#emc | Logs for 2007-09-12

Back
[00:20:59] <eric_u> do I have to go to CVS to get pyVCP to work?
[00:25:48] <eric_u> I am trying to get Anders Wallen's pyVCP widget for mesa to work
[00:25:51] <eric_u> http://www.anderswallin.net/2007/09/pyvcp-m5i20-hostmot-4-test-panel/#comments
[00:27:01] <eric_u> his file called iotest.xml doesn't make a good control panel
[00:27:31] <eric_u> and his pyiotest.hal flile runs until it gets to his variables, and says that they aren't defined
[00:27:37] <eric_u> that happens on line 24
[00:28:11] <SWPadnos> it appears that the pyvcp with TRUNK has the same features as the one in 2.1.7
[00:28:21] <SWPadnos> there are a couple of bugfixes in the TRUNK version
[00:28:59] <SWPadnos> modify the .hal file to show what's there
[00:30:32] <jlmjvm> i am trying to get the pyvcp dro to work
[00:30:39] <eric_u> can I just use halmeter?
[00:31:20] <SWPadnos> eric_u, no. you were saying that some pins weren't found - add "halcmd show pin" just before all those connections to tell you what pins are there and what they're called
[00:32:24] <eric_u> one thing I'm not clear on -- how do I stop all this stuff from running?
[00:32:30] <SWPadnos> fwiw, I just tested it with trunk, and it works fine. I'm trying to figure out how to test it with the installed version on that machine :)
[00:32:37] <SWPadnos> stop what stuff from running?
[00:32:42] <eric_u> lsmod shows that I started a batch of modules
[00:33:10] <SWPadnos> if you're talking about the HAL and RTAI modules, you prevent them from running by not using EMC
[00:33:21] <eric_u> how do I kill them?
[00:33:29] <SWPadnos> don't do that
[00:33:34] <SWPadnos> use realtime stop
[00:38:06] <eric_u> there is no reason not to just issue a halcmd show pin from the command line, correct?
[00:38:22] <eric_u> it shows only m5i20.xxxxxx pins
[00:38:25] <eric_u> lots of them
[00:38:55] <SWPadnos> the reason to put it in the script is so you can see the state while the script is running, not after it's done
[00:39:20] <SWPadnos> I think I just got the same problem you're seeing though, so I'll say that yes, you need TRUNK or 2.2 when it comes out
[00:40:00] <SWPadnos> I could be wrong - someone else may know how to fix it
[00:40:58] <eric_u> that's not something I can do in isolation is it?
[00:41:07] <SWPadnos> which?
[00:41:17] <eric_u> update pyVCP
[00:41:44] <SWPadnos> I don't think that would be easier than just using TRUNK
[00:44:18] <eric_u> if I just cvs .... blah blah blah ... co emc2, do I get TRUNK?
[00:44:24] <SWPadnos> yes
[00:44:54] <SWPadnos> you need the emc2-dev and build-essential packages as well
[00:45:16] <SWPadnos> (you'd need those for vcp as well - it's both python and C, and would need to be compiled)
[00:45:31] <eric_u> apt-get install emc2-dev?
[00:45:46] <SWPadnos> apt-get install emc2-dev build-essential
[00:45:53] <SWPadnos> sudo apt-get install emc2-dev build-essential
[00:45:55] <SWPadnos> :)
[00:46:41] <jlmjvm> SWPadnos:I am trying to use the pyvcp dro from the wiki,put the above files into your emc root directory (eg: /home/tomp/emc2-head/)
[00:47:04] <jlmjvm> question is i dont have a emc2-head folder
[00:47:30] <eric_u> jlmjvm, I think those instructions are probably to specific
[00:48:15] <eric_u> but I think we are in the process of establishing that it's not going to work anyway
[00:48:30] <eric_u> I may be able to tell you in a little bit
[00:48:40] <jlmjvm> k
[00:49:41] <eric_u> do you have a link to those files?
[00:50:28] <jlmjvm> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PyVCP
[00:53:54] <eric_u> i miss the ssh terminal program I use on windows
[00:54:38] <SWPadnos> I think pyvcp works at some level for emc 2.1.x, I just don't know the level :)
[00:55:06] <eric_u> how do you tell what version?
[00:55:27] <SWPadnos> no, I meant I'm not sure of the level of functionality in pyvcp for packaged versions of emc
[00:55:27] <eric_u> I am running?
[00:55:58] <eric_u> I'm up to the latest version that the updater knew about
[00:56:06] <SWPadnos> err - I don't remember the dpkg / apt query incantation
[00:56:18] <SWPadnos> then you've got 2.1.7
[00:56:27] <eric_u> sounds right
[00:56:47] <eric_u> are there later packaged versions?
[00:58:03] <SWPadnos> no, that's the latest
[00:58:29] <eric_u> then I'm not holding my mouth right or clicking crooked or something
[01:00:19] <jepler> eric_u: I can verify that the .xml file from anderswallin.net works with CVS TRUNK, but not with 2.1.7.
[01:00:36] <SWPadnos> I thought I had said that earlier - it's the same here
[01:00:41] <jepler> not sure why the 2.1.7 pyvcp just gives a tiny blank window instead of a useful error message though
[01:01:12] <jepler> SWPadnos: ok -- I saw that you were able to test it in TRUNK but that you couldn't figure out how to test in 2.1.7 or something to that effect
[01:01:40] <SWPadnos> ah - I figured it out :)
[01:01:52] <jepler> oh
[01:01:56] <jepler> I always miss stuff in scrollback
[01:04:16] <eric_u> so I'm stuck building EMC2 from CVS?
[01:04:26] <jepler> the correct way to prepare a system to compile emc is to apt-get install build-essential, then apt-get build-dep emc2 (to install the packages needed to rebuild the installed version of emc) or apt-get build-dep emc2-cvs-build-dep (the packages needed to install the development / CVS HEAD version of emc)
[01:04:50] <tomp> i have all the pre-requisites, and just ran the big dro using those instructions... (see jimjvm's post 3 lines up )
[01:04:56] <jepler> you may also need to apt-get install cvs if you choose to get the source using the cvs command
[01:04:58] <tomp> but i'd add cd ~/yourruninplaceemcdirectory
[01:05:03] <tomp> . scripts/emc-environment
[01:05:05] <tomp> scripts/halrun pyvcp-dro2.hal
[01:05:22] <tomp> NOTE the '.' is followed by a space
[01:07:33] <tomp> hmm, that's on emc-2.1.6
[01:07:44] <jlmjvm> eric_u:the instructions were to specific,put files in home folder and it worked
[01:08:53] <eric_u> great
[01:09:02] <tomp> did goslo-jimbo ever come back and get that big dro we made for him?
[01:09:56] <SWPadnos> I thought he said something on the user list
[01:09:58] <SWPadnos> or something
[01:10:23] <tomp> btw: that page has another m5i20 test panel i scratched up, still untested ( my bad )
[01:10:46] <eric_u> which page?
[01:11:06] <tomp> same page ;) http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?PyVCP
[01:11:38] <tomp> hers the pkg look mid page http://wiki.linuxcnc.org/uploads/m5i20-pyvcp-tab-panel.tgz
[01:12:40] <eric_u> I see why I didn't see it, it's an eye test :)
[01:12:50] <eric_u> I fail those usually
[01:13:24] <tomp> heh, i still havent figgered how to load a thumbnail, then get a full size pic with that wiki
[01:13:46] <eric_u> what I can see looks nice
[01:15:08] <jlmjvm> this full size dro is pretty cool,have my z axis reading now
[01:15:26] <eric_u> jealous
[01:15:43] <tomp> great, its now useful :)
[01:16:25] <jlmjvm> can it run when emc is running,wanna see if my counts are right
[01:17:14] <jlmjvm> i only change the a and b pin numbers in the hal file
[01:17:35] <eric_u> should be able to run
[01:18:37] <jlmjvm> jammed my computer up
[01:19:39] <eric_u> how are you typing here then?
[01:19:55] <jlmjvm> had to do a power off reboot
[01:19:55] <eric_u> sample rate too fast?
[01:20:03] <jlmjvm> 2 computers
[01:20:24] <jlmjvm> 1 on machine,other on net
[01:20:31] <eric_u> tomp has disappeared
[01:21:38] <tomp> did you change lines like 'net YphaseA parport.0.pin-04-in +> encoder.1.phase-A' ? ( no i didnt disappear , just slow typer )
[01:22:03] <tomp> s/+>/=>
[01:22:03] <jlmjvm> yes my a is 12 and b is 15
[01:22:48] <eric_u> since it locked things up, doesn't that suggest that the 20 microsecond period real-time thread is too fast?
[01:23:07] <eric_u> in combination with his steppers?
[01:23:42] <tomp> ok, how did the hang occur? immediately at running 'scripts/halrun pyvcp-dro2.hal' ?
[01:24:00] <jlmjvm> no,when i started emc2
[01:24:17] <jlmjvm> by itself its fine
[01:25:00] <jepler> you can't start emc while another hal setup is running
[01:25:14] <jepler> instead, you must integrate the pyvcp-dro2.hal with the other hal files of your emc config
[01:26:23] <jlmjvm> so just add it to my ini file?
[01:27:17] <jepler> no, it won't be that simple. For instance, pyvcp-dro2.hal creates threads, but so does one of the .hal files that emc is already using
[01:27:46] <jepler> (at least I assume that's how it works, I am not actually looking at the file)
[01:28:43] <jepler> in other words, you'll have to do it by understanding hal and looking at all the hal files and at the error messages you get until you get it working
[01:28:47] <tomp> yes , makes 2 threads
[01:29:44] <jepler> in the case of the threads, instead of creating them you'd just want to use the two already created in emc (servo-thread [slow] and base-thread [fast])
[01:34:31] <eric_u> or you could run the dro on a different computer
[01:35:04] <tomp> i'm not sure how to patch it into your .ini...
[01:35:12] <tomp> but here's some bits to know
[01:35:16] <tomp> the ini will have a 'HALFILE = yourhalfile.hal'
[01:35:25] <tomp> the halfile will 'loadusr -W halvcp yourpyvcpgui.vcp' and will describe the pins
[01:35:25] <tomp> the .vcp file will have the xml describing the gui
[01:35:53] <eric_u> ls
[01:35:57] <eric_u> oops
[01:36:03] <tomp> the wiki page we were just talking about is a whole world of info
[01:38:25] <eric_u> the only downside to adding computers is trying to balance all those monitors on top of one another
[01:39:19] <cradek> 12" height gauge, Enco brand $85.92; Mitutoyo $722.57
[01:40:02] <SWPadnos> I wonder if there's - you know - a difference
[01:40:05] <SWPadnos> :)
[01:40:24] <eric_u> that's not an indicator stand type thing is it?
[01:40:47] <user_> I'd rather have a name brand second hand from fleebay
[01:40:50] <eric_u> that's where ebay comes into the picture
[01:41:09] <tomp> forget that tgz of the m5i20 panel until its fixed :-[
[01:41:25] <eric_u> there's no hal file in there
[01:41:34] <eric_u> izzat what you're talking about?
[01:41:47] <tomp> yep, just xml, but even the giu is borqed
[01:42:00] <tomp> gui
[01:42:34] <eric_u> oh man, I downloaded it onto my computer, put it on a flash drive, walked down to the basement from the second story, and now you tell me it's not going to work?
[01:43:26] <tomp> sorry, i was trying to test it as you were walking, why didnt you tell me you were walking ;)
[01:43:49] <eric_u> I didn't, I used sftp
[01:43:51] <user_> sneakernet 2.0
[01:43:58] <eric_u> but that didn't sound as good
[01:44:30] <tomp> dunno why, as you can see it use to run...
