#linuxcnc | Logs for 2012-12-24

[03:57:39] <DJ9DJ> moin
[05:02:23] <r00t4rd3d> http://i.imgur.com/x3fkd.jpg
[05:02:30] <r00t4rd3d> anyone know what that is?
[05:06:31] <archivist> bed not made
[05:29:30] <TekniQue> r00t4rd3d: could it be a musical instrument?
[05:29:48] <TekniQue> does it make a notable sound when you knock on the side of it?
[05:30:10] <p0g0> It appears to be a spalted maple and black walnut stash box.
[06:11:11] <Thetawaves> 'stash' box
[06:11:14] <Thetawaves> XD
[06:22:03] <TekniQue> Thetawaves: in the only channel on freenode where THC means something other than cannabis
[06:42:10] <roh> *g*
[08:18:04] <jthornton> rob_h, http://gnipsel.com/linuxcnc/gui/gui08a.html
[08:19:05] <archivist> Gremlin is the python equivalent to what the Axis GUI uses. "what"
[08:20:10] <archivist> what-> graphics window or something
[08:20:19] <jthornton> working on it
[08:22:00] <jthornton> do a refresh and see if it makes any sense now... I just wrote it and have not read it :)
[08:23:57] <archivist> you still dont say what it is, you say its a equivalent to something another gui used to show a plot
[08:24:46] <jthornton> help me out
[08:25:38] <archivist> I would be guessing as I have not looked at exactly what gremlin does for the user
[08:28:52] <jthornton> I don't know how to explain it better...
[08:29:17] <archivist> is it an opengl graphics window
[08:29:55] <jthornton> I don't know, I'll see if I can see any clues in the gremlin.py file
[08:31:18] <jthornton> refresh the page and see if it is better
[08:32:07] <jthornton> it seems to be a gtk.gtkgl.widget.DrawingArea
[08:33:23] <archivist> so it is not obvious to me why you use that over the one axis used, I dont see a differentiation
[08:33:39] <jthornton> axis uses tcl
[08:35:25] <archivist> I thought axis itself was python
[08:36:32] <archivist> equivalent of the tcl graphics window
[08:42:46] <jdh> gremlin is also tcl?
[08:45:52] <jthornton> no it is python
[08:53:43] <jthornton> with gremlin and the python interface you can create a gtk python GUI from scratch
[08:55:57] <jthornton> Axis seems to be a mix of python and tcl and other stuff
[08:57:47] <awallin_> axis.py isn't exactly easy to read.. I tried once or twice
[08:58:23] <jthornton> I tried several times to read it
[09:00:44] <jthornton> it's only 3343 lines long...
[09:02:49] <jthornton> it uses a bunch of root_window.tk.call to draw in the canvas I think
[09:45:11] <jdh> what gui toolkit does gremlin use?
[09:46:24] <jdh> looks like gtk
[09:51:49] * JT-Shop just tapped the garage with the backhoe boom :(
[09:51:58] <JT-Shop> I think it is all gtk
[09:51:59] <jdh> oops
[09:52:24] <jdh> axis is python/tcl, gremlin and friends are python/pygtk
[09:52:32] <JT-Shop> don't look too bad, a bit of facia and a little straighening up
[09:53:11] <jdh> err.. python/tk
[09:53:18] <JT-Shop> and some magic I think
[09:53:25] <jdh> magic is often helpful
[09:53:33] <JT-Shop> somehow tcl plays a part I think
[09:54:56] <JT-Shop> I know ngcgui is tcl and it plays nice with Axis
[09:56:45] <JT-Shop> but, I've been told that you can't use ngcgui with a python gtk GUI
[09:57:23] <JT-Shop> here is a line from axis.py nf.source_lib_tcl(root_window,"axis.tcl")
[09:57:37] <jdh> oh
[09:59:04] <JT-Shop> the calibrate window is tcl as well as hal show and pickconfig
[09:59:10] <JT-Shop> and others
[09:59:41] <JT-Shop> now make my ELO USB touch screen work!
[09:59:50] <jdh> I've never been sure where tcl & tk meet
[10:02:37] <JT-Shop> me neither
[11:21:55] <IchGuckLive> hi all and merry X-mas to all in Europ
[11:22:30] <IchGuckLive> markvandenborre: did you got spammed with information
[12:31:45] <JT-Shop> hmm I set up an ELO usb touch screen on the BP mill following the instructions here https://help.ubuntu.com/community/EloTouchScreen
[12:32:11] <JT-Shop> trying to set another one up on a different computer and the touch does not work but lsusb shows the touch screen
[12:33:29] <Tom_itx> different stages of OS updates?
[12:34:25] <JT-Shop> should be identical OS
[12:35:51] <JT-Shop> guess I'
[12:36:06] <JT-Shop> ll try a hardware swap after while when I get done making parts
[12:36:51] <Tom_itx> would be my next suggestion :)
[12:38:57] <JT-Shop> I did figure out this one has APR touch technology
[12:40:19] <JT-Shop> LOL the latest ELO Ubuntu driver is for 8.04
[12:48:18] <tjb1> Afternoon all!
[14:07:29] <Tom_itx> JT-Shop, do you have a good speed/feed calc for lathes?
