#emc | Logs for 2005-11-28

Back
[00:04:53] <skunkworks> could some one walk me though where I would get the rs274 files emc2 (I need conditional statments and subrutines also)
[00:06:59] <K`zan> Hey folks!
[00:08:31] <K`zan> Picked up a pile of aluminum to start putting together the structureal mechanics for the router: http://wrlabs.shacknet.nu/~vw/CNC/CNC-Router-Mechanicals/
[00:08:40] <K`zan> Should be FUN :-).
[00:09:13] <K`zan> Now to finalize the size, I think 2'x2' overall should be good for this one.
[00:18:19] <lerman> skunkworks: the lerman-interp branch has all of the interpreter changes. (Hey, I'm glad to hear you say that you "need" them. -- I couldn't imagine trying to make most of the parts I do without those changes.)
[00:21:40] <skunkworks> stupid question - where is the lerman-interp branch? Be gental - I am new to linux and emc. (It is funny as I made a program to make a sphire in a cube using sin-cos to make the arcs and I looped it until it was done. (in turbocnc) I then tried to run the program in emc and must have took me about 2 hours to figure out there was no condisional statments in emc)
[00:22:53] <skunkworks> I have not tried this in emc2 yet - are they already intagrated in emc2?
[00:23:01] <lerman> Even stupider answer: beats the shit out of me. I just followed Alex Joni's directions to do this stuff.
[00:23:49] <skunkworks> Glad I am not the only one
[00:25:18] <lerman> You have to build the proper CVS branch. Someday, soon, someone will have to merge it into the main line. I suspect that we are waiting for the new directors to be elected since someone in authority should be deciding what to merge and when. The interpreter changes CHANGE THE OPERATOR PRECEDENCE slightly. That could be considered to break existing programs, although it's unlikely that there...
[00:25:20] <lerman> ...are many such programs out there.
[00:26:15] <lerman> If there is strong objection on that basis, we could add a .ini file option to change the precedence to the new style and leave the old style as the default.
[00:27:12] <skunkworks> ah - I guess I could wait until it is all straitened out. I would actually like the jogging bug fixed first(if it is a bug and not just me) ;)
[00:30:03] <jmkasunich> its probably a bug, and I'm probably the person to find it
[00:30:21] <jmkasunich> but I've been burning the midnight oil on a driver for the last 4 days, and need to finish that first
[00:30:30] <jmkasunich> I will look into the bug in the next week or so
[00:31:12] <jepler> skunkworks: to change to a CVS branch you issue a command similar to this one: cvs update -r name_of_branch
[00:31:22] <jepler> skunkworks: then you do a complete rebuild
[00:32:05] <jepler> skunkworks: to later get back to the "HEAD", use something like: cvs update -A
[00:33:25] <jepler> so you may want to just do this: cvs update -r lerman-interp
[00:34:07] <skunkworks> thanks
[00:34:24] <jepler> good luck
[00:34:46] <jepler> you can make a back-up of the source tree before starting if you want to have more of a safety net
[00:38:26] <skunkworks> thanks jmkasunich - can't wait.
[00:38:54] <skunkworks> can you copy and paste out of mirc
[00:38:54] <jepler> good luck
[00:39:02] <jepler> I don't use mirc, sorry
[00:39:06] <jepler> * jepler wanders off
[00:42:53] <Jacky^> :)
[00:42:56] <Jacky^> g night
[00:43:08] <Jacky^> Jacky^ is now known as Jacky^afk
[00:54:37] <CIA-5> 03jmkasunich * 10emc2/src/hal/components/stepgen.c: removed some dead code
[00:55:17] <cradek> I wonder when we will see election results...
[00:56:42] <jmkasunich> dunno, thalx is pretty busy
[01:41:29] <thalx> Results have just been tallied.
[01:42:25] <cradek> yay!
[01:42:33] <thalx> Typing up the message now to send out.
[01:42:53] <jtr> waiting, waiting...
[01:43:10] <SWPadnos_> fastest election results in the west ;)
[01:44:03] <thalx> Shall I stick to being all officious and stuff, and only announce it via email?
[01:44:10] <jmkasunich> thalx: thanks for your work on this
[01:44:13] <thalx> Or should I go ahead and let y'all know here?
[01:44:16] <jmkasunich> announce it here too
[01:44:21] <cradek> agreed, thanks
[01:44:25] <cradek> I bet it's tedious
[01:44:49] <thalx> Winner to loser? Loser to winner? With or without votes? Or just "The New Board" with no counts?
[01:45:02] <SWPadnos_> indeed -thanks, but send the email first, just to be imperious and official ;)
[01:45:03] <jmkasunich> just names, details by email
[01:45:08] <jmkasunich> right
[01:45:11] <SWPadnos_> how about as a "top ten list" :)
[01:45:24] <cradek> pace pace pace pace
[01:45:24] <jmkasunich> nah, just 5 names, keep it simple
[01:45:34] <jmkasunich> lol
[01:45:43] <thalx> Well, the first hint - Matt can breathe a heavy sigh of relief.
[01:45:59] <SWPadnos_> lucky dog!
[01:52:31] <thalx> Ray, however, will still be busy.
[01:52:58] <jmkasunich> no shock there
[01:53:23] <thalx> Ok, I have attempted to send officious email. May have fubared the cut-n-paste between windows, into a wrong window, or somesuch.
[01:53:40] <thalx> But, our Board of Directors is now... drumroll...
[01:54:00] <SWPadnos_> biddabiddabiddabiddabidda
[01:54:02] <SWPadnos_> biddabiddabiddabiddabidda
[01:54:07] <thalx> Ray Henry, John Kasunich, Alex Joni, Chris Radek, and Jon Elson.
[01:54:20] <cradek> yay!!
[01:54:22] <SWPadnos_> cool
[01:54:23] <jepler> cradek: congrats
[01:54:32] <jepler> jmkasunich: congrats
[01:54:35] <cradek> jepler: thanks
[01:54:44] <cradek> alex_joni (when you wake up): congrats
[01:54:45] <jmkasunich> thanks
[01:54:47] <jtr> congrats all
[01:54:47] <jepler> anybody else who is in the channel?
[01:54:57] <cradek> jmkasunich: congrats
[01:55:01] <cradek> thalx: thank you again
[01:55:20] <jtr> yes thalx, thank you
[01:55:49] <thalx> My pleasure to contribute, in this tiny way. Wish I had the space/time/spare brainpower to devote to more of the innards and debugging.
[01:55:55] <jepler> cradek: how sure are you that the "failure" code from i.open() is false? I must have gotten -1 from somewhere...
[01:56:01] <jepler> thalx: yes, thanks
[01:56:16] <cradek> jepler: well, I know it's not unsigned (the compiler tells me so)
[01:56:33] <cradek> jepler: and I see ==false in a dozen places in emc
[01:56:43] <cradek> emc2 at least
[01:57:43] <cradek> /* Open the ini file */
[01:57:43] <cradek> if (inifile.open(filename) == false) {
[01:57:43] <cradek> return -1;
[01:57:43] <cradek> }
[01:58:06] <cradek> (in emc2)
[01:58:10] <jepler> cradek: class INIFILE
[01:58:12] <jepler> const int open (const char *path);
[01:58:14] <jepler> - emc1
[01:58:40] <jtr> night all
[01:58:46] <jepler> const int INIFILE::open (const char *path) {
[01:58:46] <jepler> if (NULL == (fp = fopen (path, "r"))) { return -1; }
[01:58:57] <jtr> jtr is now known as jtr_AWAY
[01:59:02] <cradek> class Inifile {
[01:59:04] <cradek> ...
[01:59:07] <cradek> bool open(const char *file);
[01:59:14] <jepler> emc1 returns -1 on failure, 0 on success
[01:59:20] <cradek> yay, more gratuitious API breakage
[01:59:39] <cradek> I'll fix it - thanks for catching my stupid
[01:59:46] <jepler> good luck. Another #define?
[02:00:07] <cradek> sure, what else?
[02:00:10] <jepler> I'm glad I'm not the only one who breaks things
[02:00:19] <jepler> I dunno, fix emc2 HEAD ?
[02:01:08] <cradek> not with a ten foot pole
[02:01:23] <jmkasunich> pttthhhbbbbttt
[02:01:43] <cradek> jmkasunich: ?
[02:01:53] <jmkasunich> "not with a 10 foot pole"
[02:02:08] <cradek> jmkasunich: I don't want to change the dozen places inifile.open occurs
[02:02:18] <jmkasunich> just joking
[02:02:22] <cradek> jmkasunich: and if I did, someone would say they changed it to bools on purpose
[02:02:49] <jmkasunich> I wasn't reading the details of the problem
[02:03:02] <jepler> I'm sure they did change it on purpose
[02:03:05] <jepler> inifile = emc.ini(sys.argv[2])
[02:03:06] <jepler> emc.error: inifile.open() failed
[02:03:16] <jmkasunich> if you are talking about changes to the ini handler, that is something that Paul did - changed it to a C++ class I think
[02:03:16] <jepler> just verified that with your change, emc1(sim)/axis is quite broken
[02:04:01] <cradek> jmkasunich: yep, it's very C++ now, but not very unix
[02:04:05] <jepler> It was already C++, but the gratuitous renaming and API changing may well have been paul's
[02:04:05] <cradek> jepler: fixed (?)
[02:04:42] <cradek> I really don't mind breakage for a reason, but it's tedious to clean up after it when I don't see the reason.
[02:04:48] <jepler> cradek: ifdef AXIS_USE_EMC2 #define INIFILE_OPEN_FAILURE false #else #define INIFILE_OPEN_FAILURE -1 #endif ?
[02:04:54] <jmkasunich> the "inifind has been depracated" that pops up when you start emc2 annoys the hell out of me
[02:05:29] <cradek> jepler: if you like that better, it's fine with me
[02:06:36] <jepler> cradek: someone once told me that it was better to have #ifdefs separated from the logic, and I believed him
[02:07:24] <cradek> it does get hard to read when ifdefs mess up {} regions
[02:08:30] <jepler> + if (INIFILE_CHECK_FAILURE(self->i.open(inifile))) {
[02:08:49] <cradek> are you serious?
[02:08:55] <jepler> yes
[02:09:31] <cradek> ok then...
[02:09:32] <jepler> because comparing a bool to false is a dumb thing to write
[02:09:39] <jepler> if(f() == false) vs if(!f)
[02:09:46] <cradek> well I know that and you know that
[02:09:53] <cradek> but have you seen the emc coding standards?
[02:09:57] <jepler> so INIFILE_ABSURDLY_LONG_MACRO_NAME() can expand to the smart thing
[02:10:10] <cradek> you're a nut, but ok
[02:10:11] <jepler> sssh they're reading this
[02:10:17] <cradek> oh, oops
[02:10:35] <jepler> and you're a board member now
[02:10:44] <jepler> I hope they don't find out you used to be a wife-beater
[02:10:45] <jmkasunich> lol
[02:10:45] <cradek> be thankful it's not if (open() IS_NOT_EQUAL_TO false)
[02:10:56] <jepler> hey, that's old style
[02:11:12] <jmkasunich> what annoys me is if ( 0 == something )
[02:11:19] <jepler> surely it's the more C++-compliant emc::is_not_equal_to<bool, false>(open())
[02:11:31] <cradek> hahaha
[02:11:39] <cradek> great idea
[02:11:52] <cradek> jmkasunich: agreed completely
[02:12:20] <cradek> jmkasunich: twenty years ago, people wrote that to get a compile error on the mistaken if(0 = something)
[02:12:34] <jmkasunich> ah, so thats why
[02:12:35] <cradek> jmkasunich: today, we get an appropriate warning if we say if(something=0)
[02:12:45] <cradek> jmkasunich: yep
[02:12:51] <jmkasunich> yeah, I've seen that warming more than once ;-)
[02:12:58] <jmkasunich> warning
[02:13:08] <cradek> jmkasunich: you're not the only one
[02:13:51] <cradek> jmkasunich: but I also find that irritating, and I think it's no longer necessary...
[02:14:08] <jmkasunich> there is a lot of it in emc
[02:14:30] <jmkasunich> I tend to reverse them to if (something == 0 )
[02:14:40] <jmkasunich> I'm less fond of if (!something)
[02:14:40] <cradek> it causes me to stumble when I read it, but it's only a minor stumble
[02:15:11] <cradek> I prefer if(thing) and if(!thing)
[02:15:17] <jepler> I think if (!something) is clearest
[02:15:18] <cradek> it's totally natural to me
[02:15:37] <jmkasunich> unless "thing" is strcmp() ;-)
[02:15:52] <cradek> well I don't have any trouble writing !strcmp() but I do shiver each time I do it
[02:16:07] <jepler> #define streqv(a,b) !strcmp(a,b)
[02:16:33] <cradek> #define CHECK_STRCMP_RESULT_CODE_FOR_STRING_EQUALITY(a) (!(a))
[02:16:34] <jepler> #define strop(a,r,b) strcmp(a,b) r 0
[02:16:46] <jepler> if(strop(a, =, b)) ...
