#emc | Logs for 2004-10-23

[00:12:56] <paul_c> * paul_c reads the latest post from Jon Elson on rotary axis configs
[02:11:15] <les> hey Paul and all
[02:11:23] <les> breaking for lunch
[02:11:59] <paul_c> Hi Les
[02:12:46] <les> I saw a post to the user list just now...someone needs to tell these guys not to worry about Ferror settings on stepper emc
[02:14:27] <paul_c> True if it wasn't for the PID loop
[02:15:13] <les> Also found that the local isp here (I don't use them) is charging $100-$800/month for web hosting!!!
[02:16:18] <les> so they are just setting up $9 ipowerweb accounts and "administering" for all the extra money
[02:16:23] <paul_c> low cost hosting round here is about �10/month
[02:16:49] <les> Well these guys haven't paid their rent in 3 months
[02:16:58] <paul_c> If you have a static IP address, you can do it for much less.
[02:17:01] <les> I may grab there customer base
[02:17:48] <les> No...heres the plan... I will help them set up an Ipowerweb acount with the $9 billed to them
[02:18:11] <les> Show them how to do an FTP transfer...
[02:18:55] <les> and setup or transfer their web site for a flat fee of about $1000
[02:19:14] <les> It will save them loads of money
[02:19:43] <les> I need to find a kid or something who is an HTML geek NOW
[02:20:01] <les> ...Trying to get my son to do it
[02:21:02] <les> He is good at HTML but considers it boring and doesn't like it
[02:22:23] <les> oh and back to the Ferror... just set V feedforward at 1 and throw big numbers in the virtual PID right?
[02:22:43] <les> big Ferror numbers
[02:26:04] <paul_c> Could do with some PID values so that unity gain is achieved
[02:27:08] <les> yes I guess...never use the stepper stuff
[02:28:31] <les> Oh and also...Paul Hamler is visiting again a did a very nice stylus drag brass "chips" engraving with a powered by emc logo
[02:28:50] <les> Some folks might like one
[02:29:06] <paul_c> Don't know the name...
[02:29:23] <les> Famous miniature tool maker
[02:29:38] <les> remember?
[02:30:05] <paul_c> The only one I know is Barry Jordan
[02:30:17] <les> The 7 inch tall working Bridgeport miller
[02:31:36] <les> http://www.craftsmanshipmuseum.com/Hamler.htm
[02:33:58] <an0n_> * an0n_ has been learning how to TIG weld today
[02:35:06] <les> Just don't get bombarded by too many alpha particles from the thorated tungten tig tip :)
[02:35:21] <an0n_> heh..
[02:35:26] <an0n_> I wouldnt worry too much about that
[02:35:47] <les> naw just a joke...radiation is very low
[02:36:05] <an0n_> yep .)
[02:36:08] <an0n_> it's very hard tho
[02:36:28] <les> But the thorium reduces the work function of the metal a bunch I guess
[02:37:08] <an0n_> * an0n_ has no idea
[02:37:49] <paul_c> http://www.sherline.com/jmfJor03.htm
[02:38:46] <les> GAWD
[02:38:55] <an0n_> I just care about making neat welds ;)
[02:39:02] <an0n_> in alu or stainless
[02:39:11] <les> yup
[02:39:15] <les> tig is the one
[02:39:30] <les> I only have MIG here
[02:39:42] <les> um I mean GMAW
[02:39:57] <an0n_> I only have a shitty little mig at home.. :)
[02:40:04] <les> our Miller and Hobart divisions like to call it that
[02:40:28] <les> mine is 170 amp and not very good
[02:40:31] <an0n_> the tig is too expensive for me to have as a toy :)
[02:40:45] <les> yup
[02:40:59] <an0n_> I think the torch costs more then my mig :)
[02:41:19] <an0n_> although the tig is at my father's company so it dosent matter that much...
[02:41:30] <les> I have a box of tweeko tig torched somewhere
[02:41:54] <an0n_> :)
[02:42:12] <les> Had a gig to do some signal analysis of the arc for Miller
[02:42:34] <les> for their computer controlled supplies
[02:42:35] <an0n_> nice :)
[02:42:46] <les> was fun
[02:43:28] <an0n_> now I just need some stainless steel pipes, so I can weld a nice exhaust for myself..