[14:07:35] <Tom_itx> material guide etc
[14:46:47] <JT-Shop> I do steel and aluminum and just use G96
[14:47:24] <JT-Shop> I do have a spread sheet for drilling and milling
[14:47:40] <Tom_itx> manual lathe
[14:48:15] <Tom_itx> i cut by feel but a bud of mine just got a lathe and was asking
[14:56:50] <rob_h> the tip packet is a good place to start
[15:12:23] <userHD> Merry Christmas @ all
[15:13:27] <tjb1> This guy on cnczone keeps pming me questions, but he puts like 1 sentence per message and sends like 4 at a time. Inbox gets full pretty quick :/
[15:35:32] <gene__> Q: I have a loadrt lowpass module setp'd to a gain of 8, input is around +2.5, is addf'd to servo thread, but output is -nan???
[15:36:44] <gene__> What did I miss?
[15:42:05] <cradek> gene__: you may have to load an initial value by toggling the load pin
[15:42:38] <cradek> it looks like lowpass doesn't have an initial state that's forced to make sense otherwise
[15:44:33] <gene__> On restart, the output is 0 before I move the spindle, move the spindle and it goes to -nan, and the pointer of the tachometer I just made in pyvcp disappears. I didn't have to toggle it 2 weeks ago.
[15:45:01] <gene__> lemme setp the load pin to true just for S&G, BRB
[15:45:18] <cradek> well if it's how you describe, it's not an initial state problem
[15:46:03] <gene__> setp true, its following, apparently in real time
[15:46:58] <cradek> once you feed it a nan, ever, it will stick that way, unless you load a new value
[15:47:18] <gene__> back to false and its out in the south 40 someplace called -nan
[15:47:22] <cradek> because the past is always averaged in, and a single drop of nan is enough to always give you a nan
[15:48:16] <cradek> so your problem is probably before the lowpass input
[15:49:02] <gene__> so what good is a lowpass that doesn't pass? The spindle is currently turning at about 2.8 rps
[15:49:18] <gene__> the lowpass input is encoder.0.velocity
[15:56:56] <cradek> not sure I understand what you are asking. if you have load stuck true, I agree it's no use
[15:57:26] <cradek> hmm is encoder.0 a software encoder component?
[15:57:49] <cradek> if so, it smells like it sometimes gives you nan velocity for some reason
[15:58:30] <gene__> I just ran the spindle up to about 19 rps, and I see a very small reading on the tach, roughly corresponds to the rps from the encoder which is showing 19.5 approx
[15:59:29] <gene__> And I must have a heck of a lot of Igain "windup"
[16:02:08] <gene__> So, I need to switch to another instance of limit2 for filtering, and add yet another scale for some gain. And nuke 99% of the pid.0.Igain, back when I've done that
[16:21:08] <DJ9DJ> gn8
[16:52:02] <gene__> where is the best place to put a little plus offset when the creep speeds are slow in a pwmgen driven spindle?
[16:52:29] <cradek> pwmgen has an offset input
[16:55:18] <gene__> which I just raised to .039 & its about right, but that took my 600 rpms setting up to 700 actual
[16:57:55] <gene__> and by the time I zeroed the roor with FF0, its up to 800 revs. Turn up a feedback scale I gues
[16:59:28] * frysteev puts the microphone to the speaker to get more feedbacks
[17:26:27] <gene__> Humm, an offset in pwmgen of .11 is huge, or nominal?
[17:28:10] <gene__> At any rate, its close at 60 rpm, 450 rpm and 900 rpm from an S command, I like
[17:34:09] <gene__> in fact, I think I'll quit while I'm ahead
[17:34:45] <gene__> gnight all
[17:50:40] <JT-Shop> rats I should have stuck to the Intellitouch version of ELO... at least that works
[18:12:12] <pcw_home> do you have libusb 1.0? that seems to be needed
[18:25:33] <jthornton> pcw_home, yes I have that and even ran the apt get again to be sure
[18:55:11] <kwallace1> Hello. I want to change the normal X direction on my lathe. I reversed the encoder scale and the axis moves the way I want. Gremlin also shows movement to the North as positive, so far so good. My problem is that the soft limits are reversed so I run into the South limit when moving North. Anyone have a hint on how to fix this? Thanks and Merry Christmas.
[18:55:59] <jthornton> change your sign on the limits
[18:56:12] <jthornton> are you using gscreen? just wondering
[18:56:39] <jthornton> my lathe faces East and West so I'm a bit confused :)
[18:57:34] <tjb1> I dont have a lathe :(
[18:58:18] <kwallace1> I tried reversing the sign of the limits but that didn't work either. I'm using a custom UI with Gremlin.
[19:00:28] <jthornton> you have individual inputs for each end of the axis?
[19:02:30] <kwallace1> I think so. I have -.2 for South, 4.6 for North.
[19:03:36] <jthornton> yikes gotta run
[19:04:14] <kwallace1> Thanks.
[19:04:49] <kwallace1> Have a good evening.
[19:05:02] <kwallace1> Oops. I gotta go too.
[19:53:38] <Jymmm> !!! Merry Christmas !!!
[19:56:23] <tjb1> You too Jymmmy boy.
[20:05:23] <Jymmm> heh
[23:04:06] <Jymmm> http://ak.imgag.com/imgag/product/preview/flash/bws8Shell_fps24.swf?ihost=http://ak.imgag.com/imgag&brandldrPath=/product/full/el/&cardNum=/product/full/ap/3166187/graphic1
[23:21:09] <Nick001> Nice Merry Christmas