[01:44:39] <eric_u> true
[01:44:46] <eric_u> where's that hal file though?
[01:45:58] <eric_u> I think I left my flash drive in the computer in the basement, brb
[01:46:00] <tomp> no hal file, just a gui, a demo of how to make the gui, wasnt wired to hal pins yet. doesnt matter if its borqd tho...
[01:46:13] <user_> Cradek,
[01:46:14] <user_> http://cgi.ebay.com/Mitutoyo-12-Digit-Count-Height-Gage-NO-RESERVE_W0QQitemZ110167267923QQihZ001QQcategoryZ20043QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[01:46:34] <tomp> i'll be working on it tho, i got m5i20, 7i33, breakouts and now 7i37's
[01:47:03] <eric_u> I shoulda bought 2 7i37's
[01:47:34] <eric_u> that mitutoyo has a travadial on it
[01:48:21] <eric_u> 18 bids, that may not be the final price
[01:48:34] <cradek> nice
[01:48:55] <user_> I've got a fowler like this bit 18 " the mag glass is nice. It is a nice tool fowler sells for less a lot of times
[01:49:05] <user_> http://cgi.ebay.com/Fowler-Height-gage-12-30cm-model-83103_W0QQitemZ330163994474QQihZ014QQcategoryZ20043QQssPageNameZWDVWQQrdZ1QQcmdZViewItem#ebayphotohosting
[01:50:00] <eric_u> eye chart and height gauge combo
[01:50:46] <eric_u> I should have left the travadials on my bp
[01:51:26] <eric_u> just to be silly
[01:56:50] <toast> travdials are kind of useful i guess
[01:56:58] <toast> even with a dro
[01:57:40] <cradek> on a cnc they sure mean "we don't expect this to work very well"
[01:58:01] <cradek> "we don't trust it and we bet you won't either"
[01:58:26] <eric_u> those bp's were the first cnc that their owners bought
[01:58:36] <eric_u> so it makes sense
[01:59:09] <eric_u> plus they didn't work
[01:59:22] <cradek> ha
[01:59:40] <eric_u> although, from the wear on mine, it did work fine
[02:00:42] <eric_u> I was wrong, that height gauge went for 125
[02:00:59] <cradek> I predicted that from the bidding
[02:01:02] <cradek> $5 more would have gotten it
[02:01:29] <cradek> but, still too much for me
[02:02:33] <toast> and who knows how accurate it is
[02:02:47] <toast> +/- .1 mm
[02:02:58] <Skullworks-PGAB> I bought a 24" Brown & Sharp vernier height gage on ebay for $44
[02:03:06] <toast> was it any good?
[02:03:27] <tomp> eric_u: the m5i20 test panel wont work. it uses widgets that aren't in the distributed source. they're not secret, but didnt make it into the distribution. I'll remove it from the wiki.
[02:03:36] <Skullworks-PGAB> without a dial or a place to stick batteries - nobody wants them
[02:03:49] <Skullworks-PGAB> was fine
[02:03:53] <toast> nice
[02:04:02] <toast> that's a great price for a 24"
[02:04:15] <eric_u> ok, thanks tomp
[02:04:18] <Skullworks-PGAB> lazy people don't want to read a venier scale
[02:04:42] <toast> does it have a scriber point
[02:05:00] <eric_u> not so much lazy, as they know their own limitations
[02:05:12] <eric_u> same reason I don't balance the checkbook in my head
[02:05:14] <Skullworks-PGAB> no didn't have that - but did come with the clamp
[02:05:23] <user_> vernier is the way to go
[02:05:45] <toast> it's important to have a vernier around in case the batteries run out on the digital
[02:06:01] <Skullworks-PGAB> * Skullworks-PGAB wanted the Starett 50 reading...
[02:06:08] <toast> and for scribing layout, i suppose
[02:07:16] <user_> Anyhows if anyone is looking at that fowler, I can say it is just as nice as the starrett and BS that were @ school
[02:07:29] <Skullworks-PGAB> I have several surface gages I can use for scribing.
[02:09:06] <toast> i was on an old, very manual bridgeport at work
[02:09:11] <toast> today and yesterday
[02:09:14] <Skullworks-PGAB> note: much of the new Fowler is made in chian - and it shows, my Fowler 15-0-15 indicater had the needle fall off in less than a year - and the dove tails were crap.
[02:09:18] <toast> i wanted to lay the work out before i did it
[02:09:23] <Skullworks-PGAB> china
[02:09:23] <toast> everyone looked at me like i was from mars
[02:09:34] <toast> we apparently don't have any layout apparati
[02:10:06] <user_> I haven't seen any of their new stuff , height gage looks to be older like mine though
[02:10:06] <Jymmm> /ojin #gpsd
[02:10:16] <SWPadnos> toast, don't the compooter do that for you?
[02:10:30] <SWPadnos> heh
[02:10:39] <toast> i probably could have printed out a drawing to scale
[02:10:41] <toast> and taped it to the part
[02:10:50] <toast> and just used that
[02:11:18] <Skullworks-PGAB> have a 25+ year old Fowler dial depth gage thats sweet - but I would not buy it new today after the quality issues I've seen in the last 2 years
[02:12:13] <toast> how do you like depth gauges versus depth mics
[02:12:53] <tomp> in moldmaking i had to have blade deep mikes to check ribs
[02:13:19] <tomp> round was too fat
[02:13:43] <tomp> but lotsa guys would cut off 1/2 the anvil
[02:16:34] <toast> that sounds adventerous
[02:17:54] <tomp> you can get closer to a wall to measure a depth
[02:18:53] <jmkasunich_> I have to agree about recent Fowler stuff
[02:19:14] <jmkasunich_> American name on Chinese product, with pricing that's half way between the two
[02:19:32] <SWPadnos> American profit on Chinese product
[02:20:14] <jmkasunich_> that too
[02:21:11] <user_> I think people have bout had enough of the china stuff... hope it doesn't get worse though
[02:22:13] <tomp> it'll be like Japan, originally crap tin stuff, but cheap, then later fine craftsmanship and expensive
[02:22:45] <Jymmm> user_: when you have a kid/grandkid, be sure to turn it over and look for the "Made In China" stamp.
[02:24:23] <user_> Just some of those cheap china stuff is soooo poor I can't believe they used metal and machine time to make em.
[02:24:44] <Jymmm> labor is cheaper, you can stuff 6 ppl in a press machine
[02:25:00] <toast> it was only 6 in that video?
[02:25:13] <Jymmm> in the machine, yeah
[02:25:30] <jmkasunich_> I think we made the right call _not_ going sailing this evening: http://jmkasunich.dyndns.org/pics/waves-0759.JPG
[02:25:43] <SWPadnos> heh
[02:25:55] <SWPadnos> we went out for dinner with a friend and my wife was going to walk home
[02:26:06] <SWPadnos> until she noticed the heavy rain and mega lightning
[02:26:12] <SWPadnos> nice spray
[02:26:18] <eric_u> success!
[02:26:34] <user_> Pretty there, I'm 100 miles inland..
[02:26:59] <jmkasunich_> windy
[02:27:12] <SWPadnos> has stormy eyes
[02:27:54] <jmkasunich_> it's remarkably hard to hold a camera steady when the wind is strong enough that you have to lean into it
[02:28:23] <SWPadnos> get a monopod
[02:28:33] <jmkasunich_> I had a tripod
[02:28:38] <SWPadnos> oh
[02:28:40] <eric_u> so I built emc2 from cvs with run-in-place set, that's a little more work to run things
[02:28:43] <jmkasunich_> (and the first shot I posted was taken with it)
[02:28:54] <jmkasunich_> but later I went up in a lighthouse
[02:29:02] <jmkasunich_> the wind up there would have blown the tripod away
[02:29:51] <SWPadnos> eric_u, once it's done, it's basically done. you can either install that version or set up icons/shells so that's the one that gets run
[02:30:05] <SWPadnos> it's not like you have to recompile every time you want to run it :)
[02:30:29] <tomp> theres some old buildings like that, not quite light houses, in the harbors in Chicago. I'd love to live in one :) They have big doors to slide emergency boats out into the water.
[02:32:33] <jmkasunich_> this one is among the few sharp ones from atop the lighthouse: http://jmkasunich.dyndns.org/pics/waves-0767.JPG
[02:32:52] <jmkasunich_> the lake is a bit testy tonight
[02:33:21] <tomp> perch fishin !
[02:33:25] <jepler> cradek: use 'set log' in gnuplot to get log scale plots
[02:34:03] <jepler> for example, 'set log xy' gives a log-log plot
[02:37:45] <jmkasunich_> http://jmkasunich.dyndns.org/pics/flag-sky-0784.JPG
[02:38:58] <tomp> thanks, esp for today 9-11
[02:39:55] <jmkasunich_> http://jmkasunich.dyndns.org/pics/lighthouse-steps-0786.JPG
[02:40:31] <jmkasunich_> If I recall the sign correctly, that flagpole was removed from the first US navy all-iron ship
[02:50:12] <Jymmm> http://www.johnq.com/vod.htm
[02:54:02] <tomp> lighthouse... one in Chicago harbor... a postcard that you cut up and fold.. kinda nice http://www.wurlington-bros.com/BuildYour/howcharbor.html
[03:09:27] <jmkasunich_> GIMP is pretty cool
[03:09:35] <jmkasunich_> http://jmkasunich.dyndns.org/pics/lighthouse-on-hill-0795.JPG
[03:09:45] <jmkasunich_> that was shot from the bottom of a hill looking up
[03:10:01] <jmkasunich_> so the flagpole and lighthouse seemed to be leaning toward each other at the top
[03:10:24] <jmkasunich_> perspective filter in gimp straightened them right up
[03:11:45] <Skullworks-PGAB> I don't have patience enough to work with GIMP
[03:12:08] <Ziegle1> wow... washing machine motor
[03:12:16] <Ziegle1> bet that ball mill goes really really fast
[03:12:52] <Ziegle1> oops wrong window
[03:18:25] <tomp> some of the photo stitching software ( like panorama stuff in quicktime) allows perspective corrections, and can be used on individual frames. http://hugin.sourceforge.net/
[03:18:53] <jmkasunich_> it took me a while to find the perspective stuff in Gimp, and a bit longer to figure it out
[03:19:04] <jmkasunich_> (undo is the greatest invention ever!)
[03:20:52] <tomp> crazy use of stitcher http://www.bruno.postle.net/panorama/barkers-pool/
[03:21:21] <jmkasunich_> heh
[03:22:14] <Jymmm> jmkasunich_: CTRL+C, CTRL+V, CTRL+Z =)
[03:22:47] <jmkasunich_> jymmm in Gimp the first two are much less used than the last one
[03:23:22] <Jymmm> jmkasunich_: I have my trackball programmed for copy and paste (top left and right buttons)
[03:23:34] <Jymmm> I use them constantly
[03:23:44] <jmkasunich_> for photo editing?
[03:23:56] <Jymmm> Yeah, but for everything
[03:24:08] <skyfox00> who all here has sucesfully compiled rtai from source?
[03:24:09] <Jymmm> marque select, copy, paste into new image
[03:24:29] <SWPadnos> I just did
[03:24:38] <SWPadnos> (compile RTAI from source)
[03:24:48] <skyfox00> and it all worked?
[03:24:53] <SWPadnos> yes
[03:24:59] <skyfox00> what version?