[02:43:45] <an0n_> and well, a couplf of 100 hours until the welds start to look ok
[02:43:45] <les> I could use that too
[02:44:24] <les> well back to work for me (after I finish lunch)
[02:44:28] <an0n_> yeah
[02:44:29] <les> later!
[02:44:31] <an0n_> lates
[02:44:33] <an0n_> err
[02:44:34] <an0n_> laters
[02:44:51] <paul_c> A couple of tips when weldin 1.6mm wall s/s tube
[02:45:27] <paul_c> use 1mm dia wire, and run argon through the pipe to purge it
[02:47:00] <an0n_> got that tip off the instructor, however it's a e
[02:47:08] <paul_c> Oh, and stainless steel has a habit of cracking when used on an exhaust system
[02:47:09] <an0n_> exhaust system, I dont care if the welds get blue-ish
[02:47:38] <an0n_> it depends on which kind dosent it?
[02:47:56] <paul_c> * paul_c used 316
[02:48:29] <an0n_> and if you use filler :)
[02:49:03] <paul_c> the filler wire was 316 as well...
[02:49:51] <an0n_> what kind does the exhaust shops work with?
[02:50:59] <paul_c> the two common grades of stainless steel are 303 & 316
[02:51:19] <paul_c> both suffer from cracking
[02:51:35] <an0n_> I read somthing about what to use earlier
[02:52:02] <an0n_> for headers it was..
[02:52:07] <an0n_> err extractors..
[02:53:10] <an0n_> 2333
[02:53:42] <an0n_> or 2343, 2353
[02:54:06] <paul_c> these are DIN numbers ?
[02:54:48] <an0n_> I think so
[02:55:10] <an0n_> http://www.engineeringtoolbox.com/6_32.html
[02:55:16] <an0n_> there you have the conversion
[02:55:54] <paul_c> Looking in a Sandvik bok...
[02:56:03] <paul_c> Swedish numbers
[02:56:09] <an0n_> :)
[02:56:14] <an0n_> yeah
[02:56:17] <an0n_> brb... need to eat a bit
[02:56:36] <paul_c> 2333 is BSi 304
[02:56:54] <paul_c> and 2343 is BSi 316
[02:57:47] <paul_c> 316 (2343) is the standard grade for food & chemical work
[02:58:10] <paul_c> and the easiest to find in thin wall tubing.
[03:11:05] <asdfqwega> Ooo
[03:11:21] <asdfqwega> I have the script finished
[03:11:32] <asdfqwega> Now, I need something to preview the output
[03:11:54] <asdfqwega> Too bad I've never gotten emcplot3d to work
[03:12:17] <asdfqwega> Anyone care to preview my .ngc file and send me a picture?
[03:12:37] <paul_c> DCC it over...
[03:12:45] <asdfqwega> Haha very funny
[03:12:59] <asdfqwega> Remember last time?
[03:13:13] <asdfqwega> I don't think I'm able to send out over DCC
[03:13:27] <asdfqwega> but let's try anyway
[03:13:35] <paul_c> I had indeed forgotten..
[03:16:23] <asdfqwega> Will you take a .tgz through email?
[03:16:44] <paul_c> yes
[03:16:54] <asdfqwega> addy?
[03:17:35] <asdfqwega> address?
[03:18:08] <paul_c> Will have to look after Tea.
[03:19:43] <asdfqwega> 37kb on it's way
[03:20:58] <asdfqwega> * asdfqwega needs to go to the hardware store for some setscrews.
[03:21:43] <asdfqwega> Paul, thanks. When you get to it, you can just dcc the preview to me, or email it back.
[03:59:58] <paul_c> * paul_c runs grapres.ngc
[04:00:07] <paul_c> Hi Ray
[04:00:16] <rayh> Hi Paul
[04:00:22] <rayh> How you doing?
[04:01:18] <paul_c> Waiting for a file to transfer
[04:01:47] <rayh> I found a small issue with the mode widgets. They don't want to pack correctly.