[02:16:58] <jepler> else if(strop(a, <, b)) ...
[02:16:59] <jmkasunich> sick, I tell you, just sick..
[02:17:11] <jepler> cradek can tell you that I'm a big-time preprocessor abuser
[02:17:28] <jmkasunich> I think there is a 12-step program for that
[02:17:28] <cradek> no kidding
[02:17:34] <cradek> I just waded through some of that tonight
[02:17:43] <jmkasunich> the interp is full of it
[02:17:44] <jepler> oh yeah
[02:18:11] <jepler> #define GLCALL2V(name, fmt, t1, t2) static PyObject *py##name(PyObject *s, PyObject *o) { t1 p1; t2 p2; if(!PyArg_ParseTuple(o, fmt, &p1, &p2)) return NULL; name(p1, p2); CHECK_ERROR; Py_INCREF(Py_None); return Py_None; }
[02:18:17] <jepler> that's clearer if it's on multiple lines, honest
[02:18:27] <jepler> http://unpy.net/cgi-bin/viewcvs.cgi/axis/extensions/minigl.c?rev=1.3;content-type=text%2Fplain
[02:18:46] <SWPadnos_> if you really want to see preprocessor abuse, look at the metrowerks libraries for the Motorola DSP56800 series
[02:18:52] <jmkasunich> yuck, did you get any on yourself?
[02:18:55] <SWPadnos_> they're scary
[02:19:26] <cradek> jmkasunich: you may not be able to tell it, but that file is actually C
[02:19:53] <jepler> cradek: just be thankful that I didn't put most of that in "minigl.h" and use the double-inclusion trick to cut even more lines out of minigi...
[02:20:10] <cradek> fwiw, I like minigl
[02:21:13] <jepler> thanks
[02:21:43] <jepler> I am sure that it's easy to be uncomfortable with the way it's written
[03:39:54] <CIA-5> 03jmkasunich * 10emc2/src/hal/drivers/hal_ppmc.c: PPMC driver progress - step pulse generators now work. Only encoders left. Need to look into shutting things off on driver exit... if enabled, the watchdog will do it, but if not enabled, it just keeps running.
[04:38:50] <CIA-5> 03paul_c * 10emc2-auto/wiki/ (diff_log page/E/EMC2_FileHeader.db): "Auto update wiki from a cron job. Mon Nov 28 05:30:01 GMT 2005 "
[05:53:20] <CIA-5> 03swpadnos * 10emc2/src/hal/drivers/hal_ppmc.c:
[05:53:21] <CIA-5> Added encoder functions, changed step output direction so that it is correct when step outputs are fed back to the encoder inputs.
[05:53:21] <CIA-5> This driver is now fully functional.
[08:29:45] <chinamill> Hello everyone
[08:31:25] <chinamill> Does anyone know how to set the active parport with emc2?
[08:32:19] <chinamill> Am I looking in the right direction if I browse threw HAL documentation?
[08:36:38] <alex_joni> chinamill: yes
[08:36:56] <alex_joni> active parport is the one selected in configs/standard_pinout.hal
[08:37:22] <alex_joni> there is a line 'loadrt hal_parport cfg=0x378' (written from memory, so it might be a bit different
[08:37:54] <chinamill> alex_joni: thanks for the directions...
[08:38:11] <alex_joni> no problem
[11:34:24] <cradek> alex_joni: congratulations on your election
[11:34:37] <alex_joni> cradek: ditto
[11:49:47] <Jacky^afk> Jacky^afk is now known as Jacky^
[11:50:08] <Jacky^> morning
[11:51:28] <alex_joni> morning Jacky^
[11:54:15] <Jacky^> planning another week of work ..
[11:55:58] <Jacky^> id like to try these patches on my Ipaq http://www.imec.be/rtlinux/
[11:57:23] <Jacky^> lets see what friends on #handhelds says ..
[12:00:53] <Jacky^> seems obsolete , 2002 :/
[12:05:31] <samcoinc> samcoinc is now known as skunkworks_wrk
[12:07:35] <Jacky^> anyone got a parport working on Ipaq ?
[12:07:43] <Jacky^> ops ..
[12:07:45] <Jacky^> :)
[12:11:09] <alex_joni> * alex_joni goes away
[12:11:12] <alex_joni> later guys
[14:02:07] <anonimasu> hey
[15:21:16] <alex_joni> evening all
[15:21:22] <SWPadnos_> morning
[15:21:32] <alex_joni> hey stephen
[15:21:36] <SWPadnos_> hey
[15:21:46] <SWPadnos_> congrats on your newly elected status :)
[15:21:54] <alex_joni> thank you ;)
[15:23:30] <jepler> is there a swearing-in ceremony or anything?
[15:23:48] <SWPadnos_> fsk sh!t godd@mn!
[15:23:51] <jepler> "I, alex_joni, affirm that I will be an iron-fisted dictator"?
[15:23:52] <SWPadnos_> how's that?
[15:24:09] <alex_joni> jepler: did you ever doubt that?
[15:24:17] <alex_joni> it will be the hardest year for developers
[15:30:41] <Jacky^> oh.. good
[15:30:58] <Jacky^> congrats to all , i dont know nothing :/
[15:31:24] <alex_joni> Jacky^: you're not subscribed to the user list?
[15:31:36] <Jacky^> alex_joni: i think not
[15:31:46] <alex_joni> too bad
[15:31:50] <alex_joni> missing all the fun ;)
[15:31:59] <Jacky^> i just have an old account on wiki, but seem died
[15:32:07] <Jacky^> ok ..
[15:32:16] <alex_joni> not on the wiki, on sourceforge
[15:32:28] <Jacky^> understood
[15:32:56] <Jacky^> jepler: have been elected too ? :P
[15:33:58] <Jacky^> jepler: are you corruptable ?
[15:34:22] <Jacky^> if not yet, try to came here some week well teach you !
[15:34:28] <Jacky^> hahaha kidding :D
[15:34:52] <Jacky^> good,
[15:35:07] <alex_joni> morning les
[15:35:10] <Jacky^> morning les_w
[15:35:24] <les_w> hi jacky and alex
[15:35:28] <SWPadnos_> Hi Les
[15:35:37] <les_w> new board member alex!
[15:35:42] <les_w> Hi SWP
[15:35:49] <les_w> just read your post
[15:35:51] <alex_joni> seems like it
[15:35:52] <les_w> very good
[15:35:53] <SWPadnos_> The new board is: Ray Henry, John Kasunich, Alex Joni, Chris Radek, and Jon Elson.
[15:36:00] <SWPadnos_> ah - thanks
[15:38:26] <les_w> I could fill out an "absolutely must, should, could" easily enough
[15:38:33] <les_w> others might not agree
[15:38:49] <SWPadnos_> yeah - so could everyone else, and I guarantee they'd be different
[15:39:06] <les_w> yup
[15:44:35] <Jacky^> I like Ray philosophy, he think with free software in mind, thats is for me the voice of wisdom
[15:45:20] <Jacky^> John is an Hacker, nothing to say ..
[15:46:50] <Jacky^> Alex is a young man with great knowlegde
[15:47:08] <Jacky^> dont know much abot the other guys
[15:47:59] <Jacky^> les_w: got the snow there ?
[15:49:37] <Jacky^> here rain, damn rain , in one day all rain that fall in a year :/
[15:51:55] <Jacky^> we got a sea on the head
[15:52:49] <Jacky^> i'm really thinking to join greenpeace group in contribuite to save the planet !
[15:54:08] <Jacky^> E. 120 for a year.. are not few for my budget ghghgh
[15:54:35] <alex_joni> * alex_joni goes away for a couple of hours
[15:56:47] <Jacky^> * Jacky^ relax playng some mp3 guitar song
[15:59:06] <anonimasu> hello
[15:59:19] <Jacky^> hello anonimasu
[15:59:43] <anonimasu> what's up?
[15:59:48] <anonimasu> I am on my way back home now
[15:59:59] <Jacky^> planning what to eat this evening :P
[16:00:30] <Jacky^> its cold enough, just need something of strong
[16:00:56] <anonimasu> hm ok
[16:00:58] <Jacky^> like mexican food ..
[16:01:01] <anonimasu> I am waiting for my flight
[16:01:08] <anonimasu> doing code for work :/
[16:01:23] <anonimasu> I am going to order ballscrew for the Z axis tomorrow
[16:01:30] <Jacky^> good luck then
[16:01:35] <Jacky^> :)
[16:01:42] <anonimasu> oh I dont need luck
[16:01:46] <anonimasu> I need more $
[16:01:47] <anonimasu> :D
[16:01:50] <Jacky^> hehehe
[16:01:59] <Jacky^> just ask to your bank
[16:02:09] <Jacky^> i'm sure they will understand
[16:02:23] <Jacky^> especially if theyre italian bank affiliated :P
[16:02:34] <Jacky^> how much you need ?
[16:02:57] <anonimasu> enough?
[16:02:59] <anonimasu> ^_^
[16:03:00] <anonimasu> not really
[16:03:23] <Jacky^> 80 million ?
[16:03:29] <anonimasu> yeah
[16:03:31] <anonimasu> that'd be nice
[16:03:34] <Jacky^> good, np
[16:03:45] <Jacky^> you can get it in a week
[16:04:16] <Jacky^> the bank have a really nice printer
[16:04:26] <Jacky^> 2000 copies/min
[16:04:31] <Jacky^> :D
[16:05:07] <anonimasu> haha
[16:05:12] <anonimasu> you think they'll print it for me?
[16:05:12] <Jacky^> :)
[16:05:16] <Jacky^> sure
[16:05:35] <Jacky^> just firm a piece of paper
[16:06:29] <anonimasu> it just depends on how much force you apply
[16:06:40] <Jacky^> no problem
[16:07:06] <Jacky^> we have a lot of criminal guys arond italian and european parliament !
[16:07:10] <Jacky^> just ask
[16:07:13] <Jacky^> :D
[16:07:49] <anonimasu> hm
[16:07:51] <Jacky^> at the moment about 23
[16:07:52] <SWPadnos_> "excuse me , Italian and European parliament member - are you a criminal?"
[16:07:53] <anonimasu> I wonder if mastercam x is good
[16:08:07] <SWPadnos_> hm - no answer
[16:08:12] <Jacky^> SWPadnos_: yeah.. dont you know ??
[16:08:30] <SWPadnos_> yeah, but I'm sure you wouldn't get the correct answer if you just ask ;)
[16:08:37] <Jacky^> http://www.beppegrillo.it/immagini/beppe_ht.pdf
[16:08:44] <Jacky^> read.. and learn :P
[16:09:59] <Jacky^> so anonimasu dont worry.. there's a solution for all
[16:10:50] <Jacky^> you can learn more about Fazio (our bank national minister)
[16:11:02] <Jacky^> money are not a problem !
[16:11:24] <anonimasu> haha
[16:11:27] <Jacky^> :)
[16:11:50] <les_w> hmm I was just reading a paper about latencies in older pcs vs new faster ones with os like RTlinux
[16:12:19] <SWPadnos_> what paper?
[16:12:40] <les_w> as suspected, an old few hundred Mhz pent pro and a 1.6 Ghz P4 are about the same
[16:12:47] <les_w> just a sec
[16:13:10] <Jacky^> les_w: you mean compared to centrino ?
[16:13:31] <SWPadnos_> no - more or less any modern CPU sucks
[16:13:40] <Jacky^> I think they use different unit to measure performance..
[16:14:22] <les_w> os.inf.tu-dresden.de/papers_ps/rtss2002.pdf
[16:14:32] <SWPadnos_> yes - new machines are rated on "throughput" - how fast can you make mp3's or render video frames
[16:14:49] <Jacky^> its just a commercial strategy
[16:14:56] <Jacky^> they change the words
[16:15:03] <SWPadnos_> not on latency - any machine nowadays is fast enough to provide realtime latencies for multimedia applications
[16:15:06] <Jacky^> to confuse the users
[16:15:16] <les_w> well interrupt latency seems to bottom at about 20-30 microseconds
[16:15:36] <SWPadnos_> but "realtime latencies" means 16 milliseconds for a video (running at 60 FPS)
[16:15:42] <Jacky^> the secret is to CHANGE THE WORDS
[16:15:56] <Jacky^> changing the words you can change the world
[16:15:57] <SWPadnos_> it may mean 10 ms or more for audio (since you can send blocks of data to the audio card)
[16:16:02] <les_w> so every message from rt to user prob gets the 20 to 30
[16:16:07] <SWPadnos_> no - it's just different measurements
[16:16:39] <SWPadnos_> there was a recent papaer talking about the number of CPU cycles it takes to create a process (not the same, but still interesting)
[16:16:49] <les_w> yeah audio and video (if recorded) are deterministic
[16:16:58] <SWPadnos_> Linux, which was by far the best, was just under 1 million clock cycles for process creation
[16:17:03] <les_w> pid motion control is not
[16:17:05] <SWPadnos_> !
