00:54:27 danielbr has joined #emc 00:56:56 danielbr has quit 02:11:42 ffff 02:14:15 Jymmm has joined #emc 02:14:19 Jymmm has left #emc 02:14:34 Jymmm has joined #emc 02:17:23 jmkasunich_ is now known as jmkasunich 02:21:54 tomp2 has joined #emc 02:24:13 petev has joined #emc 02:24:39 Arrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrggggggggggghhhhhhhhhhhhhhhhhhhh! 02:24:48 Damn, thought I'd feel better after that. 02:25:13 what are your guys thoughts on how a jog button should work in incremental mode? 02:25:26 should you have to hold the button pressed until the move finishes? 02:25:35 no 02:25:40 or should the move finish even if you let it go? 02:25:44 but, not a bad idea. 02:25:49 safety wise 02:25:51 you don't think it's a safety issue? 02:25:57 user option maybe? 02:26:09 if the inc is .01" not so much 02:26:23 I was thining maybe the increment was too large, and you're thinking oh shit, I'm going to hit the fixture or something like that 02:26:32 hmm, loosing .01" of your finger is .01" too much 02:27:05 petev: will it repeat if I hold the button down? 02:27:08 no 02:27:29 so if I want to move .03" I cna't hit the button three times huh? 02:27:48 yes, you press it 3 times, but holding it down would not auto-repeat 02:28:11 No, I dind't mean press and hold, press and hold, press and hold. 02:28:23 I meant press press press HOLD on the final 02:28:45 petev: I believe all the current user interfaces allow you to release the button during the remainder of an incremental jog, no matter how much is left 02:29:03 oh, I guess that could be done, but it would require some type of timer and could not stop the move immediately when you released the button 02:29:32 jepler, it does, but I was thinking that might not be the safest thing 02:29:45 petev: Well, I like both options actually. If it was a LARGE move, then press and hold makes sense, if I'm trying to nudge into place a press and release works too. 02:30:03 yeah, I hear what your saying 02:30:29 what about if it's a LARGE move the press and hald works, if it's a small move than tap tap tap works 02:30:31 I think it's up to each user interface to decide whether to offer one behavior, the other, or make it a preference 02:31:04 in practice I only use the very small increments (e.g., 0.01inch and smaller) so I would not have time to exercise my "oh shit" reflex anyway 02:31:05 Jymmm, that could be done, or it could be a preference from the INI file 02:31:10 petev: I'm asuming here that the move is not at 4000 IPS either =) 02:31:14 a GUI that offers a 1 inch incremental jog is probably not giving any value to its user 02:31:44 maybe 1" or large jogs are best left to MDI mode, but they do have value 02:33:09 petev: Well 1" tap tap tap would be three inches if ever needed. that's kinda how my cruise control works, one press for every inc/dec MPH 02:36:08 petev: will you tell me about a situation where you find long (e.g., 1") incremental jogs useful? 02:36:58 jepler: I've had that 02:37:00 if you don't want to loose position with continuous mode when needing to move a certain relative distance 02:37:14 but mdi mode is ok for this too, just a little more effort 02:38:14 jepler: when I've wanted to run a proggy real quick and run it on multiple moutned workpiece 2" over. 02:38:36 I'd be mad if I moved less than my incr distance and lost position, just because I didn't hold the key quite long enough 02:39:06 cradek: Hmmm, didn't think of that. I'd be too. 02:39:07 cradek, I would be more mad though if I broke my carbide endmill on a clamp 02:39:21 I do sympathize with the "oh shit that increment was larger than I wanted" though 02:39:26 I'm not sure what's best 02:39:32 same here 02:39:34 petev: user option? 