#emc | Logs for 2012-01-10

Back
[00:03:48] -!- theorb [[email protected]] has joined #emc
[00:04:03] -!- The_Ball [[email protected]] has joined #emc
[00:04:37] -!- theorbtwo has quit [Ping timeout: 248 seconds]
[00:04:46] theorb is now known as theorbtwo
[00:09:26] -!- robin_sz [[email protected]] has joined #emc
[00:09:46] <robin_sz> off topic, anyone recommend a good CRM system
[00:13:53] <andypugh> ~What sort of CRM? http://en.wikipedia.org/wiki/CRM_114_(device)
[00:17:51] <A2Sheds> STL to bitmap, skeinforge slices stl files and produces a giant gcode file for all the layers, I'd like to slice a 3d model and convert to bitmaps slices
[00:19:03] <A2Sheds> maybe POV-Ray to slice and dice to bitmaps?
[00:19:54] <andypugh> I have some Octave code that converts STL to postscript?
[00:19:55] -!- Nick001-Shop has quit [Quit: ChatZilla 0.9.88 [Firefox 8.0/20111104165243]]
[00:20:34] <andypugh> A2Sheds: http://pastebin.com/tt80SFui
[00:21:16] -!- bilboquet has quit [Ping timeout: 240 seconds]
[00:21:28] <andypugh> You need octave, of course: http://www.gnu.org/software/octave/ but I think that is just sudo apt-get install octave
[00:21:43] -!- bilboquet [[email protected]] has joined #emc
[00:22:00] <andypugh> (Octave is Matlab for cheapskates) :-)
[00:22:12] <A2Sheds> heh
[00:24:26] <A2Sheds> very interesting
[00:24:32] <andypugh> But it's particularly good for this sort of work, the whole thing is done in a few slice-and-dice sorts, the test of the code is concerned with parsing EPS and HPGL. You probably want to lose the HPGL and serial port stuff, I wrote that to produce 3D-models using my flatbed plotter.
[00:24:50] <A2Sheds> ah, ok
[00:25:40] <andypugh> Lines 40 to 78 do all the STL slicing. No, I don't understand it any more either, it was 6 years ago.
[00:25:43] <A2Sheds> I want to produce images of each slice to send to the projector
[00:26:12] <JT-Shop> WEEEEE! GladeVCP buttons that change color based on the state of the button is easy now :)
[00:26:47] <andypugh> A2, well, that ought to produce a bunch of EPS files.
[00:27:11] <A2Sheds> I'm not super familiar of whats inside or the format of STL, I've just been using them for years
[00:27:36] <andypugh> It's a list of vertices of triangles. That's it.
[00:27:41] <A2Sheds> 3d shorthand for polygons
[00:29:11] <andypugh> But in Octave [F V C] = stlread(filename) reads the STL file into arrays F V C of Faces, Vertices and (something my code didn't use)
[00:29:25] <A2Sheds> skeinforge slices and produces a vector graphics SVG of each slice... ahaha
[00:29:47] <andypugh> I bet my code is more impenetrable though.
[00:30:01] <A2Sheds> I can take each svg slivce, convert to bmp!
[00:30:14] <A2Sheds> I'll try it all
[00:30:32] -!- sumpfralle has quit [Read error: Operation timed out]
[00:30:49] <andypugh> Octave is worth knowing about, it's spectacularly good at pulling out subsets of data files.
[00:33:03] <A2Sheds> Iv'e only played with it. I have MathCAD as well
[00:33:50] <A2Sheds> dad! How do you do this?.... here use MathCAD
[00:34:49] -!- rob_h has quit [Quit: Leaving]
[00:36:01] <andypugh> I think the logic in the Octave code is that any triangle with 1 or two vertices > slice level Z spans the slice. Lines 45 and 46 find them and store their indices in I1 and I2. Then two similar lumps of code find the x.y coordinates of the point where the z plane slices each edge. That gives you a bunch of pairs of XY coordinates definining lines, and you then just walk the list looking for shared points (it doesn't matter w
[00:36:01] <andypugh> you start) to create a continuous path.
[00:38:02] <andypugh> I have just concluded that I was cleverer 7 years ago than I am now :-(
[00:40:10] <A2Sheds> over 45? http://www.newscientist.com/blogs/shortsharpscience/2012/01/cognitive-decline-sets-in-as-early-as-45.html
[00:45:57] -!- BobGreen [[email protected]] has joined #emc
[00:49:20] -!- sumpfralle [[email protected]] has joined #emc
[00:51:49] <A2Sheds> huh, one of the dependencies for Octave is graphicsmagick
[00:57:55] -!- ries has quit [Ping timeout: 240 seconds]
[01:01:42] <A2Sheds> brilliant! I can use octave for fancy scaling as well
[01:10:15] <A2Sheds> andypugh, but you are much wiser now :)
[01:10:45] <andypugh> A2Sheds: That was quick then, I have only been 45 for a bit over a week!
[01:11:31] <A2Sheds> happy birthday
[01:12:03] <A2Sheds> was that your party on the 31st?
[01:12:41] <andypugh> Aye I have never worked out why I am so popular.
[01:13:22] <andypugh> (Actually, it was the 28th, and even I barely noticed)
[01:14:36] <A2Sheds> when you were a kid did you get double the presents or only combined xmas+birthday?
[01:16:22] <andypugh> combined, mainly.
[01:18:36] <andypugh> Sleep time.
[01:18:38] <andypugh> G'night
[01:18:45] -!- andypugh has quit [Quit: andypugh]
[01:18:53] -!- motioncontrol has quit [Ping timeout: 244 seconds]
[01:19:56] -!- sumpfralle has quit [Read error: Operation timed out]
[01:21:42] -!- sumpfralle [[email protected]] has joined #emc
[01:30:55] -!- ries [[email protected]] has joined #emc
[01:34:42] -!- Valen has quit [Quit: Leaving.]
[01:36:57] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #emc
[01:37:38] -!- Tom_L has quit [Client Quit]
[01:59:25] -!- skunkworks_ [skunkworks_!~chatzilla@str-bb-cable-south2-static-6-425.dsl.airstreamcomm.net] has joined #emc
[02:07:25] <clytle374> So I get called out because a lathe was indexing the turret to the wrong tool and alarming out. Funny thing was it always went the right number of positions, but the wrong way.
[02:07:42] <clytle374> Come to find out they moved and rewired it, phase was backwards
[02:10:37] -!- factor [[email protected]] has joined #emc
[02:18:55] -!- ries has quit [Read error: Connection reset by peer]
[02:28:17] <gene76> Well, it runs ok I think, but I didn't realize "master-rt" would get be a 2.6.0-rc-something. There are some fusses when it exits, but at least it doesn't take the machine down.
[02:29:27] <gene76> Thanks Guys & good night.
[02:29:42] -!- gene76 has quit [Remote host closed the connection]
[02:34:32] -!- memleak [memleak!~memleak@unaffiliated/memleak] has joined #emc
[02:34:39] <memleak> Hi. RTAI problems? :)
[02:35:30] <clytle374> do you read minds?
[02:36:29] <clytle374> Rumor is you are the RTAI expert
[02:37:00] <memleak> I wouldn't go THAT far..
[02:37:49] <memleak> What's wrong with RTAI?
[02:38:15] <clytle374> Any thoughts on why I can build 4 different kernels that will run emc 2.4, but none will run 2.5
[02:38:37] <clytle374> I'm running git bisect now
[02:38:55] <clytle374> 8 revisions left
[02:39:21] <memleak> Is the same kernel that works on 2.4, different than the kernel you're trying to get 2.5 running on?
