#linuxcnc-devel | Logs for 2015-03-03

[00:45:04] <adam3999_> welp, just spent several more hours troubleshooting my asrock 1900m parallel port input issue.. no joy yet
[08:58:10] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 955a94b 06linuxcnc 10(8 files in 4 dirs) tests: removed bad characters from TP auto-test programs * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=955a94b
[08:58:10] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 069f228 06linuxcnc 10(32 files in 5 dirs) tests: reorganized TP test folder to be able to test-all script * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=069f228
[08:58:10] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 2c3423a 06linuxcnc 10tests/trajectory-planner/circular-arcs/nc_files/arc-intersections/arc-arc-planeswitch.ngc 10tests/trajectory-planner/circular-arcs/nc_files/arc-intersections/arc-arc-sawtooth.ngc tests: Added missing M2 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2c3423a
[08:58:12] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 26828ef 06linuxcnc 10tests/trajectory-planner/circular-arcs/build-debug.sh tests: tweaked build script for TP debug output in TP tests * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=26828ef
[08:58:17] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 6580308 06linuxcnc 10(6 files) tests: fix for missing hal files due to recent refactor * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6580308
[08:58:20] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 95e21db 06linuxcnc 10src/emc/tp/tp.c tp: Improved handling of low-queue state * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=95e21db
[08:58:24] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 14e9153 06linuxcnc 10src/emc/tp/blendmath.c 10src/emc/tp/tc_types.h 10src/emc/tp/tp.c tp: polish and tweaks for low queue fix * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=14e9153
[08:58:28] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 575f416 06linuxcnc 10src/emc/tp/blendmath.c tp: fixed spindle atspeed overrun due to prev line consumption * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=575f416
[08:58:32] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 0802ed7 06linuxcnc 10src/emc/motion/command.c motion: catch non-fatal error during new segment and ensure that atspeed is not ignored * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0802ed7
[08:58:36] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 969e866 06linuxcnc 10src/emc/tp/tp.c tp: cleanup of blend velocity computation to eliminate unneeded planning mode * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=969e866
[08:58:40] <KGB-linuxcnc> 03Robert W. Ellenberg 052.7 9db0cad 06linuxcnc 10src/emc/tp/tp.c tp: fix for "climbing" behavior in rigid tapping * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9db0cad
[09:27:02] <skunkworks> zlog,
[09:27:04] <cradek> wow, he must really be testing
[09:27:38] <skunkworks> cradek, with pathpilot releasing - It is crunch time I bet.
[09:28:09] <skunkworks> however they are going to update.
[09:28:24] <cradek> I very much hope they have products using css, fpr, tapping, etc :-)
[09:28:44] <skunkworks> well - the lathe videos shows css
[09:29:40] <cradek> yay!
[09:29:40] <skunkworks> and tapping
[09:29:48] <skunkworks> *rigid tapping
[09:30:33] <cradek> I think I was early enough to successfully squash an idiotic thread about licenses on the forum
[09:30:45] <skunkworks> great!
[09:33:19] <skunkworks> Honestly - I can't believe how lucky they (and we) are to find rob.
[09:47:24] <skunkworks> The thread on cnczone is mostly positive..
[09:47:49] <skunkworks> 'ran a job all day without crashing...' poor poor people..
[09:51:22] <skunkworks> people really like the soft limits..
[09:53:38] <kwallace> The 770 I'm using doesn't have the switches adjusted properly, so I hit the Y switch before the soft limit. It's a pain the recover.
[09:54:06] <skunkworks> seems like something you should fix ;)
[09:54:33] <skunkworks> that was mentioned on the forum.. (how to recover if the machine is already on the limit)
[09:55:47] <cradek> mach doesn't keep you off the limit switches?
[09:56:02] <kwallace> I wanted to use the factory procedure so my machine would match what is shipped, but nothing is published.
[09:56:24] <skunkworks> cradek, sounds like it wasn't implimented very well.
