#emc | Logs for 2008-09-02

[00:00:03] <jmkasunich> hot dang - I just found something I didn't even know I had
[00:00:18] <jmkasunich> in Dad's toolbox, a 1/8" dia x 2" long carbide boring bar
[00:00:46] <jmkasunich> can't bore a 1/8" hole with a 1/8" bar, but it will do nicely for 3/16
[00:00:49] <stustev> be very careful - breaks very easily
[00:01:20] <jmkasunich> yep
[00:01:59] <jmkasunich> strange plan for holding it - I have a boring block, fits on top of my toolpost, and has 1/2" and 3/4" holes in it
[00:02:19] <jmkasunich> I also have an ER collet chuck with a 3/4" straight shank, and a 1/8" collet
[00:02:31] <stustev> that will do it
[00:04:27] <jmkasunich> descisions decisions - do I get sidetracked converting this one to 3/16" or do I stick with the program and work on the 1/8 that I actually need
[00:04:44] <jmkasunich> should do the latter I think
[00:05:39] <stustev> when you get to the same point with both it is a simple matter to bore 1/8 and then 3/16 - just before you slit them
[00:05:59] <jmkasunich> yeah
[00:06:11] <jmkasunich> assuming I do the 1/8 the same way I did the first one
[00:06:40] <jmkasunich> I don't have anything that will bore that small, even my smallest endmills are 1/8"
[00:06:45] <stustev> do you need a 5/32 and 7/32 also :)
[00:06:52] <jmkasunich> no
[00:07:11] <jtr> will making the 3/16 help with technique for the 1/8?
[00:07:23] <stustev> find a dull (broken) 1/8 end mill and grind clearance on one side
[00:07:46] <jmkasunich> jtr: depends - if I do the 1/8 by making the hole first and then turning the OD concentric to the hole, no
[00:07:57] <jmkasunich> if I make the outside first, and then bore it, yes
[00:08:19] <jmkasunich> the other problem with boring is it will be only as concentric as my lathe spindle, which isn't great
[00:08:21] <jtr> gotcha
[00:09:05] <stustev> what is the collet going in?
[00:09:16] <jmkasunich> if I put it on an arbor with a center hole, I can do the work down near the tailstock on a dead center
[00:09:28] <jmkasunich> 400W 48,000 RPM spindle
[00:09:53] <jmkasunich> originally a die-grinder kind of thing, I'm hoping to do both toolpost ID grinding, and fine milling like PC boards
[00:09:58] <stustev> I see - you cannot drill and bore and ream the hole in the spindle you will use it in
[00:10:23] <jmkasunich> I can turn the spindle down some, maybe to 10,000 RPM
[00:10:25] <jmkasunich> http://ukptocs.bosch-pt.com/boptocs-uk/Product.jsp;jsessionid=0EFEF165822E3CFA01AD71A6326F7580?division=iw&ccat_id=9749&prod_id=6328
[00:10:38] <jmkasunich> that is the spindle
[00:11:17] <jmkasunich> and this is the business end: http://jmkasunich.com/pics/bosch-collet-1-2611.jpg
[00:11:37] <JymmmEMC> Anyone know anything about AMD CPU's? 1.90 GHz AMD Athlon 64 X2 Dual-Core Processor TK-57 --versus-- 1.90 GHz AMD Athlon X2 QL-60 Dual-Core Processor
[00:12:14] <jmkasunich> darn - even at 10K and 1/8", that is 327 sfpm
[00:12:21] <jmkasunich> I'd need carbide to bore it
[00:13:58] <stustev> I bet a high speed end mill will bore it taking a very light cut and not very deep - you could then ream it by turning the spindle by hand or rig up a pistol drill and rubber band
[00:14:29] <stustev> bbl - got to go
[00:14:31] <jmkasunich> can I do "floating" reaming, where I hold the reamer in my fingers and let it self-align?