[02:39:27] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #emc
[02:39:46] <clytle374> No, all the same kernels.
[02:39:56] <memleak> also are you following git master or v2.5_branch?
[02:40:13] <clytle374> v2.5 branch
[02:40:47] <memleak> Are you sure it's an RTAI problem? What is the problem you are having?
[02:41:33] <memleak> It kind of sounds like an EMC bug if 2.4 works but 2.5 doesn't.
[02:41:38] <clytle374> I get a watchdog bit message in dmesg during or after the pinout output loading the 5i20
[02:41:58] <clytle374> You said it, not I ;)
[02:42:06] <clytle374> back in 1 minute
[02:43:44] -!- LawrenceSeattle has quit [Quit: LawrenceSeattle]
[02:43:44] <clytle374> another bad one, 4 revisions left.
[02:44:05] <memleak> Well, a watchdog problem..... Watchdog support shouldn't even be on in the kernel config.. It can mess RTAI up. RTAI runs best with a very small kernel config, only having the important stuff turned on.
[02:44:32] <memleak> The more things you have on, the higher a chance of something breaking. I never tried a custom kernel with watchdog support or any of that with RTAI + EMC
[02:45:06] <clytle374> this message hm2/hm2_5i20.0: Watchdog has bit! (set the .has-bit pin to False to resume)
[02:45:30] <memleak> I was able to get 2.5 working on RTAI btw
[02:45:31] <clytle374> My kernel is stripped down pretty far
[02:45:48] <clytle374> what kernel are you using?
[02:46:16] <memleak> I used 2.6.37 and 2.6.36.something rather. one moment
[02:46:34] <FinboySlick> memleak: Yo dude :)
[02:46:40] <memleak> 2.6.38.8 is up though
[02:47:05] <memleak> Hi FinboySlick :)
[02:47:16] <FinboySlick> Still busy being awesome, I see.
[02:47:22] <clytle374> The watchdog message isn't the one you were talking about was it?
[02:47:51] <memleak> scratch that. 2.6.35.7, 2.6.35.9 and 2.6.37
[02:48:26] <clytle374> I'm using 2.6.35.9
[02:48:51] <memleak> Those were the ones I was able to get running. The watchdog message I was talking about was the same one.
[02:49:59] <memleak> IPIPE 2.8-02 -03 or -04? -02 and -03 are broken
[02:50:07] <memleak> Can you post your kernel config?
[02:52:06] <clytle374> I'm using rtai cvs vulcano, tried magna, and 3.8.1
[02:52:10] <clytle374> here is the config
[02:52:21] <clytle374> http://pastebin.com/JkFRVrh6
[02:52:39] <memleak> Volcano has been depreciated by magma
[02:52:46] <memleak> that tree hasn't been touched in years
[02:53:06] <clytle374> The patches are being updated
[02:53:13] -!- FinboySlick has quit [Quit: Leaving.]
[02:53:16] <clytle374> I was told it is the stable tree
[02:53:29] <clytle374> But, I've been wrong... A LOT
[02:55:52] <clytle374> Yep, I had watchdog on
[02:56:21] <memleak> Turn off ACPI and all that too.
[02:56:27] <memleak> CONFIG_PM is on
[02:56:53] <memleak> that option automatically turns on CONFIG_CPU_IDLE=y
[02:57:50] <clytle374> I didn't think I had config_pm on. Somewhere I read acpi needs to be on.
[02:58:28] <clytle374> Not arguing at all, I'm sure you know what your talking about. But a google search will contradiction itself badly
[02:59:28] -!- sumpfralle has quit [Quit: Leaving.]
[02:59:42] <memleak> ACPI can cause interrupt problems and make RTAI go nuts. It's one of the very first things I turn off.
[02:59:51] <clytle374> great
[03:00:20] <memleak> There was only one case I couldn't, and it was on an AMD 785 board with an ACPI PATA controller. Couldn't access my HDDs without it..
[03:02:11] <memleak> turn off CONFIG_DEBUG_KERNEL and frame pointers too in the kernel hacking section
[03:03:28] <memleak> earlyprintk will tell you everything you want to know, if you ACTUALLY want to debug the kernel. kernel debugging just increases overhead and such. I just use earlyprintk
[03:04:21] <clytle374> I didn't want it, but was desperate
[03:04:41] <memleak> also turn off CPU sets and group schedulers and such
[03:04:47] <memleak> that can mess things up too.
[03:05:46] <clytle374> CPU sets?
[03:06:07] <memleak> CPU control sets.. I forget what they call it...
[03:06:45] <memleak> Anything to do with control groups, group schedulers, cpu group schedulers can make matters worse
[03:06:48] <clytle374> Are you seeing group schedulers as on?
[03:07:24] <memleak> CONFIG_CPUSETS for example
[03:08:06] <memleak> CONFIG_CGROUP_NS pretty much anything that has to do with grouping in the CPU / main system section can go.
[03:08:54] -!- capricorn_one [[email protected]] has joined #emc
[03:09:37] <memleak> their for virtual servers and clustering and whatnot. if this is a single machine running EMC, it should be off. RTAI really doesn't play nicely with like.. 75% of the config screen
[03:10:39] <clytle374> so what about Virtualization. I found an obscure post somewhere saying it should be on
[03:12:35] <clytle374> Should I build it?
[03:13:17] <memleak> virtualization should be off too.
[03:13:57] <memleak> multi-core scheduling (not SMP. SMP is different) and hyper-threading and such should be off as well.
[03:14:25] <memleak> speaking of which, if you have an SMP system, use make -j3 instead or -j4. It'll build a lot quicker.
[03:15:23] <clytle374> I use -j3 as long as isocpus is not set
[03:15:55] -!- the_wench has quit [Ping timeout: 240 seconds]
[03:16:15] <clytle374> With some of the stuff that is on, this config was one I pulled out of the wiki and then turned off a bunch of drivers
[03:16:36] -!- archivist has quit [Ping timeout: 268 seconds]
[03:18:02] <clytle374> yeah, I didn't turn on nsa selinux
[03:18:37] <memleak> Turning on Selinux in the kernel doesn't do much unless you have the userspace stuff installed anyway.
[03:19:26] <clytle374> I've built kernels for all my machines for about 6 years, and I still don't know 1% of the kernel stuff
[03:19:32] <memleak> It doesn't make anything more secure with it on or off if you have no selinux policies set.
[03:19:43] <clytle374> makes sense.
[03:20:36] -!- packrat has quit [Ping timeout: 240 seconds]
[03:20:55] -!- packrat [[email protected]] has joined #emc
[03:21:31] <memleak> The only reason I know the configs well was to squeeze every last bit of performance I could out of my old athlon XP machines.
[03:22:24] <memleak> Stock ubuntu install would use 1/2 of my 1 GB stick. custom kernel trimmed that down to >200.
[03:22:52] <clytle374> yep, that and build time.
[03:23:23] <clytle374> I built a .config off the wiki and couldn't believe it took hours
[03:25:26] <memleak> you won't be proud when you'll be able to tell your friends you trimmed the kernel under a single meg in 10 minutes...
[03:25:42] <clytle374> So 2.6.37 is the newest you've used, or got working?
[03:26:01] <memleak> i forget if .37 worked with EMC or not.
[03:26:23] <memleak> it might have.. it might not have.
[03:26:45] <memleak> 2.6.38.8 though I'm curious about.