[09:56:40] <cradek> heh, it's surprising what people are surprised by
[09:58:32] <skunkworks> cradek, http://www.cnczone.com/forums/tormach-personal-cnc-mill/261218-software.html
[10:01:32] <cradek> message #8 and someone is already bitten by "we picked one screen resolution to support, and used images for everything"
[10:03:41] <cradek> I wonder if their gui doesn't have the override-limits
[10:04:02] <cradek> > Seems like you can't manually run the axes into the limit switches as you could before
[10:04:05] <cradek> haha
[10:06:33] <skunkworks> I can't imagine not having that feature..
[10:06:42] <skunkworks> is that right?
[10:07:12] <kwallace> The thought was to match the screen available with a system, and it saves a ton of work and testing. Maybe UI scaling will come later, but the UI would need a major overhaul.
[10:08:47] <skunkworks> kwallace, do you think they will add an 'align vise' probing routine? Somone asked about it. (seems like a lazy way to machine...)
[10:09:24] <cradek> I can sure see probing for the corner of the jaw
[10:09:34] <cradek> for rotation, not so much
[10:10:05] <skunkworks> sounds like it would be - slap vise on table - probe and rotate coordinate system.
[10:10:44] <skunkworks> I though coordinate rotation in mach was pretty sketchy too
[10:10:44] <cradek> sure until you use a jog to cut something
[10:10:50] <skunkworks> heh
[10:11:27] <kwallace> In the Setting screen one can chose to use the limits or not, but there is no temp override as there is in AXIS.
[10:12:29] <cradek> on these machines can you easily get to the screws and just turn them with your hand?
[10:13:41] <cradek> my hnc has allen sockets on the ends of the ballscrews for this purpose
[10:13:50] <cradek> you don't need override limits, then
[10:14:08] <cradek> although of course it's still nice to have...
[10:14:27] <kwallace> No, it's not easy. My HNC has a hex socket on the end of the screws which is handy.
[10:14:38] <cradek> haha we had the same thoughts
[10:14:54] <cradek> you guys'll have to add the override-limit checkbox then
[10:16:14] <skunkworks> That is how the K&T is - hex head sockets to engauge the leadscrews
[10:16:35] <skunkworks> (although the conversion removed the y axis one...)
[10:18:12] <pcw_home> That reminds me, someone posted a link to a rotary broach, I had not seem one before, thought it was pretty clever
[10:18:13] <pcw_home> (shows how up to date I am, it was invented in 1914)
[10:18:47] <skunkworks> the wobbley kind?
[10:18:53] <pcw_home> yeah
[10:19:21] <cradek> I read that as "shows how up to date I am, I was invented in 1914"
[10:19:46] <pcw_home> not quite that old
[10:20:09] <kwallace2> Damn ISP.
[10:21:49] <kwallace2> There is a push to remove buttons or DROs if possible, let alone adding any.
[10:23:05] <pcw_home> rescue tab?
[10:23:30] <kwallace2> ?
[10:24:09] <pcw_home> fixit screen?
[10:24:38] <pcw_home> maintenance mode?
[10:25:22] <seb_kuzminsky> when you switch to that screen you should be greeted by this: http://reactiongifs.me/wp-content/uploads/2013/09/black-dynamite-you-done-fucked-up-now.gif
[10:27:24] <cradek> no, http://i1.kym-cdn.com/photos/images/newsfeed/000/234/765/b7e.jpg
[10:27:49] <kwallace2> When I crash PPi a window pops up and offers to connect you to customer support, and another to upload a log, or something similar. I haven't played with it.
[10:29:24] <pcw_home> seb_kuzminsky: lol!
[10:35:05] <skunkworks> kwallace2, really - it should yell at you and give you a non descriptive error. (you know - how everyone thinks linux is)
[10:35:50] <kwallace2> http://wallacecompany.com/tmp/Screenshot-error_1.png http://wallacecompany.com/tmp/Screenshot-error_2.png http://wallacecompany.com/tmp/Screenshot-error_3.png
[10:36:59] <kwallace2> Selecting Remote Support doesn't seem to do anything, yet.
[10:37:12] <cradek> what's an update package?
[10:38:14] <kwallace2> Dunno. Maybe a download or USB stick.
[11:09:22] <pcw_home> I'm a bit surprised they didn't use a USB stick for install
[11:25:45] <kwallace2> My guess is that DVD's are more common than USB sticks with existing customers.