[16:17:06] <Jacky^> id try with a true benchmark apps
[16:17:38] <SWPadnos_> yes - even recording video is deterministic - you only get one frame every 24/25/30/60 of a second
[16:17:57] <SWPadnos_> (or 1/50)
[16:18:48] <anonimasu> hm
[16:18:49] <les_w> I was trying to call the opencnc guys
[16:18:55] <les_w> wanted to read up first
[16:18:55] <anonimasu> ^_^
[16:19:00] <Jacky^> :))
[16:22:17] <SWPadnos_> I'd like to see some of these tests done with an external measuring device - trigger a scope on the interrupt-causing signal, then output a bit in the handler
[16:22:41] <Jacky^> yeah, woul be interesting
[16:22:43] <SWPadnos_> they always measure with internal timers, which seems "icky" to me
[16:23:38] <anonimasu> hm
[16:23:41] <anonimasu> I just belive my scope
[16:23:53] <Jacky^> but, talking about cpu performance, i wonder about HD instead ..
[16:23:57] <SWPadnos_> me too, but they almost never use a scope when measuring latencies
[16:24:05] <anonimasu> hm
[16:24:08] <anonimasu> laters guys
[16:24:12] <SWPadnos_> see you
[16:24:15] <anonimasu> I am leaving for my flight in a couple of minutes
[16:24:17] <Jacky^> bye anonimasu
[16:24:21] <anonimasu> I hope I'll land safely ;)
[16:24:29] <Jacky^> ;)
[16:24:36] <SWPadnos_> keep up those arm exercises ;)
[16:24:52] <anonimasu> um honestly, I am more afraid of getting up in the air..
[16:24:59] <anonimasu> you always come down ;)
[16:25:37] <les_w> Yeah, I always promise my passengers that I will get them back on the ground.
[16:26:09] <anonimasu> hehe
[16:26:11] <Jacky^> I prefer no comment about planes ..
[16:26:17] <anonimasu> Jacky^: afraid?
[16:26:26] <Jacky^> enough
[16:26:44] <Jacky^> some week ago, 2 guys from an italian university
[16:26:45] <SWPadnos_> I love to fly
[16:27:04] <SWPadnos_> got 300k miles in the air (on a single airline)
[16:27:05] <Jacky^> they get a 'special software' from US
[16:27:27] <Jacky^> and discovered something about DC9 crash in ustica
[16:27:34] <Jacky^> a lote of years ago ..
[16:27:43] <Jacky^> but its annoing topic
[16:27:59] <anonimasu> Jacky^: what do you mean?
[16:28:01] <anonimasu> I dont get you
[16:28:08] <Jacky^> anonimasu: nothing :)
[16:28:18] <Jacky^> dont worry ! damn
[16:28:39] <anonimasu> I am not :D
[16:28:41] <anonimasu> I were kidding
[16:28:47] <Jacky^> hehe right ;)
[16:30:14] <Jacky^> les_w: id like to fly with you, really
[16:30:23] <Jacky^> bust just with you*
[16:30:27] <Jacky^> but
[16:30:39] <anonimasu> laters
[16:30:56] <Jacky^> anonimasu: later
[16:31:23] <Jacky^> les_w: should I pay E. 250 to fly from rome to milan ?
[16:31:27] <Jacky^> to do what ?
[16:31:33] <Jacky^> to buy a pilot ?
[16:31:45] <Jacky^> would be better an hostess then :P
[16:32:56] <Jacky^> at least for a week :D
[16:33:05] <les_w> I have no idea about rates
[16:33:41] <Jacky^> les_w: yes, I can corfim
[16:33:47] <les_w> I just know all money is getting worthless...just as I am starting to make some
[16:33:50] <Jacky^> Alitalia is died too ..
[16:34:04] <Jacky^> :((
[16:34:48] <Jacky^> les_w: is it possible to calculate the cost of fuel for km ?
[16:35:12] <Jacky^> I mean with a small airplane
[16:35:20] <les_w> just guesses...it is a grear deal
[16:35:43] <les_w> 100 's of kg per passenger is common
[16:36:03] <les_w> let me check
[16:36:08] <Jacky^> yeah
[16:36:41] <Jacky^> theyre buildin planes with 800 passengers up ..
[16:36:49] <Jacky^> building*
[16:37:18] <Jacky^> that seem to me absurd
[16:38:47] <les_w> found it
[16:39:14] <les_w> 150-300 kG per passenger for transatlantic flight
[16:39:48] <SWPadnos_> it's the exponential fuel usage that kills you ;)
[16:39:54] <les_w> was about 500 kG for concorde
[16:40:54] <les_w> 747 and airbus 300 most efficient
[16:40:58] <Jacky^> okay
[16:41:40] <SWPadnos_> the 747 has a max takeoff weight of 394 tonnes, and roughly the same number of passengers, so about one (metric) ton per passenger, it seems
[16:41:56] <SWPadnos_> including fuel
[16:42:01] <les_w> yeah
[16:42:22] <SWPadnos_> you'd think they could make the seats bigger, if we're all about a ton ;)
[16:43:19] <les_w> funny, the light planes I fly often use standard economy class highback seats
[16:43:39] <les_w> but more legroom
[16:44:07] <SWPadnos_> Some of those are pretty comfortable. it's the 350-pounder in the seat next to you that makes it really difficult.
[16:44:19] <les_w> right
[16:44:23] <les_w> 19 inches wide
[16:45:42] <SWPadnos_> I think the 777 was an extra 3/4 inch - really makes a difference
[16:47:04] <Jacky^> SWPadnos_: im for the small planes
[16:47:09] <les_w> well, I have a bit of medium range travel to do soon. I'm gonna drive mostly
[16:47:15] <les_w> I love to fly too...
[16:47:20] <SWPadnos_> I like big ones - they're a lot more comfortable in bad weather
[16:47:30] <les_w> but cannot stand the security and waiting
[16:47:36] <SWPadnos_> nope - that's no fun
[16:47:51] <SWPadnos_> you should look into using a small plane, and fly into the regional business terminals
[16:47:53] <les_w> So that is why I am car shopping lately
[16:48:12] <Jacky^> SWPadnos_: yeah, I know ..
[16:48:24] <les_w> swp I flew small planes to business meeting some
[16:48:30] <les_w> but it is dangerous.
[16:48:34] <Jacky^> but 800 passengers plane is wrong too ..
[16:48:38] <SWPadnos_> it's the "last mile" that gets you ;)
[16:48:56] <SWPadnos_> yeah - they need about 6 extra doors fro an 800-passenger plane
[16:49:08] <SWPadnos_> just for normal loading / unloading
[16:49:08] <les_w> because you can't fly through weather, and can't depend on an arrival time
[16:49:12] <SWPadnos_> true
[16:49:15] <Jacky^> * Jacky^ nods
[16:49:37] <SWPadnos_> for short-ish hops, it should be OK (over 60 miles, less than 350 or so)
[16:49:44] <les_w> In very nice weather I would fly things like chicago to indy...stuff like that
[16:50:02] <les_w> 3 or 4 hour drive...45 min in a small plane
[16:50:38] <les_w> usually only on clear winter days
[16:50:46] <les_w> summer is bumpy air
[16:51:06] <SWPadnos_> obviously, you haven't lived in Vermont in the winter ;)
[16:51:13] <les_w> haha
[16:51:45] <les_w> summer unstable air can really close in on you fast
[16:52:02] <les_w> storms in 15 minutes
[16:52:05] <SWPadnos_> as can a nor-easter
[16:52:21] <les_w> yeah.
[16:52:41] <les_w> I got stuck VFR in a snow shower once.
[16:52:50] <les_w> np...just flew out of it.
[16:53:08] <SWPadnos_> at least you didn't fly into a mountain
[16:53:20] <SWPadnos_> reminds me of a great Far Side comic
[16:53:43] <les_w> I was lower than the highest terrain feature...forced down by class B airspace
[16:53:50] <les_w> so not a good feeling
[16:54:12] <les_w> ?
[16:55:23] <les_w> I think i remember the far side one
[16:56:02] <SWPadnos_> "Wha - what's that mountain goat doing up here in a cloud bank?"
[16:56:09] <les_w> yeah
[16:56:11] <les_w> haha
[16:56:25] <Jacky^> :P
[16:56:34] <les_w> well the feature in my case was the WGN radio tower
[16:56:48] <les_w> I was below the top and 2 miles west
[16:56:59] <les_w> continued on a heading of 270
[16:57:28] <SWPadnos_> easier to avoud than a 4400-foot mountain peak, I'd wager
[16:57:32] <Jacky^> les_w: waht abou flylng up to a jail ?
[16:58:11] <les_w> flying up to a jail?
[16:58:19] <Jacky^> yeah
[16:58:22] <les_w> haven't doe that...yet
[16:58:26] <Jacky^> is there a cross ?
[16:58:47] <Jacky^> how can you understand that
[16:58:56] <les_w> hmmm
[16:59:31] <Jacky^> in italy they use a cross
[16:59:44] <les_w> for what?
[16:59:44] <Jacky^> its called 'croce di san andrea'
[16:59:52] <Jacky^> as signal
[17:00:06] <Jacky^> you cant fly up there
[17:00:11] <les_w> ohhhh
[17:00:19] <les_w> we just have maps
[17:00:24] <Jacky^> otherwise they shot you
[17:00:24] <les_w> and gps
[17:00:30] <Jacky^> oh ok ..
[17:00:59] <les_w> yes we have many areas now where you will get shot down if you enter and continue
[17:01:18] <Jacky^> how you knoe that ?
[17:01:24] <Jacky^> just with the gps ?
[17:01:59] <les_w> The FCC informs me because I am a pilot
[17:02:14] <les_w> yes some gps with databases will inform you
[17:02:17] <Jacky^> oh. ok
[17:03:52] <Jacky^> I think to get pilot license is expensive then
[17:04:18] <Jacky^> and for what is happened not so simple too now
[17:04:34] <les_w> it is expensive.
[17:04:42] <les_w> I did it 10 years ago.
[17:05:23] <Jacky^> :-)
[17:05:27] <Jacky^> good
[17:06:25] <Jacky^> les_w: have you seen the head of my robot ?
[17:06:41] <Jacky^> the robot i'm building :P
[17:07:11] <Jacky^> http://www.frappr.com/emctheenhancedmachinecontroller/photo/686052
[17:07:30] <Jacky^> it just start to move the head touching it
[17:07:30] <Jacky^> very funny :)
[17:07:33] <les_w> looking
[17:07:55] <Jacky^> it should use a Linux kernel too
[17:08:40] <Jacky^> i think ill finish it in the 2006
[17:08:44] <les_w> heh
[17:08:47] <Jacky^> :)
[17:09:25] <Jacky^> dont you have any pic o frapp yet ?
[17:09:52] <les_w> frapp?
[17:10:09] <Jacky^> yeah, why dont you upload at least one ?
[17:10:19] <Jacky^> http://www.frappr.com/emctheenhancedmachinecontroller
[17:10:30] <Jacky^> youve a lot of nice photos :)
[17:11:24] <les_w> oh, yeah. I'll try to put something up
[17:11:30] <Jacky^> anna is joined too :)
[17:11:54] <Jacky^> dinnertime
[17:12:08] <Jacky^> Jacky^ is now known as Jacky^dinner
[17:12:33] <les_w> later
[17:12:39] <les_w> oh I see anna
[17:12:44] <les_w> haha
[17:12:47] <Jacky^dinner> :)
[17:12:57] <Jacky^dinner> nice photo but its not naples
[17:13:13] <Jacky^dinner> she was in florence ;)
[17:13:22] <Jacky^dinner> later
[17:20:37] <etla> hi imperator, any progress with the mesa cards lately ?
[17:20:59] <Imperator_> Hi etla
[17:21:15] <Imperator_> yes, the counter board is working well
[17:21:41] <etla> you made an external encoder counter ?
[17:21:49] <etla> or on the mesa itself ?
[17:22:15] <Imperator_> the DAC Board comes next, but i have to figure out how to program the FPGA
[17:22:52] <Imperator_> there are four (optional six RS422 recivers on that board. Counting does the FPGA
[17:23:07] <etla> peter w at mesa will probably help you if you ask
[17:23:31] <Imperator_> ok, good to know
[17:23:45] <etla> can't you take the pwm+dir and drive the DAC with that ?
[17:23:56] <Imperator_> changing the FPGA Software was easy, but for the DAC I need a new part
[17:24:26] <Imperator_> you can use PWM and Dir to produce a analog signal
[17:24:51] <Imperator_> the DAC gets directly 16 bit Words from the FPGA
[17:25:25] <Imperator_> a real DAC is much more acurate and much faster than using PWM to get a analog signal
[17:25:38] <SWPadnos_> Imperator_: what software are you using to edit and synthesize the FPGA?
