#emc | Logs for 2009-01-09

Back
[00:38:14] <jmkasunich> ack - brain age 34 and then 51
[00:38:39] <cradek> 26 from the one and only time I tried it
[00:38:56] <cradek> I didn't find anything interesting about the 'puzzle'
[00:39:49] <jmkasunich> its not a puzzle
[00:40:09] <jmkasunich> its a memory test
[00:40:27] <cradek> yeah, boring
[00:41:53] <tomp> someone was asking for raster laser plot via gcode.
[00:42:00] <tomp> they assumed gcode would do it if they could get the conversion.
[00:42:02] <tomp> best check that synchronized laser on-off gets acceptable results. may need exact stop mode.
[02:58:10] <skunkworks> I had a little time to make a pic program to output a square wave into the h-bridge. At really low current - I can see the current limit working - it trips - stays low until the next rising edge. (cool) as I increase the current limit - it starts to oscolate at the 400khz range. (when tripping the current limit). I hope to dig deeper this weekend.
[03:00:14] <SWPadnos> is that current sense resistor wirewound?
[03:05:36] <skunkworks> http://www.ohmite.com/catalog/pdf/10_series.pdf
[03:06:18] <skunkworks> Low Inductance (non-inductive
[03:06:32] <SWPadnos> ok
[03:06:35] <skunkworks> below 0.2503a9)
[03:51:52] <steves_logging> steves_logging is now known as steve_stallings
[04:10:54] <skunkworks> ok - I played a bit - thru the proccess of elimination - I was disabling the enable (setting the pin at the flipflop high so it isn't going thru the logic) setting other gates low or high so there was only a few paths..
[04:11:38] <skunkworks> looks like the noise is getting into the first gate (between the opto and the inverter)
[04:12:11] <skunkworks> a .001uf cap across the output of the op-amp to ground seems to make it stop for the current I was testing.
[04:12:41] <skunkworks> sorry - not op amp (I am getting tired) output of the opto isolator and ground
[04:13:04] <skunkworks> if that made sense.
[04:14:00] <SWPadnos> some optos only need a few uA to trigger - could hte noise be getting into the input side of the opto?
[04:14:24] <skunkworks> I think these optos need a brick to turn them on.
[04:14:28] <SWPadnos> heh
[04:16:41] <skunkworks> oh well - time for bed. I am happy :) (so far) felt good to get back into this again.
[04:16:57] <SWPadnos> heh. night
[04:40:50] <steve_stallings> steve_stallings is now known as steves_logging
[10:55:34] <ian_macintosh> Hi Neil
[10:56:04] <neil_macintosh> Hi Ian
[10:59:42] <archivist> more brits!
[11:13:12] <neil_macintosh> I am trying to find some information with regards to setting up my lathe with a spindle index signal that I can cut threads
[11:14:59] <archivist> there are some examples on the wiki
[11:17:11] <neil_macintosh> i set up everything via stepconfig and i get the spindle speed panel on the left hand side of the axis window but when I connect my encoder i do not get any reaction is there something else that i need to do
[11:18:10] <archivist> I have not used stepconf for setting up threading
[11:18:47] <neil_macintosh> ok
[11:19:12] <archivist> you can check if data can be seen with halscope
[11:20:15] <archivist> I need to add a spindle encoder to my mill one day
[11:25:41] <neil_macintosh> i have the complete opposite problem when it comes to cnc's than most people that i have spoken to ,ie very limited electronics knowledge but very wide knowledge of programming and machining
[11:26:34] <archivist> a lot in here have a reasonable mix
[11:26:59] <archivist> Im an electonics engineer/programmer now working as a clockmaker
[11:27:37] <archivist> note with some encoders you need a pullup
[11:28:11] <archivist> do you have an oscilloscope
[11:28:19] <neil_macintosh> yes
[11:30:08] <archivist> ok does the encoder signal change with rotation and does its voltage go above 3.3v and less than 5 for the high signal and down to 0 to less than about 1 for low
[11:37:59] <archivist> you need better string for your internet connection
[11:38:48] <alex_joni> I'm not sure if stepconf handles an encoder for the spindle
[11:39:05] <archivist> nor I
[11:39:16] <alex_joni> neil_macintosh: stepconf is a simple interface for creating a configuration, and only supports a very small subset of emc's capabilities
[11:39:39] <archivist> dunno if the par port can handle the datarate
[11:39:39] <alex_joni> it's fine for starting confings (simple one), but for more advanced stuff you want to get down & dirty and do it yourself
[12:05:39] <pjm1> archivist that 4'th axis part is on its way to me!
