#emc-devel | Logs for 2009-03-20

[00:12:24] <CIA-2> EMC: 03cmorley 07TRUNK * 10emc2/src/emc/usr_intf/stepconf/stepconf.py: Fix for 'all limit' signals
[01:29:41] <cradek> yay, halscope is fixed?
[01:34:09] <jmkasunich> I think so
[01:34:16] <cradek> wheee
[01:34:21] <cradek> then we can branch :-)
[01:51:55] <jepler> any week now
[02:05:07] <jepler> seriously, this cleared my personal backlog of "stuff that needs to be fixed before branch", so we can do it anytime unless somebody says otherwise
[02:05:22] <SWPadnos> hmmm.
[02:05:30] <SWPadnos> can I add the G540.stepconf file?
[02:05:47] <SWPadnos> oh. I guess I should see if it loads with 2.3 stepconf :)
[02:05:54] <jepler> that'd be a nice test
[02:06:00] <jepler> even better if you can test it working
[02:06:09] <SWPadnos> um. yeah, I can, but not tonight
[02:06:22] <SWPadnos> by (the end of) this weekend though
[02:07:44] <jepler> I was hoping to make beta2 before the end of the weekend, but that's life
[02:08:00] <jepler> you can put it on the branch, though, as long as you test the installation and all that
[02:08:21] <SWPadnos> it's just a stepconf file - sort of compact, and peiople can't run it unless they step through the config
[02:08:27] <SWPadnos> I figure that's safer (marginally)
[02:09:47] <jepler> stepconf doesn't hand-hold people choosing files from /usr/share/emc2, or whatever the file would have to go
[02:09:50] <jepler> hmm
[02:10:00] <cradek> jepler: I'm similarly clear
[02:10:13] <SWPadnos> yeah, I don't know where it should go
[02:10:34] <SWPadnos> it'll be on the gecko site for sure, we can have it on the wiki, and otherwise - uh
[02:19:53] <SWPLinux> jepler, I get this:
[02:19:54] <SWPLinux> Traceback (most recent call last):
[02:19:55] <SWPLinux> File "/usr/bin/stepconf", line 1380, in on_pport_prepare
[02:19:57] <SWPLinux> self.widgets[p].set_active(hal_output_names.index(self.data[p]))
[02:19:59] <SWPLinux> ValueError: list.index(x): x not in list
[02:20:29] <SWPLinux> going from the basic machine information page to the port/pin mapping page
[02:20:58] <SWPLinux> and all pins are unused
[02:21:19] <SWPLinux> should I build from cvs to see if this is fixed? (this is from 2.3~betawhatever)
[02:22:23] <jepler> SWPLinux: yes, I fixed it on TRUNK today
[02:22:28] <SWPadnos> ah, ok
[02:35:03] <SWPadnos> ok, it loaded with a fresh build
[02:35:31] <SWPadnos> I didn't actually test the hardware though, since I've disconnected some things (like the power supply)
[03:28:00] <dgarr> added a little more validation and arrow keys which broke some stuff, that i then fixed (i think) http://filebin.ca/gbjbo/tooledit.tcl
[08:17:54] <Tottish> Anyone awake in here?
[08:19:06] <micges_emc> yep
[08:19:37] <alex_joni> slightly
[08:21:46] <Tottish> Goodie! In this line: "net spindle-speed-cmd motion.spindle-speed-out => pwmgen.0.value" what is the first 'spindle-speed-cmd' good for? I know that net ties these things together but wouldn't it suffice with the latter two?
[08:24:08] <micges_emc> its the name of that signal
[08:27:04] <micges_emc> when you create signal with net command then if you want use that signal later in hal to connect to something more, you can only access it by name
[08:27:33] <Tottish> ok, and motion.spindle-speed-out is the pin? so why do we need to tie the signal to the pin? Don't understand this yet.
[08:27:43] <Tottish> OKOK
[08:27:48] <Tottish> Now I see
[08:29:08] <Tottish> Were you in here yesterday when we discussed my laser-issues?
[08:30:41] <micges_emc> ah lasers :)
[08:30:46] <micges_emc> moment
[08:32:57] <micges_emc> Tottish: whats the problem ?
[08:34:10] <Tottish> Have you heard any of the discussions yesterday and before that about solving a problem with turning a laser on/off while moving at constant speed in X/Y?