[03:27:01] <clytle374> My normal use kernels are 4.5M
[03:27:06] <clytle374> no modules
[03:27:34] <clytle374> Haven't used initrd since I switched to Gentoo yearsback
[03:27:34] <memleak> It's only available on the adeos site though. it's not in RTAI yet.
[03:28:01] <memleak> i really dont like initrds.
[03:28:28] <memleak> so much easier to fix if something goes wrong with just a single bzImage
[03:28:54] <memleak> simpler grub.cfg / menu.lst
[03:29:05] -!- archivist [[email protected]] has joined #emc
[03:29:43] <clytle374> you do use the rtai patches from the x86, right? even for x64?
[03:29:52] <memleak> yes
[03:29:59] <memleak> never use the i386 dir
[03:30:09] <memleak> i don't even know WHY that's in there..
[03:30:20] <clytle374> I got one thing right, yea
[03:30:24] <memleak> XD
[03:31:03] <clytle374> I started on 64bit, then went 32 in desperation. Then kept backing up kernels
[03:31:18] <clytle374> I even used a config or 2 off the wiki
[03:31:30] <memleak> if you examine the patches, you'll see that all the files in the x86 dir make changes to the x86_64 arch and also 32-bit
[03:31:54] <clytle374> Of course a Gentoo install isn't quite as quick as ubuntu
[03:32:45] <memleak> i'm kind of glad it isn't. helps widen the gap between ubuntu and gentoo.
[03:33:09] <clytle374> Don't say that out loud
[03:33:44] <clytle374> distcc is your friend
[03:34:31] <memleak> no ccache?
[03:34:42] <clytle374> yes.
[03:34:55] <clytle374> I mean I do use ccache
[03:35:32] <clytle374> You know it's bad when you have a gigabit card and a null cable on you desktop for large updates
[03:36:07] <memleak> didn't gentoo once have pmake as it's main compiler?
[03:36:47] <clytle374> It must have been before me
[03:37:13] <memleak> parallel make.. did a great job at parallel processing for make
[03:37:30] <memleak> hence pmake. is it emake now?
[03:38:06] <clytle374> I thought make just funneled thought distcc >> gcc
[03:38:52] <clytle374> Nope it is emake
[03:41:24] <memleak> i wonder what people did before make.. did they really spend a lot of time typing out gcc strings?
[03:41:25] <clytle374> emerge shields you from the internals quite well
[03:41:59] <clytle374> I don't know. At least there was a lot less stuff back then
[03:42:22] <clytle374> I used make on RH5.1. Been a noob ever since
[03:43:12] <clytle374> probably a lot of scripts
[03:43:14] <memleak> noobs don't use RH. ;)
[03:43:47] <memleak> most of the software in every distro originated from the RH dev team.
[03:44:02] <clytle374> I know that I don't know a big scarry lot of stuff
[03:44:11] <clytle374> yeah, it was a pioneer
[03:44:33] -!- the_wench [[email protected]] has joined #emc
[03:45:20] <memleak> fedora is one of the most contributing distros nowadays. gentoo pulls in patches from the fedora repos sometimes, because they were the ones to come up with the fixes.
[03:46:02] <memleak> arch does the same thing. oh boy what a train wreck that is now..
[03:46:10] <Jymmm> I still with CentOS
[03:47:24] <clytle374> what does cat Module.symvers do? and why do I have to add the stuff from /user/realtime's file?
[03:47:48] <clytle374> Ignore the "cat" copy and paste error
[03:48:07] <memleak> module.symvers is module sub-versioning support
[03:48:19] <memleak> basically, it makes a bigger pain to load modules..
[03:48:36] <clytle374> That explains why I need it when before I didn't :)
[03:49:00] <memleak> /usr/realtime ? what file? what stuff? the FIFO stuff?
[03:49:21] <memleak> RTAI [ 0 - 9 ] FIFO.. something like that?
[03:50:28] <memleak> it allows RTAI access to certain parts of memory
[03:50:34] <clytle374> I've had to execute cat /usr/realtime/modules/Module.symvers >> Module.symvers in /usr/src/linux
[03:51:03] <memleak> why?
[03:51:36] <clytle374> I get symbol errors when building emc without it. got it from the wiki.
[03:51:59] <clytle374> I'm buiding now, so we'll see
[03:52:32] <memleak> i never use version support so i wouldn't know about that.
[03:53:04] <clytle374> Lucky you, I've read everything a could find on rtai.. And got more confused
[03:53:57] <clytle374> like the consensus on hpet timer seems be it should be off, but most .configs have it on
[03:54:06] <memleak> i turn HPET off too
[03:54:13] <memleak> not just in the kernel but in BIOS as well
[03:54:15] <clytle374> some in the wiki have cpu freq scaling on too
[03:54:25] <memleak> that goes off too
[03:54:27] <clytle374> yeah, I got it off both places.
[03:54:50] <clytle374> I hope this works and I keep my under 4000 jitter
[03:56:41] <clytle374> after finding the 2.4 worked I gave up on the rtai kernel and started looking elsewhere for the proble
[03:57:21] <clytle374> I only had 2 more bisects to go.... Really neat feature in git
[03:58:33] <memleak> i love git :)
[03:58:43] <clytle374> moment of truth, gotta walk into the cold room to test it out. working through ssh
[04:00:50] <clytle374> I guess I have module versioning support on, reuilding emc
[04:02:03] <clytle374> I'd never used git before, well I did 2 years ago when I started this project
[04:02:10] <clytle374> got busy and put it aside
[04:02:39] <clytle374> Now I, and bossman, got too much money invested to give up
[04:03:17] <memleak> git is a lot more powerful and easier to manage than mercurial, svn, cvs, bzr (ewwww) i've used them all
[04:03:43] <clytle374> I've used svn, cvs.. But never as a developer
[04:08:07] <clytle374> what happened now I'm getting rtai_hal: disagrees about version of symbol module_layout
[04:08:09] -!- n2diy has quit [Quit: Ex-Chat]
[04:08:36] <memleak> module versioning support ring a bell? :)
[04:09:10] <clytle374> yeah, but never have that after the piping the files together.
[04:09:21] <clytle374> I'm shutting it off and trying again
[04:12:14] <clytle374> these atoms aren't real good at compiling
[04:12:56] <memleak> i've compiled on an atom before. it does take awhile.
[04:12:56] -!- danimal_laptop has quit [Read error: Connection reset by peer]
[04:13:23] -!- danimal_laptop [[email protected]] has joined #emc
[04:13:43] <clytle374> So how do you get a kernel under 1M?
[04:14:16] <memleak> if you send me a dmesg and lspci i could get you a working kernel probably
[04:14:23] <memleak> small too
[04:14:28] <memleak> roughly 1MB but not under
[04:14:45] <memleak> no point in that anyway other than just to do it.
[04:15:02] <clytle374> yeah, but still neat
[04:15:40] <memleak> Smallest I've gotten was 900K with no modules
[04:15:54] <memleak> i can compile the kernel from my end
[04:16:04] <clytle374> You tried gnome3 in gentoo? Someone said it's a lot worse then ubuntu
[04:16:17] <Jymmm> http://www.linfo.org/mulinux
[04:16:20] <clytle374> A good config would make me happy
[04:16:23] <memleak> ubuntu switched to gnome 3 too didnt they?
[04:16:35] <clytle374> yeah
[04:16:37] <memleak> config it is. lspci and dmesg and i'll be set
[04:17:44] * memleak gets ready
[04:18:01] <clytle374> for emc?