[12:05:58] <archivist> :)
[12:06:02] <pjm1> good morning btw !
[12:06:39] <archivist> you could'nt wait for ally pally
[12:06:51] <pjm1> hehh nope!
[12:07:05] <pjm1> i've been cleared by the missus to go to that if I want!
[12:07:54] <archivist> ally pally could be quicker than post, but the bits we got from chronos were quick enough and good enough
[12:08:34] <archivist> I have morse collets for the rotary centre hole
[12:08:54] <pjm1> its MT2 init?
[12:09:14] <archivist> some I got from chronos and some locally from arc euro trade
[12:09:24] <archivist> mine is mt2
[12:10:33] <archivist> my spindle is mt2 as well so its all swappable about from blank holding to cutter holding
[12:10:45] <pjm1> yeah thats a good idea, i'm mt3 throughout here
[12:11:04] <pjm1> but i plan on bolting a chuck to the rotary table and leaving it on there
[12:11:20] <pjm1> i could get rid of my dividing head once i have this rotary table
[12:11:30] <archivist> I made an adapter as its off and on
[12:11:57] <archivist> heh why not mod the dividing head
[12:12:33] <archivist> put dividing head on vertex 5axis !
[12:12:46] <pjm1> hahh ah its not one with a handle to crank round
[12:13:00] <pjm1> it has some notches and u can index it around in 15 deg lumps
[12:13:09] <pjm1> perhaps dividing head is the wrong word
[12:13:14] <archivist> ah steam dividing
[12:15:10] <archivist> thats the older method boss uses
[12:16:08] <pjm1> its this http://pjm.dyndns.org/lump.jpg
[12:17:57] <archivist> some of our lathes have something similar built into the headstock
[12:18:56] <pjm1> this is made by marlco
[13:14:25] <skunkworks_> Ok - I didn't know the camera wasn't focusing until after I was done.. But this is the pwm in and enable_not. http://imagebin.ca/view/oQ-Fyl2A.html
[13:14:31] <skunkworks_> you get the drift I hope..
[13:14:49] <skunkworks_> well - I guess it is just enable.
[13:15:41] <skunkworks_> the taller trace is the enable. the smaller duty cycle trace is the pwm in.
[13:16:11] <archivist> focus? /me would separate for clarity
[13:16:25] <skunkworks_> you can see the enable goes low at the right and then doesn't go high until the next rising edge of the pwm input
[13:16:45] <skunkworks_> well - it would have looked a lot better if it was in focus ;)
[13:17:18] <skunkworks_> totally different feel now.. no whistle when in current limit.
[13:18:13] <archivist> hmm dinner time /me should go fetch his new bubble for the clinometer
[13:21:33] <tomp> i cant idenitfy a pwm signal in that photo. are they overlapped? its an interesting idea to use the drive signal as the enable tho
[13:21:53] <jepler> skunkworks_: could be an I Ching reading for all I can tell
[13:22:26] <jepler> actually it's a fish .. you can see top of head, eyes, mouth, and flippers
[13:22:31] <jepler> (flippers are horizontal)
[13:22:43] <tomp> hah chines fortunetelling
[13:24:29] <skunkworks_> heh
[13:24:31] <skunkworks_> sorry about that
[13:24:49] <skunkworks_> the pwm signal is in the middle vertically - the enable is bigger
[13:25:19] <skunkworks_> I must have put the camera in short focus mode by accident./
[13:25:52] <skunkworks_> (I don't think I was awake this morning_
[13:26:28] <tomp> tektronix? post the printput?
[13:26:37] <tomp> printout?
[13:26:48] <skunkworks_> no - b&k cheap scope.
[13:26:52] <skunkworks_> and it is at home..
[13:28:44] <tomp> ah, hook it to hal and screenshot it :)
[13:29:03] <skunkworks_> heh - hal isn't that fast... :)
[15:17:56] <archivist> get a digital scope
[15:18:23] <archivist> dont break bubbles on clinometers £80
[18:19:29] <JanVanGilsen> I've got my new toy today! http://imagebin.org/35409
[18:19:33] <JanVanGilsen> http://imagebin.org/35410
[18:19:37] <JanVanGilsen> http://imagebin.org/35411
[18:20:31] <archivist> I wonder if you have my pcb design in it
[18:20:43] <archivist> do you have the terminal
[18:23:32] <JanVanGilsen> yes we have the terminal, but there's a problem with loading the software, We're going to put an EMC2 controll on it =)
[18:24:24] <skunkworks_> JanVanGilsen: how many axis?