[08:38:21] <micges_emc> writting in hal is the only resolution for now
[08:39:04] <micges_emc> better: I must write such module soon so you could chatch sources of simmilar one
[08:39:16] <Tottish> Oh really? Yesterday I was to understand from SWPadnos that I would have to change the sourcecode
[08:40:35] <micges_emc> Tottish: to be sure: I don't have time now to dig logs, can you shortly describe what you want to do and problems do you have up to now
[08:40:57] <micges_emc> I'm at work now :)
[08:43:59] <Tottish> Hmmm, I guess it's pretty complicated but the problem seemed to be that I wanted to stream values from a textfile. Then, they said, I have to use something called halstreamer and userspece and for that I had to change the sourcecode. Can't say I understood the half of it and it seems like a lot of work to me so the reason I asked you if you knew about my problems was that I wanted to try a new idea.
[08:44:47] <Tottish> One which I would think is possible to solve without digging deeper from HAL.
[08:46:57] <Tottish> Are you up for a little Idea-Bouncing? Really, It's ok if you have to work... =P
[08:47:19] <micges_emc> not now
[08:50:02] <Tottish> OK, one quick question, then: Is there a pin or something containing the current velocity of an axis?
[08:50:56] <Tottish> never mind, found it!
[08:57:38] <alex_joni> micges_emc: Tottish wants to do laser engraving/burning
[08:57:44] <alex_joni> but doing it with raster scanning
[08:57:59] <alex_joni> moving the laser left-right on a line, then advance to the next line, etc
[08:58:15] <alex_joni> the issue is to turn the laser on at the exact time during the raster scan
[08:58:23] <Tottish> do you have time for a little chat alex?
[08:58:42] <micges_emc> alex_joni: funny but I'm designing same machine here :D
[08:58:54] <Tottish> cool=)
[08:59:03] <micges_emc> same problems
[08:59:23] <alex_joni> Tottish: yup, I have at the moment
[08:59:40] <alex_joni> micges_emc: I suggested a possible solution, which I'm sure you can implement ;)
[08:59:52] <micges_emc> but it will be about month to write that module :)
[09:00:00] <alex_joni> I don't think so
[09:00:04] <alex_joni> it should be a day or so
[09:00:21] <alex_joni> especially since you can compile emc2 for simulator and debug it
[09:00:52] <micges_emc> err I can start write it in may ;)
[09:04:06] <Tottish> Nice, So I'm thinking about speedcompensating. What about putting the value from motion.current-vel in the the pwm-value for a pwmgen and use a digital-out to enable the PWMgen. Invert the pin and you should have a decent inverted value of the speed to the power of the laser. You shuld probably put a SCALE in between but a testrun would do without it. What do you think?
[09:05:24] <Tottish> update in servothread and makepulse in base thread
[09:05:35] <Tottish> should give a nice responsetime
[09:05:38] <Tottish> ?
[09:05:50] <micges_emc> Tottish: I have that module
[09:05:55] <Tottish> oh?
[09:06:03] <micges_emc> it's working nicely
[09:06:23] <Tottish> OK? gimmegimmegimme!
[09:06:28] <Tottish> =)
[09:06:55] <micges_emc> Tottish: I must work, remember me soon
[09:06:58] <micges_emc> bbl
[09:07:51] <Tottish> alex:why do you figure he needs this new module if the old one workes fine?
[09:09:23] <alex_joni> I'm not sure what module micges has
[09:09:34] <alex_joni> but probably the one for pwmgen
[09:10:42] <Tottish> Sounds like it, but why would a full-line rasterscan be preferrable? If this new idea is working it sounds like a better way to me...
[09:10:54] <Tottish> new for me, that is... =)
[09:11:24] <alex_joni> er.. I don't understand
[09:11:49] <Tottish> ok =)
[09:12:07] <alex_joni> please try to be more explicit when talking about things than "this new idea"
[09:12:28] <Tottish> What I wrote in the long post about 10 posts up...
[09:12:32] <alex_joni> there have been at least half a dozen new ideas lately, and my psychic skills are pretty low atm
[09:12:38] <Tottish> thats 'the new idea
[09:12:48] <alex_joni> speedcompensated pwm ?
[09:12:53] <Tottish> yeah
[09:13:02] <alex_joni> then write that.. not that hard :)
[09:13:29] <alex_joni> I have my doubts (just gut feelings) you'll get much better results
[09:13:52] <Tottish> I will try, hav never done HAL before. I was just wondering if you had a clue about why a rasterscan would be preferable. ==)
[09:14:00] <alex_joni> better speed
[09:14:42] <Tottish> Yeah, may be so. I'll give it a shot. Wish me luck!