[17:25:48] <etla> ok.. so you want your dac to get the voltage command with 16 parallel lines
[17:25:50] <Imperator_> Hi Stephen
[17:25:59] <Imperator_> the Xilinx Software
[17:26:04] <SWPadnos_> oh yeah - hi :)
[17:26:42] <SWPadnos_> ok - I just bought the OpenTech CD set from OpenCores.org, but unfortunately, most of the software is for Windows
[17:26:45] <Imperator_> btw. m unhappy that you got not enough votes to join the board
[17:26:47] <SWPadnos_> it's kind of nnoying
[17:26:56] <etla> xilinx has the compiler on the web
[17:27:04] <etla> yes I think it's only for windows
[17:27:15] <Imperator_> jep
[17:27:18] <SWPadnos_> heh - no sweat - this leaves me free to say things that a respectable board member shouldn't :)
[17:27:30] <Imperator_> Spephen you have also a Mesa Card, right ?
[17:27:38] <Imperator_> :-)
[17:27:57] <Imperator_> Stephen
[17:28:00] <SWPadnos_> yes
[17:28:19] <SWPadnos_> I haven't plugged it into my main work machine yet - I'll probably do testing in a separate computer
[17:28:27] <SWPadnos_> (which I need to fix)
[17:28:28] <Imperator_> good, any expiriance with programming the FPGA ?
[17:28:42] <Imperator_> ok
[17:28:54] <SWPadnos_> none
[17:29:08] <SWPadnos_> I will try to get the Altera and Xilinx tools installed under Linux
[17:29:27] <SWPadnos_> I have their dev boards, so I can test some stuff anyway
[17:29:39] <Imperator_> ok, but the Xilinx Software for Linux is not for free
[17:30:03] <SWPadnos_> I may have the Linux version with the devkit
[17:30:16] <Imperator_> ok
[17:31:00] <Imperator_> i have no problems with windoze, i use always the OS that is best for what i want to do
[17:31:17] <SWPadnos_> nope - at least I have a license number for the software - I can probably convince them to let me use Linux
[17:31:29] <Imperator_> cool
[17:31:30] <SWPadnos_> yeah - that's why I have VMWare
[17:31:47] <Imperator_> hehe, very good
[17:31:50] <SWPadnos_> I have a BDI install and Windows 2000 in virtual machines
[17:31:59] <Imperator_> nice
[17:32:20] <Imperator_> but i need windows also for CAD, so i need maximum performance
[17:32:23] <SWPadnos_> of course, realtime isn't really realtime in a virtual machine ;)
[17:32:53] <Imperator_> BDI inside Windows or vise versa
[17:33:10] <SWPadnos_> BDI and Windows 2000 inside of Ubuntu AMD64
[17:33:15] <SWPadnos_> also puppy
[17:33:39] <Imperator_> ok
[17:33:56] <etla> do you know anyone who uses the mesa card to drive a machine already ?
[17:34:23] <SWPadnos_> I don't
[17:34:34] <SWPadnos_> not that I would know if someone were
[17:34:53] <etla> would be nice to hear som experiences from someone...
[17:35:07] <etla> alhtough it's so cheap that I can just buy one and try it
[17:35:18] <SWPadnos_> that was my view as well
[17:35:32] <SWPadnos_> I have Jon Elson's USC (which now works with emc2)
[17:35:44] <SWPadnos_> that's what I expect to use on my machine
[17:35:57] <etla> I'm going with servos for next machine
[17:36:07] <SWPadnos_> Mariss sent me a G-Rex because I wrote a downloader for it
[17:36:13] <SWPadnos_> and I have the 5i20 as well
[17:36:28] <SWPadnos_> so it should be fun doing all sorts of development ;)
[17:36:35] <etla> time to start buying linears and ballscrews instead of electronics ??
[17:37:18] <SWPadnos_> nah - Ive got a Bridgeport with chrome ways, and already did the ballscrew retrofit
[17:37:55] <etla> ok
[17:38:59] <etla> as soon as the servos I've ordered arrive I'll try to hack together a drive for them.
[17:39:14] <etla> then when the drive works it's time to order a servocard, probably the mesa
[17:40:03] <SWPadnos_> note that you'll want some of the isolation cards as well (were we discussing this before?)
[17:40:18] <etla> yep!
[17:40:25] <SWPadnos_> OK - thought so :)
[17:40:38] <etla> probably one for the drives
[17:40:49] <etla> and an extra one for misc I/O maybe
[17:41:24] <SWPadnos_> yep - check the speed of the optos though - they may not be fast enough for PWM (of course, an analog card would be)
[17:41:43] <etla> hopefully by then you will all be up to speed on programming the fpga so I can benefit from that :)
[17:42:05] <SWPadnos_> hopefully
[17:42:19] <etla> yeah, 8000 count encoders with 3000rpm motors is something like 400kHz pulse rate
[17:42:33] <etla> that should go through an optoisolator unmodified ??
[17:42:57] <SWPadnos_> possibly, but that's actually fairly high speed for an optoisolator
[17:43:08] <etla> the torque command would be maybe 10kHz PWM so lower requirement
[17:43:19] <SWPadnos_> though you only have 2000 cycles (or 4000 transitions) per phase
[17:43:40] <skunkworks> where is the history from this irc channel stored? i remember seeing a link somewhere but can't find it this morning. thanks
[17:43:43] <SWPadnos_> most encoders aren't that fast, by the way
[17:43:51] <SWPadnos_> logger_aj: bookmark
[17:43:51] <SWPadnos_> See http://solaris.cs.utt.ro/irc/irc.freenode.net:6667/emc/2005-11-28#T17-43-51
[17:43:58] <SWPadnos_> try there :)
[17:44:24] <SWPadnos_> go up onw level, and you'll get an index of all the logs
[17:44:26] <SWPadnos_> one
[17:44:49] <skunkworks> thank you - that is what I was looking for
[17:45:04] <Imperator_> Stephen that extra boards are exactly what we are making here
[17:45:15] <skunkworks> it opened to where I wanted to be also. spooky ;)
[17:45:19] <etla> hmmm looking at the mesa optoisolator board manual
[17:45:21] <SWPadnos_> heh
[17:45:30] <etla> outputs turn on 2us off 6us
[17:45:33] <Imperator_> but we want real industrial quality, thats why we don't bought the breakout boards form Mesa
[17:45:38] <SWPadnos_> right
[17:45:39] <etla> inputs on 6us off 25us
[17:45:49] <SWPadnos_> that's the problem I have with those boards as well.
[17:46:11] <SWPadnos_> there should be an external "breakout box" (like you get with video editing hardware)
[17:46:17] <Imperator_> we don't use optos for the encoders, that is not needed
[17:46:30] <SWPadnos_> that should have opto-22 (or equivalent) sockets for outputs, and isolated inputs
[17:46:34] <Imperator_> but for IO. But ther you have no speed problems
[17:47:20] <SWPadnos_> if you want isolation of controls from the machine, then you have to isolate the encoders as well
[17:47:27] <SWPadnos_> but full isolation may not be necessary
[17:47:46] <Imperator_> the encoders have no conection to the machine
[17:48:28] <etla> but when something goes wrong and you get the 200VDC on the encoder line ?? what will blow ?
[17:48:36] <SWPadnos_> that should be true, but the wires are still long antennas
[17:48:40] <Imperator_> also the DAC outputs, they are differential, that is enough. But for the DAC it would make much more sense as for the encoders
[17:49:28] <Imperator_> right, we use differential signals for the encoders. and on version two we ad also much more stuff for noise reduction
[17:51:54] <etla> hmm.. a lot of work to build external counter cards and DAC...
[17:52:14] <etla> I'm planning on using the mesa card builtin encoder counters and pwm+dir for torque command
[17:52:23] <Imperator_> we have now some special coils for suppressing noise on differential lines. Cant descripe it in english well. See http://www.tdk.co.jp/tefe02/e971_zjys.pdf
[17:52:23] <etla> nothing but the servodrives to build myself
[17:52:28] <etla> + isolation in between
[17:53:23] <Imperator_> buy profesional servo drives at ebay !!! They are much better than all the homebuild stuff
[17:53:44] <SWPadnos_> cool - nice parts.
[17:53:48] <Imperator_> building a DAC and counter card is a much easyer task
[17:53:49] <SWPadnos_> what is the cost in quantity?
[17:54:31] <Imperator_> don't know, my companion got them
[17:54:45] <SWPadnos_> nevermind - as little as $0.45 in quantity at DigiKey
[17:55:07] <SWPadnos_> closer to $1.50 for low quantities
[17:55:18] <Imperator_> we want to have it realy save
[17:55:52] <Jymmm> Imperator_: go find a broken down truck
[17:56:12] <Imperator_> and then ?
[17:56:24] <icee> etla: how are you going to do commutation?
[17:56:27] <Jymmm> Imperator_: you'll save a WHOLE LOT!
[17:56:46] <etla> hi icee !
[17:56:54] <icee> to do good commutation/vector control you need the encoder value
[17:57:12] <icee> hey there
[17:57:20] <etla> icee: I have not started on the project yet. waiting for my servos to arrive. sitting back and looking at what others are doing for the moment
[17:57:38] <etla> yes the drive would read the encoder and get the rotor angle from there
[17:57:51] <Imperator_> vector control is needed for asyncronous motors, the one used for driving the joints are syncrounous
[17:58:24] <Imperator_> synchronous
[17:58:29] <etla> or, if youre crazy enough you would do away with the drives and write the commutation code for the fpga and have the fpga output the phase pwm's directly...
[17:58:56] <icee> doing commutation is easy enough
[17:59:04] <icee> I have a cordic implementation I wrote in HDL if you want it
[17:59:13] <icee> VHDL
[17:59:28] <etla> thanks, I'm not at that stage yet.
[17:59:47] <etla> I'll try to get the motors moving with the dsPIC+IRAMS combination first
[17:59:53] <icee> lawrenceg has the encoder counting on the DSPIC working
[18:00:06] <icee> etla: something we've learned is that heatsinking of the IRAMS is exceptionally critical
[18:00:16] <etla> burned any yet ?
[18:00:17] <Imperator_> driving a motor is easy, but a servo drive is much more !! That is very precise stuff
[18:00:21] <icee> to do 10A, for instance, on the IRAMY, it dissipates about 130W -and- case temperature needs to be kept below 80C
[18:00:27] <icee> no, purely from FEA/thermal modelling
[18:00:51] <icee> e.g. i'm planning to build special ductwork and use long heatsink extrusions to get acceptable heat removal
[18:01:10] <etla> it's like a modern cpu then...
[18:01:10] <Imperator_> build a good current controller, the rest can be done with the FPGA on the mesa card
[18:01:36] <icee> the thing is, if you have a dspic already, and you're doing current control there..
[18:02:03] <icee> it's not much more work to use the internal quadrature decoder to get rotor position and to use that to do a sine table lookup
[18:02:18] <etla> yes.
[18:02:21] <icee> if you just want the dspic to be a current controller, iMO you would be better off building analog circuitry to do the same thing
[18:02:23] <icee> it'll work better
[18:02:39] <etla> my only concern at this planning stage is the resolution of the current feedback
[18:02:51] <etla> people were suggesting the allegro acs704
[18:03:02] <etla> and digitize that with dsPIC 10-bit ADC
[18:03:11] <icee> is that a hall?
[18:03:15] <etla> yes
[18:03:21] <icee> We've moved a little bit to putting halls on two motor phases
[18:03:21] <etla> you get isolation for free...
[18:03:34] <icee> I don't think that was the one i selected, but it was like $2 at digikey
[18:03:55] <icee> we've changed our isolation strategy too
[18:03:56] <etla> yes, two sensors keeps the algorithms for the future open (svm etc.)
[18:04:02] <icee> we've moved to putting the micros on motor ground potential
[18:04:06] <icee> and talking to them over an opto'd SPI
[18:04:29] <etla> ok, with one dsPIC doing the epp stuff
[18:04:33] <icee> so 1 communications micro on a floating ground and an independent floating 5V supply
[18:05:03] <icee> and 4 axis micros on the motor common
[18:05:04] <icee> yah
[18:05:20] <icee> the main reason to do that isn't performance (it's actually worse for latency, etc), but it saves a lot of money and complexity on isolation
[18:05:50] <Jacky^dinner> Jacky^dinner is now known as Jacky^
[18:05:57] <etla> did you get some epp comm working already ?
[18:06:11] <icee> oh yah, that's been working for a long time now
[18:06:20] <icee> lg has built a board that's run my software
[18:06:30] <icee> and pc software has issued address/data commands to turn leds on and off and to read values
[18:06:32] <icee> from memory
[18:06:39] <etla> and you can hang all the axes off the same spi bus ?
[18:06:42] <Jacky^> hey guys
[18:06:44] <icee> sure
[18:07:04] <etla> cool. hope you will release the source and schematics !
[18:07:07] <icee> 2 opto'd lines for the SPI from EPP->all axes, plus 1 select opto'd line per axis
[18:07:21] <icee> I will.. I'm not going to make the power stage schematics generally available though
[18:07:28] <icee> because i'm concerned about people cooking themselves
[18:07:48] <etla> yeah.