[11:31:16] <pcw_home> Yeah but an order of magnitude less reliable
[11:35:45] <seb_kuzminsky> i'm surprised they didn't use the existing debian packaging, and the -dev package to build their gui against
[13:38:05] <cradek> > One good thing is that cutter comp works within subroutines now. A major plus.
[13:40:11] <cradek> > no hiccup in the contour that I saw, Mach would stop and start at certain points during the long contour in a random fashion, each time it makes a small hickey in the finish
[13:40:41] <cradek> seems like a lot of good news
[13:53:13] <skunkworks> cradek, yes
[13:53:58] <skunkworks> The consensus so far seems to be - hey - not so bad..
[13:56:21] <skunkworks> wow - 6 laptops that won't run a newish linux.. (video issues - intel i915, Trident Microsystems XGI Volari XP5)
[13:59:38] <cradek> skunkworks: thanks for pointing people at the docs so I didn't have to
[14:00:52] <cradek> > Warning: If the X Y coordinates specified are not the current coordinates when calling G33.1 for tapping the move will not be along the Z axis but will rapid move from the current location to the X Y location specified.
[14:00:57] <cradek> this is totally wrong, isn't it?
[14:02:40] <PCW> must resist asking forum poster what specific part of the GPL is Tormach violating
[14:03:31] <cradek> well I see "vmax" didn't take my advice
[14:04:37] <cradek> PCW: I think that would be a very useful question, if you're bold enough to step in
[14:05:09] <PCW> here goes....
[14:05:34] <cradek> he calls a link directly to the faq entry answering his exact question "not having a clue either"
[14:05:44] <cradek> I'm really sad for him
[14:25:08] <cradek> hm, when testing that doc bug, I find tapping is not quite right in 2.7
[14:26:01] <seb_kuzminsky> uh-oh, what's wrong with it?
[14:26:11] <cradek> I'm making a bug
[14:26:26] <seb_kuzminsky> 
[14:27:08] <seb_kuzminsky> err
[14:37:35] <cradek> https://sourceforge.net/p/emc/bugs/420/
[14:38:24] <seb_kuzminsky> oh man
[14:39:47] <cradek> I broke several air-taps. fortunately they're very inexpensive.
[14:49:54] <skunkworks> I have broke a lot of mills lately...
[14:50:01] <skunkworks> air mills
[14:50:38] <skunkworks> same breakage in robs branch (just checking)
[14:50:57] <skunkworks> has anyone emailed rob?
[14:51:09] <cradek> surely he sees our bug reports
[14:51:10] <skunkworks> (poor guy - he touched it last..)
[14:51:12] <skunkworks> ok
[14:51:38] <seb_kuzminsky> this is what happens when we dont have automated tests of motion :-/
[14:52:56] <skunkworks> damn it! damn it all to hell!!
[14:54:16] <cradek> jeez
[14:56:16] <skunkworks> sorry
[14:56:19] <skunkworks> :)
[14:56:38] <skunkworks> so is that a 1/2 inch of movmemnt in x and z
[14:56:56] <skunkworks> no
[14:56:58] <skunkworks> wait
[14:57:33] <cradek> it's moving only 1" toward the endpoint, instead of moving to x1z1 (which is more than 1")
[14:57:47] <skunkworks> right
[15:01:04] <cradek> should probably test diagonal g33 too
[15:01:24] <skunkworks> it works
[15:01:27] <skunkworks> (jsut did it)
[15:01:34] <cradek> can you tell for sure that the pitch is right?
[15:02:16] <skunkworks> no..
[15:02:37] <PCW> tap wax?
[15:02:48] <skunkworks> about 15in/min - at 300rpm
[15:04:23] <skunkworks> that is 20 pitch - right?
[15:04:36] <skunkworks> yah - that seems right
[15:05:31] <cradek> yay thanks
[15:05:49] <cradek> I bet we'd see if it was off by sqrt2
[15:21:30] <skunkworks> heh - we have a laser at work here using gecko servo drives.. When they get hot - the are under danmpened. (change in direction shows a nice ring..)
[16:20:34] <Tom_itx> ok PCW any updates on the J1900 boards? or is there a better one out there to grab?