[03:25:12] <SWPadnos> magma CVS from a couple of days ago
[03:25:26] <skyfox00> which kernel?
[03:25:27] <SWPadnos> hmmm - no, wait. this is the source from the experimental SMP RT kernel
[03:25:39] <SWPadnos> 2.6.20.14?
[03:25:49] <Skullworks-PGAB> I have the "SpaceNavigator 3Dconnexion" (a division of Logitech) - its the bomb for Google earth.
[03:26:03] <Skullworks-PGAB> also for Solidworks
[03:26:07] <SWPadnos> go to www.linuxcnc.org/experimental and download anything with 2.6.20 in the name
[03:26:21] <SWPadnos> those are provided with really no support :)
[03:26:43] <skyfox00> ok, I need take notes... I just spent all day trying to get linux-2.6.15/rtai-3.3 working...
[03:27:06] <SWPadnos> I think you need magma, and I believe the only way to get any recent version of it is to use cvs
[03:28:59] <skyfox00> hmmm, this only gets more complicated...
[03:30:36] <SWPadnos> it's pretty complicated if you want to do it all yourself. if you want it simple, then you need to use the precompiled packages somehow
[03:31:23] <skyfox00> emc2 is best suported on a ubuntu computer connected to the internet and the computer I am trying to get it working on runs slackware(my preference) and cannot be connected to the net...
[03:31:40] <SWPadnos> yep - that makes it more complicated for you :)
[03:32:03] <SWPadnos> you should try to follow the instructions on the wiki for installing to other distributions
[03:32:24] <SWPadnos> there's also a page called RTAISteps which should have some good information, though it is a bit outdated
[03:32:39] <SWPadnos> (ie, most of the versions will have changed by now)
[03:34:59] <skyfox00> do you know when the next binary version will be release?
[03:35:10] <SWPadnos> "in the next few months"
[03:35:23] <SWPadnos> err - of emc2?
[03:35:30] <skyfox00> yeah, emc2
[03:35:32] <SWPadnos> ok
[03:35:53] <skyfox00> will that include a liveCD iso(for installing)?
[03:36:38] <SWPadnos> there will probably be a new ISO, but until at least April of next year, we're thinking it will still be based on the Ubuntu 6.06 LTS distribution
[03:36:56] <SWPadnos> so EMC will be different but the kernel probably won't be
[03:38:06] <skyfox00> well, thats a long time, what is involved in upgrading the current Ubuntu liveCD to the CVS version of emc2?
[03:38:30] <SWPadnos> I don't understand the question
[03:38:48] <SWPadnos> when the next EMC2 is released, the liveCD will be updated (soon after anyway) with the new EMC2
[03:39:02] <skyfox00> can Ubuntu install sources(packages) from disk(usb thumb drive)?
[03:39:10] <SWPadnos> or are you asking what it takes to compile CVS EMC2 on an Ubuntu install?
[03:39:17] <SWPadnos> yes, it should be able to
[03:39:41] <SWPadnos> you can always use `dpkg -i <package-name.deb> to install a package
[03:39:49] <SWPadnos> +/- some other command line options
[03:40:08] <skyfox00> what I am wondering is how hard it is to install the EMC2 Ubuntu liveCD then upgrade to the CVS version of EMC2
[03:40:40] <SWPadnos> the siki installing emc2 page tells how to do that: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_5_10_and_6_06_from_source
[03:40:45] <SWPadnos> s/siki/wiki/
[03:40:57] <skyfox00> ok, will take a look
[03:41:18] <SWPadnos> it does assume a net connection, but you can use CVS on a connected PC and then copy the tree over to the EMC2 PC (use a flash disk or something)
[03:41:42] <SWPadnos> installing the development packages would be much easier with a net connection
[03:41:43] <Skullworks-PGAB> any vague ideas as to next release date?
[03:41:52] <SWPadnos> "in the next few months"
[03:42:08] <Skullworks-PGAB> thats 2.2 - or 2.1.8?
[03:42:13] <SWPadnos> yes :)
[03:42:23] <SWPadnos> or no
[03:42:31] <skunkworks> yes and no
[03:42:39] <jmkasunich_> 2.2 in a few months, to get all the shiny new features currently only available from CVS
[03:42:53] <jmkasunich_> 2.1.8 will come as if and when needed, but will contain only bugfixes
[03:42:54] <Skullworks-PGAB> ( yeah I did ask for vague... )
[03:43:08] <SWPadnos> and a 2.1.8 if there are any major bugs found before 2.2 is "ready"
[03:43:10] <Skullworks-PGAB> ok
[03:44:50] <Skullworks-PGAB> I'm still working on the hardware end - so it really makes no big dif to me.
[03:45:40] <skyfox00> anyone know where I can find a list of files that are needed for installing the development packages for EMC2?(so I can put them on my thumb-drive)
[03:46:11] <Skullworks-PGAB> but I was going to start trying to port many of my Macro-B formated marco subs to work in EMC and post a library
[03:46:56] <Skullworks-PGAB> have a 2.1.7 box here to run sim for compatability testing with interp
[03:48:10] <Skullworks-PGAB> the really nice one will generate the tool path to threadmill female NPT threads.
[03:48:09] <fenn> skyfox00: are you sure you arent just making things difficult for yourself?
[03:49:11] <fenn> that sounds pretty useful actually
[03:49:23] <skyfox00> I am just trying to get the reset capability for the parport in EPP mode for use with the pluto_servo board.
[03:50:49] <skyfox00> I had the CVS version running with a buggy kernel and rtai-3.5 and it was working and could recover the parport after an error... but that was with a nvidia tainted buggy kernel.
[03:51:55] <skyfox00> what is the recomended PCI/ISA EMC2 compatible servo-motor contoller board?
[03:53:34] <Skullworks-PGAB> Is there any USB multi-channel I/O board that could use a userspace driver for non-realtime critical events like coolant on etc. ?
[03:54:27] <skyfox00> if you cannot find one, it would be easy to make one...
[03:55:37] <skyfox00> if you want a closed-loop realtime usb interface, put on a flame retardant suit first...
[03:57:28] <Skullworks-PGAB> no - no need for true realtime for stuff like coolant pumps, E-stop -OFF amp enable, etc.
[03:59:10] <skyfox00> are you looking for somthing that can drive motors, etc, directly or somthing that puts out logic signals?
[03:59:43] <Skullworks-PGAB> but I would prefer something like an 8255
[03:59:54] <SWPadnos> Skullworks-PGAB, look into the arcade game controllers that let you hook up buttons - those should be supported by hal_input
[04:00:09] <SWPadnos> at least for inputs. I'm not sure how many outputs they usually have
[04:00:31] <Skullworks-PGAB> been there done that - good for buttons - but not for switches
[04:00:40] <SWPadnos> ok
[04:01:19] <skyfox00> not likely what your looking for....: http://www.sparkfun.com/commerce/product_info.php?products_id=7841
[04:01:20] <Skullworks-PGAB> handles momentary fine - but will trigger a stuck key response if a switch is used
[04:01:33] <SWPadnos> even the USB ones?
[04:01:42] <SWPadnos> I cn imagine that from the PS/2 style
[04:01:44] <SWPadnos> can
[04:02:08] <Skullworks-PGAB> don't know about those - I was looking mainly at the PS/2 type
[04:02:40] <Skullworks-PGAB> but think the signal gen is same even if I/O method differes
[04:03:14] <SWPadnos> could be if it registers as a keyboard
[04:03:43] <Skullworks-PGAB> It would work fine if people wanted to put alot of extra bells & whisles in the GUI front ends.
[04:04:18] <Skullworks-PGAB> so that status of these extra options were shown
[04:04:20] <skyfox00> those FT245RL boards show up as a serial port and the 8 data lines are just 8 bits... not sure how to interface a virtual com port to emc2 though...
[04:05:41] <Skullworks-PGAB> I'm still thinking that tripple 8255 board would be the hot ticket for a nice hardwired control panel
[04:05:58] <skyfox00> what is a 8255?
[04:05:58] <SWPadnos> that;s a cool little board. too bad it has zero protection on the I/Os
[04:06:08] <SWPadnos> it's a 24-I/O chip fromthe old skool
[04:06:13] <Skullworks-PGAB> NEC
[04:06:20] <Jymmm> Anyone know of a mobo the size of a deck of cards or smaller that has ethernbet, and maybe usb or serial, and a CF for storage?
[04:06:40] <Skullworks-PGAB> yes was often used on 1st gen ISA cards
[04:06:42] <SWPadnos> yes, but I haven't been able to find it online (saw it at a trade show)
[04:06:58] <Jymmm> SWPadnos: remember the brand?
[04:06:59] <SWPadnos> we designed one such card, ages ago
[04:07:07] <SWPadnos> Tokyo Electric, I think
[04:07:16] <SWPadnos> it was in the AMD booth as a sample of a tiny computer
[04:07:30] <Skullworks-PGAB> is the heatsink larger than the board?
[04:07:49] <SWPadnos> no heatsink needed - it used a 3W eden or something similar
[04:08:16] <SWPadnos> http://www.via.com.tw/en/initiatives/spearhead/pico-itx/
[04:08:19] <SWPadnos> those are close
[04:09:15] <Skullworks-PGAB> 8255 - http://www.futurlec.com/PCI8255.shtml
[04:09:34] <Jymmm> It doens't even need to be that fancy, just want to install a minimilistic nix on it.
[04:09:38] <SWPadnos> I really don't understand why anyone would buy that card when the MEsa 5i20 is only $200
[04:10:22] <Jymmm> SWPadnos: Hmmmm C7 cpu on that thing
[04:10:24] <fenn> Skullworks-PGAB: phidgets is a usb i/o thing, seems crappy but that's all i know off the top of my head
[04:11:28] <Jymmm> $230 for that board
[04:11:39] <SWPadnos> that's not bad
[04:12:37] <Jymmm> In itself, no. But I wasn't looking for as many features as it has, just a minimilistic nix install (not BusyBox, necessarily)
[04:12:52] <LawrenceG> http://gumstix.com/ Jymmm
[04:12:53] <fenn> ts7200 :)
[04:12:54] <Jymmm> ~$40 range
[04:13:01] <SWPadnos> hahahaha
[04:13:31] <fenn> http://www.embeddedarm.com/epc/ts7200-spec-h.html
[04:13:35] <SWPadnos> maybe grab a cheap and small wireless access point and install Linux on it
[04:13:39] <LawrenceG> Jymmm: netgear wgt634u
[04:13:43] <SWPadnos> or a wired access point - either way
[04:13:44] <fenn> what are you doing jymmm?
[04:13:45] <Skullworks-PGAB> does mesa make a more rudamentry card - it would seem to me the 5i20 is overkill for simple switching I/O
[04:13:57] <Jymmm> SWPadnos: That's what I have now, just a tad bulky
[04:14:12] <SWPadnos> Skullworks-PGAB, they do have a 96-I/O dumb card, but I think it's PC/104 only
[04:14:23] <SWPadnos> Jymmmtake the case off ;)
[04:14:28] <Jymmm> fenn: Bummer $100/100qty pricing
[04:14:40] <fenn> its like $129 single qty
[04:14:39] <Skullworks-PGAB> yeah seems most there line is PC104 based
[04:14:46] <fenn> is that too much?