[04:02:10] <rayh> We are going to have to get you a t1.
[04:02:31] <paul_c> drool....
[04:02:48] <rayh> Tell me!
[04:02:50] <paul_c> with a spur to Crystal Falls I reckon.
[04:03:06] <rayh> You bet. Direct.
[04:03:24] <rayh> You seen mike around at all today?
[04:03:36] <paul_c> * paul_c prods asdfqwega
[04:03:58] <paul_c> Not seen or heard Mike for a while
[04:06:55] <paul_c> The two tcl scripts live in emc/tcl ?
[04:07:48] <rayh> Yep. I tried to weed out a lot of the extra libraries by sticking stuff right into smini.
[04:08:05] <rayh> The only source is buttonde now.
[04:08:23] <rayh> Didn't see a good reason to leave it three directories deep.
[04:08:35] <paul_c> OK, so all the tcl/stkemc/ is redundant then ?
[04:09:07] <rayh> There are a few references to scripts but I plan to take all of them out as well.
[04:09:27] <rayh> Not needed with this app.
[04:09:50] <paul_c> OK, I'll leave the stuff in place and let you delete when you are ready.
[04:10:38] <rayh> The only thing I'm thinking is bringing IO_Show.tcl forward alongside these two. The rest can go.
[04:11:42] <rayh> That should clean up 500k or so.
[04:14:27] <rayh> If we leave the tickle stuff in emc/src/emcmot we can always troubleshoot from tkemc.
[04:18:16] <paul_c> OK - (most) of the tarball added to cvs
[04:19:44] <rayh> Okay.
[04:20:40] <paul_c> Just the run file ommitted
[04:20:48] <rayh> I discovered last night or early this morning that the command to move files wrapped the text. Damned unhandy with tickle code.
[04:21:26] <paul_c> you mean scp ?
[04:21:29] <rayh> I suppose I should have made a tarball on 32 and then passed that here. We get to soon old and to late smart.
[04:21:33] <rayh> Yep.
[04:22:05] <paul_c> it shouldn't have done...
[04:23:09] <rayh> Tried it several times with the same result.
[04:23:35] <rayh> perhaps there was some sort of filtering going on between.
[04:24:44] <paul_c> perhaps the file was trashed at the other end - scp should treat all files the same.
[04:25:59] <rayh> Might be. I never did find the stuff that I was running while I was there.
[04:29:02] <paul_c> Is there a backplot in smini ?
[04:29:29] <rayh> Yes. You're 3d version with four preset angles.
[04:29:43] <rayh> Pretty much the same as mini.
[04:31:02] <paul_c> Got it..
[04:39:47] <paul_c> One comment to make on smini
[04:43:27] <rayh> Go. Got the nomex handy.
[04:43:52] <paul_c> No A axis readout
[04:44:10] <rayh> There is if you define a in the ini.
[04:44:53] <paul_c> Doh.
[04:44:54] <rayh> It stretches that top panel down a bit. I think that I'll have to make a little more space for the position displays.
[04:45:39] <rayh> It certainly does not handle extra axes as well as mini does.
[04:47:04] <paul_c> I see 0,0,0 is set at 0.25" from the limit switches
[04:47:24] <paul_c> Smithy's decision ?
[04:47:29] <rayh> Right. That is just about the diameter of the pickup ring.
[04:47:49] <rayh> on the prox.
[04:49:01] <rayh> a rotary will have to have a much larger number in there. It will depend upon where they place the prox for a.
[04:52:18] <paul_c> Getting a bad return from emcAxisSetCycleTime as soon as I enable A axis
[04:53:31] <rayh> With the ini I sent?
[04:53:41] <paul_c> yes
[04:55:25] <acemi> acemi is now known as acemi_
[04:55:52] <rayh> Don't think that's right.
[04:55:59] <rayh> Let me look a bit.
[04:57:17] <acemi_> acemi_ is now known as acemi
[04:59:54] <rayh> I lifted the rotary pretty much from my Sherline but it looks like I didn't finish it up right. Let me send a new ini.