[18:08:11] <Imperator_> we will make our stuff public
[18:08:15] <icee> I have only one big issue to work still before I can start wrapping things up and ordering boards
[18:08:30] <etla> what clock does the spi run at ? is there time for torque command+ encoder count signaling for all axes
[18:08:32] <icee> and that's the power supply-- inrush limiting, whether to do PFC, etc
[18:09:09] <Imperator_> Imperator_ is now known as Imperator_Away
[18:09:14] <icee> et: I'm thinking it'll probably be roughly 4MHz, full duplex. so enough to do a 40 bit update, each axis at 25khz
[18:10:10] <icee> since it'll be asynchronous to the EPP though, with EPP at 2.5KHz and it at 10-25KHz, we'll lose some performance, but it should be acceptable
[18:10:20] <icee> there's no real way to couple them together tighter
[18:11:02] <etla> btw what current sensor are you planning on using ?
[18:11:17] <icee> let me see if i can find that easily in my parts folder
[18:11:21] <icee> i can't look at the schematic from down here
[18:12:30] <etla> did you model the emc-suppression caps for the DC bus voltage
[18:12:48] <icee> EMC suppression isn't that bad
[18:12:58] <etla> really?
[18:13:00] <icee> dealing with ripple currents, etc, requires a lot of bulk capacitance though
[18:13:15] <icee> the IGBTs don't switch that fast-- the edge rates are not that high
[18:13:26] <etla> but those big caps are in the PSU ?
[18:13:48] <icee> the big caps need to be close to the power supply for stability, yes
[18:13:58] <icee> undetermined whether the power supply is on the amplifier board or not at this point
[18:14:12] <icee> if i do a simple rectifier, it will be; if I do a fancy PFC boost converter, it won't
[18:14:38] <etla> I'm thinking I don't want to learn chopper design right now also so maybe linear psu ?? any thoughts ?
[18:14:56] <icee> linear? for the bus?
[18:15:07] <etla> well, unregulated
[18:15:13] <icee> if you want a 170V bus, you just rectify 110V and put a bunch of caps on it
[18:15:24] <icee> (i mean, there's a little more to it than that, but..)
[18:15:31] <etla> yep, that what I was thinking.
[18:15:32] <cradek> and if you're smart, you isolate it
[18:15:41] <SWPadnos_> oh no - not this again ;)
[18:15:46] <etla> I have 230V so an 230/110 transformer inbetween, that will isolate
[18:15:48] <cradek> uh-oh, did I come in late?
[18:15:54] <alex_joni> no, not if you're smart.. if you want to live :D
[18:15:58] <alex_joni> hi guys
[18:16:00] <SWPadnos_> about 6 days late for that conversation
[18:16:00] <cradek> hello
[18:16:05] <SWPadnos_> hi Alex
[18:16:14] <alex_joni> * alex_joni is half-dead :(
[18:16:18] <icee> swp: well, i finally found a cheap isolation transformer, so, mine will be isolated
[18:16:23] <alex_joni> think I catched a cold
[18:16:23] <SWPadnos_> phew ;)
[18:16:26] <cradek> your remaining half must include the typing parts
[18:16:38] <SWPadnos_> bummer - have some chicken soup, and you'll feel better
[18:16:44] <SWPadnos_> or throw up
[18:16:52] <icee> etla: you will need to deal with a lot of ripple current though
[18:16:57] <alex_joni> I think the fingers experience some weird reflexes..
[18:16:58] <icee> and you'll need to size your caps based on that
[18:17:12] <icee> and then, because you have a ton of capacitance, the conduction angles will be small, and so your power factor will be low
[18:17:14] <icee> 0.4-0.6
[18:17:48] <etla> supid q: and why is that bad ?
[18:18:09] <icee> well, it is inefficient and makes you need to use a much bigger transformer, etc, and depending on the harmonics might upset other stuff in your house
[18:18:27] <icee> or shop or whatever
[18:18:52] <icee> you also need to deal with the inrush currents in some way or another
[18:19:11] <etla> damn... designing a DIY chopper is one more task I'm not sure I want to do...
[18:19:23] <etla> inrush when the motor starts ?
[18:19:24] <icee> there's a very simple IRF PFC part
[18:19:27] <SWPadnos_> http://www.geckodrive.com/
[18:19:28] <icee> i'm still not going to do it necessarily
[18:19:35] <icee> no, inrush to charge the capacitors on the bus
[18:19:43] <etla> SWP: not for 3-phase AC servos
[18:19:48] <icee> in my case, 13 millifarads * 170 volts is 2.24 coloumbs
[18:19:49] <SWPadnos_> not yet, anyway
[18:19:59] <SWPadnos_> in that case, http://www.rutex.com/
[18:20:12] <etla> rutex: not sinusoidal commutation
[18:20:51] <icee> most of the charging happens in one AC cycle because the input is pretty low impedance
[18:20:57] <SWPadnos_> ok - those are for BLDC, not brushless AC?
[18:21:00] <icee> 13200 microfarads * 170 volts / 16 milliseconds = 140 amps
[18:21:03] <icee> with much higher peak currents
[18:21:12] <icee> e.g. enough to pop the magnetic portion of a circuit breaker
[18:21:26] <etla> so some kind of soft start to charge the caps
[18:21:33] <icee> yah. there's a few options
[18:21:55] <icee> complicated active circuits with SCRs, a big resistor with a relay set to fire when the bus voltage gets over a certain level, or a varistor
[18:22:42] <icee> you also need to design in bleeder resistors to discharge the capacitors
[18:22:51] <icee> and be sure to make them redundant.. -and- not to trust them
[18:22:56] <icee> and calculate the appropriate time constant for them
[18:23:11] <etla> one more darwin award nominee :)
[18:23:21] <icee> http://www.ametherm.com/Inrush_Current/MegaSurge%20Inrush%20Current%20Limiters.htm
[18:24:18] <icee> didn't find the halls, sorry, i'll get back to you on that when i have the design handy
[18:24:53] <etla> hmm, how does the resistance vary with voltage/current for that inrush limiter ?
[18:25:10] <icee> when it gets warm, resistance drops
[18:25:21] <icee> that means you need to wait a minute after turning the drive off to turn it back on
[18:25:50] <icee> the varistor inrush limiting isn't as good for that reason as using a relay
[18:25:58] <icee> and because the resistance doesn't vary enough
[18:26:10] <icee> But i'm still heavily considering it because of its simplicity
[18:26:33] <etla> so some kind of comparator looking at the bus voltage and switching out a big limiter resistor
[18:27:03] <icee> well, or you just have a relay set up so that when the coil voltage-- proportional to bus voltage gets high enough-- you close it
[18:27:08] <icee> and bypass the limiter resistor
[18:27:27] <etla> ok, i see.
[18:27:41] <etla> then there's the powerfactor problem still left.
[18:27:42] <icee> or there's SCRs for soft start, with a zero crossing detector
[18:28:01] <icee> see the IR1150S data sheet if you want to deal with PFC
[18:28:07] <icee> what power levels are you thinking of?
[18:28:27] <etla> well I have 3x 400W surpluscenter/sany servos for the axes.
[18:28:32] <etla> and a 1kW for spindle
[18:28:37] <icee> hmm.. a couple kilowatts
[18:28:48] <icee> so it may be cheaper for you to worry about power factor than to buy a bigger isolation transformer
[18:29:19] <icee> IRF has a nice calculator on their site that does most/all of the component selection for the IR1150S
[18:29:31] <icee> the hard part is finding/buying the massive power inductor you need
[18:29:36] <etla> 1kW was biggest 230/110 transformer I found with a quick look, maybe two can be put in parallell ?
[18:29:51] <icee> they'll load share well, if they're the same transformer
[18:30:09] <icee> if your power factor is .5, though, and you want to run 2 kilowatts of motors, you'll need 4 kilowatts of transformer
[18:30:11] <etla> heavy stuff, something like 7kg per toroid
[18:30:55] <icee> (though people -do- cheat on this stuff.. and use smaller rated transformers, i don't think it's a good idea)
[18:31:40] <etla> the 1kW models were around 80eur from farnell
[18:32:14] <icee> you'd never want to ship something like this, but just for reference, for about $200 i've found some 3000VA transformers
[18:32:38] <icee> about 20 kilograms
[18:32:59] <etla> with the PFC chip, do I need the same amount of bus/ripple caps ?
[18:33:07] <icee> just about, yah
[18:33:13] <icee> you don't get a 170V bus voltage with the PFC chip though
[18:33:20] <icee> it's basically a 'boost' converter
[18:33:25] <icee> so you get 200-210V depending on your design
[18:33:43] <icee> it also has soft start, which for various reasons is ineffective in our design
[18:33:53] <icee> so it doesn't solve the inrush problem
[18:34:15] <etla> any idea of how much for the dc bus voltage is neccessary with the IRAMS to get the rated 100vac to the servo ?
[18:34:33] <icee> oh, not much over 100vac
[18:34:48] <icee> er, not much over 100vdc, maybe 120vdc
[18:35:04] <icee> higher voltages make control more difficult
[18:35:18] <icee> you know the motor time constant? that's L/R, and they specify R
[18:35:32] <icee> so you can figure out the motor inductance and come up with current ramp rates at 0RPM
[18:36:07] <icee> so I've done some simulation, and if you say, apply 170VDC across the motor for 1/15000 of a second-- 1 PWM cycle, on my 10A motors that translates to a 3A increase in current
[18:36:19] <icee> decay is a bit slower but still pretty fast
[18:37:27] <etla> where did you get the 1/15000 ?
[18:37:47] <icee> the PWM frequency that the IRAM* parts like
[18:38:22] <etla> ok, so that would be max command. with a really small duty cycle the current change would be much less ?
[18:38:30] <icee> yes
[18:38:51] <icee> what i'm saying is, the gain per cycle isn't ridiculous
[18:39:07] <icee> btw: be careful how you design your control loop
[18:39:20] <icee> you need it to be safe if somehow you lose track of motor position
[18:40:05] <etla> yes, check to see that the encoder count updates frequently enough ?
[18:40:12] <icee> not so big of an issue with halls on two phases, but it's still possible to do stupid things that cause current to run away to infinity
[18:40:34] <icee> well, your control loop should still be safe-- not blow up the motors-- even if you have the wrong encoder position
[18:40:46] <etla> what sw are you using for simulation ?
[18:41:09] <icee> i'm doing mechanicals and thermals in pro/mechanica, and the circuit sim in macspice
[18:41:24] <icee> if you want, i can clean up my 'power supply' sim and send it to you
[18:41:34] <LawrenceG> good morning gentlemen
[18:41:39] <etla> hi
[18:41:42] <alex_joni> good morning
[18:41:43] <alex_joni> evening
[18:41:46] <icee> hey LG :)
[18:41:47] <alex_joni> whatever ;)
[18:41:48] <etla> evening I believe :)
[18:42:11] <LawrenceG> still morning here...just barely
[18:42:52] <icee> lg: i was just talking with etla about power supply issues
[18:43:02] <icee> I'm thinking about doing power factor correction, what do you think?
[18:43:25] <icee> IRF has a part that makes it relatively easy, but there's a massive inductor required
[18:43:32] <LawrenceG> testing is going slowly... work keeps getting in the way...PF... for my 300w units, I'm not going to worry about it
[18:43:44] <icee> * icee nods
[18:43:52] <icee> so you'll size your isolation transformer about double peak load then?
[18:43:54] <etla> is that 1x300W or 3x100W ?
[18:44:01] <icee> double peak sustained load
[18:45:36] <LawrenceG> 3 * 300w motors.... I will look for a transformer with 1 to 2kva rating, hopefully with split primary and split secondary so that I can have a choice of 2 bus voltages
[18:46:18] <LawrenceG> and 120/240 supply options
[18:46:30] <icee> * icee nods
[18:46:40] <icee> with a PF of 0.5, a 1KVA transformer provides 500W
[18:47:07] <icee> i assume you'll end up using the smaller IRAM* modules than me, so i don't know what switching dissipation is like
[18:47:11] <icee> but you need to account for that too
[18:47:14] <etla> so it's 1.5D machining then :)
[18:47:31] <icee> well, you never have all axes doing peak torque and peak speed all the time
[18:47:35] <LawrenceG> servo duty cycle is much less than 1
[18:48:09] <LawrenceG> I have a low mass system (small benchtop lathe/mill)
[18:48:27] <icee> as things are now for me, with PF of 0.5, and switching losses of 70-100W, 4 axes doing nothing could require 800VA of transformer
[18:49:14] <LawrenceG> yea 4kw of motors is nothing to sneeze at
[18:49:46] <icee> the thermal issues i have from all the switching losses-- even when axes are doing nothing -- is substantial, too
[18:49:56] <LawrenceG> I am replacing single stack nema 34 steppers!