[00:14:37] <stustev> no
[00:48:53] <jepler> hm, I see that "split nut" is not what I meant ... jtr is probably right
[01:04:47] <stustev> there is more than one type of 'split nut' - one with an axial split - clamped tighter to reduce the pitch diameter
[01:05:19] <stustev> and another that is actually two nuts mounted together to be able to adjust the pitch between them
[01:05:22] <cradek> that's how the plastic acme nuts I have work
[01:05:45] <stustev> plastic works well that way - it will conform nicely
[01:06:59] <stustev> the cincikins comps on all positions on the table
[01:07:07] <cradek> yay
[01:07:17] <stustev> I am a happy camper
[01:07:33] <stustev> I will post video tomorrow
[01:07:44] <cradek> neat
[01:08:09] <stustev> I am working on interp_convert.cc to get the W and Z working like I want
[01:08:39] <cradek> I saw that when I read back - not sure what the right answer is
[01:09:00] <stustev> :) the right answer is 'what I want' :)
[01:09:01] <cradek> the interp itself really doesn't know about TLO
[01:09:12] <cradek> I mean the best way to get what you want
[01:09:18] <stustev> oh
[01:09:40] <stustev> rats - I want cheap - easy - fast
[01:10:50] <stustev> the interp make decisions about w-tool-length so it has some knowledge of TLO
[01:11:16] <cradek> yep I was wrong
[01:11:38] <stustev> I will mark this day on the calendar :)
[01:11:55] <cradek> the TLO is in the interp state (settings->tool_Xoffset)
[01:12:38] <stustev> I can see this "should" work pretty easy with A0 B0
[01:13:10] <cradek> are you going to have convert_tool_length_offset() issue an extra move?
[01:14:12] <stustev> no I was planning on changing the assigns at the end of the function according to the status of the w-tool-length
[01:15:35] <cradek> I have a bad feeling about that
[01:15:54] <stustev> an extra move may be a better deal - the way it works now is you command a -w and +z and the machine does not move just the displays. I haven't tried this with AB not zero. I will try that tomorrow
[01:16:21] <stustev> you can g90 or g91 and g01 or g00
[01:16:36] <stustev> the machine does not move - just the displays
[01:16:37] <cradek> right, with AB zero, it won't move because Z,W are the same direction
[01:16:51] <cradek> with A or B nonzero you will get motion
[01:17:09] <stustev> probably
[01:17:16] <stustev> I will try it and see
[01:17:42] <stustev> I am going to the shop right now to try it - bbl
[01:17:52] <cradek> heh
[01:17:55] <cradek> don't you ever go home?
[01:21:53] <jmkasunich> I wonder if "to the shop" was "from my office" or "from home" ?
[01:22:16] <cradek> iirc, shop and home were not at all near each other
[01:22:21] <jmkasunich> ah
[01:45:30] <stuste1> i was at home
[01:47:04] <stuste1> I got ZX motion with B at -30 - the axes moved over and down and then back to the same vector with the W and Z exchanged
[01:48:16] <stuste1> the way it is now I can make it work but I will need to know the exact tool length and have that value in the program - if I change tools I will have to change the program.
[01:49:24] <stuste1> if I can make it work with A0B0 that will work for now. I will work on being able to do offset changes at any angle later
[01:49:35] <cradek> I don't see why that would be
[01:50:28] <cradek> if you load the tool somewhere above the work, then apply G43, then back to the work with G0 X... Y... Z0 W0, you'll get straight motion that applies the offset during the move
[01:50:53] <cradek> the program doesn't need to know the tool length
[01:51:53] <stuste1> yes - I agree - won't I get motion if I don't match the tool length to the motion command - I will try it
[01:52:16] <cradek> yes you will get motion, but it will end up with the tool tip in the right place
[01:58:15] <stuste1> yes - I get motion - for a normal mill with more Z travel and usually shorter tools this is not a problem. This machine has 14 inches of Z travel and many times I have tools more than 10 inches long in it. This requires three or four commands to move Z down a little bit, move W up a little bit, move Z down and then move Z up again.