[05:18:52] <learath> lilyj: richlowe is trying to make me cry
[06:03:39] <acemi> acemi is now known as acemi_
[06:26:17] <asdfqwega> paul_c: You prodded, sir?
[06:27:02] <paul_c> That was hours ago...
[06:27:22] <paul_c> A suggestion for your script
[06:27:28] <asdfqwega> Yes, I always read the backlog
[06:27:32] <asdfqwega> I'm listening
[06:27:43] <paul_c> Reduce the M03/M05 calls
[06:28:38] <asdfqwega> That's easy to say, but that's the nature of turning a raster image into gcode
[06:29:31] <asdfqwega> Did it cause a problem?
[06:29:53] <paul_c> yup
[06:30:15] <asdfqwega> There are guys who do what I'm doing, but they use a trip switch on their Z-axis to do the same thing :)
[06:31:12] <paul_c> Are you using M03/M05 to control a laser ?
[06:31:19] <asdfqwega> yes
[06:32:01] <paul_c> In that case, you'd be wanting to use another M code
[06:32:27] <asdfqwega> what do you suggest?
[06:33:04] <asdfqwega> I know that the bridgeportio has a wait_time setting - and I have mine set to 0
[06:33:21] <asdfqwega> I thought about using M07/M09
[06:33:26] <paul_c> Trying to remember what the functions are...
[06:33:50] <asdfqwega> M07 - mist, M08 - flood, M09 - coolant off
[06:37:34] <asdfqwega> So what happened with my program? I know all the little pauses would add up fast...
[06:37:49] <asdfqwega> Unless it choked the interpreter
[06:40:10] <paul_c> OK, found it...
[06:41:11] <paul_c> to turn a bit ON on the parallel port, use M62 P[0-11]
[06:41:30] <paul_c> and OFF, M63 P[0-11]
[06:41:58] <paul_c> however, bits 0 to 7 will be used by the stepper motors...
[06:42:13] <paul_c> so P is in the range of 8 to 11
[06:42:35] <asdfqwega> I have 2 parport on a card - and bridgeportio has spare output on the second port
[06:43:07] <paul_c> The advantage of M62/M63 is it is coordinated with motion
[06:43:21] <asdfqwega> where can I read about M62/M63?
[06:43:36] <paul_c> so M62 will turn a bit ON at the start of a move,
[06:43:43] <paul_c> and M63 OFF at the end.
[06:44:45] <asdfqwega> That would definately help prevent that little bit of dwell time...which is bad, with a laser
[06:44:47] <paul_c> Not sure if the Handbook has been updated, but I do remember adding some notes to the LyX master some time back.
[06:45:28] <paul_c> Search the archives @ SF for a post from Fred Proctor
[06:54:22] <paul_c> The only mention of M62/M63 in the Handbook(s) is one short paragraph.
[06:55:55] <les> I was not aware of M62/M63...looks handy
[06:56:30] <paul_c> Also M64/M65 for non-realtime IO
[06:56:40] <paul_c> and M101-M199
[06:57:15] <les> so 62 and 63 are RT?
[06:57:57] <paul_c> The ON/OFF is syncronised in realtime
[06:58:09] <les> cool
[06:59:00] <asdfqwega> checking out cvs of documents/ <--- didn't know it was there
[07:01:19] <paul_c> You'll need LyX >= Ver 1.3
[07:01:24] <asdfqwega> 62/63 are RT?!
[07:01:45] <asdfqwega> * asdfqwega 's mind starts to whirl with possibilities...
[07:01:54] <les> yeah...
[07:02:14] <paul_c> I'll try to expand on earlier comments...
[07:02:40] <paul_c> M62 P8 G1 X2.75 F100
[07:03:03] <paul_c> Will place two commands on the RT motion stack...
[07:03:15] <paul_c> Turn ON Bit 8
[07:03:31] <paul_c> Move X2.75 @ F100
[07:04:08] <paul_c> Bit 8 will not turn ON untill the X move commences
[07:04:44] <asdfqwega> Do this still work with the limit of 4 m-codes per block/line?
[07:05:07] <asdfqwega> * asdfqwega starts thinking about 4-bit output
[07:05:09] <les> hmmm large format emc inkjet printer....