02:39:38 well a jogwheel is best 02:39:41 true 02:39:46 a wheel is nice 02:39:59 NUMPAD FTW!!! LOL 02:40:01 right now, I think you can hit escape and stop an incr jog (if you have time) 02:40:11 but you won't have time... 02:40:43 oh that's right, emc doens't have to have jog "enabled" 02:41:58 under TCNC you had to be on the correct screen to jog. 02:42:33 then cradek's cat can't run across the kybd and crash the tool into the workpiece =) 02:43:03 that's why there's a door on this room 02:43:17 or cradek can't run across the kybd and crash the tool into the workpiece =) 02:43:18 bbl 02:48:48 petev: yeah I see your position but I imagine myself using mdi 02:49:05 (I'd set incremental distance mode, mdi over the desired amount, and then forget to set absolute distance mode .. my "oh shit" would come at a later moment) 02:49:27 I've definitely incremental-jogged into the table/work before 02:49:40 now, I'm very careful to always check the increments 02:50:01 what if the screen changed colors depending on what mode 02:50:13 cradek: now that axis lets you choose the units available, have you? 02:50:29 I've never changed them 02:50:40 bah 02:50:44 you're no power user 02:50:49 lol 02:50:56 no, I never hooked them to the jogwheel either 02:51:32 btw I think I want to let external buttons set axis's idea of the "current axis" for jogging 02:51:49 why? 02:52:08 because a minute ago I wanted the "click in" on my jogwheel to let me swap between jogging X and Z 02:52:27 I haven't figured out all the details -- certainly the external button has to be something momentary.. 02:52:35 interesting idea 02:52:59 X---Y---Z---Back to the beginning 02:53:02 maybe jogwheel + vcp to show the jogwheel's axis is better 02:53:36 That's how my radio does it to switch baack and fourth between bands. 02:53:50 the existing jogwheel logic supports external button axis select 02:54:11 except it does a high or low beep too so you dont have to look at the dispaly and start transmitting on the wrong thing. 02:54:11 just turn on the appropriate enable bit 02:54:11 jmkasunich: yes indeed 02:54:27 but you are talking about changing axis's axis ;-) 02:54:59 set incremental to 0.001, click click click X, the change to Y, click, click click some more? 02:55:28 CLICK_HOLD and TURN to change increments 02:55:36 eww 02:55:37 I guess I'm not sure whether this would be useful or gimmicky 02:55:44 increments should NOT change easily 02:55:52 don't think it can't be both 02:55:54 my brainstorm was certainly "I imagine this gimmick" not "this would actually be useful" 02:55:53 jmkasunich: that's why it's CLICK AND HOLD 02:56:42 if you have a physical button next to your physical jogwheel, might as well just use that button to change the increment 02:57:14 I've found that I don't need to change the increments 02:57:29 maybe if I often cared about < .001, I would 02:57:30 of you are talking about a USB funny thing that is sort-of a jogwheel, and has 8 buttons around it, ditto - use one of them to set 0.001, one to set 0.01, one to set X active, one for Y, etc 02:57:38 but even .001/click, I can jog it at a good clip pretty easily 02:57:47 jmkasunich: Ew 02:57:59 right - thats the whold point of a jogwheel - wheels have enormous dynamic range 02:58:15 Jymmm: why Ew? 02:58:44 its a machine, not a PC, doing everything with one button is a royal pain in the rear 02:59:06 night all 02:59:11 night jeff 02:59:16 bye 02:59:25 jmkasunich: No no, not a mac mouse (so to speak) but the one button per axis thing. 