[01:58:34] <stuste1> the tool tip is in the right spot then but it has been a hassle
[01:58:47] <cradek> no, maybe I am missing something, but if you move them together it is not a problem
[01:58:59] <cradek> say clearance plane is Z0.1
[01:59:14] <cradek> if you G43 then move to Z0.1 W0, it will go there in one shot, in a straight line
[02:00:30] <stuste1> I will try it - btw: feed hold doesn't work in MDI - how do I stop an MDI command?
[02:00:50] <cradek> hitting escape will abort it
[02:01:02] <stuste1> ok
[02:01:11] <stuste1> bbiam
[02:20:20] <jtr> jepler: if you can add a second block and nut insert, you will have the second type of split nut stustev mentioned. you'll have differential threads, too.
[02:21:54] <stuste1> yes - it goes to Z0.1 in a straight shot - canceling is a little different - if your Z move is not more than the tool length the motion is down - not a good thing - if your Z move is too much the machine doesn't move and gives an axis over travel warning
[02:23:15] <stuste1> having the axes invert position by the exact and opposite magnitude is much better - no motion
[02:23:51] <stuste1> cradek: enough about me - how is your lathe coming this evening?
[02:24:19] <cradek> going ok - I made a few more test parts
[02:24:39] <stuste1> are these imaginary or do you have pictures?
[02:24:50] <cradek> it was fun to see it run a whole program - drill, bore, turn, part
[02:25:04] <cradek> they are very uninteresting - aluminum rings
[02:25:36] <stuste1> I like watching lathe work the best - seeing the shape of the parts emerge from the stock is nice
[02:25:37] <cradek> also it extends the part chute, parts, part drops onto ways below part chute, then the part chute retracts
[02:25:53] <cradek> it is adjustable :-)
[02:26:03] <stuste1> isn't the chute supposed to catch the part?
[02:26:17] <cradek> no I think it's just for decoration, while the part drops next to it
[02:26:38] <stuste1> nice - I have seen lathes like that before
[02:27:18] <jmkasunich> argh
[02:27:20] <cradek> internal and external indexable threading kits cost a fortune.
[02:27:28] <cradek> uh-oh
[02:27:48] <stuste1> yeah - I heard comments like that before
[02:27:56] <stuste1> jmk - what happened
[02:28:00] <stuste1> ?
[02:28:07] <jmkasunich> I made three blanks with nice accurate 0.125 holes in them - that part is fine and I haven't ruined them yet
[02:28:11] <jmkasunich> then I started making an arbor
[02:28:28] <jmkasunich> got it almost turned down to 0.126, and I oopsed and the tool dug in
[02:28:36] <jmkasunich> now I have a nice bent arbor, and need to start over
[02:29:00] <stuste1> that is for non concentric parts
[02:29:49] <jmkasunich> I'll save it for the next time I have a batch of those
[02:29:56] <stuste1> paste an insert on it and you will have a fly cutter - always a positive
[02:30:10] <stuste1> :)
[02:30:39] <jmkasunich> got any inserts that will fit on a 0.140 diameter bar?
[02:31:04] <jmkasunich> the ease with which that bent doesn't fill me with confidence in this approach
[02:31:21] <cradek> 1/8 is pretty small
[02:31:26] <jmkasunich> yeah
[02:31:34] <stuste1> it will bend just as easy when it is carrying the collet blank
[02:31:39] <jmkasunich> exactly
[02:31:51] <stuste1> will you have a center supporting it on the end?\
[02:32:00] <jmkasunich> yeah, thats the reason for the arbor
[02:32:03] <cradek> have some drill rod? 1/8 dowel pins?
[02:32:21] <cradek> a dowel pin 'dead arbor' like you mentioned earlier?