[09:17:37] <alex_joni> good luck :)
[10:58:07] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/hal/images/halscope-11.png: add images
[11:08:49] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/hal/tutorial.lyx: finished updating this one afaik
[11:25:40] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/hal/tutorial.lyx: minor edits
[11:31:58] <alex_joni> errr..
[11:37:40] <Tottish> Hey, Can I tie a float from a parameter to a signal? I want to use the motion.current-vel in the pwmgen.0.value... It seems the net-command only tie pins and signals?
[11:44:21] <BigJohnT> if it is RW I think you can
[11:50:21] <Tottish> Then maybe it's not because I get an error "pin 'motion.current-vel' does not exist" But I kind of already knew that since it's a parameter... =)
[11:52:31] <BigJohnT> ok I missed the parameter part
[11:52:56] <Tottish> OK, but this must be doable, right?
[11:53:15] <alex_joni> not really
[11:53:22] <alex_joni> you can't connect parameters to anything
[11:53:31] <alex_joni> except using halscope to see them
[11:54:02] <alex_joni> there is a component which watches parameters and outputs the same value on a pin, but that is not realtime, and won't help you
[11:54:11] <Tottish> bah, then how will I get the Velocity into the PWM?.... grrrr...
[11:54:18] <alex_joni> otoh, look at the sim config, there is another way to get the velocity
[11:54:22] <alex_joni> using ddt blocks
[11:57:31] <Tottish> hmmm, ok, I'll be honest: I have no idea what that means. =) Does it mean I will have to (re)compile the source in the end? Were should I start looking?
[11:57:42] <Tottish> where
[11:58:19] <alex_joni> Tottish: you do know what a configuration is.. right?
[11:58:30] <Tottish> yeah..
[11:58:37] <alex_joni> and that there are sample-configs we distribute
[11:58:46] <alex_joni> and along those there is a folder called sim/
[11:58:55] <alex_joni> and inside there are some "sim" configs
[11:59:05] <alex_joni> composed from .hal + .ini + ...
[11:59:17] <alex_joni> if you look in the hal files you'll see a current vel
[12:00:25] <Tottish> OK, I created my configs with stepconfwizard is there a velocity component in one of those?
[12:00:39] <alex_joni> I'm quite sure there isn't
[12:00:50] <alex_joni> http://cvs.linuxcnc.org/cgi-bin/cvsweb.cgi/emc2/configs/common/core_sim.hal?rev=1.15
[12:01:07] <alex_joni> there is a hal file which contains the ddt part you'll want to use
[12:01:34] <Tottish> Thanks alot, I'm sorry I'm such a noob here. Thanks alot for having patience with me...
[12:01:46] <alex_joni> but if you want to make this work, I would rather try to learn what a full configuration looks like, how thins work, and how they interact .. instead of using stepconf
[12:02:06] <alex_joni> stepconf is nice & easy, but at one point it's limited, and more of a problem than a help
[12:02:37] <Tottish> yeah I suppose thats what I'm doing now... but I use the stepconf generated files as a base...
[12:04:37] <Tottish> I've read the Integrator manual and most of the HAL tutorial so I'm starting to get the picture...
[12:40:02] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/hal/ (hal-examples.lyx tutorial.lyx): add file
[12:45:24] <CIA-2> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/ (Master_Integrator.lyx Submakefile docs.xml index.tmpl): add hal-examples to integrator and html
[14:22:16] <Tottish> It seems I succeeded in making the speedcompensating HAL-script! I have yet to try it out on the machine, though. In any case it's a small victory to me. Thanks alot for the help alex!
[14:25:32] <BJT-Work> Tottish: what did you end up doing?
[15:30:47] <Tottish> Speedcompensating, with the XY-velocity input into the PWM for the laser. And ON/OFF on a M64-Digital output thingy. I'll probably test later today to see how even the result gets. I figure I migh be able to SCALE the value to match my lasers output/dutycycle-curve but this is a good start!
[16:44:23] <Tottish> Jippie! It seems to work nicely.. still needs more testing though...
[19:38:02] <CIA-2> EMC: 03tissf 07TRUNK * 10emc2/docs/src/hal/tutorial_fr.lyx: french translation update
[20:51:05] <CIA-2> EMC: 03tissf 07TRUNK * 10emc2/docs/src/ (6 files): french translation update - to follow John
[20:51:05] <CIA-2> EMC: 03tissf 07TRUNK * 10emc2/docs/src/hal/ (hal-examples_fr.lyx tools_fr.lyx tutorial_fr.lyx): french translation update - to follow John
[22:15:05] <Guest374> Guest374 is now known as skunkkworks