03:00:17 Eh, it's what folks are used to is all. 03:00:20 yeah 03:00:31 re 1" jog increment: today installed machine and had a guy laser it. He used 1" jogs and a renishaw laser to record leadscrew errs in X Y & Z. Then we mounted a 3R (tm) Refix plate and measured it's 100mm ctr to ctr holes using 100mm jogs. So large jogs are useful, tho rare. 03:00:33 real machines have one button per axis with a light in the active one 03:00:47 Butm the ability to let a user define what they want would be pretty cool. 03:01:41 g0g91x1 03:01:40 x1 03:01:42 x1 03:02:55 jmkasunich: hey, did you have a url for the usb thingy you were talking about. or do you mean that fancy doohicky usually used for AV stuff? 03:03:54 jepler is the one who mentioned using a usb thingy, dunno if he has the AV thing or something else 03:04:32 I was referring to whatever he has, but thinking of the AV thing 03:04:41 Ah, no biggy. I'm still planning on using my numpad for a while still. 03:04:55 i'm using a wireless joypad 03:13:23 tomp2 has left #emc 03:33:05 much closer to being done with my machine! 03:33:25 http://images.myonlinesite.com/cnc/index.php 03:33:36 see very bottom link 03:34:08 actually here: http://images.myonlinesite.com/cnc/construct/20070606/3axis.jpg 03:34:11 neat 03:34:16 what will the spindle be? 03:34:45 hehe... not completely sure yet 03:34:54 I have my eye on a few trim routers 03:35:06 but I also have a machine spindle at the shop I could try 03:36:01 any... off to sleep I go 03:36:05 Ziegler has quit 03:51:59 petev has quit 04:43:32 ports_ has joined #emc 04:43:42 hey 05:44:51 howdy 05:53:36 augnob has joined #emc 06:33:45 ports_ has quit 07:00:12 Twingy has quit 09:09:23 Bo^Dick has joined #emc 09:49:54 acemi has joined #emc 09:52:32 Jymmm has left #emc 10:50:32 DusanC has joined #emc 10:53:49 DusanC is now known as dusanc 10:55:03 The_Ball_ has joined #emc 10:57:46 dusanc_ has joined #emc 11:02:59 dusanc_ has left #emc 11:16:48 dusanc2 has joined #emc 11:17:23 hello 11:18:04 dusanc has quit 11:20:39 DusanC has joined #emc 11:23:20 DusanC_ has joined #emc 11:25:15 Hi. DOes anybody know how could I setup adaptive control with EMC@? 11:26:15 Something like arc sensing, and using it to control arc height for plasma cutter. 11:27:35 http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Torch_Height_Control 11:27:47 I think this information may be a little out of date 11:28:08 Dallur, who is often on irc, has made an advanced torch control setup 11:28:26 I think you can find it in the sample configurations: dallur-thc 11:29:41 Looking at it now. 11:30:24 It works like setting Z height and then controlling the height with ladder? Right? 11:31:03 sorry I haven't studied it to know the details 11:31:39 Ok. 11:33:16 I have to get ready for work, stay around and maybe you can talk to him later 11:33:52 DusanC__ has joined #emc 11:36:34 I have a college assigment to design a 3axis plasma table and 5 axis welding robot, and I plan to use EMC2 11:37:24 Now do you think that I could make seam finding with trough the arc sensing easily with emc2? 11:38:11 i know how to do it with ladder, but the robot has to move and not to have any fixed axis 11:40:05 dusanc2 has quit 11:40:20 DusanC__ is now known as DusanC2 11:41:32 Sorry I have bad connection. I'll try later. 11:42:11 DusanC2 has left #emc 11:42:41 DusanC has quit 11:54:56 DusanC_ has quit 12:13:10 03jepler 07TRUNK * 10emc2/scripts/emc-environment.in: LD_LIBRARY_PATH is necessary on my amd64 machine, otherwise python's emcmodule doesn't find libemchal.so 12:17:59 Guest900 has joined #emc 12:18:11 03jepler 07TRUNK * 10emc2/scripts/runtests: at exit, print the name of any test that failed 12:18:22 Guest900 is now known as skunkworks_ 12:56:42 Unit41 has joined #emc 13:54:59 Bo^Dick has quit 14:00:54 xemet has joined #emc 14:01:09 Hi 14:01:16 Alex are you there? 