[02:32:38] <jmkasunich> unfortunately a pecularity of the shoptask is that you can't get the tailstock very close to the headstock without trapping the table
[02:32:46] <cradek> argh
[02:32:48] <jmkasunich> a real lathe runs the tailstock on its own set of ways
[02:33:16] <jmkasunich> my part is only 1" long, but I can't just make a 1.2" long arbor and run it between centers
[02:34:09] <stuste1> going home - see you later
[02:34:30] <jmkasunich> my plan was to have an arbor about 3" long, with most of the length a bit larger (I started with some 7/32 drill rod)
[02:34:55] <jmkasunich> collet blank on the tailstock end - the tailstock has less runout than the spindle
[02:35:03] <jmkasunich> live center is a couple tenths, dead center zero
[02:35:27] <jmkasunich> I have to turn the arbor, so its OD is concentric with the center hole
[02:35:33] <jmkasunich> unless I use a dead arbor
[02:55:40] <jmkasunich> one of my blanks isn't horrible as is - the hole is centered within 0.0003 at one end, and about 0.0007 at the other
[02:55:47] <jmkasunich> the other two stink
[02:56:01] <jmkasunich> 0.003 and 0.007 on the one, and 0.005 and 0.009 on the other
[02:56:48] <cradek> wow, why are they so bad?
[02:57:18] <jmkasunich> they were just drilled and reamed - drills don't center, and they definitely don't go straight
[02:57:44] <jmkasunich> I center drilled, then drilled 0.109, 0.120, then reamed 0.125
[02:58:08] <jmkasunich> my goal was just to get accurate size holes, then spin around the hole and turn the OD concentric
[02:58:45] <jmkasunich> if I had bored a starting hole they would have been much better at the start end, and somewhat better at the exit end
[03:00:13] <jmkasunich> stuart was probably right - make the blanks (OD) first, then bore them in the actual spindle
[03:00:57] <jmkasunich> which means I need to make a mount for the spindle (which I was going to have to do anyway), and come up with a freaking tiny "boring bar"
[03:01:30] <cradek> how deep does the bore (grippy part) need to be?
[03:01:47] <cradek> 1/4?
[03:01:48] <jmkasunich> about 1/2"
[03:01:54] <cradek> woo that's deep
[03:02:19] <jmkasunich> I could make the nose shorter I suspect - the only thing out there is wrench flats
[03:02:58] <jmkasunich> I could probably get the bore depth down to about 0.3"
[03:03:13] <jmkasunich> which is less than 3:1 depth/diameter
[03:05:49] <cradek> I don't know what collets usually have
[03:06:16] <jmkasunich> these are weird collets (with the wrench flats out front)
[03:06:37] <jmkasunich> the 3:1 ratio comment wasn't about collets, it was about boring bar stiffness - 3:1 is doable
[03:06:45] <cradek> ah
[03:07:24] <jmkasunich> I have a feeling I'm going to be disappointed when I get the collets made
[03:07:50] <jmkasunich> the tool was intended for hand use, and I think the spindle itself might have close to 0.001 runout
[03:08:21] <jmkasunich> it only has two bearings, and they aren't angular contact or anything special like that
[03:08:43] <jmkasunich> they are lightly preloaded by a spring
[12:14:18] <stuste1> is there a way to compile modified .c and .cc files without recompiling the whole of EMC? I modified interp_convert.cc. It seems as if I have to do 'make -B' to compile it. I am impatient.
[12:15:49] <stuste1> I think I know what I need to do. I will try it in a minute.
[12:15:57] <archivist_ub> make noemally does just compile the needed, but it still has to link the whle pile
[12:18:02] <jepler> yes, normally after modifying one or more files you just type "make" and it rebuilds what is necessary
[12:18:17] <jepler> is there a reason you don't think that make without "-B" (which I don't think I've ever used!) doesn't do what you want?
