#emc | Logs for 2009-05-08

Back
[01:16:00] <dgarr> introductory robotics course videos (stanford univ). each lecture includes an interesting robot demo. http://academicearth.org/courses/introduction-to-robotics
[03:12:55] <slick-> hello
[03:28:08] <tomp> tomp is now known as tomp3
[03:55:39] <slick-> what does it mean when I turn on my computer, with my parallel port connected to my stepper drivers and the drivers are "holding" the motors iow both coils are powered on
[03:57:52] <eric_unterhausen> it probably means that you shouldn't power your steppers until your computer is on and emc can control them
[03:59:22] <eric_unterhausen> holding doesn't mean that both coils are powered on
[04:00:05] <tomp3> archivist: any tutorials for heekscad?
[04:00:48] <eric_unterhausen> didn't he say he was leaving?
[04:02:21] <tomp3> dunno just got here
[04:02:24] <slick-> hmmm well my driver has LEDs for each coil and both are lit...
[04:03:15] <slick-> even after EMC starts with test config I made with the wizard the drivers don't change
[04:03:35] <eric_unterhausen> are you sure it's connected?
[04:04:40] <slick-> to the parallel port?
[04:04:55] <eric_unterhausen> yes
[04:04:55] <slick-> yeah, but I'm not sure it is connected right...
[04:05:16] <eric_unterhausen> do you have a breakout board?
[04:05:33] <slick-> well sorta
[04:05:50] <eric_unterhausen> what drivers?
[04:06:08] <tomp3> does your driver turn on a coil with a high or a low? (check with halscope and a meter.. halscope will tell you what emc is trying to do, meter will tell you what is happening )
[04:06:59] <slick-> ok ... whats halscope?
[04:07:55] <slick-> I'm pretty sure my drive turns on with HI I have controlled the steppers with step/dir coming from an arduino
[04:08:21] <eric_unterhausen> so your drivers are step/dir?
[04:08:47] <slick-> yes
[04:09:13] <slick-> step dir and the need an enable pin set low to enable
[04:10:19] <slick-> in the wizard I checked the invert box for the pins I set to enable was that right?
[04:13:47] <slick-> argh I have to do a lot more reading ...
[04:13:56] <slick-> thanks for the help
[06:57:15] <micges> good morning
[07:06:09] <micges> I've installed eric's smp packages on Pentium Dual Core E5300 and it hang up with "(initramfs)" prompt
[07:55:45] <Lerman> Lerman is now known as lerman
[08:04:48] <micges> lerman: hello
[11:29:11] <BigJohnT> is there a linux scandisk?
[11:35:39] <obinou> BigJohnT: check photorec & testdisk
[11:36:20] <obinou> (photorec is not only for pictures - don't be fooled by the name)
[11:36:27] <tomp> tomp is now known as tomp3
[11:36:33] <BigJohnT> ok thanks
[11:40:30] <BigJohnT> I have one computer that will not boot up all the time and makes it to the ubuntu progress bar then stops
[11:41:32] <skunkworks405> isn't there a 'nosplash' you could put in the boot line maybe to see where it hangs?
[11:41:41] <skunkworks405> (I am not a linux person)
[11:42:14] <BigJohnT> I don't know
[11:43:22] <micges> BigJohnT: when grub is showing stage 1.5 then hit eascape
[11:43:56] <BigJohnT> ok thanks micges
[11:44:19] <micges> then hit e to edit boot params and then remove word SPLASH or simmilar
[11:46:03] <micges> will be at second line with KERNEL word
[11:48:20] <tomp3> if its running, then edit /boot/grub/menu.lst and do as micges suggests.
[11:48:20] <tomp3> on mine the line is 'kernel blahblahblah splash', so make it kernel blahablahblah'
[11:50:59] <tomp3> maybe remove 'quiet' also... that makes for a lot of info, everything that dmesg would have
[13:11:24] <Optic> moo
[13:55:59] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[15:10:05] <NewType> NewType is now known as NewType_away
[15:48:18] <tomp> tomp is now known as tomp3
[16:17:23] <gene> Cradek: is it possible to use the tool table to make a font like dejavu-sans look like straight stick by offsetting the dig motion to the edge of the bit?
[16:23:33] <cradek> I don't think that would be successful.
[16:24:03] <gene> Or to use the g410.0625 to cancel most of a 5/64's drill bit being used to engrave this stuff?
[16:24:48] <gene> That I can try with an already working code for a B :)
[16:24:57] <cradek> ??
[16:30:47] <gene> The tracing is following the outside of the character. I want the tracing to follow the inside edge of the character by offsetting the bit by its radii.