[16:21:01] <Tom_itx> i'm about to get something right soon and don't wanna get stuck with a non working parallel port
[16:21:33] <Tom_itx> i haven't heard anything else except adam3999's issues with it
[16:47:10] <Tom_itx> does that G3258 come in a small footprint board?
[16:49:40] <skunkworks> zlog:
[17:08:16] <kwallace_T770> Probing and vises was mentioned earlier. Was probing for adjusting the vise or adjusting the coordinate system to match the vise?
[17:13:29] <PCW> The ASROCK one I mentioned works fine
[17:14:02] <Tom_itx> i will probably go with that one
[17:14:41] <PCW> You can get a LGA1150 Mini-ITX MB for a G3258 but its more $ of course and needs a fan
[17:14:56] <Tom_itx> yeah i like the idea of fanless too
[18:19:18] <KGB-linuxcnc> 03Dewey Garrett 052.7 5d168a0 06linuxcnc 10configs/sim/axis/xhc-hb04/README xhc-hb04 sim README: improve instructions * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5d168a0
[21:29:29] <skunksleep> I don't think I got my point across with Jon
[21:52:55] <cradek> nope
[21:52:59] <cradek> and he didn't read the docs
[21:54:53] <cradek> > Please quote the section of the GPL that states this
[21:54:59] <cradek> heh
[21:56:38] <skunksleep> Saw that
[21:56:42] <cradek> the GPL is made up of words that have meanings, not feelings
[21:57:17] <Tom_itx> somebody get their toes stepped on?
[21:58:12] <cradek> nah, there's just a thread with more noise than signal
[21:58:34] <cradek> ob. https://xkcd.com/386/
[22:00:18] <Tom_itx> PCW your generosity of 'samples' is apprecitated but it cost me a asrock board case & memory to test it...
[22:00:40] <seb_kuzminsky> the motion described by the g33.1 docs sounds really safe
[22:01:28] <skunksleep> seb_kuzminsky: yes. Never had an issue
[22:03:25] <adam3999_> hey tom did you end up getting an asrock j1900 motherboard?
[22:03:30] <Tom_itx> yup
[22:03:51] <adam3999_> which model?
[22:04:37] <Tom_L> http://www.newegg.com/Product/Product.aspx?Item=N82E16813157565
[22:04:48] <Tom_itx> ^^
[22:05:50] <adam3999_> gotcha, i have the Q1900M which is damn close except mine doesn't have the lpt port in the atx back panel
[22:06:06] <adam3999_> PCW tested a few J1800/1900 motherboards yesterday and confirmed that his lpt inputs are fine
[22:06:15] <adam3999_> i'm still troubleshooting why mine don't work
[22:06:35] <Tom_itx> yeah i noticed, that's why i went ahead and pulled the trigger
[22:06:55] <Tom_itx> either way, i can use it
[22:07:25] <adam3999_> are you going with a mesa card or similar?
[22:07:46] <Tom_itx> i use mesa
[22:07:51] <Tom_itx> already
[22:07:54] <adam3999_> gotcha
[22:08:17] <adam3999_> if you have a few min, i'd be interested to see if your motherboard lpt inputs work as expected
[22:08:19] <Tom_itx> the 7i90 is in the control right now
[22:08:29] <Tom_itx> i've got a 7i24 to test in this new box
[22:08:53] <adam3999_> i can get mine to work with a separate parallel port testing app but not with the hal-driven ptest application
[22:08:56] <Tom_itx> 7i90 is parport
[22:09:13] <Tom_itx> i've never used that
[22:09:32] <adam3999_> just a simple tool for parport troubleshooting
[22:09:33] <adam3999_> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Parallel_Port_Tester
[22:09:51] <adam3999_> i used this when i started troubleshooting to try and isolate where my issue was
[22:10:07] <adam3999_> if i run that tool and short any of the input pins (such as pin 10) to ground, no joy
[22:10:22] <adam3999_> but its fine with a separate utility that doesn't rely on hal
[22:11:00] <Tom_L> http://tom-itx.ddns01.com:81/~webpage/temp/parport/parport.mp4
[22:11:08] <Tom_L> i had a c program that did that to test io
[22:11:21] <Tom_L> i'm not sure where it is now...