[12:18:43] <archivist_ub> ive never used -B either
[12:20:40] <jepler> (after "touch"ing interp_convert.c it takes about 2 seconds to finish a 'make' here)
[12:20:51] <jepler> argh I should go in to the office
[12:36:30] <stuste1> jepler: good morning - I figured it out - I work in interp_convert_stuart.cc - then I copy it to interp_convert.cc - I forgot to copy it - make didn't touch it until I copied it over - the last statement of my first post says it all. I should have waited to think about it a little before crying wolf. Thanks - takes about 15 seconds here (if done correctly)
[12:55:08] <jepler> stuste1: yes, that's a somewhat atypical way to work, and 'make' can get confused by it.
[12:57:04] <stuste1> make wasn't confused - I was - I copy the original to XXXXX_orig.XX - that way I can easily return to the original file after I have 'worked' on the system
[13:34:43] <skunkworks_> odd http://pastebin.ca/1191508
[13:38:45] <alex_joni> skunkworks_: indeed
[13:39:01] <skunkworks_> I think I will point him to the email list.. ;)
[13:53:43] <jepler> sounds like he means he wants kinematics
[13:54:17] <jepler> .. but it's hard to tell
[13:54:28] <archivist_ub> just a bit
[13:55:38] <archivist_ub> "Crossbar engines" is a very strange term
[13:57:36] <skunkworks_> he is from Uzbekistan
[13:58:13] <alex_joni> jepler: or tool comp.
[13:58:14] <cradek> gantry motors?
[13:58:58] <cradek> I see A is the distance from the origin, but I don't recognize B,C equations
[13:59:11] <cradek> I can't guess what he's trying to do
[14:05:36] <renesis> heh crossbar engines, awesome
[14:06:27] <BigJohnT> skunkworks he pm'd me the same question
[14:49:44] <skunkworks_> BigJohnT: interesting.. I pointed him to the emc_user list
[14:50:17] <BigJohnT> I told him I didn't even have clue what he was asking :)
[14:51:51] <skunkworks_> same here
[14:52:13] <skunkworks_> but maybe someone on the list will ask the right questions :)
[14:53:44] <BigJohnT> :)
[16:23:25] <Stinger_> are there any other good general machining/metalwork channels out there?
[16:26:43] <archivist_ub> dunno but we do natter about machining as well in here
[16:32:19] <BigJohnT> practical machinist is my fav forum dunno about any IRC channels
[16:33:39] <BigJohnT> Stinger_: http://www.practicalmachinist.com/vb/index.php
[16:35:26] <skunkworks_> BigJohnT: has the search quit working on the practicalmachinist forum?
[16:35:29] <skunkworks_> or is it just me
[16:36:24] <BigJohnT> skunkworks_ dunno I haven't used it in a while
[16:37:15] <BigJohnT> doesn't work for me either
[16:38:16] <skunkworks_> ok :)
[16:57:48] <tomp> does Jon E ever come here? he may be interested in IBM's OpenCV from sourceforge http://sourceforge.net/projects/opencvlibrary
[17:15:25] <archivist_ub> nice tomp, I may need that for tool positioning
[18:03:13] <SWPadnos> heh: http://www.asktheramguy.com/v3/attachment.php?attachmentid=4045&d=1219938210
[18:26:12] <stuste1> cinci ball test on youtube: http://www.youtube.com/watch?v=Nn1bJ3YAQdI
[18:27:37] <skunkworks_> holy crap.. that is awesome.. I was really wondering what you guys where talking about..
[18:27:54] <jepler> is it supposed to make that banging noise? bam bam bam bam
[18:28:05] <jepler> (I assume that noise is something else going on in the shop..)
[18:28:47] <stuste1> riveting
[18:29:14] <jepler> I know the video is riveting, I was asking about the noise
[18:29:27] <skunkworks_> heh ;)
[18:29:36] <skunkworks_> I think that should go into the quotes
[18:29:41] <stuste1> a part is being riveted behind the camera :)
[18:29:59] <anonimasu__> stustev: nice!
[18:30:03] <cradek> wow
[18:30:07] <stuste1> thanks
[18:30:32] <anonimasu__> you need to show off some 5 axis parts being cut with it later ^_^
[18:30:56] <stuste1> we will - for sure
[18:31:19] <skunkworks_> stuste1: spindle fixed?