[04:14:56] <SWPadnos> $40 - that's why I laughed
[04:15:10] <Jymmm> fenn: I think the better deal would be the VIA board
[04:15:15] <fenn> good fast cheap - pick two
[04:15:31] <SWPadnos> good fast cheap compact - pick at most two
[04:15:42] <fenn> in this case, compact is part of "good"
[04:15:50] <SWPadnos> ok, I can accept that
[04:16:25] <fenn> and fast refers to how long you have to surf ebay
[04:16:48] <Jymmm> fenn: I have a dawg slow connection here at work
[04:17:05] <SWPadnos> musta changed jobs
[04:17:05] <fenn> i thought you were mister 100MB/s
[04:17:16] <SWPadnos> 1000MB/s ;)
[04:17:35] <Jymmm> * Jymmm hands SWPadnos a 0
[04:17:45] <SWPadnos> well, that's per pipe ...
[04:17:47] <Jymmm> or a comma, whatever you prefer
[04:18:24] <SWPadnos> maybe I should try to get hired by them and have an OC48 installed at my house, so I can effectively telecommute
[04:18:37] <SWPadnos> or telecommute effectively
[04:18:53] <Jymmm> SWPadnos: dark fiber to NY?
[04:19:15] <SWPadnos> I'm a ways from NY, and there's a lake in the way
[04:19:25] <Jymmm> TX ?
[04:19:32] <SWPadnos> I can get to a SONET ring pretty close by though
[04:19:41] <Jymmm> there ya go
[04:21:15] <skyfox00> well, I think I'm going to try linux-2.6.17/rtai-3.5 next...
[04:21:26] <Jymmm> that is a nice looking board though... http://us.st11.yimg.com/us.st.yimg.com/I/yhst-27518546784426_1965_109064608
[04:21:48] <SWPadnos> yeah - it is
[04:22:10] <Skullworks-PGAB> hmm MESA ISIO
[04:22:27] <SWPadnos> that's just an isolator for parports
[04:22:40] <SWPadnos> useful, but doesn't give you any more I/O than you had
[04:23:32] <Skullworks-PGAB> true
[04:24:04] <Jymmm> Hmmm optional A/d converter http://www.embeddedarm.com/epc/ts7200-spec-p.php
[04:24:07] <Skullworks-PGAB> but it looks much more robust
[04:24:09] <Skullworks-PGAB> The ISIO has 4 relay or 4 transistor outputs and 8  opto-isolated inputs. Relay outputs are rated at 2A, 24VDC or 1A, 120VAC.  Transistor outputs are rated at 1A, 48VDC. Inputs are 10 to 48VDC.  All isolated I/O uses screw terminals.  Outputs are disabled at powerup until a specific 12  bit code is sent to the ISIO
[04:24:12] <SWPadnos> Skullworks-PGAB, the reason I'd get the 5i20 (or better) over a simple I/O board is that i removes the burden of faster I/Os from the parallel port. I'd shift step/PWM generation and encoder reading to the 5i2x, and use the parport I/Os for the slow stuff
[04:24:23] <SWPadnos> s/i/it/
[04:25:45] <Skullworks-PGAB> I may end up buying Jon's PWM amps and controller
[04:25:46] <SWPadnos> even for software step or PWM generation, the 5i20 is better. it has a lower overhead since outptus can be written in larger chunks, and it has 33 MHz bus cycles instead of ~1 MHz, so wait states are much shorter
[04:26:29] <skyfox00> anyone want a good deal on a new pluto board?
[04:26:33] <SWPadnos> heh
[04:27:02] <skyfox00> how well is the 5i20 suported/work in emc2?
[04:27:21] <SWPadnos> the hostmot4 configuration is completely supported
[04:27:35] <Skullworks-PGAB> I have yet to find a PWM amp with current limiting for less -
[04:27:57] <SWPadnos> that's 4 PWM (pseudo-DAC) outputs, 4 wuadrature inputs, and 32 out+16 inputs (or 32 in + 16 out - don't remember)
[04:27:59] <skyfox00> I guess I should ask if can run servo motors
[04:28:20] <SWPadnos> Skullworks-PGAB, what does Jon charge?
[04:28:24] <SWPadnos> ~$159 or something
[04:28:53] <Skullworks-PGAB> $125 /ea for amps + $250 for the PWM inferace
[04:29:01] <SWPadnos> oh, that's pretty darned good
[04:30:44] <Skullworks-PGAB> add the in-rush limiter and E-stop breaking card the package will be about $1000
[04:31:05] <SWPadnos> well, for me it should be less
[04:31:34] <SWPadnos> I think all I need is the EEPROM to make my USC a UPC, and I already have the gecko servo interface
[04:32:08] <fenn> $1k for an inrush limiter?
[04:32:19] <fenn> and a resistor
[04:32:23] <fenn> feh
[04:32:24] <SWPadnos> I think the PWM in the new mesa configs will be able to run his amps though - I recall some discussions about that (but I don't recall the outcome)
[04:32:26] <Skullworks-PGAB> (4axis)
[04:32:46] <SWPadnos> no, $1000 for the whole thing, including the inrush limiter and E-Stop stuff (plus UPC and drives)
[04:32:58] <fenn> oh
[04:33:04] <skyfox00> I guess cnc is not for people with thin wallets...
[04:33:18] <Skullworks-PGAB> well if they get the hostmot8 functional and could drive his amps...
[04:33:28] <fenn> skyfox00: not if you are moving big chunks of metal around fast
[04:33:47] <SWPadnos> hostmot8 isn't all that hard to get going
[04:34:01] <SWPadnos> but I'm not sure it will work with Jon's amps. they have an odd bootstrap requirement
[04:34:07] <fenn> skyfox00: but there's a lot of uses for cheap slow cnc too
[04:34:32] <Skullworks-PGAB> key thing is I have a 130 VDC (peak) 18A DC spindle motor
[04:35:14] <SWPadnos> hmmm. 160V / 20A is pretty perfect for my motors
[04:35:19] <Skullworks-PGAB> that is still within the specs for his amps - to run the spindle as a servo
[04:35:45] <skyfox00> I dont have a job right now, so cheap/slow cnc is thinning my wallet execivley...
[04:36:24] <Skullworks-PGAB> though I'm not looking forward to building a powersupply for that motor :~
[04:37:43] <skyfox00> and getting free multiple-party software/hardware/firmware to all get along without killing eachother has proved to be quite a nightmere...
[04:38:00] <Skullworks-PGAB> I bet
[04:38:25] <Skullworks-PGAB> I'm getting fired on friday
[04:38:30] <toast> ...?
[04:38:51] <Skullworks-PGAB> ( lucky I start a new job on Monday)
[04:39:08] <toast> did you quit
[04:39:12] <toast> or are you seriously getting fired
[04:39:15] <skyfox00> what will you be doing for your new job?
[04:39:16] <Skullworks-PGAB> nope
[04:39:25] <Skullworks-PGAB> company got sold
[04:39:29] <toast> oh
[04:39:45] <skyfox00> did they sell to your new company?
[04:39:49] <Skullworks-PGAB> I got an offere from the new company
[04:40:02] <Skullworks-PGAB> many people didn't
[04:40:33] <Skullworks-PGAB> bout 1/3 are getting serverence pkg
[04:41:14] <Skullworks-PGAB> maybe closer to half overall - but 1/3 in my dept.
[04:41:46] <skyfox00> well, thats better than nothing, when I left my paper route, one of my old customers sent me a check for $20 several months afterward...
[04:42:00] <skyfox00> I must have done a better job than the new guy...
[04:42:23] <Skullworks-PGAB> I'm slow
[04:42:32] <Skullworks-PGAB> about 50% eff
[04:43:06] <Skullworks-PGAB> but I have the best quality record in the company for last 3 yrs
[04:43:27] <skyfox00> what kind of company?
[04:43:51] <Skullworks-PGAB> manufacturing
[04:44:01] <Skullworks-PGAB> I make fuser rollers
[04:44:16] <skyfox00> like for printers/copiers?
[04:44:29] <Skullworks-PGAB> the imaging drums that transfer images in laser printers and high speed copy machines
[04:46:33] <skyfox00> how do people generate g-codes without using expencive(non-free) software?
[04:47:39] <fenn> gcam is a start: http://gcam.js.cs/
[04:47:48] <fenn> er, http://gcam.js.cx/
[04:48:31] <skyfox00> I have it and was quite disapointed with the lack of user interface....
[04:48:35] <fenn> it would be nice to just take a DXF and "convert" it but that doesnt seem to work for me
[04:49:01] <Skullworks-PGAB> http://www.skullworks.net/include/DSCF0272.JPG
[04:49:36] <Skullworks-PGAB> Haas VF-0E fed by robot
[04:49:40] <skyfox00> ok that looks like fun!
[04:49:44] <fenn> i recognize that robot
[04:49:57] <Skullworks-PGAB> just another job I have to setup
[04:50:11] <fenn> http://fenn.dyndns.org/sketches/emc-splash/Puma.png
[04:50:29] <skyfox00> does the robot feed a lathe?
[04:51:24] <Skullworks-PGAB> no - it loads rollers between 4th axis centers to have the drive feature added ( flats or drill & tap )
[04:52:15] <Skullworks-PGAB> fenn : these are Stauubli robots
[04:52:40] <Skullworks-PGAB> We did have some pumas - those are much older
[04:52:49] <Skullworks-PGAB> pumas are older
[04:52:54] <skyfox00> oh, on a mill?(I dont know what a 'Haas VF-03' is)
[04:52:55] <fenn> its not actually a puma, but the same kinematics
[04:53:16] <fenn> i just thought real pumas were ugly
[04:53:54] <Skullworks-PGAB> also the machine next to this one has a GMF / Carrol Fanuc with yello cap 4 brush motors
[04:54:00] <fenn> looks like that e-itx board should have a blower, not an axial fan
[04:54:16] <Skullworks-PGAB> yes - even the paint color was ugly
[04:54:56] <Skullworks-PGAB> these Staubli are scary fast
[04:55:30] <Skullworks-PGAB> we had a open house for halloween last year
[04:55:54] <fenn> and you put costumes on the robots and had the guests walk through the haunted factory?
[04:56:52] <Skullworks-PGAB> robot would reach down inside a bucket in a lathe, grab some candy, whip around and toss it into the kids pail that was sitting in a little circle on the floor
[04:57:08] <skyfox00> lol!
[04:58:01] <Skullworks-PGAB> think they were jumbo tootsie rolls
[04:58:09] <skyfox00> where does your company operate?
[04:58:35] <Skullworks-PGAB> company firing me is based out of MN
[04:59:34] <Skullworks-PGAB> new company is 1/3 local here in Colorado Springs, 1/3 a spinoff of Minolta, 1/3 in china
[04:59:45] <skyfox00> oh, I thought the US outsourced all manufacturing years ago... nice to know there behind schedule...
[04:59:56] <Skullworks-PGAB> No china owners
[05:00:18] <Skullworks-PGAB> 60% Japan 40% US
[05:00:39] <Skullworks-PGAB> Japan group own 4 factories in china
[05:01:33] <skyfox00> I am a bit biased, but I say we need to stop all the outsourcing and bring manufacturing back to the US before everybody forgets how to make stuff...
[05:01:37] <Skullworks-PGAB> part of why they bought us was to help fix all the quality issues in the china plant
[05:01:59] <Skullworks-PGAB> China still has not learned good tool and die
[05:02:18] <Skullworks-PGAB> so things don't work as envisioned
[05:02:47] <skyfox00> and while they learn good tool and die, we forget... madening I tell ya....
[05:03:24] <Skullworks-PGAB> yep - our best old timers are helping design tools for them
[05:03:51] <skyfox00> * skyfox00 wishes he could learn machining....