[18:50:06] <icee> i need to keep cases below 80C to do 10A, and I need to keep them below 80C when dissipating as much as 130W
[18:50:36] <LawrenceG> I liked your idea of turbo mode.... changing the pwm freq as load changes
[18:51:01] <icee> figuring it'll be 40C in the case, that's .3 degrees/watt-- for each part
[18:51:02] <icee> hehe
[18:51:05] <LawrenceG> At idle they could probably run at 1-2 khz
[18:51:12] <icee> Yah, I plan on doing that, but the gains you can get from that are limited
[18:51:19] <LawrenceG> and cut switching losses in half
[18:51:35] <LawrenceG> or 1/10th
[18:51:38] <icee> well, the switching losses are pretty much proportional to frequency.. so if you cut it down to 5KHz, they'.. yah
[18:52:56] <icee> it's also possible to nail the 'lowest voltage' motor phase to the negative rail
[18:52:59] <icee> and not switch it at all
[18:53:03] <icee> as an optimization
[18:54:42] <icee> still, i need to size things for what the IRF tools tell me
[18:54:47] <LawrenceG> hmmm... that maybe possible using the hardware overrides in the dspic
[18:54:56] <icee> even though it may be a worst case with some PWM frequency slewing etc
[18:55:07] <etla> do you have the irams version that lies flat agains the pcb ?
[18:55:12] <icee> oh, it's certainly possible. it just makes calculating the sine values a little screwy
[18:55:17] <icee> etla: I haven't been able to find that
[18:55:25] <etla> i got the ones that stand up... harder to heatsing maybe
[18:55:25] <icee> and that would make the thermals even harder to deal with
[18:57:15] <icee> my current thoughts are something like: 18" of 4" high aluminum extrusion, with a long slot milled at the bottom to provide clearance for the edge of the board
[18:57:35] <icee> screwed to the bottom of the case for stability, and then with the parts attached with M4 screws tapped into it
[18:57:50] <icee> and then tapering ductwork built out of sheet metal around the sink, with a big fan on one side
[18:57:58] <LawrenceG> I have been doing some test on the encoder.... seem to get some noise on the position value... trying to figure out if it is internal to the dspic or in the epp comm stuff
[18:58:14] <icee> in FEA, the results are -almost- acceptable- for that setup
[18:58:14] <K`zan> too cool: http://forum.saabturboclub.com/viewtopic.php?t=113993
[18:58:33] <icee> what do you mean by noise?
[18:58:46] <icee> is it dithering by a couple points, or.. just getting radically different values
[18:59:52] <LawrenceG> big glitches when reading from the pc in hog mode... while (1) read_encoder();
[19:00:21] <icee> I'm very concerned about the way your (PC) parallel port hardware doesn't report timeouts
[19:00:39] <icee> it means that it isn't implementing the whole EPP state machine
[19:00:56] <LawrenceG> I need to set up a 5khz read rate... I may have to got to rt or dosfor testing
[19:01:42] <LawrenceG> I think my PP has a pullup on the wait line not a pull down....
[19:01:57] <LawrenceG> adding the external pull down causes waits
[19:02:07] <icee> it should timeout whether wait is high or low
[19:02:24] <icee> it should delay starting a transaction until wait is low, and then not finish it until wait is high
[19:02:47] <LawrenceG> I tested it by making the dspic delay before answering and I started geting timeouts
[19:03:20] <icee> well, verify you get timeouts when you delay after answering but before deasserting wait
[19:03:32] <icee> because i really really wouldn't trust it unless you do
[19:04:15] <icee> i'm guessing some, but the fact that you don't get timeouts with no board connected worries me a bit about it
[19:04:32] <LawrenceG> can do..... also need to implement timouts on the dspic side so I know when something goes amiss in the protocol
[19:05:11] <LawrenceG> blip the fault led so I can see what is going on
[19:05:14] <icee> one thing you could try
[19:05:26] <icee> is putting big sleeps inbetween reading different encoder bytes
[19:05:29] <icee> and seeing if that makes it better
[19:05:58] <icee> yah. you could have a command to grab the encoder count, and blink it out in morse code
[19:05:59] <icee> ;)
[19:06:15] <etla> anyone of you done morse code ?
[19:06:18] <icee> i've done a lot of troubleshooting on atmel parts exactly that way
[19:06:30] <etla> I learned 60chars/min in navy B)
[19:06:43] <icee> i'm very very slow, just for amateur radio licensing
[19:06:46] <icee> i can copy 7-8 wpm ;)
[19:06:53] <icee> but it's exceedingly useful for troubleshooting this stuff
[19:07:04] <icee> i can provide routines if anyone wants them
[19:07:24] <etla> youre actually using it for debugging ?
[19:07:38] <icee> sure, if you have code running on there, and a single LED, but don't have better comms like RS-232 working yet
[19:07:40] <icee> it's ideal
[19:07:45] <cradek> I've also considered that... pretty funny
[19:07:54] <LawrenceG> well, I do some morse... this chip does have 2 serial ports that could be used to fire out status messages
[19:08:09] <LawrenceG> or I can expand the epp stuff to read error strings
[19:08:27] <icee> might make sense to develop a console/logging over EPP facility
[19:08:28] <alex_joni> lol..that's funny ;)
[19:08:30] <icee> i could write that for you
[19:08:54] <icee> the other thing the morse is nice for is if you're doing robotics or something
[19:09:23] <icee> you can have it stream out telemetry via morse, when hooking up a serial cable would be impractical
[19:09:41] <cradek> I think reading morse by LED is much harder than by tone
[19:09:46] <cradek> of course I suck at both
[19:09:47] <LawrenceG> not a problem... I have been hacking in all kinds of tests on the epp comm channel
[19:10:24] <icee> i'm also happy to review your code whenever, to see if there's any problems i can catch
[19:10:25] <alex_joni> cradek: how about voice synthesizing on a small speaker :D
[19:10:28] <icee> i know reading the encoder count is tricky
[19:10:41] <icee> cradek: well, you could always put a piezo onboard ;)
[19:11:02] <LawrenceG> comm speed seems to be about 10us/byte with all the handshakes
[19:11:07] <alex_joni> icee: and drive it by pwm
[19:11:12] <cradek> alex_joni: the mp3 player in my car uses voice synthesis for all menus
[19:11:20] <alex_joni> nice
[19:11:23] <cradek> it is nice to not have to look at anything
[19:11:54] <icee> pretty tough to fit decent voice synthesis on a micro though
[19:11:59] <icee> though-- it has been done to a limited extent
[19:12:00] <LawrenceG> I have a nice led show watching the 3phase pwm outputs
[19:12:05] <alex_joni> icee: there are chips that do that
[19:12:24] <cradek> ha, I used Festival
[19:12:59] <jepler> and a full-size PC
[19:13:14] <alex_joni> jepler: why do you spoil the fun :D
[19:13:21] <cradek> I'm not too meek to use brute force
[19:14:07] <cradek> especially if it's very cheap
[19:14:27] <alex_joni> right, and you can do a lot with it too
[19:15:06] <icee> i got your messages, btw, LG, about the index pulse. if it's high for several cycles, we may have to deal with it ourselves rather than using the quadrature decoder to do the right thing
[19:16:16] <LawrenceG> icee: the encoder seems very stable when reading it at 10 times/sec.... I am also using X4 mode at the momemnt and I dont think we need that... I want to try some of the different options that the hardware supports
[19:17:14] <icee> OK, so it could be parallel port state machine related when two requests are adjacent
[19:17:14] <Jacky^> hey Jymmm
[19:17:20] <Jymmm> hi Jacky^
[19:17:27] <Jymmm> les_w you awake?
[19:17:30] <Jacky^> :PP
[19:20:35] <Jymmm> This is pretty interesting/informative --> http://www.hanita.com/hanita_protected/techinfo_contents.htm
[19:23:01] <Imperator_Away> Imperator_Away is now known as Imperator_
[19:23:16] <Imperator_> Hi Alex congratulation !
[19:23:26] <alex_joni> hey martin
[19:23:48] <alex_joni> thanks
[19:23:55] <Jymmm> What... is alex_joni the new dictator or something?
[19:24:01] <Imperator_> now you are one of the oficials :-)
[19:24:09] <etla> board of dictators.
[19:24:14] <alex_joni> Jymmm: somehow
[19:24:22] <alex_joni> no go do some work
[19:24:27] <alex_joni> now go do some work
[19:24:30] <Imperator_> thake care what you are saying etla :-)
[19:24:31] <Jymmm> RECOUNT!
[19:24:48] <Jymmm> ReVOTE this time w/o drugs and boose in the voting booth!
[19:24:49] <alex_joni> Jymmm: what do you wanna count? that 12%
[19:25:06] <alex_joni> I would have hoped a LOT more results
[19:25:45] <Imperator_> hehe, the good thing about america is that the are doing all elections automaticaly, there you can select the result befor voting :-)
[19:25:49] <Jymmm> alex_joni but it's you. what do you expect
[19:25:57] <cradek> I prefer people abstain if they don't care/don't know
[19:26:08] <cradek> it's better than everyone guessing
[19:26:23] <alex_joni> cradek: agreed, I would just have hoped that there are more that care/know
[19:26:29] <cradek> sure
[19:26:34] <cradek> can't argue with that
[19:26:35] <Jymmm> Imperator_ (dont get me started on the american votig system)
[19:26:45] <cradek> yeah, let's avoid that discussion
[19:26:47] <Imperator_> hehe
[19:27:04] <cradek> I trust our emc board results
[19:27:06] <cradek> it's refreshing
[19:27:08] <Imperator_> * Imperator_ is quiet
[19:27:22] <Imperator_> me too
[19:27:43] <Jymmm> Well, lets see how they do the first 30 days, THEN we'll hang em with a new rope
[19:28:05] <cradek> obviously the first thing to do is get rid of Jymmm
[19:28:30] <Imperator_> :-)
[19:28:35] <CIA-5> 03swpadnos * 10emc2/src/hal/drivers/hal_ppmc.c: added index sense pins, also fixed some comments
[19:28:50] <Jymmm> cradek: Not even in your wet dreams!
[19:29:09] <Jymmm> cradek *I* am your worse nightmare >:)
[19:29:21] <SWPadnos_> a nightmare on EMC street
[19:29:39] <alex_joni> SWPadnos_: you've been busy, that's a nice change ;)
[19:29:42] <SWPadnos_> Jymmmmy kruger
[19:29:49] <SWPadnos_> yeah yeaqh - tell it to the judge ;)
[19:29:49] <Jymmm> SWPadnos I prefer Hell Raiser
[19:30:04] <Jymmm> Just call me pinhead
[19:30:41] <Jymmm> http://www.airbrushexcellence.com/pinhead/pinhead%20036.jpg
[19:30:45] <SWPadnos_> well - ppmc seems pretty complete now - there are some setup options that aren't there, but they're not really necessary
[19:32:07] <SWPadnos_> I suppose I should finnish hacking the stg hal file into a suitable ppmc/usc hal file
[19:32:13] <SWPadnos_> finish, even
[19:38:30] <icee> cool, so i'll have something to test against
[19:38:36] <icee> i need to get my butt in gear
[19:52:57] <Jymm> boingy boingy boingy
[19:53:14] <alex_joni> Jymm: lost something?
[19:53:27] <Jymm> alex_joni yep... my connection
[19:53:35] <alex_joni> and an 'm'
[19:53:36] <alex_joni> :)
[19:53:48] <Jymm> Jymm is now known as Jymmmm
[19:53:53] <Jymmmm> hows that?
[19:54:19] <alex_joni> try /ns GHOST Jymmm yourpasswd
[19:54:25] <alex_joni> then /nick Jymmm
[19:54:28] <alex_joni> and you'll be set ;)
[19:54:38] <Jymmmm> oh... you mean F10 =)
[19:55:00] <les_w> blah. Was fly cutting some aluminum, and a chip flew up my nose.
[19:55:09] <Jymmmm> ouch
[19:55:26] <alex_joni> how far up?
[19:55:27] <cradek> you should have worn your ... noseguards?
[19:55:46] <les_w> had eye protection but not nose...
[19:55:55] <alex_joni> aluminum chips usually don't get past titanium nose-hair
[19:56:00] <les_w> guess it went down my throat.
[19:56:17] <les_w> I can feel the "oldtimers" comming on already
[19:56:33] <Jymmmm> * Jymmmm hands les_w one of these --> http://www.bouncemart.com/FRONTPAGE%20WEBSITE%20BOUNCEMART/Sumo-Action(1).JPG
[19:56:45] <Jymmmm> for next time
[19:57:00] <alex_joni> still lots of the face exposed
[19:57:22] <les_w> haha
[19:58:02] <Jymmmm> les_w ok, you get one of these too --> http://www.redtail-intl.com/images/10252.jpg
[19:58:34] <les_w> heh
[19:58:44] <les_w> have an auto dark helmet
[19:58:54] <les_w> do need welding jacket
[19:59:19] <les_w> caught clothes on fire once.
[19:59:31] <les_w> several times really bad sunburn too
[19:59:53] <Jymmmm> never thought about the sunburn aspects
[19:59:53] <alex_joni> les_w: tell me about it ;)
[20:00:04] <alex_joni> Jymmmm: don't wanna go there .. trust me
[20:00:07] <les_w> oh yeah....it can get really bad
[20:00:27] <alex_joni> I remember a few years ago, I was finishing some program on the robots
[20:00:45] <alex_joni> and I was testwelding (buttweld, about 3-400 Amps), and it was summer (40+ degrees)
[20:01:01] <alex_joni> so I was wearing only a t-shirt :(
[20:01:03] <alex_joni> bad idea
[20:01:18] <les_w> red arms!
