#emc-devel | Logs for 2008-11-09

[00:14:36] <jmkasunich> If I knew anything about the interp, I'd be half tempted to try to implement G90.1/G91.1 (abs/relative I&J for arcs)
[00:21:10] <cradek> it would be a good first project
[00:21:23] <jmkasunich> heh
[00:21:38] <cradek> add them to _gees in interp_array.cc
[00:22:20] <jmkasunich> what the heck, might as well give it a shot
[00:22:47] <cradek> handle them in convert_distance_mode (set a new flag in "settings")
[00:22:53] <jmkasunich> they'll become a modal group, right
[00:23:00] <cradek> yes
[00:23:20] <cradek> well first, how do you want them to work?
[00:23:34] <cradek> can you have g91 and g90.1 both in effect?
[00:23:46] <cradek> or can you only have one of the four?
[00:23:46] <jmkasunich> good question
[00:24:01] <cradek> if the former, new group - latter, same group as g90/g91
[00:24:10] <jmkasunich> * jmkasunich navigates to the quick-ref
[00:24:47] <jmkasunich> I was originally thinking of them as a distinct pair
[01:01:27] <skunkworks> no smoke yet.
[02:00:24] <BigJohnT> what are we testing?
[02:02:00] <skunkworks> this h-bridge http://www.youtube.com/watch?v=fGF07oVkl6M
[02:02:43] <skunkworks> hooked to this servo. http://electronicsam.com/images/KandT/DSC_0242.JPG
[02:03:53] <BigJohnT> cool, I can't look at it now I'm downloading open office for windoz's on dial up :)
[02:04:17] <skunkworks> dial up? what is that? ;)
[02:05:00] <BigJohnT> it connects to the string that goes out to the highway and there is a tin can on the other end
[02:05:53] <skunkworks> and most of the time the string is broken
[02:06:39] <BigJohnT> no it is reliable but slow
[02:07:07] <BigJohnT> I can get high speed if I install a 75' tower to see the antenna
[02:08:50] <skunkworks> no close neighbors>
[02:08:52] <skunkworks> ?
[02:09:36] <BigJohnT> no I'm down in a valley on a dead end dirt road off of a dirt road...
[03:59:19] <cradek> with cutter compensation in effect, some dwell will occur for blocks that are executed in less than .25 seconds. colinear moves cannot be programmed.
[04:00:05] <cradek> M codes cannot be used until the block after cutter compensation is turned off.
[04:01:23] <SWPadnos> BOSS?
[04:01:34] <cradek> yes
[04:01:39] <SWPadnos> phew
[04:01:53] <cradek> Z axis moves will not be compensated.
[04:01:58] <cradek> (I don't know what that means)
[04:02:15] <cradek> Z motion must not be programmed in either the first or last three compensated blocks.
[04:02:26] <cradek> Multiple Z moves must be separated by at least two blocks of XY data.
[04:02:50] <cradek> most of these are design compromises - but the colinear thing seems like a 'documented bug'
[04:32:06] <CIA-38> EMC: 03jmkasunich 07TRUNK * 10emc2/src/emc/rs274ngc/ (7 files): initial work on G90.1/G91.1 - so far all it does is parse the g-codes and sets an internal variable
[08:33:44] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[10:49:28] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[13:22:16] <F1514> hi.. i need a bit answer on:
[13:22:17] <F1514> acording im reading on HOSTMOT2.c, PC communication behave on Mesa is based on 3 Major variable:
[13:22:17] <F1514> "Encoder", "StepGen", "PWMGen" ... and PID calc being done on PC, based on encoder information, is my perception Correct?
[13:22:17] <F1514> mesa do the counting and send the value of pos to pc
[13:22:17] <F1514> if u dont mind, please explain anything on this, to enrich my knowledge(or curious)... any commment would be very usefull for me
[13:22:19] <F1514> |_____|``|_____| ''=is the length of PWM and stepgen is how much |``| pulse being generate
[13:24:05] <F1514> cos i must disconnect my internet connection about couple minute from now... thats above is my homework
[14:48:43] <BigJohnT> alex_joni: did we do the 2.2.7 yet?
[14:53:01] <alex_joni> not yet
[14:53:15] <alex_joni> I hope to get word back from max about the stg fix on TRUNK
[14:53:27] <alex_joni> if he tests it and it's ok, I'd definately want to stick it in 2.2.x
[14:53:39] <BigJohnT> ok, just wondering
[14:53:41] <alex_joni> but it's probably not that important to hold off 2.2.7
[14:54:10] <BigJohnT> what is stg?
[14:59:25] <alex_joni> www.servotogo.com
[15:00:37] <BigJohnT> ok, thanks
[15:02:15] <alex_joni> I think it was one of the first (if not the very first) hardware that worked with emc1
[15:31:43] <skunkworks> logger_dev: bookmark
[15:31:43] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2008-11-09.txt
[19:19:02] <robh> hi all, would fanuc,siemens etc operator & machine manuals be of any use as reference material? if so im happy to donate some copys your way
[19:33:43] <jmkasunich> paper or electronic?
[19:34:25] <jmkasunich> electronic would be more convenient, but probably would run afoul of copyright if we posted them somewhere, or made copies for every emc developer
[19:35:14] <jmkasunich> paper would avoid that, but murphy says the guy who has them isn't the guy who needs to refer to them
[19:59:33] <robh> i have siemens in electronic, as downloaded that from there website,
[20:00:42] <robh> also have a fadal one also on CD,
[20:25:26] <jmkasunich> we need to follow copyright
[20:26:07] <jmkasunich> is the seimens website open to anybody? if so, let us know the URL, we can put it on our wiki or somewhere, and if we need to look at the manual we can go there
[20:26:34] <jmkasunich> if the website is only open to seimens customers, etc, then things are more complicated
[20:26:53] <alex_joni> SWPadnos (but not exclusively): http://www.xtratum.org/
[20:27:06] <alex_joni> that is _really_ interesting to watch during the following period
[20:28:03] <alex_joni> it seems RTLinux/GPL is evolving into XtratuM which is both a layer ontop of a regular linux (like the old RTLinux patching), but also a system designed for embedded archs (if I read it properly)
[20:32:37] <alex_joni> seems it's similar to ADEOS in some aspects
[20:36:27] <robh> ok jmkasunich ill add the seimens link onto wiki some where ill have to dig it out my bookmarks for ya, if there ever anything u need looking up just let me know ill be than willing to help out any way i can
[21:09:59] <alex_joni> hi Roguish