[18:31:33] <stuste1> yes - spindle is running
[18:31:46] <stuste1> not in this video but it is working
[18:33:22] <anonimasu__> hm..
[18:33:28] <anonimasu__> I just saw the fastest toolchange on a video I've ever seen
[18:35:11] <renesis> if you could see it i cant imagine it was that fast
[18:35:13] <renesis> =D
[18:35:31] <anonimasu__> haha
[18:35:50] <archivist_ub> * archivist_ub notes the drip onto the dti at about 3.15 secs in vid
[18:36:08] <anonimasu__> http://www.youtube.com/watch?v=4XYakTeQahA
[18:37:30] <cradek> stuste1: that's very nice. congratulations.
[18:37:42] <stuste1> thank you very much
[18:38:38] <archivist_ub> the dancing cinci !
[18:38:41] <stuste1> archivist_ub: this is an old slow messy oil leaking coolant flinging hunk of iron
[18:39:20] <stuste1> that was all done in rapid traverse feed rate
[18:39:54] <archivist_ub> thats going to eat metal soon
[18:40:01] <anonimasu__> * anonimasu__ has been looking forward to seeing emc make some impellers
[18:40:08] <cradek> heh, 'rapid'
[18:40:19] <stuste1> rapid is a relative term
[18:40:46] <cradek> being fast is not its job I think.
[18:40:58] <anonimasu__> hehe
[18:41:05] <anonimasu__> no more like being heavy is it's job
[18:42:06] <stuste1> bbl
[18:42:32] <archivist_ub> seeing the spindle motor on top, I would stand well back when it throws chips around
[18:43:03] <archivist_ub> they will be large and hot
[18:57:49] <tomp> wow, thats a helluva machine tool. i still dont get the ZAB construction, looks like a backwards Jhead in the slot, and the slot pivots on that fan shaped backplate, but cant see the Z clearly ( i can see the spindle move up but not how it moves up )
[19:09:14] <renesis> anonimasu__: hahaha wtf was that
[19:10:07] <renesis> like prob half the toolchange time its waiting for the changer doors to open =\
[19:15:55] <anonimasu__> yep :)
[19:15:59] <anonimasu__> it's the fastest I've seen
[19:19:21] <archivist_ub> http://www.youtube.com/watch?v=L5aC5Kpvibc&feature=related
[19:37:54] <skunkworks_> stuste1: http://www.cnczone.com/forums/showthread.php?t=57550
[19:41:15] <Jymm> As this has always bugged me.... http://www.youtube.com/watch?v=G5bXb8l52yk
[19:41:55] <Jymm> I'll have to check it out when I get home
[19:51:33] <stuste1> tomp: the Z axis is behind the fan way the B axis rotates on
[19:54:06] <skunkworks_> stuste1: is it setup the way you want it?
[19:57:06] <stuste1> yes - this will make parts as it is - and do a very nice job - the point of diminishing returns is very close
[19:57:43] <stuste1> I want to do a few things but what I want to do has more to do with the control side and not the machine
[20:03:47] <skunkworks_> neat
[20:05:25] <hielke> wie heeft er ervaring met emc2 en blender?
[20:05:52] <SWPadnos> check out the CNC TOolkit
[20:06:40] <SWPadnos> It doesn't work with Blender, but I think it may be close to what you're asking for
[20:06:48] <SWPadnos> (I don't speak or read Dutch, so I'm not sure ;) )
[20:08:47] <hielke> there is an .stl out in blender.
[21:10:31] <skunkworks_> the word is getting out.. ;) http://www.cnczone.com/forums/showthread.php?t=24817&page=5
[21:13:28] <anonimasu__> HEH
[21:15:51] <anonimasu__> it's nice that alot of people throw random things aroun..
[21:15:57] <anonimasu__> that they have no real clue about :p