[20:01:29] <alex_joni> half a meter away from the weld.. took the whole summer to get better
[20:01:39] <alex_joni> still had marks in winter :D
[20:01:47] <les_w> ooooh
[20:02:10] <Jymmmm> * Jymmmm places a 'Eat At Joes' clear sticker on alex_joni forehead!
[20:02:40] <les_w> this was silly...my .375 tooling plate did not come in, so I had to fly cut some 0.5 down
[20:02:46] <alex_joni> Jymmmm: well I learned my lesson the hard way :D
[20:03:36] <Jymmmm> alex_joni no doubt, thought the stencil sunburn would have really rubbed it in
[20:04:25] <les_w> like writing nasty things in front yard of someone you are mad at with fertilizer or roundup
[20:05:04] <Jymmmm> les_w bleach
[20:05:09] <les_w> haha
[20:05:13] <Jymmmm> table salt
[20:05:32] <les_w> or murine....on carpets...for black light
[20:05:38] <Jymmmm> some punk ass kids did it to my neighbor
[20:05:47] <Jymmmm> the bleach thing
[20:08:16] <Jymmmm> Jymmmm is now known as Jymmm
[20:08:21] <Jymmm> I'm back!
[20:09:57] <les_w> so I'm making this floating dust shield thing for the new spindle...it's on thomson bearings
[20:10:05] <les_w> just thought of something
[20:10:10] <les_w> it's gonna rattle
[20:11:32] <les_w> I had a real chip collection problem before with the old nylon brush fixed shield
[20:11:41] <Jymmm> no vac?
[20:11:49] <les_w> It matters when you are making thousands of pounds of chips
[20:12:29] <les_w> yeas...big dust collector. But the longer tools resulted in a gap between the nylon brush shield and the work
[20:12:31] <Jymmm> just blow the chips to the center of the shop
[20:12:40] <les_w> haha
[20:12:59] <les_w> so...I am making a floating one
[20:13:26] <Jymmm> is this still cherry?
[20:13:52] <les_w> 4 kw concentrated in 0.5 inch diameter + tons of fluffy chips on table ==fire
[20:14:07] <les_w> looks like mostly curly maple this year
[20:14:46] <Jymmm> les_w go get yourself some mini burlap bags printed up and sell the chips for (meat) smoking
[20:14:59] <les_w> which is fine. We get 3 times the price
[20:15:26] <les_w> oh all bbq cooking is done on cherry scraps
[20:15:50] <Jymmm> sell each bag for $9.95
[20:16:00] <les_w> good idea
[20:16:17] <les_w> I have a pile of cherry chips
[20:16:21] <les_w> a big pile
[20:16:25] <Jymmm> just have them soak the whole thing.... burlap and all, then wrap in foil and place on/near coals.
[20:16:29] <les_w> many cubic meters
[20:17:12] <les_w> I have been putting it around fences to stop weed growth
[20:17:14] <Jymmm> * Jymmm is hoping butlap is non-toxic =)
[20:17:22] <Jymmm> burlap
[20:17:24] <les_w> but now
[20:17:35] <les_w> I am pulling up a lot of the fence
[20:17:42] <les_w> I hate barbed wire
[20:17:56] <les_w> there is over a mile of it just here at the house
[20:18:00] <Jymmm> replace with razor wire instead =)
[20:18:45] <SWPadnos_> barbed wire == teh suck
[20:19:01] <SWPadnos_> (or teh suxx0rz)
[20:19:05] <les_w> I guess I will never be a hillbilly. I hate barbed wire fence, cows, and cow poop.
[20:19:15] <les_w> Golf course!!!
[20:19:31] <SWPadnos_> actually, convertina (razor) wire is really annoying
[20:19:37] <SWPadnos_> concertina
[20:19:41] <les_w> i'm sure
[20:19:48] <Jymmm> les_w and have a cow as trap =)
[20:20:13] <SWPadnos_> I fell into some in Basic training - it wasn't the happiest moment in my life
[20:20:45] <les_w> I was gonna get some emus just to look weird...but nah...pull the fence. Golf course.
[20:20:54] <Jymmm> Left 2, 3, 4, leave the poop at the door!
[20:21:13] <SWPadnos_> military + poop = error!
[20:21:18] <les_w> haha
[20:21:28] <Jymmm> SWPadnos whats the diff?
[20:21:39] <SWPadnos_> poop is shit in the military
[20:22:01] <Jymmm> SWPadnos whats the difference between poop (shit) and the military?
[20:22:35] <SWPadnos_> military == shit, poop == shit, but military != poop - it's nonilinear math
[20:24:24] <les_w> I just hate ruining a nice meadow with cow and cow poop
[20:24:30] <les_w> lots here
[20:24:36] <Jymmm> les_w Got Goats?
[20:24:46] <SWPadnos_> yeah - don't allow dogs either
[20:24:58] <les_w> and excersize in trying to turn ammonium nitrate into meat. Bad deal.
[20:25:12] <les_w> goats next door
[20:25:19] <les_w> about 30
[20:26:11] <Jymmm> * Jymmm needs a 'dunkable' clear coat system =(
[20:26:46] <les_w> I am doing a dip coating sealer on the calls this year
[20:26:51] <les_w> not sure how
[20:26:58] <Jymmm> les_w what chem?
[20:27:04] <les_w> tell you when I figure it out
[20:27:07] <Jymmm> lol
[20:27:30] <Jymmm> les_w did you say you use something on wood to help prevent fuzzies?
[20:27:47] <Jymmm> a sealer, or preserver, or ??????
[20:27:57] <les_w> not really
[20:28:00] <Jymmm> filler maybe ???
[20:28:04] <Jymmm> ok,
[20:28:06] <Jymmm> .
[20:28:11] <les_w> I will not penetrate far enough
[20:28:15] <les_w> it
[20:28:23] <les_w> haha
[20:28:49] <Jymmm> I think we have just discovered les_w divorce tendancies
[20:28:59] <Jymmm> =)
[20:29:16] <les_w> all wrong!!
[20:29:23] <les_w> heh
[20:29:36] <Jymmm> lol
[20:29:58] <les_w> I get divorced cause I am too bossy and won't do what women say.
[20:30:03] <les_w> but the other bits
[20:30:14] <les_w> haven't heard any complaints there.
[20:30:21] <SWPadnos_> hmm - bits won't penetrate far enough
[20:30:22] <les_w> Of course I'm old now.
[20:30:28] <Jymmm> les_w: so you'ld be gay if it wasn't for the whole sex thing huh?
[20:30:41] <les_w> uh...
[20:31:10] <Jymmm> think about it.... same interests, get to go out with the guys all the time, etc.
[20:31:19] <les_w> Well you said I must be gay when I was looking at audi a4 cabriolets
[20:31:22] <les_w> so ok...
[20:31:30] <les_w> I'm getting an A6
[20:31:33] <les_w> ok?
[20:31:40] <Jymmm> I did? dont remember saying that.
[20:31:51] <Jymmm> musta been someone else.
[20:31:52] <les_w> someone did.
[20:32:00] <les_w> I'ts a girl car.
[20:32:03] <les_w> I guess.
[20:32:17] <Jymmm> No clue, not into Foreighn cars too much myself.
[20:32:18] <SWPadnos_> If it's not a veyron, it's not a man car
[20:32:42] <les_w> I figure a6. Old man car.
[20:32:52] <les_w> ditch that bmw.
[20:33:07] <alex_joni_> ok.. night guys
[20:33:10] <alex_joni_> * alex_joni_ goes to bed
[20:33:13] <SWPadnos_> see ya
[20:33:17] <Jymmm> night dictator!
[20:33:18] <alex_joni_> gotta get that cold out of me :/
[20:33:25] <Jymmm> =)
[20:33:44] <alex_joni_> Jymmm: I'm a nice guy..
[20:33:48] <alex_joni_> till you get me started :D
[20:33:56] <Jymmm> alex_joni So was Hitler =)
[20:34:06] <alex_joni_> right
[20:34:27] <Jymmm> I dont think he's taking my teasing too well.
[20:35:14] <les_w> heh
[20:36:00] <Jymmm> as far as sports cars go, no clue. I like muscle cars.
[20:36:32] <les_w> I have a lot of traveling to do
[20:36:32] <Jymmm> or vintage
[20:36:44] <les_w> I want something quiet and comfortable
[20:36:56] <skunkworks> lexus?
[20:36:57] <Jymmm> get a minivan!
[20:37:12] <ValarQ> * ValarQ likes ugly and weak cars
[20:37:16] <SWPadnos_> http://commons.wikimedia.org/wiki/Image:Bugatti_new_front.JPG
[20:37:17] <les_w> skunk I looked at lexus closely
[20:37:47] <les_w> seems to be mostly camrys with better interiors
[20:38:01] <SWPadnos_> actually, you should check out a Honda or Acura
[20:38:14] <SWPadnos_> not a Civic, of course
[20:38:44] <les_w> neat bugatti
[20:38:48] <ValarQ> http://upload.wikimedia.org/wikipedia/en/6/6c/1987volvo244small.jpg
[20:38:55] <SWPadnos_> heh - yep. 1000 HP
[20:39:03] <skunkworks> I like the porche suv - just looks mean. (I guess I could say the same about the bmw suv)
[20:39:31] <les_w> I looked at lexus, acura, infiniti, MB, audi, Volvo, saab
[20:39:59] <les_w> many have really silly styling that I cannot handle
[20:40:06] <les_w> really goofy looking
[20:40:09] <Jymmm> viper? Corvette?
[20:40:57] <les_w> I like simple elegant lines. Dead quiet at highway speeds. Nice interior.
[20:41:35] <skunkworks> stealth Twin turbo? ;) We had a jetta tdi - I really liked that little car (diesle) 50 mpg
[20:41:48] <Jymmm> les_w coffin on wheels?
[20:42:28] <les_w> this is my idea of a car with the qualities I want. agree?
[20:42:30] <les_w> http://lostworld.pair.com/a6.html
[20:42:51] <SWPadnos_> unfortunately, it's a duck-butt
[20:43:05] <SWPadnos_> not too too bad, but still a duck-butt
[20:43:36] <les_w> pre 05 is more rounded
[20:43:49] <SWPadnos_> Try a Honda Accord Hybrid / Leather
[20:43:55] <les_w> and I am considering cpo
[20:44:06] <les_w> 4 yaer warranty...half the price
[20:45:02] <les_w> but you see what I mean...simple lines....no goofy nurbs gone wild stuff....comfortable on a long trip
[20:45:16] <les_w> plenty of power but nort a race car
[20:45:22] <SWPadnos_> we have a new Accord EX (non hybrid)
[20:45:30] <les_w> like?
[20:45:41] <SWPadnos_> it's pretty comfortable, and it has plenty of power, even with the 4 cylinder
[20:45:44] <SWPadnos_> yes - we like it
[20:46:01] <SWPadnos_> the V6 has 240 HP, so it's spirghtly enough
[20:46:07] <SWPadnos_> sprightly
[20:46:23] <les_w> my 325 has about the best handling of any car I have owned
[20:46:24] <SWPadnos_> the Hybrid actually adds 15 HP, while increasing economy by 50%
[20:46:31] <les_w> but it is work to drive it
[20:46:37] <les_w> it's a race car.
[20:46:42] <les_w> not a trip car
[20:47:04] <les_w> Hybrid s a good idea
[20:47:27] <SWPadnos_> it gets 29-32 MPG - not super, but not bad for a 255 horse luxury car
[20:47:40] <les_w> plenty
[20:48:01] <SWPadnos_> it's $30k new, the only option is the satellite nav system (also has a steereo upgrade, I think) for an extra $2k
[20:48:42] <les_w> I could live without sat nav
[20:48:49] <SWPadnos_> yep
[20:48:52] <les_w> in a plane...yeah I like it
[20:49:02] <les_w> might have even saved me once
[20:49:24] <SWPadnos_> the nav systems are pretty cool, but not really necessary (this coming from a technophile :) )
[20:49:43] <les_w> lost oil pressure...hit "nearest airport" on the gps...gave me vectors and I flew straight there
[20:50:26] <SWPadnos_> I've often wondered why the commercial pilots sill carry a briefcase full of charts - they should have a tablet PC or the like (though low batteries could pose a problem)
[20:50:55] <les_w> to this day we do not depend on gps
[20:51:19] <Jacky^> :)
[20:51:22] <les_w> any knee jerk reaction from some incursion and norad turns it off
[20:51:40] <SWPadnos_> yep - radio ain't the most reliable thing anyway
[20:51:43] <SWPadnos_> btw - have you considered the Chrysler 300C?
[20:52:02] <les_w> yeah. looked at it. Can't handle the styling
[20:52:05] <SWPadnos_> that's a grandpa car
[20:52:06] <Jacky^> * Jacky^ playng around paltak channels: Italiani e Brasiliani abbracciati nel mondo !
[20:52:09] <Jacky^> LOL
[20:52:12] <Jacky^> :)
[20:52:32] <les_w> a6 is a grandpa car too
[20:52:39] <SWPadnos_> yep - I know Chrysler used to be the luxury car people (along with Oldsmobile - the 88 was a biggie)
[20:52:49] <skunkworks> along those lines - dodge charger?