14:25:26 03jepler 07TRUNK * 10emc2/src/hal/hal_lib.c: on 64-bit machines, give 8-byte alignment for stuff in hal shared memory when appropriate 14:26:45 03jepler 07TRUNK * 10emc2/src/libnml/buffer/tcpmem.cc: fix warnings on 64-bit systems 14:26:45 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/usrmot.c: fix warnings on 64-bit systems 14:26:46 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: fix warnings on 64-bit systems 14:26:47 03jepler 07TRUNK * 10emc2/src/hal/classicladder/ (arrays.c classicladder_gtk.c symbols_gtk.c): fix warnings on 64-bit systems 14:26:48 03jepler 07TRUNK * 10emc2/src/hal/drivers/ (hal_motenc.c hal_vti.c): fix warnings on 64-bit systems 14:26:49 03jepler 07TRUNK * 10emc2/src/hal/user_comps/devices/hal_joystick.c: fix warnings on 64-bit systems 14:26:51 03jepler 07TRUNK * 10emc2/src/hal/utils/ (halrmt.c scope_horiz.c scope_vert.c): fix warnings on 64-bit systems 14:28:38 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: oops, did not mean to commit this -- revert to rev 1.100 14:42:36 xemet has quit 14:51:15 http://www.cnczone.com/forums/showthread.php?p=306129#post306129 14:52:41 The_Ball_ has quit 15:04:10 Bo^Dick has joined #emc 15:08:25 DusanC2 has joined #emc 15:10:01 Hi. I have some questions about EMC2 and adaptive control. Can anyone help? 15:10:20 ask away 15:11:29 A problem for my master work: 5 axis weld robot 15:11:39 3 axis gantry + 2 rot 15:11:59 I need to make through the arc seam tracking 15:12:49 Measuring arc current ant based on that deciding how to change path 15:13:11 s/ant/and 15:14:05 i looked at ladder but I don't know how to make corrections while the robot travels down the programmed path 15:14:46 Ofcourse I choose EMC2. Any ideas? 15:15:45 path corrections would be difficult, I think 15:15:48 offtopic=> I may suddenly go offline because of lousy connection 15:15:53 you can certainly change speed easily 15:17:33 The idea is that robot should follow not so quality manufactured seam 15:18:40 with HAL, you can add in an offset to the comanded position from the motion controller (in fact, I think the motion controller may have an offset input now) 15:19:03 SWPadnos: no, I don't think it does 15:19:15 ok. I thought that was added. I guess it was just discussed 15:19:31 there is a component called "offset" which may be useful for applying an offset that emc doesn't see 15:19:35 DusanC2_ has joined #emc 15:19:35 so if you have the hardware to measure arc current, and can write a HAL component to calculate the correct offsets, they can be added in HAL 15:19:37 ok 15:19:42 It should have some weave pattern he should follow and should modify the amplitude of the weave based on the avg. current. 15:20:10 HW to measure should go through motenc's ADC 15:21:27 Missed that offset talk because of reconnection 15:21:33 10:19:29 there is a component called "offset" which may be useful for applying an offset that emc doesn't see 15:22:03 But can it work in 3D RT? 15:22:54 i mean translating through space simultanously? 15:23:06 or is it per axis? 15:23:13 you use one per axis 15:23:18 one per joint 15:23:24 right - per joint :) 15:23:31 joint. Correct 15:23:45 unfortunately the outputs are all in joint space, so an operation like "move one inch in X" is tough to do 15:24:01 It's ok. 15:24:29 It would have to move in space. So some vector would have to be the input 15:25:41 And could the weave pattern be added as some subfunction of G-code? Or it should be postprocessors work? 15:28:45 DusanC2__ has joined #emc 15:39:49 Bo^Dick has quit 15:41:14 could you point me to the reference of that joint "offset" in html documentation? 15:41:33 http://linuxcnc.org/docs/2.1/html/man/man9/offset.9.html 15:41:34 Currently reading it but just can't find it 15:42:27 that component continually computes 'out' and 'fb_out': out = in + offset; fb_out = fb_in - offset 15:43:04 ahh. I looked at the higher level :) 15:43:17 so you can place it between emc's motor position commands (e.g., axis.0.motor-pos-cmd) and the pid that controls your servo; and the feedback goes between the encoder that gets feedback from your servos, and axis.0.motor-pos-fb 15:43:33 But that would change the speed, not position, correct? 