[04:18:32] <memleak> with lspci and dmesg i can make you a (hopefully) very good config
[04:18:44] <memleak> which you can use for RTAI + EMC
[04:18:46] <clytle374> http://pastebin.com/h1FRc0P9
[04:18:58] <memleak> would you be willing to try 2.6.38.8 or stick with 35.9?
[04:20:09] <memleak> RTAI will need to be recompiled along with EMC, not just the kernel
[04:20:35] <memleak> after you reboot into the new kernel i mean
[04:20:43] <clytle374> http://pastebin.com/X6017iUA
[04:20:59] <clytle374> yeah, compile rtai on new kernel.
[04:21:40] <clytle374> 38.8 is great. I'm only on 35 because I was backing up trying to get it to work
[04:21:50] <memleak> ok use magma this time though
[04:23:43] <memleak> from cvs
[04:23:48] <clytle374> yep.
[04:24:00] <memleak> ok :)
[04:24:04] <clytle374> I had sound on too augh!
[04:24:14] <memleak> downloading kernel source. ill let you know when the config is done
[04:24:28] <memleak> won't take too long. i fly through the entire config quite quickly..
[04:25:08] <clytle374> I usually do, but when you've changed so many things your taking notes... You slow down
[04:26:45] -!- n2diy [[email protected]] has joined #emc
[04:29:46] <memleak> For the RTAI use adeos-ipipe-2.6.38.8-x86-2.11-01.patch http://download.gna.org/adeos/patches/v2.6/x86/
[04:30:09] <memleak> 2.11-01 not 2.10-02
[04:30:45] <memleak> apply using patch -p1 < ../adeos-ipipe-2.6.38.8-x86-2.11-01.patch
[04:31:16] <memleak> 32 bit right?
[04:32:04] <clytle374> yeah, I might switch back to 64 bit. But that will take some time
[04:32:13] <clytle374> adeos patch?
[04:32:37] <memleak> also what version of acct do you use? 6.5.3 6.5.5 etc
[04:32:56] <memleak> yes RT systems use adeos patches for the kernel
[04:33:19] -!- BobGreen has quit [Read error: Connection reset by peer]
[04:33:27] <memleak> but then they pull in their (adeos) patches into their own tree
[04:34:13] <clytle374> so use the patch from adeos, not magna
[04:34:22] <memleak> right
[04:34:25] <clytle374> sys-process/acct-6.5.5-r1
[04:34:31] <memleak> ok thanks!
[04:36:16] -!- n2diy has quit [Remote host closed the connection]
[04:36:17] <clytle374> memleak, you gonna be around tomorrow? gonna take 45min to download the kernel and I have to milk the cow in 7 hours
[04:38:15] <memleak> yes i will. intel atom right?
[04:38:42] <clytle374> yes. Thanks a bunch for the help.. It was driving me nuts.
[04:38:59] <clytle374> I know I could 'use' the live cd... But it turned personal
[04:39:45] <memleak> XD
[04:40:25] <clytle374> good night all
[04:40:45] -!- clytle374 has quit [Quit: Leaving]
[04:40:57] <mozmck> memleak: I thought the rtai folks modified the adeos patch somehow...
[04:42:22] <memleak> IPIPE is IPIPE
[04:42:45] <memleak> RTAI might but nothing too significant
[04:43:15] -!- BobGreen [[email protected]] has joined #emc
[04:43:21] -!- Tom_L has quit []
[04:44:07] -!- BobGreen has quit [Quit: User pushed the X - because it's Xtra, baby]
[04:45:01] <mozmck> memleak: do you know if there is any work yet on a 3.x patch for x86?
[04:45:09] -!- packrat has quit [Ping timeout: 248 seconds]
[04:45:17] <memleak> x86 no i do not. powerpc yes
[04:45:29] <memleak> already working on PPC
[04:46:36] <mozmck> I saw that. We will need one for 3.2 for the next LiveCD of EMC
[04:47:18] <mozmck> I wonder how hard it is to port a patch to a newer kernel. More work than I need to tackle right now I'm sure.
[04:51:02] <memleak> I think you should email mante and try getting his help or something.
[04:51:32] <memleak> mante does most of the kernel work and deep core of RTAI. i talked to him before. pretty nice guy.
[04:52:16] -!- hackrat [[email protected]] has joined #emc
[04:52:32] <mozmck> Might do that. I'd like to learn a little more about it anyhow. Looking at a diff between the adeos and rtai patches now.
[04:53:31] -!- psha[work] [psha[work][email protected]] has joined #emc
[04:58:54] <KimK> I saw clytle374's post (yesterday?) about the RTAI folks posting updates again after 10 months of none, that sounds good to us uninformed out here. Hi mozmck, how are you? Are you going to do the kernel for 12.04 again? (And are we using 12.04?)
[04:59:42] -!- memleak has quit [Ping timeout: 240 seconds]
[05:00:08] -!- A2Sheds has quit [Ping timeout: 240 seconds]
[05:03:19] -!- LawrenceSeattle [[email protected]] has joined #emc
[05:05:11] -!- hackrat has quit [Ping timeout: 244 seconds]
[05:06:30] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #emc
[05:10:55] -!- packrat [[email protected]] has joined #emc
[05:11:48] -!- memleak [[email protected]] has joined #emc
[05:11:55] -!- memleak has quit [Changing host]
[05:11:55] -!- memleak [memleak!~memleak@unaffiliated/memleak] has joined #emc
[05:11:58] -!- memleak [memleak!~memleak@unaffiliated/memleak] has parted #emc
[05:12:03] -!- memleak [memleak!~memleak@unaffiliated/memleak] has joined #emc
[05:17:10] -!- mhaberler [mhaberler!~mhaberler@host25-14-static.23-87-b.business.telecomitalia.it] has joined #emc
[05:17:35] -!- packrat has quit [Ping timeout: 244 seconds]
[05:23:38] <memleak> mozmck, find anything conclusive on those diffs?
[05:26:24] -!- packrat [[email protected]] has joined #emc
[05:29:07] -!- factor has quit [Ping timeout: 260 seconds]
[05:36:01] -!- hackrat [[email protected]] has joined #emc
[05:36:01] -!- packrat has quit [Read error: Connection reset by peer]
[05:44:00] -!- n2diy [[email protected]] has joined #emc
[05:55:05] <memleak> im logging off now. take care guys!