[05:04:04] <Skullworks-PGAB> machining will be a cottage industry outside of defense work in 20 years
[05:04:37] <Skullworks-PGAB> Production will move offshore - OSHA and others will force that
[05:05:13] <skyfox00> time to string OSHA up with some bright orange safety tape...
[05:06:45] <Skullworks-PGAB> I'd settle for some budget cuts, enough to be sure they don't leave the lights on in the restrooms ( at OSHA)
[05:07:04] <Skullworks-PGAB> (make it hurt)
[05:07:50] <skyfox00> and we all know the red comunists who we call the government would never do that...
[05:08:29] <Skullworks-PGAB> likewise gut the "clean air act" - while it was a good idea - many of the current regs cause more damage than good - like oxy-fuel
[05:08:54] <Skullworks-PGAB> oxy-fuel made cars with carburators run cleaner
[05:09:58] <Skullworks-PGAB> but the whole Oxly fuel thing is redundant with computer controlled fuel injection systems that use oxygen sensors.
[05:10:25] <Skullworks-PGAB> not that many cars with carbs now days
[05:10:36] <skyfox00> well, when the government breaks up due to total economic destruction, we can always work up from being cavemen agian... re-envent government as a non-opressive entity.
[05:12:46] <skyfox00> Thanks agian for all the help and cya later!
[05:12:51] <Skullworks-PGAB> well - bed time for me - still going to milk the OT from the old company - another 12hr shift comming up.
[05:22:33] <Jymmm> Skullworks-PGAB: Mooooooooooooooo
[05:22:56] <Jymmm> Skullworks-PGAB: Just dont milk the cow with horns
[05:37:58] <user_> http://www.youtube.com/watch?v=N8TXMUaC9Os&NR=1
[05:38:18] <Jymmm> O_o
[06:00:23] <skyfox00> does anyone recognize this emc2-cvs compile error: WARNING: "rt_task_wait_period" [/EMC2/emc2-trunk/src/rtapi.ko] undefined!
[06:02:45] <skyfox00> it claims it compiled and the file is there, so never mind I guess...
[06:21:44] <toast> blur
[06:21:49] <toast> wtf
[06:21:53] <toast> where did that come from
[06:22:11] <fenn> deep within your subconscious
[06:22:32] <toast> frightening
[06:22:58] <Jymmm> toast: Not really, we've grown used to it.
[12:15:24] <cradek_> cradek_ is now known as cradek
[12:42:19] <skunkworks> so my 89 honda has just enough brake fluid resevoir that the brake lite comes on when the pads are almost to the wear tab.
[12:42:36] <skunkworks> must be a feature by design ;)
[12:46:26] <jlmjvm> i am trying to do some of the pyvcp examples from the wiki,near the end of the page it says pyvcp wont be released untill 2.2,i
[12:46:31] <jlmjvm> 2.2
[12:46:55] <jlmjvm> does that mean the examples wont work with ver 1.7?
[12:48:54] <jepler> in the end, we ported an early version of pyvcp into emc 2.1.x. The one that will be in 2.2.0 has many more capabilities, and when you give its .xml files to the one in 2.1.7 it will not do what you want.
[12:52:47] <jlmjvm> i was able to make the big dro work,but when i tried the panel with led indicator,no luck
[12:54:26] <jepler> this documentation should match what is possible in 2.1.7's pyvcp: http://linuxcnc.org/docs/html/hal/vcp/index.html
[12:55:04] <jlmjvm> k,thats the page i was reading
[12:55:04] <jepler> I tried the first example with one label and one led on emc 2.1.7's pyvcp and it worked
[12:55:32] <jlmjvm> k,that lets me know the problem is on my end
[12:55:52] <jlmjvm> as so often the case,lol
[12:56:28] <jepler> I guess the advice on how to run it (just 'pyvcp ...') is not right, though
[12:56:32] <jepler> you need to: halrun
[12:56:40] <jepler> then at the halcmd prompt, 'loadusr pyvcp ...'
[12:59:25] <jlmjvm> i did that and it says this halcmd: Usage: pyvcp -c hal_component_name myfile.xml
[13:00:38] <jlmjvm> but no panel,i put the stuff in a file called tiny.xml,does it need to be changed to myfile.xml?
[13:01:00] <jepler> halrun
[13:01:00] <jepler> halcmd: loadusr pyvcp -c mypanel tiny.xml
[13:01:00] <jepler> halcmd: pyVCP: Creating widgets from tiny.xml ... Done.
[13:01:13] <jepler> ... and the panel popped up above my terminal ...
[13:03:20] <jepler> the "-c mypanel" is required; if I leave it out I get the error message you did.
[13:04:12] <jepler> one last thing -- when you want to incorporate this into a .hal file, you have to add even more words to the command:
[13:04:16] <jepler> halcmd: loadusr -Wn mypanel pyvcp -c mypanel tiny.xml
[13:04:19] <jepler> Waiting for component 'mypanel' to become ready......pyVCP: Creating widgets from tiny.xml ... Done.
[13:04:22] <jepler> halcmd:
[13:04:44] <jepler> "loadusr -Wn mypanel" forces halcmd to wait (-W) for a new component named mypanel (-n mypanel) so that a following 'net' or 'link' statement can succeed
[13:11:11] <jlmjvm> every time i reopen terminal and type halrun it says realtime already running
[13:11:24] <jlmjvm> then command not found
[13:12:06] <jlmjvm> is there a command to stop realtime from terminal
[13:12:12] <jepler> if you left emc or halrun going in another terminal, or made one of them exit in an odd way, you will get that message
[13:12:24] <jepler> "halcmd unload all" followed by "/etc/init.d/realtime stop"
[13:14:23] <jlmjvm> hey,it worked
[13:14:40] <jlmjvm> have a led
[13:15:10] <jlmjvm> thanks jepler
[13:15:32] <jepler> you're welcome
[13:16:33] <jlmjvm> now i can work on the other examples
[13:18:24] <jepler> great
[14:06:38] <tomp> the mesa boards 7i33, 7i37 have a T variant, and that means "The -T version uses 3.5 mm screw terminal compatible plugs". This sounds like it would eliminate the need for breakout boards. anybody use 'em?
[14:07:11] <SWPadnos> yes
[14:07:15] <skunkworks> that came out after I had orderd a few :)
[14:07:23] <SWPadnos> beware the cost and accessibility of the terminal plugs
[14:07:25] <cradek> beware, they don't come with the screw terminals, only the socket to plug them in. after getting the mesa board, you have to go hunt for the right connectors.
[14:07:48] <SWPadnos> each 24-position plug is over $20 (there are two per board)
[14:07:56] <skunkworks> pete doesn't have a good source?
[14:08:07] <SWPadnos> they'd still be in the $20 range
[14:08:31] <SWPadnos> unless he got thousands
[14:09:17] <SWPadnos> there are also accessibility problems. the two terminal strips are only 3/16 " apart - there's no room for the wires to exit the inner connector if you want them both to face the close edge of the board
[14:09:44] <SWPadnos> there are left, top, and right facing plugs, so that problem can be worked around
[14:10:34] <SWPadnos> I suggested to Pete that he use the right-angle header for the outer connector, but I don't know if he
[14:10:39] <SWPadnos> he'll do that
[14:29:17] <tomp> thanks
[17:24:21] <adam2> Hello
[17:29:47] <skunkworks> hello
[17:29:54] <skunkworks> how is the conversion coming?
[17:30:16] <skunkworks> if you are the adam I am thinking of..
[17:57:47] <xemet> hi
[17:58:24] <alex_joni> hi
[17:58:28] <xemet> just compiled the trunk version run in place, I'm trying to use one of my configuration files that work with the official version
[17:58:33] <xemet> and I get this error
[17:58:49] <xemet> Traceback (most recent call last):
[17:58:50] <xemet> File "/home/mill/develop/emc2-nurbs/bin/axis", line 3260, in ?
[17:58:50] <xemet> root_window.tk.call("set_slider_min", step_size*30)
[17:58:50] <xemet> _tkinter.TclError: domain error: argument not in valid range
[17:59:01] <xemet> any suggestion?
[17:59:30] <xemet> with the sim/axis configuration everything works
[18:00:10] <alex_joni> wonder if that isn't related to INPUT_SCALE vs. SCALE
[18:00:40] <alex_joni> try changing INPUT_SCALE to SCALE in your ini file
[18:00:45] <xemet> ok
[18:01:50] <xemet> nope
[18:01:50] <xemet> HAL:70: Ini variable not found.
[18:01:51] <xemet> HAL config file /home/mill/emc2/configs/COBRA/core_stepper.hal failed.
[18:01:51] <xemet> Shutting down and cleaning up EMC2...
[18:02:06] <xemet> maybe I've to change in the core stepper
[18:02:10] <xemet> the name
[18:02:29] <jepler> if you change the inifile you have to change references in the hal file as well
[18:02:58] <jepler> but it should not be necessary to do that -- only if I introduced a bug in axis with a recent change ..
[18:03:07] <jepler> and it's better for me to fix the bug than to require everyone to change .ini files
[18:04:38] <xemet> nope, if I change INPUT_SCALE to SCALE in ini and core_stepper, everything remains the same
[18:04:38] <xemet> Traceback (most recent call last):
[18:04:39] <xemet> File "/home/mill/develop/emc2-nurbs/bin/axis", line 3260, in ?
[18:04:39] <xemet> root_window.tk.call("set_slider_min", step_size*30)
[18:04:39] <xemet> _tkinter.TclError: domain error: argument not in valid range
[18:05:22] <alex_joni> wonder why sim/axis works then?
[18:05:33] <xemet> don't ask to me :)
[18:06:57] <jepler> mind putting your inifile on pastebin?
[18:08:11] <jepler> ah I bet you have a negative SCALE
[18:09:08] <xemet> yes...negative
[18:09:18] <alex_joni> the abs got lost in the change :)
[18:09:31] <jepler> I just committed a fix .. update again
[18:09:40] <CIA-8> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: work properly for negative scales
[18:09:57] <xemet> http://www.pastebin.ca/694266
[18:13:16] <alex_joni> http://www.lizo.de/Japaner_0461248ee351d
[18:28:08] <acemi> http://www.osnews.com/story.php/18596/QNX-Opens-Neutrino-Source-Code-Interview-QNX-CEO/
[18:29:11] <alex_joni> this is so freakin cool http://www.lizo.de/Crazy_Drummer_5432490c97381
[18:30:16] <xemet> ok, now it works
[18:30:22] <xemet> thanks
[18:46:03] <adam2> skunkworks - I am wiring up my Mesanet boards, Building a pendant, and waiting on a VFD.
[18:47:13] <jlmjvm> alex joni: i have an encoder on my z axis and its hooked up to my breakout board with a and b channel,is there a place on the wiki that deals with adding an encoder
[18:47:29] <jlmjvm> wanting to get ferror
[18:48:02] <alex_joni> jlmjvm: search the configs for the component encoder
[18:48:17] <alex_joni> you need to loadrt encoder num_chan=1 (if my memory serves me right)
[18:48:32] <alex_joni> then add the 2 functions of encoder to base_thread and to servo_thread
[18:48:40] <alex_joni> you need to get the scale right
[18:48:59] <alex_joni> and the output from encoder needs to get connected to the axis feedback position
[18:51:37] <jlmjvm> configs?do you mean where the hal and ini file is?dont have encoder component in my configs,
[18:52:15] <jlmjvm> may not be looking in the right place
[18:54:31] <skunkworks> jlmjvm: look at the etch-servo hal for a starting point for setting up the encoder module
[18:54:43] <jlmjvm> will do
[18:57:50] <jlmjvm> looks like what i need to get started,thanks
[19:09:59] <skunkworks> could I get a hal module written for this? http://www.electronicsam.com/images/imageset.jpg
[19:10:26] <skunkworks> sorry - phone camera
[19:15:51] <alex_joni> skunkworks: heh
[19:19:28] <cradek> what is it?