15:43:46 "offset" comes from the component you must write, which computes the desired weave shape 15:44:11 Ok. Got it now. 15:44:52 for instance, if you hook offset to siggen.0.sine, you would get a 1 motor unit sine wave applied to the position that emc finds from the gcode 15:45:08 DusanC2 has quit 15:46:09 lost connection 15:47:08 The problem is that it has to move like sine but in space and not per joint. 15:47:46 So i would have to have some thread that would recalculate the correct "offset" values for every joint? 15:48:02 Very fast thread :) 15:48:04 why can't this sine shape be in gcode? does it have to change dynamically? 15:48:06 DusanC2_ has quit 15:48:19 10:19:42 It should have some weave pattern he should follow and should modify the amplitude of the weave based on the avg. current. 15:48:24 Well no\ 15:48:34 oh ok 15:48:43 Jepler is correct 15:49:33 Well it should modify the average position and the amplitude :) 15:49:45 s/average/middle 16:11:37 DusanC2__ has quit 16:29:05 xemet has joined #emc 16:29:09 hi 16:46:32 lerneaen_hydra has joined #emc 16:52:23 xemet has quit 17:19:33 lewin1 has joined #emc 18:00:47 robin_sz has joined #emc 18:31:04 meep? 18:31:41 fff 18:31:54 * ds2 pours birdseed 18:51:03 lewin1 has quit 19:21:56 xemet has joined #emc 19:22:18 hi 19:22:41 is there anyone who has access (and can modify the content) to the linuxcnc.org website? 19:30:52 mschuhmacher has joined #emc 19:44:16 JymmmmEMC has joined #emc 20:17:44 wake up! 20:37:31 xemet, yes, yes there are 20:42:06 ok 20:42:16 robin, do you can? 20:49:31 * ds2 does an improper gear change on a bridgeport gear and and leaves it running ;) 20:52:21 hahaha. 20:52:30 i really would like a GOOD manual/cnc. 20:52:41 goodbye gear changes, hello "s1000 m3" 20:59:40 VFD 20:59:59 well i'd also like to do other neat CNC stuff with the one-off ease of a manual machine 21:00:38 i hear some of the cincinati hydrotel mills have a cnc tracer 21:00:48 which would probably be the ultimate in both cnc and in manual control 21:03:34 MDI mode =) 21:06:02 it's not quite the same 21:06:14 i'd like to have some feed buttons with macros on them 21:06:30 so i can reprogram the buttons to do things like feed at a certain rate to a predetermined stop 21:06:34 things like that 21:08:35 keyboard macro + MDI 21:09:15 lerneaen_hydra has quit 21:09:52 xemet has quit 21:12:53 this 21:12:56 http://sourceforge.net/projects/stellarium/ 21:13:04 controlls my telescope. pretty cool 21:25:10 ds2: i think if i hooked a couple extra buttons, like a mcdonalds-style cash register keyboard, plus some levers/knobs/dials 21:25:19 that i could improv a decent interface 21:26:28 also i did some investigation and found out why saws cost 30k 21:26:30 for an automatic 21:30:45 why? 21:30:53 gears made of solid gold? ;) 21:31:09 nah 21:31:20 the rigidity to handle 15 hp cuts 21:31:28 plus the bearings to saw all day long 21:31:37 then you get into features - controllable saw pressure and head approach 21:31:44 15HP cuts?! Hmmm 21:31:44 automatic part gauging, rapid stops 21:31:52 three way vices 21:31:55 for bundle cutting 21:32:02 infeed and outfeed tables 21:32:11 the columns that support the sawhead 21:33:31 skunkworks_ has quit 21:33:42 chip removal 21:33:48 coolant and cooling for the motor 21:33:57 plus all the different modes the saw can run in 21:34:40 acemi has quit 21:35:30 like, these saws will run 2 tpi blades 21:35:38 and can run them on thin walled parts 21:51:42 sounds like I can use it to make thin slices of