[18:24:27] <archivist> hmm if it has a 6502 and djc on the pcb
[18:26:16] <JanVanGilsen> 6, iirc the puma 560 has 6 joints and the puma 550 has 5 (they both have written 500 on them[500 series])
[18:26:30] <archivist> JanVanGilsen, if scrapping the terminal then I would like the pcb for my history (I was an external subcontractor on the replacement pcb)
[18:26:49] <skunkworks_> ok - I thought mine said 560 on it.. but I could be wrong
[18:27:13] <skunkworks_> that is the way I understand it.
[18:28:48] <JanVanGilsen> archivist, I'll still need to move the terminal home, then I can check if it has your pcb...
[18:29:10] <skunkworks_> cool mechanics.. Lots of places to adjust the slack out of it.
[18:29:10] <archivist> ok thanks was a long time ago :)
[18:29:13] <JanVanGilsen> please contact me via e-mail with a description where 2 look
[18:29:51] <archivist> I never had a complete terminal, only did pcb and software design
[18:30:07] <archivist> and got the prototype working
[18:30:51] <archivist> I will dig out the prototype and take a pic in the next week or so
[18:32:35] <JanVanGilsen> I've got this:Unimation Visual Display Terminal Model 2D65568G01 Equipment Manual ...
[18:33:43] <JanVanGilsen> I could scan you a copy =)
[18:34:09] <archivist> I had a spec to work to and an old pcb for outline
[18:34:57] <archivist> took me best part of a year to get paid
[18:45:02] <JanVanGilsen> Is it possible to dynamicly change the home offset parameters via hal? The puma robot homes to the next index pulse and calculates the offset from the absolute position of a pot-meter.
[18:48:32] <medo_gcode> sddd
[18:49:47] <skunkworks_> same day different day?
[18:50:10] <archivist> lift cat off keyboard?
[18:50:54] <archivist> http://icanhascheezburger.files.wordpress.com/2008/12/funny-pictures-cat-threatens-to-edit-your-face.jpg
[18:52:44] <skunkworks_> This is howard. http://imagebin.ca/view/Ln5dwQ.html
[18:53:04] <skunkworks_> (at the grandparents.. (our next door neighbors that love on our cats)
[18:53:44] <archivist> that needs a diet
[18:54:06] <skunkworks_> na - he is just big boned
[18:54:17] <archivist> my late puss was never large
[18:54:18] <skunkworks_> 22lb iirc
[19:17:16] <motioncontrol> jepler you see me
[19:18:16] <archivist> we all see you
[19:18:42] <motioncontrol> i have one problem with emc228 and rtai
[19:19:12] <motioncontrol> i have speack yesterday with jepler
[19:19:21] <motioncontrol> help me
[19:19:43] <archivist> ask clear questions so we can
[19:19:54] <motioncontrol> thanks
[19:20:17] <motioncontrol> i have istalle linux debian etch 4.05
[19:20:48] <archivist> better to use the live cd
[19:20:58] <motioncontrol> becouse
[19:21:02] <motioncontrol> because
[19:22:01] <motioncontrol> i want compile the emc and on ubuntu not is possible
[19:22:17] <archivist> I compile on ubuntu
[19:23:02] <motioncontrol> but on ubuntu more library for compile not is present.
[19:23:37] <archivist> there is a wiki page that tells you what you need to install
[19:25:27] <motioncontrol> please send me the web addres for compile with ubuntu.i have install first ubuntu 8.04 live cd, but you command ./configure --enable run in place, more librury not present(G++, python.dev eccc.)
[19:25:51] <archivist> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_6_06_or_8_04_from_source
[19:30:01] <skunkworks_> http://imagebin.ca/view/gjfhU9E.html
[19:30:55] <archivist> http://icanhascheezburger.files.wordpress.com/2008/12/funny-pictures-cat-waits-for-you-to-rub-his-belly.jpg
[19:32:58] <fragalot> http://www.27bslash6.com/overdue.html
[19:33:03] <skunkworks_> neighbor calls him 'the gentle giant'
[19:33:20] <fragalot> I consider my link to be superior in terms of merrit
[19:34:47] <motioncontrol> thanks
[21:45:05] <skunkworks_> oh my - mach can only handle 10000000 lines of gcode.. ;)
[21:46:26] <archivist> entire lifetime of mach use?