[05:55:15] -!- memleak has quit [Quit: Leaving]
[05:57:12] -!- misfitsu has quit [Read error: Operation timed out]
[06:02:25] -!- steves_logging has quit [Ping timeout: 252 seconds]
[06:07:12] -!- factor [[email protected]] has joined #emc
[06:26:04] -!- misfitsu [[email protected]] has joined #emc
[06:35:53] -!- packrat [[email protected]] has joined #emc
[06:37:30] -!- hackrat has quit [Ping timeout: 252 seconds]
[06:44:39] -!- packrat has quit [Ping timeout: 252 seconds]
[06:47:39] -!- packrat [[email protected]] has joined #emc
[06:52:32] -!- packrat has quit [Ping timeout: 260 seconds]
[07:10:35] -!- packrat [[email protected]] has joined #emc
[07:18:10] -!- capricorn_one has quit [Remote host closed the connection]
[07:26:39] -!- n2diy has quit [Quit: Ex-Chat]
[07:29:44] -!- mhaberler has quit [Quit: mhaberler]
[07:32:56] -!- packrat has quit [Ping timeout: 252 seconds]
[07:38:09] -!- packrat [[email protected]] has joined #emc
[07:44:36] -!- packrat has quit [Ping timeout: 240 seconds]
[07:49:31] -!- rooks has quit [Quit: Ex-Chat]
[07:49:52] -!- packrat [[email protected]] has joined #emc
[07:53:41] -!- theorbtwo has quit [Remote host closed the connection]
[07:58:47] -!- theorbtwo [[email protected]] has joined #emc
[08:00:18] -!- e-ndy [e-ndy!jkastner@nat/redhat/x-eijmqhxiwziocgws] has joined #emc
[09:25:20] -!- theorbtwo has quit [Ping timeout: 240 seconds]
[09:39:39] -!- rob_h [[email protected]] has joined #emc
[09:44:32] -!- theorbtwo [[email protected]] has joined #emc
[09:49:25] -!- theorbtwo has quit [Ping timeout: 268 seconds]
[09:51:34] -!- theorbtwo [[email protected]] has joined #emc
[10:05:39] -!- mhaberler [mhaberler!~mhaberler@host25-14-static.23-87-b.business.telecomitalia.it] has joined #emc
[10:56:53] <awallin> nice progress by tom3p on HAL-gui based on 'vision' !. The graph editor is shown here: http://mgltools.scripps.edu/flash/vision
[11:03:37] -!- crib has quit [Remote host closed the connection]
[11:11:27] -!- mhaberler has quit [Quit: mhaberler]
[11:25:39] -!- mhaberler [mhaberler!~mhaberler@host25-14-static.23-87-b.business.telecomitalia.it] has joined #emc
[11:32:44] -!- factor has quit [Read error: Connection reset by peer]
[11:41:07] -!- crib [[email protected]] has joined #emc
[11:50:59] -!- Loetmichel [[email protected]] has joined #emc
[11:51:02] -!- factor [[email protected]] has joined #emc
[11:53:25] -!- Loetmichel_ has quit [Ping timeout: 248 seconds]
[11:56:11] <anonimasu> can anyone have a look at a plasma/oxyfuel thc for me?
[11:58:21] <anonimasu> http://www.cnc4pc.com/Store/osc/product_info.php?cPath=70&products_id=330
[11:59:33] <anonimasu> what should I have pwm version or up/down for interfacing emc?
[12:11:45] -!- mhaberler has quit [Quit: mhaberler]
[12:15:21] -!- rooks [[email protected]] has joined #emc
[12:21:28] <jthornton> I modified the thc comp for up/down, but both would benefit from using a PID. You could test the up down comp as is...
[12:21:42] -!- servos4ever [[email protected]] has joined #emc
[12:34:47] <skunkworks_> http://www.cnczone.com/forums/mach_software_artsoft_software/120113-mach_3_cv_mode_stalling.html
[12:40:39] <anonimasu> well, it has a ext input for adjustment, so I guess I could hook that back into emc and do PID...
[12:44:19] -!- zlog has quit [Remote host closed the connection]
[12:44:21] -!- Tom_itx has quit []
[12:45:49] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[12:45:50] -!- zlog [[email protected]] has joined #emc
[13:02:20] -!- ries [[email protected]] has joined #emc
[13:17:59] -!- factor has quit [Read error: Connection reset by peer]
[13:22:47] -!- skunkworks_ has quit [Ping timeout: 268 seconds]
[13:28:35] -!- ries has quit [Ping timeout: 240 seconds]
[13:29:25] -!- rooks has quit [Ping timeout: 248 seconds]
[13:35:07] -!- factor [[email protected]] has joined #emc
[13:46:02] -!- rooks [[email protected]] has joined #emc
[14:01:48] -!- mhaberler [mhaberler!~mhaberler@host25-14-static.23-87-b.business.telecomitalia.it] has joined #emc
[14:05:00] -!- mshaver [mshaver!~mshaver@173-167-237-41-ip-static.hfc.comcastbusiness.net] has joined #emc
[14:06:41] -!- FinboySlick [[email protected]] has joined #emc
[14:13:10] -!- ries [[email protected]] has joined #emc
[14:35:05] -!- mhaberler has quit [Quit: mhaberler]
[14:39:18] -!- jthornton_ [[email protected]] has joined #emc
[14:40:00] -!- robin__ [[email protected]] has joined #emc
[14:40:05] -!- jthornton has quit [Ping timeout: 240 seconds]
[14:40:05] -!- cjdavis has quit [Read error: Connection reset by peer]
[14:40:05] -!- robin_sz has quit [Ping timeout: 240 seconds]
[14:40:06] -!- misfitsu has quit [Ping timeout: 240 seconds]
[14:40:14] jthornton_ is now known as jthornton
[14:40:15] -!- mshaver has quit [Ping timeout: 240 seconds]
[14:40:15] -!- the_wench has quit [Ping timeout: 240 seconds]
[14:40:15] -!- saschi has quit [Ping timeout: 240 seconds]
[14:40:18] -!- cjdavis [[email protected]] has joined #emc
[14:40:35] -!- A2Sheds has quit [Ping timeout: 240 seconds]
[14:40:55] -!- tlab has quit [Ping timeout: 240 seconds]
[14:41:08] -!- mshaver [mshaver!~mshaver@173-167-237-41-ip-static.hfc.comcastbusiness.net] has joined #emc
[14:41:41] -!- tlab [[email protected]] has joined #emc
[14:41:55] -!- misfitsu [[email protected]] has joined #emc
[14:42:11] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #emc
[14:46:10] -!- A2Sheds has quit [Client Quit]
[14:46:29] -!- A2Sheds [A2Sheds!~ly@unaffiliated/l84supper] has joined #emc
[14:54:16] -!- rob__H [[email protected]] has joined #emc
[14:56:01] -!- mhaberler [mhaberler!~mhaberler@host25-14-static.23-87-b.business.telecomitalia.it] has joined #emc
[14:57:01] -!- rob_h has quit [Ping timeout: 276 seconds]
[15:14:16] -!- factor has quit [Quit: Leaving]
[15:18:45] -!- mhaberler has quit [Quit: mhaberler]
[15:21:42] -!- i_tarzan [[email protected]] has joined #emc
[15:38:06] -!- gene76 [[email protected]] has joined #emc
[15:38:28] <gene76> Good morning all!
[15:41:10] <gene76> First question this morning: Where do I increase the default number of screens? I need more than 4
[15:41:47] <FinboySlick> gene76: In ubuntu, I assume?
[15:42:09] -!- syyl [[email protected]] has joined #emc
[15:42:31] <gene76> yes on my new D525 machine, installed from the linuxcnc cd
[15:43:39] <gene76> 10.04 LTS with supplied emc, but emc is now a 2.6.0-pre2
[15:43:42] <FinboySlick> I think you right-click on the workspace switcher applet and choose preferences.
[15:44:33] <gene76> Got it, thanks!
[15:45:07] <gene76> One would think this is available in a pulldown, but it is not.
[15:46:10] <gene76> So far, w/o isolcpus set, all this has got be a max jitter of 11 u-secs, not too shabby.\
[15:46:37] <gene76> I have done the isolcpus bit to grub, but have't rebooted yet.
[15:47:15] <gene76> Maybe I should do that next...
[15:47:56] <gene76> brb
[15:48:32] -!- gene76 has quit [Remote host closed the connection]
[15:50:05] -!- gene76 [[email protected]] has joined #emc
[15:50:21] <FinboySlick> Yeah, you really do need to reboot for kernel params to take effect.