[16:31:41] <gene> But I just found that G41 only accepts integer value for the tool carousel socket #
[16:34:04] <gene> Do I have to edit the tool.tbl and then load a dummy tool position? Seems hard to do for one off stuffs.
[16:34:33] <cradek> g41.1
[16:36:09] <gene> My docs are old?
[16:36:28] <steves_logging> steves_logging is now known as steve_stallings
[16:37:09] <cradek> gene: my favorite doc is this one: http://www.linuxcnc.org/docs/html/gcode.html
[16:37:21] <cradek> each gcode is a link to more details about how to use it
[16:37:51] <gene> and it just fussed about the dot. This is 2.2.8 from the repos.
[16:38:30] <cradek> I'm pretty sure 2.2.8 has g41.1
[16:38:43] <cradek> the docs say so, anyway
[16:40:30] <gene> Looks like I was missing the D, I'll try that. Thanks
[16:41:39] <cradek> with 2.2's cutter compensation, it's virtually impossible to use it on a path that wasn't specifically written for compensation.
[16:41:58] <cradek> with 2.3 you may have better luck, but without writing entry/exit moves by hand, it probably still won't be quite what you want.
[16:42:34] <gene> Now its bitching about a concave corner & gouging. Axis shows only a few moves before the error.
[16:43:33] <cradek> yep that's what I meant above
[16:46:14] <gene> and continues to fuss even when the g41.1D0.0001 is set, talk about picky!
[16:46:28] <cradek> haha
[16:46:37] <cradek> I bet D0.0 will work right, though
[16:46:58] <gene> So I have to write an alphabet for ngc. Ugg.
[16:47:38] <gene> yup 0.0 works.
[16:49:04] <cradek> did you try http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Simple_EMC_G-Code_Generators#Text_Engraving_Software
[16:49:19] <cradek> I think some of the qcad fonts ARE single line
[16:49:35] <gene> No it doesn't, code loads, but won't run
[16:50:01] <gene> But the last time I looked at qcad, it didn't output .ngc?
[16:50:17] <cradek> check out that link
[16:50:17] <BJT-Work> nope
[16:50:27] <BJT-Work> dxf
[16:50:38] <gene> and what converts?
[16:50:44] <cradek> it has nothing to do with qcad except it for how it uses qcad's fonts
[16:51:31] <gene> and I don't have it installed here, be back eventually...
[16:59:18] <gene> insstalling, but I had to stop & restart emc to get rid of the cutter raduis error even with that line replaced with a G40. Bug?
[17:00:28] <cradek> if you get an error in the middle of a program, you may end up with cutter compensation still on. A G40 or better, M2 in MDI mode will fix it.
[17:03:53] <gene> I see. Ok, I just redid the B from dejavu-sans, and the motion is all climb cut.
[17:05:12] <gene> I'm using a scale of 0.0013 for a char about .350 high, and the bit is actually drawing two groves, How can I offset that to put the bit in the middle of the char on both passes?
[17:08:24] <gene> It is a 5/64's bit, seem like I could use a small offset since the cut it always climb, and get what I want..
[17:09:29] <gene> updatedb is running, so I can find the qcad fonts.
[17:11:00] <gene> Humm, those are all .cxf, can ttt use them?
[17:26:28] <gene> Looking at them in qcad, those aren't very good fonts.
[17:32:30] <gene> and ttt spits them out. line 225 unk font error
[18:53:36] <steve_stallings> steve_stallings is now known as steves_logging
[19:49:35] <NewType> I am back. Lunch time.
[19:49:37] <NewType> :)
[19:56:28] <NewType> I am getting confused with the SCALE for B and C axis
[19:56:39] <SWPadnos> counts per unit, just like the other
[19:56:45] <SWPadnos> only units are degrees of rotation
[19:56:59] <SWPadnos> (unless you've gone and used radians or something)
[19:57:35] <NewType> what should be the unit? step/deb?
[19:57:43] <NewType> step/deg? opps
[19:57:56] <NewType> OK, so it is in deg.
[19:58:03] <SWPadnos> B and C are, by definition, rotations
[19:58:34] <NewType> yep. I am using stepconf, and I got a bit confused because I am thinking about the gear reduction I have on the stepper.