[21:46:34] <skunkworks_> heh
[21:46:52] <skunkworks_> per file
[21:48:13] <skunkworks_> That's the nail in the coffin.. I definatly using emc now. ;)
[21:48:38] <fragalot> skunkworks_: found that out by personal experience?
[21:48:42] <archivist> you needed nails to change
[21:49:21] <skunkworks_> fragalot: no - someone had mentioned it on cnczone.
[21:49:42] <skunkworks_> http://www.cnczone.com/forums/showthread.php?t=71138
[21:50:00] <archivist> has anyone tested emc for similar limits
[21:50:44] <fragalot> I wonder what limits the number of gcode in mach
[21:51:13] <archivist> memory probably
[21:51:45] <fragalot> so, why would it then be a mach issue.
[21:51:59] <fragalot> If i'm honest,... even NOTEPAD.EXE can't handle that much
[21:52:00] <fragalot> lol
[21:52:18] <fragalot> (which is a really,... really sad example, to be fair)
[21:56:32] <alex_joni> skunkworks_: 10M ?
[21:56:58] <alex_joni> fragalot: notepad is one HUUUUGE POS
[21:57:36] <alex_joni> I'm not sure how they did it, but it seems they somehow managed to implement all the slowest possible algorithms
[21:58:00] <alex_joni> try replacing , with ; in a smallish (~400k) csv file
[21:58:05] <alex_joni> it'll take at least 10 minutes
[21:58:06] <fragalot> alex_joni: yeah, and put exploits into wordpad.
[21:58:18] <alex_joni> pn is pretty ok
[22:02:21] <skunkworks_> alex_joni: 10m
[22:02:23] <skunkworks_> M
[22:07:08] <jepler> the AXIS preview will be useless far before 10 million movements, and maybe even when loading 10 million non-motion lines into the text preview.
[22:08:02] <skunkworks_> heh
[22:09:08] <archivist> I killed axis once with some bum gcode :)
[22:12:33] <skunkworks_> bottom of my h-bridge is a 446kb file. 19305 lines
[22:13:20] <jepler> xemc copes just fine with a 1-million-line file with two motions and all the other lines just reading "n(some number)".. it takes about 10 seconds on my machine to run it
[22:14:42] <fragalot> lol jepler
[22:15:04] <SWPadnos> I've loaded a 60000 line file just fine
[22:15:13] <SWPadnos> can even spin the preview nice and fast
[22:15:18] <fragalot> That's nearly simular to saying "My compiler doesn't work, it's as if it's completely ignoring my comments"
[22:15:38] <SWPadnos> (on my dual dual-core Opteron system with Quadro FX3500, in sim)
[22:16:37] <SWPadnos> hmmm. actually, it may have been more lines than that. it's a 20M+ file
[22:17:40] <alex_joni> less | wc -l ;)
[22:17:57] <jepler> it runs this file at a few thousand lines a second: (echo g0 x0 y0 z0; echo g20; echo g64p.1 f100; seq -f "g1x[%.0f/100000]" 1000000; echo g0 x1; echo m2) > /tmp/big.ngc
[22:17:59] <SWPadnos> it's more a matter of finding where the file is ;)
[22:18:02] <jepler> again in xemc
[22:18:17] <archivist> I was trying to remember the quicky line counter as well
[22:19:46] <jepler> 650794: ERROR: Linear move on line 940313 would exceed joint 0's positive limit
[22:19:49] <jepler> sheesh
[22:20:03] <skunkworks_> heh I went from notepad which doesn't display the file correctly (no returns) to wordpad which I couldn't see a way to show the number of lines - to ms word which I ended up having to turn on line numbering and look at it in print preview.. jeeze
[22:20:16] <SWPadnos> get programmers notepad
[22:20:20] <SWPadnos> 2
[22:20:29] <jepler> get vim
[22:20:30] <jepler> :set nu
[22:20:41] <SWPadnos> it lets you use either type of line ending and convert between them
[22:20:54] <SWPadnos> hmmm. vim for Windows. novel
[22:22:16] <archivist> I use syn on windaz, it deals with either line ending type, but it does not like long lines much
[22:22:38] <jepler> SWPadnos: http://www.vim.org/download.php#pc and it's quite good
[22:22:54] <SWPadnos> I noticed it
[22:23:02] <SWPadnos> though "good" is more of a religious question ;)
[22:23:05] <SWPadnos> or answer
[22:23:16] <jepler> it's an excellent implementation of vim
[22:23:25] <SWPadnos> ah, ok then ;)
[22:23:42] <archivist> I still type vi, vim is to new fangled