[19:24:25] <jlmjvm> heres my hal file,crashes when emc starts
[19:24:30] <jlmjvm> http://pastebin.ca/694392
[19:25:33] <jlmjvm> im only using 1 encoder right now,and the a and b pins are right
[19:25:43] <cradek> uh what's the error?
[19:26:03] <cradek> HAL:20:error message here
[19:26:06] <cradek> (20 is the line number)
[19:26:30] <jlmjvm> want me to paste error msg
[19:26:36] <alex_joni> encoder.2 is wrong
[19:26:42] <alex_joni> you need encoder.0
[19:26:55] <alex_joni> and loadrt encoder num_chan=1 is enough for now
[19:27:24] <alex_joni> if you want encoder.2.* then you need num_chan=3 (0,1 and 2)
[19:27:59] <jlmjvm> my z axis is number 2
[19:28:09] <jlmjvm> encoder on z axis
[19:30:33] <skunkworks> cradek: it is a interface card for an old image setter
[19:31:43] <jepler> jlmjvm: 'Z axis' doesn't mean you automatically put .2. everywhere
[19:32:07] <jepler> jlmjvm: in this case, this is the first and only thing you are hooking up to 'encoder', so you use encoder.0
[19:32:34] <jlmjvm> just tried it with 0
[19:33:01] <jlmjvm> still no go
[19:33:13] <jepler> yeah but the error is almost certainly different now :-P
[19:33:27] <jlmjvm> prolly so
[19:33:42] <jlmjvm> want me to post it
[19:34:26] <jepler> well that's the only way anybody here is going to be able to help you
[19:38:51] <skinnypuppy1334> Hey guys are there any other IRC channels where machinists hang out regularly?
[19:39:17] <anonimasu> there's #machinist too, but there arent lots of people around there
[19:40:26] <skinnypuppy1334> right about not a lot of people there
[19:40:44] <anonimasu> well, it's the only place I know
[19:40:53] <anonimasu> cnczone probably has a chat somewhere..
[19:40:58] <anonimasu> if you like white noise
[19:41:11] <tomp2> the drummer juggler is great
[19:41:19] <anonimasu> drummer juggler?
[19:41:24] <skunkworks> I have never been in the cnczone chat when other poeple where actually there
[19:41:41] <skinnypuppy1334> I hadn't either but it was a long time ago
[19:41:47] <anonimasu> I expect it to be like the rest of cnczone
[19:42:29] <tomp2> is 'white noise' racist?
[19:42:49] <skinnypuppy1334> Forums are cool, but I always feel like I am chasin my tail..
[19:43:00] <jepler> Results 1 - 10 of about 159,000 for "white noise" racist
[19:43:03] <jepler> </google>
[19:43:21] <anonimasu> um
[19:43:25] <skinnypuppy1334> Is pink noise feminist ?
[19:43:27] <anonimasu> white noise as the stuff you see on tv..
[19:43:40] <anonimasu> random stuff with no coorelation with anything else
[19:44:37] <anonimasu> tomp2: I wouldnt know, about them, it's all random noise.. ;)
[19:44:37] <tomp2> wow, its a genre, and a musical
[19:44:46] <anonimasu> heh, that's more then I knew
[19:46:06] <adam2> Question: I am unsure of the pinout for my glass scales, I am trying to wire them up to my MeSA 7I33. Any suggestions for testing this?
[19:46:21] <anonimasu> use a oscilloscope
[19:46:46] <skinnypuppy1334> Jeez I was just looking for a replacement scale for my sargon/enco
[19:47:23] <anonimasu> or contact the maker.. and ask for a pinout
[19:47:37] <jlmjvm> http://pastebin.ca/694421,thats the changes i made
[19:47:59] <jlmjvm> how do you send an error,cant copy
[19:48:37] <tomp2> somebody check me on these alternative names... the 7i33 can accept 'RS244" (aka differential ) and TTL ( aka single ended )
[19:50:09] <skinnypuppy1334> adam2 do you know about the cadcamdro list on yahoo?
[19:53:59] <alex_joni> jlmjvm: start emc from a terminal
[19:54:06] <alex_joni> then you can select and copy the error
[19:54:36] <alex_joni> jlmjvm: but the error is probably :
[19:54:42] <alex_joni> newsig Zpos-fb float
[19:54:46] <jlmjvm> ahhh
[19:54:47] <alex_joni> that one should already be defined
[19:54:49] <SWPadnos> tomp2, RS422
[19:55:04] <alex_joni> jlmjvm: the linksp should be enough
[19:55:12] <adam2> skinnypuppy no i dont, what is it?
[19:55:36] <alex_joni> adam2: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/
[19:58:41] <skunkworks> hmm - for about a $1198 you loose 3-4 inches of table travel.. :) http://littlemachineshop.com/video/ToolChangeDemo.mpg
[19:59:01] <skinnypuppy1334> Is that the spider looking tool changer?
[19:59:02] <skunkworks> cool though
[20:00:58] <skunkworks> sort of a tool chain - maybe and upside down spider
[20:01:32] <skinnypuppy1334> I'll have to look at it later, no codecs on the 6.06 box
[20:01:51] <alex_joni> apt-get install w32codecs
[20:04:34] <skinnypuppy1334> alex, thanks would you look at this pastebin ?http://pastebin.com/m7d2a9cee
[20:04:51] <jlmjvm> http://pastebin.ca/694451
[20:05:01] <jlmjvm> finally got the error msg
[20:05:31] <alex_joni> jlmjvm: don't see the actual error
[20:06:20] <alex_joni> skinnypuppy1334: did you add a repository?
[20:06:22] <jlmjvm> k,on the way,that was the dmesg
[20:06:28] <alex_joni> skinnypuppy1334: try http://onlyubuntu.blogspot.com/2007/03/install-mplayer-and-multimedia-codecs.html
[20:07:53] <skinnypuppy1334> Alex, thanks. I didn't add repo, which is the w32 coming from?
[20:08:42] <alex_joni> deb http://mirror.ubuntulinux.nl dapper-seveas all
[20:09:01] <alex_joni> skinnypuppy1334: it's in the previous link I posted
[20:09:04] <jlmjvm> http://pastebin.ca/694459
[20:09:15] <jlmjvm> that should be the error
[20:09:45] <JymmmEMC> Greetings people of Emc!
[20:09:46] <alex_joni> jlmjvm: I already told you that you need to remove the newsig
[20:09:52] <skinnypuppy1334> Yeah those aren't in my sources, thank you
[20:10:18] <jlmjvm> gives another error when i did
[20:10:39] <alex_joni> jlmjvm: what's the error?
[20:10:42] <alex_joni> already has a writer?
[20:11:17] <jlmjvm> lemme remove the newsig and resend error
[20:11:26] <alex_joni> jlmjvm: no need
[20:11:31] <alex_joni> remove the newsig
[20:11:38] <alex_joni> then open core_stepper.hal in the same dir
[20:11:46] <alex_joni> and look for the following line 'linksp Zpos-fb <= stepgen.2.position-fb
[20:11:53] <alex_joni> and comment that line out
[20:13:36] <jlmjvm> k,will give it a try
[20:15:06] <skunkworks> your going to have to scale the encoder module correctly to work correctly with you input scale
[20:16:01] <skunkworks> I think I could have squeezed another 'correctly' into that sentence
[20:16:40] <jlmjvm> that worked
[20:17:20] <jlmjvm> emc came on and if i turn the motor the numbers change
[20:17:28] <skunkworks> Sweet :)
[20:17:31] <jlmjvm> motor poer is off
[20:17:36] <jlmjvm> power
[20:17:54] <jlmjvm> where do i set the cpr amount at?
[20:18:18] <jlmjvm> 1024cpr quadrature
[20:18:27] <jlmjvm> 20480 counts per inch
[20:21:33] <skunkworks> scale I think
[20:22:09] <SWPadnos> it's up to you actually
[20:22:10] <skunkworks> setp encoder.0.position-scale
[20:22:21] <SWPadnos> you can stick a line like that in the .hal file
[20:22:22] <skunkworks> setp encoder.0.position-scale 20480
[20:22:40] <SWPadnos> you can use any name you want in the ini file, and reference that in the .hal file
[20:22:44] <skunkworks> I ment scale pin on the encoder.
[20:22:48] <SWPadnos> yes
[20:23:01] <SWPadnos> so setp encoder...scale 20480 is one way
[20:23:03] <skunkworks> SWPadnos: he is doing steppers with encoders.
[20:23:08] <SWPadnos> rigth
[20:23:10] <SWPadnos> right
[20:23:19] <jlmjvm> right
[20:23:30] <skunkworks> so he has the scale in the ini for the stepper which will be different than his encoder scale.
[20:23:42] <skunkworks> * skunkworks is thinking out loud
[20:23:45] <SWPadnos> you can add any variables you want to the ini, and reference tem in the HAL file, so it really depends on how you want to set it up
[20:23:55] <skunkworks> really - cool
[20:24:03] <SWPadnos> so you can add a section called [MY_STEPPER_ENCODER_FEEDBACK]
[20:24:04] <skunkworks> really? - cool
[20:24:29] <SWPadnos> and a variable in it called THE_NUMBER_OF_COUNTS_PER_INCH_ON_THE_X_AXIS
[20:24:45] <jlmjvm> so add that setp line to my hal file?
[20:25:01] <SWPadnos> and you can set that by using "setp encoder.0.scale [MY_STEPPER_ENCODER_FEEDBACK]THE_NUMBER_OF_COUNTS_PER_INCH_ON_THE_X_AXIS" in the hal file
[20:25:08] <jlmjvm> setp encoder.0.position-scale 20480
[20:25:21] <SWPadnos> that'll work too, of course :)
[20:30:16] <skunkworks> hmm - wasn't there an issue with the stepgen with using encoders as feedback - like it doesn't reset it position to the 'encoders' position after an estop or something like that?
[20:30:22] <jlmjvm> changed the hal file,still have joint error,was i supposed to put something in the ini file also?
[20:30:48] <skunkworks> what is the error? following error?
[20:30:55] <jlmjvm> yep
[20:31:19] <jlmjvm> encoder is doing its job already
[20:32:18] <skunkworks> with it off - does the readout count correctly? like 20 turns equal an inch on the readout?
[20:32:35] <SWPadnos> are they going in the same direction?
[20:32:36] <skinnypuppy1334> That is a cool little tool changer, hadn't seen that one .
[20:32:39] <jlmjvm> i can check
[20:32:59] <skinnypuppy1334> This is the one I considered spider looking,
[20:33:01] <skinnypuppy1334> http://home.insightbb.com/~joevicar3/default.htm
[20:33:40] <skunkworks> SWPadnos: did you read what I wrote above about stepgen?
[20:33:58] <SWPadnos> yes. that could be a problem, I don't know
[20:34:42] <jepler> this is interesting: http://home.insightbb.com/~joevicar3/Air_Spindle_Plans.htm
[20:35:10] <jepler> I wonder how fast it spins .. doesn't seem to say
[20:36:39] <SWPadnos> he does say 50 IPM taking 0.06" cuts, but he doesn't say how many flutes the cutter is
[20:36:56] <SWPadnos> or if "deep" means chip load or cut depth
[20:37:16] <cradek> guys I have some kind of tapping head that appears to just be a clutch with settable max torque. It also has a (tiny) bit of give axially. Is that used by just feeding too slow for the rpm and letting it constantly slip?