[15:50:33] <FinboySlick> At least when you pass them as kernel params.
[15:51:54] <gene76> Humm, ran latency-test, got 6 u-secs, switched back to konversation's screen, then back, now 17 u-secs, worse
[15:52:54] <gene76> That sequence got me 11 u-secs without isolcpus=1
[15:53:00] <gene76> ????
[15:53:42] <FinboySlick> gene76: isolcpus disables one of your cores. Anything that you want to run on that isolated core has to be specifically assigned. It's not a magic answer to everything.
[15:54:41] <gene76> I understand I need to set something in my .hal file, but what, or will stepper-config do that for me?
[15:58:16] <skunkworks> gene76: what kind of step generation speed do you need?
[15:59:10] <gene76> I was running the old machine at about 35 u-sec base thread, and moving a bit over 24"/min
[15:59:53] <gene76> And I just pulled in gkrellm but its display is puzzling at best!!!!
[16:01:02] <gene76> it says I have 4 cores, each showing a different load, with core 1 being idle. Atoms are dual core, or do I need to disable hyperthreading too?
[16:01:47] <FinboySlick> gene76: disable hyperthreading
[16:02:17] <FinboySlick> That's likely your main culprit.
[16:02:39] <FinboySlick> The core being idle is probably the second one, which is what isolcpus=1 does.
[16:03:35] <gene76> How do I do that?
[16:03:55] <gene76> turn off the hyperthreading that uis?
[16:04:54] <gene76> Is there a wiki page on that?
[16:07:37] <gene76> If so, I can't find it.
[16:07:59] <gene76> Or is that a bios selectionj?
[16:08:49] <gene76> brb, goona look
[16:08:59] -!- gene76 has quit [Remote host closed the connection]
[16:12:36] -!- psha[work] has quit [Quit: Lost terminal]
[16:13:14] -!- gene76 [[email protected]] has joined #emc
[16:15:32] <gene76> I'm back, it was in the bios
[16:18:17] <gene76> and that seems to have fixed the latency-test at about 7 u-secs after starting firefox
[16:18:28] <skunkworks> gene76: what is your steps per inch on your machine?
[16:18:45] <skunkworks> cool!
[16:19:07] <gene76> So we are looking better. Depends on axis, 32000 to 34xxxx
[16:20:12] <gene76> Using 20 tpi screw for xy, and 16 microsteps in mm542 drivers
[16:21:35] -!- e-ndy has quit [Quit: Ex-Chat]
[16:21:56] -!- GoSebGo [[email protected]] has joined #emc
[16:22:39] <skunkworks> wow - that is quite a bit
[16:22:51] <skunkworks> no wonder you want really low latency :)
[16:24:35] <cpresser> you should consider selecting a lower microstepping value
[16:24:52] <gene76> yeah, but you ought to see how those motors can creep, you can just barely see them move going slow :)
[16:25:00] <cpresser> without microstepping you already are as accurate as 0.007mm/step
[16:25:24] <gene76> yup, noisey though
[16:25:45] <cpresser> higher microstepping makes less torque; if you dont need torque, then its okay:)
[16:25:53] <skunkworks> remember - microstepping doesn't give you higher resolution.. You can really only count on half stepping - and that is pushing it.
[16:26:29] <gene76> true, but even half stepping is noisy
[16:27:07] <gene76> 4 is better, 8 is only tolerable, and 16 is pretty quiet
[16:27:13] -!- chester88 has quit [Ping timeout: 252 seconds]
[16:27:29] <skunkworks> sure - and you still have resonence problems.. - we have good luck with 1000 - 20000 steps per rev with the drives we have used.
[16:27:34] <skunkworks> * 2000
[16:27:50] <gene76> It was an experiment, and I could go back to 8.
[16:27:55] <syyl> i run 1/5th microstepping
[16:27:57] <skunkworks> * with half stepping
[16:28:06] <syyl> and positioning is pretty dead on ;)
[16:28:11] <cpresser> i dont care about noise; so i have 1/4 stepping now
[16:30:01] <gene76> I have resonance dampers, and they help, but so does the higher microstepping by reducing the resonance triggering motions.
[16:30:21] <syyl> good stepper drives also limit the music made by the steppers..
[16:30:38] <gene76> MM-542's?
[16:30:56] <gene76> seem to be better that Xylotex are
[16:31:00] <syyl> http://benezan-electronics.de/shop/product_info.php?cPath=8&products_id=61&language=en&osCsid=865a8cdb29352e2e4051bb05e2f393cc
[16:31:09] <syyl> i run the triple beast
[16:31:45] <syyl> i know the guy who developes and manufactures them
[16:33:30] <syyl> but i like the leadshine drives too
[16:33:39] <gene76> and at 300 euros, 6x the MM-542 a channel
[16:34:10] <gene76> I have US51$ each in these
[16:35:09] <gene76> Now, I'm going to see if I can make a better config with stepper-config, back later.
[16:48:06] <skunkworks> This is cool
[16:48:08] <skunkworks> http://www.cnczone.com/forums/1047924-post257.html
[16:51:55] <pcw_home> When we played with this we found 32 uSteps was about as good as you could get anything less was pretty noisy
[16:51:56] <pcw_home> anything more just bumped up that step rate with no noticeable advantages
[16:52:38] <gene76> back, stepconf has broken pipe, no parport visible???
[16:53:43] <pcw_home> but of course with software step generation you need to compromise on bigger uSteps
[16:56:02] <gene76> so I've rmmod'd all thje parport stuff, what do I put back in by itself?
[16:56:48] <skunkworks> yeck. http://www.cnczone.com/forums/mach_software_artsoft_software/120113-mach_3_cv_mode_stalling.html#post1048125
[16:56:55] <skunkworks> (last post)
[16:57:08] <skunkworks> wonder what the even means.
[16:57:33] <pcw_home> gene76 wheres that undo command when you need it
[16:58:16] <skunkworks> wonder if the 'faceting' he is seeing is the actual gcode path...
[16:58:53] <cradek> constant velocity is trivially easy to implement if you don't care whether you exceed vel/accel constraints and stall your motors or worse
[16:59:25] <gene76> undo, I was getting rid of lp
[17:01:42] <gene76> loaded everything else but no parport in an lspci list
[17:02:31] <gene76> Got this in dmesg
[17:02:38] <gene76> [ 2882.117823] parport_pc 00:08: activated
[17:02:40] <gene76> [ 2882.117833] parport_pc 00:08: reported by Plug and Play ACPI
[17:02:41] <gene76> [ 2882.117898] parport0: PC-style at 0x378, irq 7 [PCSPP]
[17:02:43] <gene76> [ 2882.141254] ppdev: user-space parallel port driver
[17:02:59] <gene76> but still a broken pipe & no lspci listing
[17:04:05] <gene76> coffee cup ampty, back in a few while you all cogitate on this
[17:39:02] <pcw_home> The parallel port would not show up in lspci as its not a pci device (I think its a LPC (low pin count ISA) interfaced SuperIO chip on the 525)
[17:45:15] -!- WalterN [[email protected]] has joined #emc
[18:02:41] -!- psha [[email protected]] has joined #emc
[18:03:31] <gene76> I _think_ this might be it:
[18:03:36] <gene76> 00:1f.0 ISA bridge: Intel Corporation NM10 Family LPC Controller (rev 01)
[18:03:37] <gene76> Subsystem: Intel Corporation Device 574d
[18:03:39] <gene76> Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx-
[18:03:40] <gene76> Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
[18:03:42] <gene76> Latency: 0
[18:03:43] <gene76> Capabilities: [e0] Vendor Specific Information <?>
[18:04:31] <gene76> sorta walks & quacks like that duck. :)
[18:05:58] <gene76> Anyway, I've carved out a new config, but it may need tweeking when I get the right motors hooked up, I'm in the garage with only that draggy A drive to test with
[18:06:19] <gene76> and its running about half speed because of that.