[19:58:58] <NewType> the geared down stepper is 0.036deg/step. in fact, I can just use that number in stepconf
[19:59:15] <SWPadnos> I'm nmot sure stepconf will do what you want for a machine with more than 4 axes, or for a machine that uses any axis (joint) named other than XYZA
[19:59:40] <cradek> I'm sure that's beyond the capabilities of stepconf
[19:59:53] <NewType> yeah, I've gone through the whole thing yesterday to get B and C configured. Thanks to a lot of help here. :)
[20:00:03] <NewType> I am trying to get the scale right
[20:00:06] <cradek> if your scale is actually .036deg/step, that's 27.7777778 (kind of odd)
[20:00:38] <NewType> It is a 30:1 down scale gear head, but the stepper is already giving me the scale down number.
[20:00:43] <NewType> that's why I was confused.
[20:00:57] <NewType> because it will translate to 10000 steps per revolution
[20:01:22] <NewType> but the leadscrew pitch is therefore 360 deg/rev.
[20:01:36] <NewType> which is just therefore 1 to 1.
[20:01:43] <NewType> I was getting confused why it was there.
[20:02:02] <NewType> I think it has to do if you don't have the number, 0.036deg/step
[20:03:05] <NewType> I also have another problem with the spindle... it will start turning when the program starts... and then stop
[20:03:40] <NewType> that shouldn't happen. I guess there was a 5VDC signal on the line during the initialization of the port.
[20:03:45] <NewType> I think.
[20:04:28] <NewType> cradek, how did you get 27.777778 for scale?
[20:05:15] <NewType> oh, 1/0.036
[20:05:42] <NewType> why don't I get the same number using stepconf...
[20:10:32] <jepler> the stepconf "A Axis Configuration" page's questions don't make a lot of sense
[20:11:21] <NewType> I just got an ERROR: Module probe partport is in use
[20:11:36] <NewType> may be I didn't exit stepconf correctly just now.
[20:12:07] <NewType> humm... how do I get out?
[20:14:00] <NewType> oh well, let me restart the computer... brb
[20:23:43] <NewType> back!
[20:24:23] <NewType> actually I am going away for now...
[20:24:28] <NewType> NewType is now known as NewType_away
[20:24:33] <NewType_away> oh crap.
[20:24:47] <NewType_away> never mind. I'll be back later.
[21:37:38] <Guest204> Guest204 is now known as andypugh
[21:43:15] <andypugh> Does anyone know if the 2.3 upgrade (from a .deb file) overwrites the
[21:43:37] <andypugh> ...rtai.conf file?
[21:44:18] <andypugh> running the (new) latency testing widget it seems I am getting the every-64-seconds SMI glitch.
[21:45:31] <andypugh> I am currently trying to upgrade my Virtual machine to see what happens there.
[21:54:24] <jepler> andypugh: if you modified rtai.conf, then the upgrade process should prompt you to find out whether you want to keep your changes or install the package version.
[21:54:36] <andypugh> I accidentally closed the window. Did I miss any replies?
[21:55:01] <jepler> you mean /etc/emc2/rtapi.conf, correct?
[21:55:39] <andypugh> Indeed. I had added the rtai_smi.ko module to disable the SMI interrupt
[21:56:45] <jepler> you mentioned a virtual machine -- the timings you get when running latency test in a virtual machine don't mean anything.
[21:56:48] <andypugh> I did get a prompt, but I think that was asking if I wanted to install a new RTAI module, not chnge the conf. I could be wrong, and the installer quit with a segmentation fault right after.
[21:57:22] <jepler> if you didn't destroy the error, you should show it to us
[21:57:29] <andypugh> Oh, indeed, I am just re-running the updater on the virtual machine to see if the rtai.conf modification date changes.
[21:58:01] <andypugh> Destroy the error?
[21:58:05] <jepler> yeah, close the window or whatever
[21:58:14] <andypugh> I probably did, as the machine is shut down now.
[21:58:17] <jepler> i.e., what most users do as soon as they see an error :-/
[21:59:23] <andypugh> Guilty, I am afraid. I simple-mindedly restarted and tried again.
[22:01:46] <jepler> bbl
[22:01:52] <andypugh> I was installing with dpkg using the .deb files. When dpkg stopped with a segmentation fault I just guessed I had installed in the wrong sequence, and that it needed a restart to sort out the dependencies
[22:06:27] <JymmmEMC> Has anyone ever come across a NTSC to ATSC converter? This is the opposite of every box you would see in the stores today.
[22:09:19] <andypugh> As in TV signal converter?
[22:10:04] <JymmmEMC> yes
[22:13:06] <andypugh> http://www.converters.tv/news/hdmi_to_video_cv_sv_downscaler_now_available_71.html
[22:13:33] <andypugh> Or do you want an upscaler?
[22:33:31] <andypugh> OK, time I did other stuff