[20:37:38] <cradek> (it seems like that can't possibly work very well)
[20:37:55] <anonimasu> cradek: isnt it a reversing head?
[20:37:56] <toast> the float is to keep the tap from snapping, just like in a cnc floating tap holder
[20:37:58] <SWPadnos> you're supposed to feed at the right rate, but if there's a mismatch, the tap will wait for the quill to catch up
[20:38:15] <cradek> anonimasu: no reversal stuff as far as I can tell
[20:38:20] <anonimasu> ok
[20:38:26] <SWPadnos> what brand?
[20:38:28] <toast> cradek: does the entire head rotate
[20:38:34] <jlmjvm> counting correctly
[20:38:36] <toast> or is there an apparatus to keep part of the head stationary
[20:38:46] <cradek> SWPadnos: but if you feed a tiny bit too fast it would smash it into the thread right?
[20:38:53] <cradek> the whole head rotates, there is no arm
[20:39:04] <SWPadnos> yes, I think that's probaby true
[20:39:07] <toast> then there's no reversing.
[20:39:19] <SWPadnos> right - need a static reference to reverse
[20:39:38] <toast> cradek: the tap will pull itself along
[20:39:51] <toast> no feed necessary
[20:39:59] <toast> in a BP, anyway
[20:40:13] <cradek> I think it only moves axially a tiny bit (.020)
[20:40:14] <anonimasu> toast: you mean pull the quill down?
[20:40:16] <toast> the problem is when it bottoms, it needs slack to absorb the inertia of the spindle
[20:40:19] <cradek> it's possible it's broken or something
[20:40:20] <toast> anonimasu: yes
[20:40:24] <anonimasu> ouch
[20:40:25] <SWPadnos> you probably want to program some feed with a CNC BP, else PID will try to stop the tap from moving :)
[20:40:27] <toast> and to get that clutch slipping
[20:40:27] <anonimasu> that's a f-ugly thing
[20:40:28] <cradek> oh it's cnc
[20:40:44] <cradek> yeah it won't pull anything along on this machine!
[20:40:47] <toast> ugly doesn't have much to do with how well something works
[20:41:01] <anonimasu> toast: um, that's f-ugly..
[20:41:08] <toast> and it f-works
[20:41:22] <SWPadnos> that's f-ine f-or you f-olks
[20:41:26] <skunkworks> jlmjvm: so the second you turn the machine on - you get a following error - or is it when you start to move?
[20:41:41] <toast> cradek: cnc tap holders need a lot more play than .020
[20:42:00] <cradek> toast: maybe it's not meant for this machine. I don't know anything about it really
[20:42:04] <JymmmEMC> * JymmmEMC lol @ SWPadnos
[20:42:05] <jlmjvm> when i start to move
[20:42:13] <cradek> I'll just wait and use rigid tapping with emc
[20:42:18] <skunkworks> so when you start to jog - it will error?
[20:42:24] <jlmjvm> yes
[20:42:27] <skunkworks> I would negate the scale on the encoder
[20:42:31] <toast> floating tap holders usually have like .100-.250 float in either direction
[20:42:43] <cradek> pretty sure this one doesn't have that
[20:42:51] <SWPadnos> jlmjvm, you may need to fiddle with the stepgen accel and vel limits
[20:43:04] <cradek> it seems to be just a clutch, you're supposed to set it with a torque wrench
[20:43:11] <skunkworks> setp encoder.0.position-scale -20480 - or does it count the right direction?
[20:43:11] <jepler> or use halscope to display what is happening with commanded & feedback positions during the short time you try to move -- that will give you a better idea what is going on
[20:43:17] <toast> cradek: that's probably just a manual tapping head
[20:43:23] <toast> for a drill press or BP type arrangement
[20:43:31] <cradek> toast: gotcha
[20:43:40] <skunkworks> jepler: what? systematic aproach? bah
[20:43:41] <jlmjvm> yes
[20:43:41] <skunkworks> ;)
[20:43:42] <cradek> it's QC30, I bet I could sell it to a manual BP user
[20:43:49] <toast> probably
[20:43:54] <SWPadnos> me, me!
[20:43:58] <SWPadnos> oh wait, I have R-8
[20:44:03] <SWPadnos> (and a set of tpamatics :) )
[20:44:11] <SWPadnos> and tapmatics, which work much better
[20:44:25] <toast> i'd like a collapsing tap head
[20:44:27] <toast> for work
[20:44:40] <skunkworks> jlmjvm: yes what?
[20:44:52] <SWPadnos> I'd like it if MTD would send me my f-ing servo mounts
[20:44:57] <jlmjvm> yes its reading the right way
[20:45:25] <skunkworks> ah - ok. Jepler really has the correct way to see what is going on. use halscope.
[20:45:32] <jlmjvm> and my rapids are slow right now,60 ipm
[20:45:47] <SWPadnos> try moving much much slower than that
[20:45:49] <skunkworks> what if you jog slow - like 1ipm or such
[20:45:56] <jlmjvm> same thing
[20:46:03] <cradek> jepler: I think 3CFM @ 90 needs a non-toy compressor
[20:46:04] <jlmjvm> tried that
[20:46:13] <SWPadnos> use MDI to do something like G1X1F10
[20:46:27] <SWPadnos> 3SCFM @ 90 = $100 or so compressor
[20:46:41] <cradek> SWPadnos: I think they're more even at HF
[20:46:46] <jlmjvm> was doing this same thing yesterday on mach and it could read at 100 ipm
[20:46:48] <skunkworks> jlmjvm: what is your following errir set to in the ini?
[20:46:57] <jlmjvm> lemmee see
[20:47:19] <SWPadnos> I got one at an auto parts stor (slightly on sale) for $89, it says 4.2 SCFM @ 90 PSI
[20:47:35] <cradek> huh ok
[20:47:35] <SWPadnos> it also says 3.5HP, but plugs into a 120V outlet, so it's probably wrong
[20:47:44] <cradek> that's those fake HPs
[20:47:56] <jlmjvm> ferror=.050 minferror=.010
[20:47:56] <SWPadnos> and probably fake SCFMs too
[20:48:02] <cradek> "4HP shop vac" :-)
[20:48:19] <skunkworks> I would think that would be more than enough.
[20:48:28] <skunkworks> ferror that is.
[20:48:37] <SWPadnos> jlmjvm, you have a setup error, and you'll need to use the diagnostic tools to figure out what that error is
[20:48:46] <jepler> cradek: wasn't 3@90 what that harbor freight 56krpm grinder required?
[20:48:56] <cradek> I thought it was 2@90
[20:49:04] <SWPadnos> halscope is probably the best tool at the moment, because at the moment, nobody knows exactly what's going on
[20:49:18] <SWPadnos> and without knowing what's happening, it's hard to fix it
[20:49:32] <cradek> Micro design requires only 2 cubic feet per minute at 90 PSI to deliver 56,000 RPM.
[20:49:53] <jepler> I stand corrected
[20:49:58] <jlmjvm> wasnt i supposed to add something to the ini file to point to the encoder scale?
[20:50:07] <SWPadnos> jlmjvm, not necessary
[20:50:20] <SWPadnos> but you should make sure the scale is correct before going on
[20:50:38] <skunkworks> it seems by your account - it is working correctly - 20 turns of the leadscrew is 1 inch.
[20:50:47] <skunkworks> in the display
[20:50:59] <SWPadnos> ie, manually turn the motor shaft/screw to move the table 1 inch, and verify that the display shows 1 inch of movement
[20:51:21] <jlmjvm> its doing that
[20:51:29] <cradek> does it go the right way?
[20:51:32] <jlmjvm> yes
[20:51:47] <SWPadnos> ok, and if you don't use the encoder feedback, but use the stepgen feedback, can you jog?
[20:51:57] <SWPadnos> and when you do that, does it jog in the correct direction?
[20:52:19] <jlmjvm> yes,was moving it yesterday
[20:52:20] <SWPadnos> you can experiment a little by setting ferror very large - like 1 inch
[20:53:08] <cradek> are you sure software counting can keep up with your encoder rate?
[20:53:10] <jlmjvm> will try,but was i supposed to add something to the ini file?
[20:53:13] <SWPadnos> just to be sure, you are exiting from EMC and restarting every time you change anything in the hal or ini files, right?
[20:53:25] <jlmjvm> yes
[20:53:33] <SWPadnos> ok - just had to check ;)
[20:53:38] <jlmjvm> no prob
[20:53:43] <jepler> jlmjvm: no, you did not need to modify the .ini file if you put the encoder scale explicitly in the .hal file
[20:53:53] <SWPadnos> you have the option of using an ini file setting for the scale (or anything else), but it's not required to use one
[20:53:53] <jlmjvm> k
[20:54:10] <jlmjvm> just wanted to make sure
[20:54:23] <jlmjvm> lemmee open up the ferror setting
[20:55:54] <skunkworks> open up both of them.
[20:56:42] <adam2> I called and got the pin-out for my glass scales: 1-A Channel, 2-B Channel, 3-+5v, 4-Common, 5-Ground, 6-Blank. Whats the deal with common and ground?
[21:00:35] <SWPadnos> the A and B channels may be either (a) isolated from the power supply or (b) common + instead of common -
[21:01:38] <jlmjvm> its working
[21:01:56] <jlmjvm> some big dummy had the pins backwards
[21:02:16] <SWPadnos> heh
[21:02:32] <jlmjvm> i swear when the power was off the motors it was reading right
[21:02:47] <SWPadnos> it has to be right, and also it has to match the motor
[21:03:00] <SWPadnos> it's more important for it to match the motor
[21:03:11] <SWPadnos> (as far as following errors go)
[21:03:14] <jlmjvm> swapped the a and b pin # and its working
[21:03:30] <SWPadnos> ok. you could also have negated the scale as skunkworks suggested
[21:03:37] <skunkworks> jlmjvm: negating the encoder scale I think would have done the same thing..
[21:03:38] <skunkworks> ;)
[21:04:01] <skunkworks> * skunkworks thinks now jlmjvm should now play with adaptive feed controll....
[21:04:28] <jlmjvm> lets rock n roll,lol
[21:04:34] <SWPadnos> * SWPadnos thinkd skunkworks needs to take his meds
[21:04:41] <SWPadnos> err - thinks
[21:04:43] <jlmjvm> i really appreciate the help from everyone
[21:05:06] <SWPadnos> any time. just remember to put it on the wiki if you discover anything major about steppers+encoders :)
[21:05:07] <adam2> Does anybody know if the Mesa electronics 7I37 card can output less than 48vdc? for I/O commands?
[21:05:15] <skunkworks> jlmjvm: you really need to post this on cnczone. This is always brought up.
[21:05:16] <jlmjvm> im the only guy on the block with encoders on his steppers
[21:05:25] <jlmjvm> really
[21:05:42] <skunkworks> So - now if you stall - do you get a following error?
[21:05:49] <jlmjvm> yall can have a copy of my hal and ini if you want
[21:06:00] <SWPadnos> no thanks. I'd rather have you put it on the wiki ;)
[21:06:07] <jlmjvm> im sure it will
[21:06:09] <SWPadnos> hint hint
[21:06:29] <jlmjvm> lemmee give it a big accel and stall it
[21:06:32] <jlmjvm> brb
[21:06:48] <jlmjvm> me,on the wiki
[21:07:04] <SWPadnos> just set the axis vel limit to 1000IPM and try G1X10F500
[21:07:15] <SWPadnos> but set the ferror back down first :)
[21:07:49] <skunkworks> he should be able to measure the following error - and as it increases slow the feed rate down using adaptive feedrate.. I think it would be a cool experiment..