[18:08:57] <gene76> So its about time to truck it all back up the hill to the shop & see if it will carve parts. I think it will. I have not seen a realtime error popup at all. Thanks for the help guys.
[18:09:43] <gene76> Gotta change my handle, I'm 77 nowadays. ;-)
[18:09:44] <skunkworks> cool beans!
[18:10:06] -!- gene76 has quit [Remote host closed the connection]
[18:11:34] -!- Loetmichel has quit []
[18:12:02] -!- LawrenceSeattle has quit [Quit: LawrenceSeattle]
[18:14:44] <skunkworks> http://www.machsupport.com/forum/index.php/topic,20518.0.html
[18:15:00] <skunkworks> (same guy with the contouring problem above)
[18:15:46] <skunkworks> the second post... from BR549 - His statement is 'I think' because he switched to emc2......
[18:15:55] -!- The_Ball has quit [Ping timeout: 276 seconds]
[18:16:43] <pcw_home> Thats the bridge (from PCI? to LPC) but it does not show the LPC devices hanging off the bridge
[18:20:36] <skunkworks> (yes he has been using emc for a couple of years now)
[18:22:09] <skunkworks> pcw_home: it is like arguing with a rock - isn't it?
[18:27:08] -!- Nick001 has quit [Ping timeout: 252 seconds]
[18:27:14] <pcw_home> Yeah I seems like a fair hypothesis that is you have conservative acceleration settings (and these are violated) you would get better contouring performance
[18:27:15] <pcw_home> (until you stall at the worst places) But its just a guess. It also so dependent on the actual G code vs interpreter options
[18:30:22] -!- ve7it [[email protected]] has joined #emc
[18:30:23] <pcw_home> I guess in EMC you could plot the axis acceleration and see if thats the limiting factor
[18:35:51] -!- IchGucksLive [[email protected]] has joined #emc
[18:35:56] <IchGucksLive> Hi all
[18:37:57] <pcw_home> Does EMC have a flag to tell you that it cant maintain feed rate due to acceleration bounding?
[18:39:23] <pcw_home> Hi IchGucksLive how the weather in ?Germany?
[18:42:01] <IchGucksLive> Nice
[18:42:47] <IchGucksLive> im also searching if there is a halpin to tell my python pendant that the mdi is finished
[18:42:49] <syyl> its dark over here, pcw_home ;)
[18:43:02] <syyl> and warm Oo
[18:43:07] <pcw_home> Ha!
[18:43:55] -!- Loetmichel [[email protected]] has joined #emc
[18:44:32] <Loetmichel> re @ home
[18:44:36] <pcw_home> Its a funny very dry and warm winter here
[18:44:37] <pcw_home> this is usually our coldest month but it was almost 70F yesterday
[18:45:19] -!- LawrenceSeattle [[email protected]] has joined #emc
[18:46:18] <pcw_home> no rain in the last month or so either
[18:46:36] <rob__H> its like there here in the UK too 10c today where normaly we have little snow or cold wind
[18:51:24] <skunkworks> it has been close to 50f here the last few days... (WI) go figure
[18:55:49] <skunkworks> (that is shorts and t-shirt weather here..)
[18:58:22] <mshaver> pcw_home: Any chance of getting new 8i20 firmware today?
[18:59:57] <pcw_home> Likely tommorow Gilbert not here today and he's the DSPIC guru
[19:01:08] <mshaver> OK, just getting a little desperate here :)
[19:01:39] -!- gene76 [[email protected]] has joined #emc
[19:03:07] <gene76> Back, hooked up out here, moved the backlash vaues into the config, but I can only run the spindle in reverse at bare min speed, about 25 rpm.
[19:03:23] <pcw_home> I can send you the new utilities today (I forgot last week)
[19:03:25] <pcw_home> BTW you need to be in setup mode to set the EEPROM variables
[19:03:26] <pcw_home> (this should not be required but its a bug in the current firmware)
[19:04:50] <mshaver> When you say firmware (where the bug is), do you mean the DSPIC firmware, or the bit file?
[19:05:09] <pcw_home> DSPIC
[19:05:14] <gene76> IIRC I had some ABS functions in that at one time because I do reverse with a relay called CCW, The signal pwm is always FWD in my case. So I guess I study the hal manual :)
[19:07:08] <pcw_home> you would use the UART config to access the 8I20 in setup mode (@115200 baud)
[19:07:09] <pcw_home> Or windows with a serial port or USB-serial adapter
[19:07:57] <A2Sheds> anyone know a low cost source for shaft couplings 5-10mm in the US?
[19:08:28] <pcw_home> (5I20 UART config --> 7I47 --> 8I20 (DOS/ MENV.BAT)
[19:09:08] <pcw_home> (or windows PC --> serial port--> 8I20)
[19:09:18] <mshaver> I'll have to look back in my e-mails with you to see what i did before. Oh, OK! MENV.BAT rather than ENV.BAT. Plus change the jumper.
[19:09:33] -!- IchGucksLive has quit [Quit: ChatZilla 0.9.87 [Firefox 9.0.1/20111221214946]]
[19:11:09] <mshaver> Although, with the 1.14 bit file (SS2.BIT) and the existing DOS disk I have, I was able to change NVMAXCURRENT on one of our drives. How do I check the current DSPIC firmware version?
[19:13:48] <pcw_home> You may get away with it but its buggy (sometime fails to write/crashes)
[19:13:49] <pcw_home> no problem in setup mode
[19:14:43] <mshaver> OK. I'm going to exercise some of this this afternoon so that I am ready for tomorrow!
[19:16:25] <pcw_home> it has to do with the fact the the DSPIC has no EEPROM so non-volatile variables are stored in flash progrm memory
[19:16:26] <pcw_home> this requires halting the processor and gracefully restoring interrupt tasks etc
[19:18:13] <pcw_home> and its not quite right in version 4 firmware (in operate mode)
[19:18:15] <pcw_home> I would at least reduce the PI loop gains slightly to cure your overshoot problem
[19:18:16] <pcw_home> Do you know the motor winding inductance?
[19:23:22] -!- the_wench [[email protected]] has joined #emc
[19:31:10] -!- PCW [[email protected]] has joined #emc
[19:35:44] -!- GoSebGo has quit [Quit: Bye]
[19:36:02] -!- GoSebGo [[email protected]] has joined #emc
[19:37:08] -!- ries_ [[email protected]] has joined #emc
[19:37:15] -!- ries has quit [Ping timeout: 240 seconds]
[19:37:15] ries_ is now known as ries
[19:41:42] -!- psha has quit [Quit: Lost terminal]
[19:44:52] <gene76> back with a new puzzle: I've lost my spindle speed +- buttons in axis need clue
[19:46:45] <gene76> so thye spindle can only turn dead slow, either direction
[19:52:08] <cradek> http://www.linuxcnc.org/docs/devel/html/gui/axis.html#_manual_control
[19:52:17] -!- bilboquet has quit [Remote host closed the connection]
[19:52:48] -!- bilboquet [[email protected]] has joined #emc
[19:53:16] -!- bilboquet has quit [Client Quit]
[19:54:32] -!- sumpfralle [[email protected]] has joined #emc
[19:54:46] -!- bilboquet [[email protected]] has joined #emc
[19:56:20] <gene76> my-mill-atom.hal:27: Signal name 'abs.0.out' must not be the same as a pin.