[22:12:12] <adam3999_> yeah i saw a few programs on the web for running that type of test
[22:12:17] <adam3999_> my outputs are fine though
[22:12:22] <adam3999_> i've been milling already
[22:12:25] <Tom_L> that doesn't test input though
[22:12:26] <adam3999_> no inputs, unfortunately
[22:12:44] <adam3999_> getting fed up with manually homing :)
[22:13:13] <Tom_L> send it home before you shut down
[22:14:27] <adam3999_> just got the mill converted so i'm playing around with things still
[22:15:25] <Tom_L> does that parport tester just read the registers?
[22:16:40] <adam3999_> not sure
[22:16:58] <adam3999_> i can measure 3.3V on the output pins when i enable them
[22:16:59] <Tom_L> or do you need buttons attached to the cable
[22:17:13] <adam3999_> i used a multimeter on the outputs like pins 2,3
[22:17:33] <adam3999_> and then shorted pin 10 and 15 to ground as a test for the input, as well as using the e-stop and limit switches through my MX3660
[22:17:44] <Tom_L> do you need pull up/down on the inputs?
[22:18:02] <adam3999_> i don't believe so, but not 100%
[22:18:10] <adam3999_> shorting the pin straight to ground with that app worked on my laptop
[22:18:22] <Tom_L> you might check on that
[22:18:36] <adam3999_> i can also see the input pins change state when i use my limit switches and e-stop via the MX3660 and the other parallel port app
[22:18:50] <adam3999_> http://yyao.ca/projects/ParallelPortLinux/
[22:19:08] <adam3999_> that reports pins 10 and 15 correctly when i engage them via the MX3660 or just shorting them
[22:19:23] <adam3999_> but ptest that runs through HAL and the realtime stuff doesn't
[22:19:43] <adam3999_> obviously neither does linuxcnc/halscope
[22:21:06] <Tom_L> acording to that link 10 & 15 are active low inputs
[22:21:20] <Tom_L> 11 12 13 aren't
[22:21:49] <adam3999_> makes sense, i see green lights when i run the ptest app, and then they go red when you short them
[22:22:11] <adam3999_> on my laptop... when i run ptest on my Q1900, all red
[22:22:43] <adam3999_> can't get them to change states... tried the original BIOS, new 1.50 BIOS, normal/epp 1.7/epp 1.7/bidirectional parallel port setting
[22:23:08] <adam3999_> live iso linuxcnc 2.6, 2.8-pre from the dev repo
[22:23:28] <adam3999_> default 0x378 addr, bios hard coding the lpt port to 0x278
[22:23:38] <adam3999_> no joy thus far :)
[22:24:25] <adam3999_> sorry that previous line should say that i tried the parport in normal, epp 1.7, epp 1.9 and bidirectional modes
[22:25:20] <Tom_L> what's the difference between 1.7 & 1.9?
[22:25:32] <Tom_L> besides .2 :D
[22:25:40] <adam3999_> not sure, different spec for the parallel port mode
[22:25:54] <adam3999_> it was one of four options i cycled through to see if any of them would behave differently
[22:25:57] <Tom_L> pretty sure you want epp mode
[22:26:02] <adam3999_> yep
[22:26:15] <adam3999_> i started with epp 1.9 and all the power saving modes disabled
[22:26:23] <adam3999_> the asrock bios is pretty nice
[22:26:28] <adam3999_> latency is very low
[22:26:36] <adam3999_> was not expecting to have this input issue heh...
[22:27:10] <adam3999_> micges found a thread on another forum that references the same issue, but no resolution
[22:27:18] <adam3999_> and PCW was kind enough to try some of his boards, but no resolution yet
[22:27:37] <Tom_L> i can try it when the junk gets here
[22:27:44] <adam3999_> thanks, appreciate that
[22:28:01] <Tom_L> you might have to remind me
[22:28:03] <adam3999_> i need to post on the forum to capture my environment and what i've tested so far, perhaps reach a larger audience
[22:28:18] <adam3999_> no prob, i imagine i'll be hanging out here until i get it fixed
[22:28:29] <adam3999_> when folks have enough of my yaking they'll dive in to fix it :)