[21:12:55] <jlmjvm> oh yeah,jacked the acceleratio up from 6 to 60 and it gives an immediate joint following error,doesnt even move
[21:13:41] <jlmjvm> thats gotta be working
[21:14:42] <jlmjvm> plus its showing up when i home out the axis,-.0002,.0001 etc
[21:15:21] <jlmjvm> i think this is gonna really help
[21:15:25] <skunkworks> cool
[21:15:51] <jlmjvm> im ready to have a keg party now,this was magor to me
[21:16:18] <skunkworks> what base period are you running?
[21:16:22] <jlmjvm> 10000
[21:17:17] <jlmjvm> its reading at 60 ipm no problem
[21:17:41] <skunkworks> yah - you should be able to do 150ipm no problem
[21:17:41] <jlmjvm> should be able to get 100 ipm like i did in mach yesterday
[21:17:58] <jlmjvm> 150 would be awesome
[21:18:08] <skunkworks> what is your input scale.
[21:18:11] <skunkworks> ?
[21:18:22] <SWPadnos> remember - you can output at least twice as many steps as you can read
[21:18:23] <jlmjvm> the cool thing is i had these encoders all along
[21:18:29] <jlmjvm> 10000
[21:18:35] <skunkworks> the encoder counting is going to be your real limmit
[21:19:01] <SWPadnos> you probably won't be able to count past ~90 IPM reliably
[21:19:03] <skunkworks> SWPadnos: explain? you mean doing doublefreQ?
[21:19:09] <jlmjvm> yes,they were higher count than i wanted,but they were on the motors i bought
[21:19:13] <SWPadnos> yes, doublefreq is part of it
[21:19:36] <SWPadnos> but you need some margin for reading quadrature that isn't needed for step generation
[21:19:42] <jlmjvm> and they were 250.00 encoders,wish they had an index
[21:19:48] <skunkworks> AH - but he should be able to count at 50khz shouldn't he? with base period set to 10000?
[21:20:02] <SWPadnos> at most
[21:20:09] <skunkworks> atleast ;)
[21:20:16] <SWPadnos> oh - I think I had an extra factor of two in my head
[21:20:18] <skunkworks> so that is 146ipm
[21:20:21] <jlmjvm> 100 is prolly all i will go
[21:20:52] <SWPadnos> no, at most. if latency is 0 and the encoder has perfect 90 degree phases, then 50 KHz is doable
[21:21:34] <JymmmEMC> and at 60Hz?
[21:21:34] <SWPadnos> if latency isn't zero, then the least time the encoder needs to be at one state is 10000 + latency to guarantee that the software sees it
[21:21:52] <JymmmEMC> oh, nm didn't see the K
[21:21:53] <SWPadnos> if the encoder isn't perfect, then that restriction is on the shortest phase
[21:23:06] <SWPadnos> so if it's got a "short pulse" in one quadrature state - say only 75 electrical degrees instead of 90, then that short state needs to be BASE_PERIOD + latency_max long to be noticed
[21:23:55] <SWPadnos> in that example, 75 degrees is ~1/5 of a full cycle, so a cycle of 4 counts = 5 *(base_period+lat_max)
[21:24:03] <skunkworks> huh
[21:24:08] <SWPadnos> heh
[21:24:20] <SWPadnos> so it works great, until you start looking at the fact that nothing is perfect ;)
[21:24:33] <skunkworks> heh
[21:24:48] <jlmjvm> alright lets get started on that adaptive feedrate,lol
[21:24:52] <skunkworks> heh
[21:25:00] <jlmjvm> just kidding
[21:25:03] <SWPadnos> opk - time for me to get back to paying work! ;)
[21:25:30] <jlmjvm> thanks for the help
[21:25:35] <skunkworks> I could have swarn though that I was counting encoders close to base period... but I don't know for sure..
[21:25:45] <skunkworks> sworn
[21:26:04] <skunkworks> but that is an argument for another time.. :)
[21:26:05] <SWPadnos> you could have been
[21:26:10] <jlmjvm> thats how we say it down south
[21:26:20] <SWPadnos> ahh coulda swahn
[21:26:33] <jlmjvm> lol
[21:27:16] <skunkworks> SWPadnos: this is all I have http://www.electronicsam.com/images/KandT/servostart/Fastermouse.png
[21:28:53] <skunkworks> jlmjvm: atleast get a wiki up and I will post it on cnczone
[21:29:01] <jlmjvm> btw,couldnt this same scenario be used with a gecko g320 setup by also connecting the a and b encoder wires to a 2nd breakout board also
[21:30:15] <jlmjvm> how do you get a wiki up,never done 1,and no smart remarks,lol
[21:30:34] <jlmjvm> never posted before
[21:30:54] <jlmjvm> still in intense learning mode
[21:34:38] <jlmjvm> back in a bit,gotta eat supper
[21:39:24] <alex_joni> jlmjvm: glad it works
[21:39:48] <jlmjvm> thank alex
[21:39:54] <jlmjvm> thanks
[21:40:04] <alex_joni> jlmjvm: the wiki is pretty straightforward
[21:40:21] <jlmjvm> gonna check it out
[21:40:33] <alex_joni> first thing read & try to follow this page: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?BasicSteps
[21:40:52] <jlmjvm> i think this is an awesome safety feature
[21:40:57] <jlmjvm> will do
[21:41:28] <alex_joni> once set up right, you can have unstallable steppers :P
[21:41:48] <alex_joni> basicly if the stepper starts to lag behind, you adjust the feedrate to go slower
[21:41:53] <jlmjvm> just like marris is doing
[21:41:58] <alex_joni> yup
[21:42:03] <jlmjvm> sweet
[21:42:21] <alex_joni> if you prove it working :)
[21:42:32] <jlmjvm> that would be awesome
[21:42:45] <alex_joni> we all think it should, but so far no one bothered enough to try it out *blush*
[21:43:05] <jlmjvm> i will try it
[21:43:48] <jlmjvm> gotta get my 2nd paraport installed tomorrow
[21:44:38] <jlmjvm> then i can wire it all correctly,was outta inputs,had to disable y limit switch
[21:45:37] <jlmjvm> 2 ports,many inputs
[21:46:18] <alex_joni> I would personally try to get my hands on some hardware counter for the encoders
[21:46:42] <alex_joni> I'd trust one a bit more than software counting
[21:47:25] <jlmjvm> may look in to that later
[21:48:17] <jlmjvm> just glad its working
[21:49:56] <jlmjvm> alex joni:shouldnt i be able to do this with gecko servos also,run a and b to gecko and second parport?
[21:50:32] <jlmjvm> brb
[21:55:06] <JymmmEMC> SWPadnos: have you hacked any GPS modules/antennas by chance?
[21:56:23] <JymmmEMC> SWPadnos: I was just wondering if I unsoldered the ceramic antenna from the PCB, and tossed in some RG-58 or RG-213 if I'd fubar anything?
[21:57:24] <SWPadnos> JymmmEMC, I wouldn't bet on the antenna being 50 (or 75) ohm
[21:57:30] <SWPadnos> it may be, bit you should check first
[21:57:36] <SWPadnos> s/bit/but/
[21:58:01] <JymmmEMC> SWPadnos: I was hoping it would be 50ohm
[21:58:47] <JymmmEMC> no identifying marks on it, but maybe I can find the specs. it be a pain to find a balance
[21:58:48] <SWPadnos> well, as oen of the guys in my national guard unit used to say "shit in one had and wish in the other, and see which fills up first"
[21:58:51] <SWPadnos> one
[21:59:08] <JymmmEMC> lol
[21:59:29] <SWPadnos> yeah - that's why I've remembered it for 15 years :)
[22:00:04] <JymmmEMC> Ok, if I decide to do this as a test, is there any way for me to check once I've unsoldered everything?
[22:00:17] <JymmmEMC> I got crap for T/E
[22:00:18] <SWPadnos> only with a spectrum and/or network analyzer
[22:00:22] <SWPadnos> then no
[22:00:29] <JymmmEMC> I was afraid you were gonna say that =)
[22:00:43] <JymmmEMC> The one I WANT is only $68,000
[22:00:49] <SWPadnos> it's OK. the newest spectrum analyzers are cheap - only around $9k
[22:01:02] <JymmmEMC> HP baby!
[22:01:09] <JymmmEMC> err Agenlent
[22:01:10] <SWPadnos> Agilent, thank you
[22:01:13] <SWPadnos> heh
[22:01:21] <JymmmEMC> HP to me still =)
[22:01:40] <archivist> I have one with HP on the front
[22:02:04] <JymmmEMC> I'll have to check the specs, to see if the module has an antenuation var I can pole.
[22:02:42] <JymmmEMC> see if belden.com is back online
[22:02:55] <JymmmEMC> fsck no
[22:03:06] <JymmmEMC> ColdFusion.... wth?!
[22:04:04] <cradek_> cradek_ is now known as cradek
[22:27:28] <dmess> Hi all... sure am glad those weren't MY dash 8 legs having problems in sweden and denmark
[23:02:21] <skinnypuppy1334> Twingy, I've been playin with Gcam. Pretty cool
[23:03:05] <Twingy> next release will have a few more bugs worked out
[23:03:16] <Twingy> gerber -> g-code is much improved too
[23:03:26] <Twingy> graduate school is eating up all my time though
[23:03:28] <skinnypuppy1334> Cool, do you have a set date?
[23:03:51] <Twingy> think there's one minor bug I need to take care of... soon as I get 2-3 hours free
[23:04:08] <Twingy> you can grab a sneak peak if you checkout subversion repo
[23:04:37] <skinnypuppy1334> I'll probably just wait for the iron out.
[23:04:41] <Twingy> kk
[23:04:56] <Twingy> next major features are 4th axis support and dxf/stl + contour pocketing
[23:05:13] <skinnypuppy1334> :0)
[23:05:30] <skinnypuppy1334> is 4th going to be rotary?
[23:05:48] <Twingy> those still require a fair bit of planing, but contour pocketing code is almost identical to the isolationing code I wrote for RS274X importer
[23:05:58] <Twingy> yes, rotary 4th axis, i.e. lathe
[23:06:40] <Twingy> before I jump into that stuff I want to take a step back and optimize a few sections of the code
[23:06:51] <Twingy> I'm a big code optimization zealot
[23:07:27] <Twingy> it'll probably be 12-18 months before said features are available
[23:07:38] <Twingy> the project has been going for about 18 months now
[23:08:31] <skinnypuppy1334> It's pretty usefull, kudos
[23:08:59] <Twingy> for basic stuff, yep, for complex stuff you'll probably want to use solid works
[23:09:46] <skinnypuppy1334> I use mastercam X @ school for anything too hard.
[23:10:32] <Twingy> another good app
[23:11:46] <skinnypuppy1334> Runs the hell outta the heidenhein bp, I haven't tried any McamX posts in emc but should be attempting soon.
[23:19:40] <skinnypuppy1334> Speaking of heidenhein, is there an emc gcode that will allow incremental rotation around a set pole? Like g10/g11 heidenhein?
[23:29:31] <JymmmEMC> Ok, I must be clueless today, this ceramic patch antenna only has one connection to it. Am I missing something here?
[23:29:37] <JymmmEMC> 25mm http://www.emtac.com/products/ceramic_patch_antenna/index.html
[23:33:27] <fenn> dont antennae usually have one connection?