[19:57:06] <gene76> line 27;net abs.0.out => pwmgen.0.value
[19:57:32] <gene76> I am not groking this at all
[19:57:58] <cradek> net signalname pin ... => pin ...
[19:58:51] <gene76> and pwmgen.0.value is a pin?
[19:59:14] <PCW> yep you need a wire!
[20:00:30] <PCW> signalname is the wire (net) name
[20:02:45] <PCW> net skinny_blue_wire_with_yellow_stripe abs.0.out => pwmgen.0.value
[20:03:52] <gene76> so I would say "net spindle-abs <= abs.0.ou"t and then "net spindle-abs => pwmgen.0.value" on the next line?
[20:04:07] <gene76> so I would say "net spindle-abs <= abs.0.out" and then "net spindle-abs => pwmgen.0.value" on the next line?
[20:04:17] <PCW> all in one line is fine
[20:04:41] <gene76> demo plz
[20:05:21] <PCW> net spindle-abs abs.0.out => pwmgen.0.value
[20:06:18] <bilboquet> hello, i search an dxf to gcode software. which do you recommend
[20:07:36] <gene76> custom_postgui.hal:9: Pin 'abs.0.out' was already linked to signal 'spindle-abs', because I specced a speed gui
[20:11:33] <gene76> and removing that line with a # sign doesn't fix it, same error exactly??
[20:11:59] <gene76> Do I need to flush a cache someplace?
[20:13:36] <gene76> Or, do I need to load 2 copies of abs and isolate those functions for the gui from the machine stuffs?
[20:14:02] <gene76> Using abs.1.in and abs.1.out?
[20:15:12] <PCW> what does the speed GUI do?
[20:18:11] -!- ries_ [[email protected]] has joined #emc
[20:18:11] -!- ries has quit [Read error: Connection reset by peer]
[20:18:12] ries_ is now known as ries
[20:18:40] <gene76> With the use of abs.1.in and outs, I can run slow, either direction, the gui does't show the speed, only 0 and the at speed is green in the gui.
[20:20:53] <gene76> Re-enabled that line in the custom_postgui.hal, the gui speed works, spindle doesn't
[20:21:29] <PCW> You should only need one abs, you still have a wiring problem
[20:22:11] <gene76> It comes on, the -+ sign button are unghosted, but no response to a click from the spindle
[20:22:25] <gene76> Here is that whole stanza:
[20:22:50] <gene76> net spindle-cmd <= motion.spindle-speed-out
[20:22:51] <gene76> net spindle-abs abs.1.out => pwmgen.0.value
[20:22:53] <gene76> net spindle-cmd => abs.1.in
[20:22:54] <gene76> net spindle-enable <= motion.spindle-on => pwmgen.0.enable
[20:22:56] <gene76> net spindle-pwm <= pwmgen.0.pwm
[20:22:57] <gene76> setp pwmgen.0.pwm-freq 100.0
[20:22:59] <gene76> setp pwmgen.0.scale 2500
[20:23:00] <gene76> setp pwmgen.0.offset 0.0936734693878
[20:23:02] <gene76> setp pwmgen.0.dither-pwm true
[20:23:03] <gene76> net spindle-ccw <= motion.spindle-reverse
[20:23:43] <gene76> I also need a high limit to about 97%
[20:25:28] -!- The_Ball [[email protected]] has joined #emc
[20:26:02] <PCW> and net spindle-pwm connects to the proper parallel port pin?
[20:27:04] <gene76> yes, its getting to the motor controller, but stuck at the 0.offset value
[20:27:41] <gene76> That s/b pin 16 but lemme check
[20:27:57] <PCW> Only a few nets, should be easy to trace with halmeter
[20:28:11] <gene76> yes:net spindle-pwm => parport.0.pin-16-out
[20:30:21] <PCW> so I guess i would trace from motion.spindle-speed-out on
[20:31:01] <gene76> spindle-abs has no value
[20:31:28] -!- Valen [[email protected]] has joined #emc
[20:31:37] <PCW> but spindle-cmd does?
[20:32:54] <gene76> Yes, I think missing this might be it, brb
[20:32:58] <gene76> addf abs.1 servo-thread
[20:34:20] <skunkworks> heh - I have done that a a few times - forget to add a function to a thread
[20:34:24] <gene76> That was it!
[20:34:48] <gene76> ok, now how do I put in a + limit?
[20:35:22] <gene76> If I hit the + button too many times it goes to 100% and shuts it off.
[20:35:48] <PCW> pwmgen.N.max-dc
[20:36:07] <gene76> net or setp
[20:36:22] <PCW> setp
[20:36:28] <gene76> good
[20:36:48] <PCW> man pwmgen
[20:40:49] -!- factor [[email protected]] has joined #emc
[20:41:39] <gene76> Ok works, but I need to scale the pyvcp whatever, sometime later , thanks a bunch guys.
[20:45:28] -!- LawrenceSeattle has quit [Read error: No route to host]
[20:46:05] -!- LawrenceSeattle [[email protected]] has joined #emc
[21:37:09] -!- clytle374 [[email protected]] has joined #emc
[21:42:28] <clytle374> Found out today the ProtoTrak mills are designed to be repair resistant
[21:56:42] <clytle374> You have to disconnect everything to get the electrical box out of the larger box, then hook it all back up on a bench. Just to get to a bad contactor
[22:05:03] -!- Nick001-Shop [Nick001-Shop!~chatzilla@wlk-barre-208-103-141-230.dynamic-dialup.coretel.net] has joined #emc
[22:06:59] -!- sumpfralle has quit [Ping timeout: 248 seconds]
[22:08:28] -!- sumpfralle [[email protected]] has joined #emc
[22:14:55] <clytle374> I rebuilt the kernel with memlock's instructions. Rtai tests and emc's tests all still complete without errors.
[22:15:21] <clytle374> I finished the git bisect and this is the result http://pastebin.com/PQ8uLfek
[22:15:39] <clytle374> gotta go feed the animals, back in a few
[22:17:11] <clytle374> FYI This is the watchdog biting while card(5i20) is initializing issue.
[22:38:02] -!- FinboySlick has quit [Quit: Leaving.]
[22:54:11] -!- Fox_Muldr has quit [Ping timeout: 252 seconds]
[22:56:08] -!- Fox_Muldr [[email protected]] has joined #emc
[22:58:51] -!- skunkKandT [[email protected]] has joined #emc
[23:06:15] -!- ries has quit [Ping timeout: 240 seconds]
[23:07:18] -!- syyl has quit [Quit: Leaving]
[23:22:08] -!- i_tarzan has quit [Ping timeout: 240 seconds]
[23:25:04] -!- WalterN has quit [Read error: Connection reset by peer]
[23:26:26] -!- ries [[email protected]] has joined #emc
[23:29:56] -!- robin__ has quit [Remote host closed the connection]
[23:31:24] -!- jy76 [[email protected]] has joined #emc
[23:32:49] -!- TekniQue has quit [Read error: Connection reset by peer]
[23:47:57] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust492.basl.cable.virginmedia.com] has joined #emc
[23:56:19] -!- jy76 has quit [Quit: Bye Bye]
[23:56:31] -!- jy76 [[email protected]] has joined #emc