an entire car? 21:51:54 if you had a 20"x20" car 21:52:01 the saw at work is like 10"x10" 21:52:06 it's a nice saw 21:52:16 a pleasure to use 21:52:34 the high power just means you can really put the pressure on 21:53:43 it goes through large rounds pretty quickly, a couple seconds where the saw i use in college takes many minutes 21:54:35 but you can only vary the head approach on that saw, you don't have pressure control or high horsepower 21:54:54 Ziegler has joined #emc 22:00:32 ah... sounds like itwould be great for the tougher stuff like Stainless 22:00:44 yes, it does cut stainless 22:01:07 mdynac has joined #emc 22:01:11 does cut and great are two different things... a cold chisel will cut stainless 22:02:11 well, when i say "it does cut it" 22:02:25 i'm coming at it from a production/manufacturing standpoint 22:02:31 so i do mean it does an amazing job 22:02:57 hmm that sounded more arrogant than i intended 22:03:00 i apologize 22:03:36 gotcha 22:03:41 but yeah 22:03:47 it goes through stainless/tool steel very nicely 22:04:01 not as fast as it will go through aluminum/copper 22:04:15 but definitely a product tool (vs a protoshop tool)? 22:04:35 it's a production tool 22:04:45 but it's just as easily used for single cuts 22:05:05 i would prefer using a production saw for single cuts versus a manual saw 22:05:23 because the production saw usually has great features, even for doing one or two parts 22:05:46 you can do a little skim cut off the face, then cut to +/- .005 22:05:51 sometimes less 22:05:54 but at 30K, it'd take a lot cuts per day to make the price per cut reasonably 22:05:57 yeah 22:06:09 it just takes some looking to find an inexpensive automatic saw 22:06:24 it may need minor work, but i have been able to find autos for around 4-5k 22:06:54 think there is also the problem of training people to use it 22:07:04 it's much simpler than a milling machine, but yes 22:07:23 seems that the saw at the school gets beaten to death cuz people donno how to use the hydraulic feed (they open it up 100%) 22:07:35 the saw is just a peice of crap at school 22:07:37 it's gravity fed 22:07:44 with a hydraulic damper 22:08:01 it's got no power, and is made of sheet metal instead of cast iron 22:08:04 ours can be slowed/adjusted with a hydraulic valve 22:08:18 yeah, that's what i'm calling gravity feed 22:08:24 oh 22:08:39 what do you call the import ones that just have return spring? 22:08:40 versus hydraulic saws which use hydraulic power to actualy feed the sawhead down 22:08:45 gravity feed 22:08:51 but it isn't adjustable. 22:08:57 hmmmm okay 22:09:01 adjustable vs. fixed gravity saw 22:09:09 there are some very nice gravity saws, though 22:09:46 my old school had a good gravity saw 22:10:06 easy to use, had a lot of weight in the head 22:10:25 was the speed adjustable and did people know how to adjust it? 22:10:40 yes 22:11:00 that was the only knob it had 22:11:22 was/is 22:11:57 the whole thing is thin sheet metal, no infeed/outfeed table 22:11:59 the vise is shitty 22:11:59 oh lucky... ours was a pulley/belt change. lots of students didn't bother 22:12:10 oh, speed of the BLADE 22:12:13 not sawhead 22:12:14 sorry 22:12:17 yeah, ours has a belt 22:12:27 it's set to a low speed so nobody burns the blade up 22:13:16 but since it doesn't have a seperate pressure adjustment 22:13:22 we can't use big blades 22:13:39 it's just an iffy unit in general. 22:13:55 i suppose i could add weights 22:13:56 but it beats hacksaws =) 22:14:15 aww hells yeah 22:14:26 unless it's an AUTOMATIC hacksaw! 22:15:31 automatic? 22:15:39 yeah, they make power hacksaws 22:15:41 or made 22:16:59 so it is a hacksaw with the handle attached to a motor? 22:17:24 yes 22:19:24 they're popular with the home shop crowd 22:19:34