler> after awhile it started giving different (better) tracebacks
[23:04:14] <archivist> wee it runs thanks jepler
[23:04:21] <alex_joni> I think the gcode.strerror() has some problems too
[23:04:29] <skunkworks> we require screen shots... :)
[23:04:29] <alex_joni> at least for this case it did
[23:04:44] <jepler> alex_joni: hey, you and archivist are lucky you got me to look this close
[23:04:45] <JymmmEMC> * JymmmEMC goes make to inventorying...
[23:04:57] <alex_joni> jepler: I know, I know..
[23:05:03] <archivist> :)
[23:05:11] <alex_joni> otoh, you wrote it.. so you get to fix it :D
[23:05:25] <alex_joni> * alex_joni hardly follows it
[23:07:03] <JymmmEMC> * JymmmEMC knows jepler's solution.... rm -rf *
[23:08:41] <alex_joni> jepler: since you looked closely.. can you fill me in why arc_feed() is not part of Canon() ?
[23:09:14] <jepler> alex_joni: it's provided by ArcsToSegmentsMixin, which turns it into a series of straight_feed calls
[23:09:41] <jepler> updated http://axis.unpy.net/01169521961
[23:09:53] <alex_joni> oh crap.. didn't see that in Canon (...)
[23:10:06] <alex_joni> Translated takes care of offsets & whatnot?
[23:10:18] <alex_joni> rs274.interpret.Translated I mean
[23:10:28] <jepler> yeah
[23:10:36] <jepler> it turns out to be a fragile system, but it's how it ended up
[23:11:03] <alex_joni> works here too
[23:11:18] <alex_joni> var file that caused the silence
[23:11:25] <alex_joni> it was the var*
[23:13:57] <alex_joni> well.. off to bed for me
[23:13:59] <alex_joni> good night all
[23:14:54] <archivist> jepler just run it on my 64 bit box as well
[23:16:33] <jepler> archivist: hardy?
[23:17:15] <archivist> how can one find the release name of a system
[23:17:39] <jepler> lsb_release -r
[23:17:40] <archivist> Linux hp 2.6.24-19-generic #1 SMP Wed Aug 20 17:53:40 UTC 2008 x86_64 GNU/Linux
[23:18:01] <archivist> 8.04
[23:18:05] <jepler> huh, same here
[23:18:12] <jepler> but it still doesn't work for me
[23:18:48] <jepler> you probably don't have any (TOOL,) comments in your gcode
[23:19:51] <toastydeath> live dangerouly, comment nothing
[23:19:52] <archivist> none iirc
[23:20:34] <jepler> a (TOOL) comment can switch from the tool specified on the commandline to another one
[23:20:39] <archivist> its running on truetype tracer ngc output
[23:21:17] <archivist> that I will have to do to get the vform i want
[23:21:30] <jepler> I suspect that you will see errors when you do
[23:22:12] <jepler> 'python arraytest.py' will show an AssertionError if the underlying bug exists on your system, or print nothing if your system is OK
[23:25:59] <jmkasunich> I wonder if my video link will make Olusegun rethink his plans?
[23:28:09] <archivist> jepler, I get an error ValueError: Arrays have incompatible shapes
[23:28:35] <jepler> archivist: yeah that's exactly the error I get
[23:29:11] <jepler> I am pretty sure it's a bug in the 'numarray' module that manifests on 64 bit systems only
[23:31:08] <archivist> i get the error on the p4 box as well
[23:31:47] <jepler> 32-bit os?
[23:32:13] <archivist_emc> Linux cnc 2.6.24-16-rtai #1 Tue Sep 30 22:54:33 EEST 2008 i686 GNU/Linux
[23:32:19] <jepler> huh
[23:32:37] <jepler> I have only compared 6.06/32-bits to 8.04/64-bits
[23:32:44] <jepler> obviously that's not the whole story
[23:32:51] <archivist> no
[23:39:43] <BigJohnT> jmkasunich: you should see what a 24" auger does to an underground 4160v line
[23:48:09] <archivist> jepler found another way of crashing it python gdepth.py -x "0 1" -y "0 1" -z "0 1" -r 75 -t .5 1.ngc
[23:48:36] <JymmmEMC> archivist you found it, you fix it!
[23:51:37] <JymmmEMC> jepler: Hey, now I know why you do that! It shuts em up pretty quick =)