[07:05:16] <paul_c> Good question...
[07:06:20] <asdfqwega> I'm going to have to hack a SMPS to give my laser variable control...
[07:06:56] <paul_c> There is still a limit of 4 M codes per line
[07:07:22] <paul_c> but that can be changed (at compile time) if need be.
[07:11:07] <les> ...black and white photography on wood with a defocused beam (with some gamma correction)
[08:17:51] <asdfqwega> paul_c: in your example - M62 P8 G1 X2.75 F100 - is the bit cleared when the motion stops?
[08:18:28] <asdfqwega> If not, I can see how I would need to use M63 in a line following it
[08:26:00] <paul_c> The bit is not turned OFF untill you use an M63
[08:26:58] <asdfqwega> I see. This will work perfectly for me
[08:27:15] <asdfqwega> A quick mod of my script, and I'll try it out
[08:35:39] <paul_c> Eeeww... Kernel panic on a new test build..
[16:55:08] <asdfqwega> I'm off to see the wizard.
[21:28:47] <alex_joni3> hello paul
[21:38:32] <alex_joni3> alex_joni3 is now known as alex_joni
[21:44:35] <paul_c> Hi Alex
[21:45:07] <alex_joni> what's up?
[21:50:59] <paul_c> I need a faster connection...
[21:54:50] <alex_joni3> sorry... my internet connection is pretty crappy
[21:54:52] <alex_joni3> :(
[21:55:34] <paul_c> I know the feeling
[21:55:47] <paul_c> Eight hours to upload a file last night...
[21:56:14] <alex_joni3> wasting a lot of time for trivial things...
[21:59:12] <alex_joni3> spotted a possible error in the lathe fork
[21:59:27] <alex_joni3> on BDI2.18
[21:59:55] <alex_joni3> same on emc2 HEAD branch...
[22:00:08] <alex_joni3> Compile 'rtl_ulapi.c' to tmp dir using non-realtime C rule
[22:00:19] <alex_joni3> -I-I/usr/src/rtlinux-3.0/linux/include
[22:02:31] <paul_c> Is that in the Makefile.inc ULFLAGS ?
[22:02:59] <alex_joni3> I think so... I just spotted it from the compile_farm logs
[22:03:09] <alex_joni3> I don't have BDI2.18 myself...
[22:03:23] <alex_joni3> however it seems to work ....
[22:04:27] <paul_c> It should read " -I. -I/usr/src/blah/blah"
[22:06:06] <alex_joni3> -I. follows later, maybe one -I is extra ;)
[22:06:46] <alex_joni3> just tried autoconf, ./configure
[22:06:50] <alex_joni3> works like a charm here
[22:07:09] <alex_joni3> 2.4.21-adeos with RTAI-3.0
[22:08:21] <alex_joni3> is php acceptable for the web-page?
[22:08:28] <alex_joni3> or should I stick with plain html?
[22:09:31] <paul_c> As long as you don't use java or flash
[22:10:17] <alex_joni3> no way ;)
[22:10:36] <alex_joni3> php is easier to admin...
[22:10:49] <alex_joni3> I only use <include's> and such...
[22:14:28] <alex_joni3> another thing about BDI2.18...
[22:14:58] <alex_joni3> building seems to fail because of X11 libs tcl and tk.lib depend on
[22:15:00] <alex_joni3> -L/usr/X11R6/lib -lX11 -ldl
[22:15:12] <alex_joni3> those seem to be missing from the linking
[22:15:44] <alex_joni3> they should have been extracted from tclConfig.sh and tkConfig.sh...
[22:18:03] <paul_c> yup
[22:19:32] <alex_joni3> alex_joni3 is now known as alex_joni
[22:20:33] <alex_joni> I don't have tcl8... so I can't really test it
[22:21:40] <paul_c> what distro are you running ?
[22:24:56] <alex_joni> SuSE
[22:25:01] <alex_joni> tcl8.4
[22:51:00] <alex_joni> I gotta run to work.. finish packing for tomorrow..
[22:51:09] <alex_joni> but I'll be back later on IRC
[22:52:37] <alex_joni> bye