[20:52:58] <SWPadnos_> how about the Batmobile?
[20:53:02] <les_w> spent a lot of time at CTC
[20:53:07] <Jacky^> SWPadnos_: great car, yes, just forget croma :)
[20:53:14] <Jacky^> hehehe
[20:53:17] <skunkworks> or a chev - gto
[20:53:22] <SWPadnos_> croma?
[20:53:22] <Jacky^> kidding !
[20:53:26] <les_w> I spent many years designing useless gadgets for cars.
[20:53:28] <Jacky^> Fiat croma
[20:53:35] <SWPadnos_> no thanks
[20:53:41] <Jacky^> hahhaha
[20:53:45] <SWPadnos_> I can't get into half those cars you Italians drive
[20:53:55] <Jacky^> :)
[20:53:56] <SWPadnos_> the Ape - I can almost lift it
[20:54:05] <SWPadnos_> (and I'm thin)
[20:54:09] <Jacky^> hehehe
[20:54:12] <Jacky^> really
[20:54:14] <les_w> funny thing...I think audi owns bugatti now
[20:54:36] <SWPadnos_> http://money.howstuffworks.com/bugatti.htm
[20:55:10] <SWPadnos_> 45000 l/min of air + 1.33 gallons/min of gasoline at full power
[20:55:11] <les_w> another thing...hate suvs. Too high. A lot of swaying. Not comfortable.
[20:55:26] <SWPadnos_> I prefer the height - lets you see over all the idiots in small cars ;)
[20:55:39] <SWPadnos_> (our other car is a Plymouth minivan)
[20:56:33] <les_w> holy cow
[20:56:36] <SWPadnos_> yep
[20:56:46] <SWPadnos_> have you gotten to page 2 yet?
[20:56:57] <les_w> no ...let me look
[20:57:24] <SWPadnos_> it might be a little outside your preferred budget though
[20:58:12] <les_w> I think
[20:58:28] <les_w> audis ain't cheap either
[20:58:31] <CIA-5> 03swpadnos * 10emc2/src/hal/drivers/hal_ppmc.c: Fixed bug in export_UxC_encoders. Wasn't a problem, but was still wrong.
[20:58:43] <SWPadnos_> it's that extra order of magnitude that really gets you
[20:58:49] <les_w> ought to buy a corrolla.
[20:58:59] <SWPadnos_> Honda Civic - $13k
[20:59:04] <SWPadnos_> $19k for the Hybrid
[20:59:20] <SWPadnos_> gets the same mileage as the Prius, in the real world
[20:59:47] <les_w> I am strongly swayed by interiors and NVH
[20:59:58] <SWPadnos_> NVH?
[21:00:08] <les_w> I guess cause I did a bit of that for a living, i don't know
[21:00:19] <les_w> noise vibration harshness
[21:00:22] <SWPadnos_> ah
[21:00:24] <SWPadnos_> get a Jaguar
[21:00:41] <les_w> they break a lot
[21:00:42] <les_w> and
[21:00:51] <SWPadnos_> they're actually not much more expensive than the Audis
[21:00:55] <les_w> audio is MUCH better than jag these days
[21:01:08] <les_w> audi
[21:01:10] <les_w> oops
[21:01:35] <SWPadnos_> well - I can say that a 1976 XJ6L has pretty darned good ride and noise, but the Deeco electronics left a lot to be desired
[21:01:47] <Jacky^> ferrari
[21:01:53] <Jacky^> good machine
[21:02:00] <les_w> I designed some jaguar stuff at Lucas
[21:02:04] <skunkworks> Good point - my dad had a jaguar convertable for a few years. That was the quietest engine I have ever heard - and it was a v12
[21:02:12] <Jacky^> but who have so much money
[21:02:19] <SWPadnos_> that fiber optic lighting thing wsn't yours, was it?
[21:02:26] <Jacky^> 250 horses
[21:02:38] <les_w> oh, I love jaguars.
[21:02:44] <Jacky^> bah
[21:02:46] <SWPadnos_> yeah - the V12s are exceedingly quiet
[21:03:08] <les_w> have a mobile repair van following you at all times.
[21:03:31] <SWPadnos_> Haven't they gotten better since Ford bought them?
[21:03:36] <les_w> Audis are not the most reliable either.
[21:03:36] <SWPadnos_> oh, right - Ford
[21:03:41] <les_w> SWP: some.
[21:04:04] <les_w> oh, I have a Ford truck. Pretty good.
[21:04:07] <SWPadnos_> actually, Consumer Reports shows the Audis as having the best safety record of all cars on the road
[21:04:15] <SWPadnos_> and very good reliability as well
[21:04:16] <les_w> yes I know
[21:04:43] <les_w> I owned an audi before the bmw.
[21:05:21] <les_w> it was comfortable. The BMW could run circles around it but was less comfortable.
[21:05:40] <les_w> Engine growl sounds neat though.
[21:05:40] <SWPadnos_> in that case, get a TT :)
[21:05:59] <les_w> I'm too old. I don't deserve a TT.
[21:06:03] <SWPadnos_> heh
[21:06:13] <les_w> cheaper than an a6 though...
[21:06:42] <les_w> it's a go cart like my bmw.
[21:06:57] <SWPadnos_> hm - Consumer Reports really likes the Infiniti M35 / M45
[21:07:10] <les_w> I looked closely at that
[21:07:57] <SWPadnos_> only "very good" for noise though
[21:08:33] <les_w> It looked ok...had some problems with goofy looking styling though
[21:08:39] <les_w> looks matter to me
[21:09:00] <SWPadnos_> ah well - I guess you'll just have to wait
[21:09:16] <les_w> so many cars just look ridiculous to me these days
[21:09:56] <Jymmm> dumb question.... spindle is CW, so conventional milling would be tool moving CW or CCW around the material?
[21:09:59] <les_w> I like bauhaus form over function kinda stuff
[21:10:15] <SWPadnos_> counter, I think
[21:10:24] <les_w> jymmm, depends on outside or inside cut
[21:10:33] <Jymmm> outside (cutour)
[21:10:36] <Jymmm> cutout
[21:10:40] <SWPadnos_> if the cutter is "climbing" over the work, then it's climb, elst it's conventional
[21:10:46] <SWPadnos_> else
[21:10:55] <les_w> CW yeah
[21:11:13] <Jymmm> ok, ty
[21:11:47] <SWPadnos_> if you're doing a cutout, it's both - just conventional on the part and climb on the "rest" of the workpiece
[21:12:26] <les_w> yup...unstable too
[21:12:36] <SWPadnos_> slots are fun!
[21:12:37] <Jymmm> unstable?
[21:13:16] <les_w> yeah. cut a deep groove and you will see
[21:13:38] <Jymmm> I'm just trying to help prevent deflection
[21:13:47] <Jymmm> and damn toolmarks in the piece
[21:13:55] <les_w> tool wants to pull toward the climb mill side
[21:15:17] <Jymmm> ah
[21:15:33] <les_w> but you can't do much
[21:15:44] <les_w> other than a climb milled finish pass
[21:16:59] <Jymmm> les_w I might have soem backlash too, need to setup the TDI to check.
[21:20:14] <les_w> backlash is not good!
[21:23:39] <Jymmm> No, it's not. Might need to try things manually and see how it goes.
[21:24:57] <Jacky^> hi rayh
[21:25:20] <SWPadnos_> hiya Ray
[21:26:12] <Jacky^> rayh: how is going the dialup connection ? :P
[21:30:13] <Jacky^> :-)
[21:30:26] <SWPadnos_> not so well, it seems ;)
[21:30:35] <rayh> slow slower slowest.
[21:30:49] <rayh> I think the bits are getting lost in the mud.
[21:31:00] <SWPadnos_> okwellcompressbynotusingpunctuationorspaces
[21:31:09] <rayh> Good plan.
[21:31:17] <SWPadnos_> okthenhowareyoutoday
[21:31:19] <Jacky^> nahh is looking at emails received :)
[21:31:23] <Jacky^> heheh
[21:31:34] <rayh> Testing ppmc.hal on another box.
[21:31:51] <SWPadnos_> cool - I'm constructing a default ppmc.hal file right now
[21:32:16] <SWPadnos_> any suggestions for outputs to use for spindle on / fwd / rev, etc?
[21:32:28] <SWPadnos_> I'm also adding the spindle output to the driver
[21:34:20] <rayh> Okay. I've got the start of one here but if you send me your's I be pleased to test it.
[21:35:06] <SWPadnos_> can I dcc it to you now?
[21:35:10] <rayh> I'd think the default pins could match up to what is hard coded in rc46
[21:35:19] <SWPadnos_> it's not complete enough to commit to cvs
[21:35:30] <rayh> Try it. My firewall might complain but hey.
[21:35:49] <SWPadnos_> ok - I'll look there or at the ppmc.ini from pico-systems
[21:40:09] <rayh> Try that dcc again, steve. I think I've got the connections now.
[21:40:34] <SWPadnos_> ok
[21:41:03] <SWPadnos_> trying...
[21:42:44] <rayh> Na that's gonna take some work. Can you email it?
[21:42:51] <SWPadnos_> I'll be able to commit in a few minutes anyway, now that I've beed reminded about the sample ini
[21:44:09] <rayh> That wasn't it!
[21:44:16] <SWPadnos_> heh - apparently not ;)
[21:44:33] <SWPadnos_> maybe it would be safer to wait for the CVS version ;)
[21:45:25] <SWPadnos_> one question though - Jon uses both input 15 and input 14 for estop - can you explain the reasoning for that?
[21:46:12] <rayh> Matt and I figured out why that was but I've forgotten.
[21:46:38] <SWPadnos_> heh - OK. i'll just leave it with input 15 used as estop
[21:49:18] <rayh> Safe is a relative thing. Most of my relatives are NOT.
[21:51:35] <CIA-5> 03swpadnos * 10emc2/configs/ppmc_io.hal: Initial attempt at a ppmc hal file. Should duplicate the I/O settings in the sample.ini on the pico systems website.
[21:51:36] <SWPadnos_> ok - try that
[21:52:17] <SWPadnos_> I haven't added the spindle stuff yet (the P8 output that is)
[21:56:51] <rayh> k -- cvs up on the way
[21:56:59] <SWPadnos_> cool
[21:57:29] <SWPadnos_> hm - I just read your email - have you tried the dell port fixup program?
[22:00:29] <rayh> Do I need a fixup program if the communication diagnostic works properly.
[22:00:35] <SWPadnos_> nope
[22:01:00] <rayh> I thought it was odd that both the board and the el-cheapo card worked.
[22:01:21] <SWPadnos_> ah - I think I added in a word or two there - I thought you were saying that it didn't work with the diag
[22:02:01] <SWPadnos_> wait - what did work with the diagnostic?
[22:03:17] <rayh> identified the devices and comm with them.
[22:03:43] <rayh> When I tried the out test, it did not blink any leds.
[22:03:48] <SWPadnos_> ok - so you did univstepdiag commtest, and that was OK?
[22:03:57] <rayh> jmk though it might be the watchdog.
[22:04:09] <rayh> Right the commtest was good.
[22:04:12] <SWPadnos_> nope - that worked on mine, and I had the watchdog on
[22:04:19] <rayh> Showed both at the same time.
[22:04:28] <SWPadnos_> but the command is './univstepdiags diocontinuous'
[22:04:55] <SWPadnos_> the help is wrong ;)
[22:05:10] <rayh> Oh. I'll try that.
[22:06:06] <SWPadnos_> (assuming that you have resistors or SSRs installed) you should see SSR8 on continuously, and a shifting pattern onthe other SSRs
[22:32:49] <SWPadnos_> so rayh - anything working yet?
[22:41:18] <les_w> had to run to town. Boy is it foggy! without street lights and botts dots road visibility is not more than 50 feet. Had to go very slow.
[22:59:04] <les_w> hey jymmm...btw fram damper is a fourth order systen used to reduce vibrations at resonance
[23:00:39] <les_w> http://www.kettering.edu/~drussell/Demos/absorber/DynamicAbsorber.html
[23:09:12] <les_w> It places a zero over the system pole. It however creates two new poles, but the q is lower than the original if the absorber is damped. Commonly used in skyscrapers in earthquake zones, airplane wings, and machine tools.
[23:09:46] <les_w> Stanley uses it in a hammer too.
[23:10:33] <les_w> http://www.amazon.com/gp/product/B00005A1KR/103-6232679-8821438?v=glance&n=228013&v=glance
[23:44:04] <jtr> jtr is now known as jtr_
[23:50:08] <jtr_> jtr_ is now known as jtr
[23:54:42] <rayh> SWPadnos_: Trying the sf ppmc_io.hal but servo thread not found.
[23:56:28] <rayh> I see, the ini file does that for you.
[23:58:47] <SWPadnos_> I didn't do anything with threads in the hal file
[23:59:40] <SWPadnos_> I basically duplicated the stg_io.hal file - maybe I should do the same with stg_motion.hal