#emc | Logs for 2008-12-17

[01:02:40] <toastyde1th> toastyde1th is now known as toastydeath
[03:13:36] <stustev1> I am calculating the forward and inverse kins - I see the difference - thanks
[03:19:15] <SWPadnos> heh
[04:17:39] <eric_u> I need to install a new version of linux
[04:19:31] <eric_u> I've been defeated by a missing latex package
[04:21:05] <stustev1> my mind is toast tonight - I will work on this tomorrow - I haven
[04:21:56] <stustev1> 't gone to the shop since last wednesday - looks like I will be home again tomorrow - thanks again guys :)
[09:02:17] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[13:01:04] <[1]a-l-p-h-a> [1]a-l-p-h-a is now known as a-l-p-h-a
[14:40:07] <devi0us1> what to you guys recommend as CAM software to compliment EMC? i'm working with limited budget on a home mill, so mastercam, etc are out of the question
[14:40:36] <seb_kuzminsky> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cam
[14:40:47] <seb_kuzminsky> though i think a lot of people us "vi" ;-)
[14:40:57] <seb_kuzminsky> s/us/use/
[14:41:21] <archivist> and some are developing some new toys
[14:41:22] <devi0us1> yeah.. i know. i'm doing things more complex than i want to manually generate gcode for
[14:41:42] <seb_kuzminsky> archivist: you're hacking on apt360, right?
[14:41:49] <archivist> no
[14:41:59] <archivist> well not at the moment
[14:42:28] <devi0us1> thanks.. there's more than i thought. i only knew about pycam
[14:42:30] <skunkworks_> go to #cam
[14:52:15] <cradek_> cradek_ is now known as cradek
[14:53:14] <anonimasu> zae: hm
[14:53:17] <anonimasu> err wrong place
[14:53:20] <anonimasu> * anonimasu is 3d contouring
[14:53:23] <anonimasu> it's dead slow
[14:53:24] <anonimasu> :)
[14:53:40] <anonimasu> 2 hours for this press form
[14:54:30] <anonimasu> but then it's flowline contouring at 40770lines
[14:55:32] <anonimasu> err 4069 lines.. sorry :p
[14:55:40] <anonimasu> it's incrementing in 10 per line.. -_-
[15:12:30] <skunkworks_> I think this is the final for ver 3.2 ;) http://imagebin.ca/img/BaUwDpQ.png
[15:13:06] <anonimasu> err nice
[15:13:51] <cradek> have you made one yet?
[15:13:59] <skunkworks_> no - soon.
[15:14:38] <skunkworks_> everything is setup as I started milling this weekend before I noticed the pads where too small.
[15:15:02] <anonimasu> about 2 hours left :)
[15:15:57] <skunkworks_> * skunkworks_ thinks that is how long it will take to mill the circuit board.
[15:16:51] <anonimasu> * anonimasu curses the design team that only put a 2krpm spindle on the mill
[15:28:51] <toastatwork> goddamnit leeroy.
[15:29:06] <anonimasu> hehe
[15:29:12] <anonimasu> I love carbide.
[15:29:24] <anonimasu> they dont sound funny like hss on heavy cuts...
[15:44:52] <anonimasu> bleh..
[15:48:07] <toastatwork> http://www.hustlepaintball.com/PhotoGallery.asp?ProductCode=10361234
[15:49:40] <anonimasu> hehe
[15:49:49] <anonimasu> woot only 1000 lines left!
[15:49:59] <anonimasu> -_-
[15:52:08] <anonimasu> toastatwork: I think im realizing that the only way to go is carbide for milling -_-
[15:56:03] <toastatwork> ?
[15:58:11] <anonimasu> I roughed my press tool with hss
[15:58:23] <anonimasu> it nearly killed the endmill after the passes -_-
[15:58:59] <anonimasu> 12mm endmill and 5mm deep cuts shouldnt be a problem
[15:59:26] <anonimasu> or well, 4mm in my case..
[15:59:36] <anonimasu> with flood coolant..
[15:59:44] <archivist> only rated for an hour or so in contact
[16:01:53] <toastatwork> snap
[17:57:45] <jmkasunich> today's project: http://jmkasunich.com/pics/KE01052109.pdf
[17:58:37] <archivist> heh you need a cnc hobbing machine
[17:59:21] <jmkasunich> I wish
[17:59:55] <archivist> I could also just about fit that on the ordinary hobbing machine here
[18:00:05] <jmkasunich> what it should look like when done: http://www.metalworking.com/dropbox/_2001_retired_files/Worm03.jpg
[18:00:23] <jmkasunich> gashing setup: http://www.metalworking.com/dropbox/_2001_retired_files/Worm07.jpg
[18:00:35] <jmkasunich> hobbing setup: http://www.metalworking.com/dropbox/_2001_retired_files/Worm09.jpg
[18:00:44] <jmkasunich> free hobbing, driven by the cutter running in the gashes
[18:01:01] <archivist> you made these before?, worn out already?
[18:01:34] <jmkasunich> they are part of a test jig
[18:01:45] <jmkasunich> I've sold about 10 to various labs over 8 years or so
[18:02:14] <jmkasunich> one lab had two jigs made by someone else who did a crap job, so I got the job of making them right
[18:03:11] <jmkasunich> I'm not looking forward to the gashing and hobbing, it is below 0C in the garage where that mill is
[18:03:36] <archivist> time to get a fan heater out
[18:04:05] <jmkasunich> yeah
[18:04:10] <archivist> well about 5 hours in advance of the job
[18:04:12] <jmkasunich> that will be tomorrow
[18:04:35] <jmkasunich> today my goal is to get the blanks done and mounted on the screws, and also make the matching nuts
[18:04:46] <jmkasunich> since blanks and nuts both need threaded 5/8-5 acme
[18:04:58] <jmkasunich> I'm doing two sets
[18:05:12] <jmkasunich> all four blanks are (just now) faced to length and drilled - time to setup for boring
[18:10:20] <skunkworks_> it isn't the heat - it's the humidity....
[18:10:51] <jmkasunich> it isn't the cold, it's the..... well, yeah, it is the damned cold!
[18:11:28] <skunkworks_> heh - heat sucks though - the machine starts dewing instantly.
[18:12:01] <archivist> heats sucks less than cold
[18:12:30] <skunkworks_> we just went for our 1.5 mile walk we do at lunch - better than that last few days.. (11deg here today (and sunny))
[18:13:15] <archivist> walk! please explain, I have no idea what that is
[18:14:06] <skunkworks_> it is my anti-weight control plan.
[18:14:17] <skunkworks_> wait - is that a double negative?
[18:14:43] <SWPadnos> gotta control those anti-weights
[18:30:03] <jmkasunich> boring boring boring
[18:30:27] <SWPadnos> boring is pretty boring, isn't it
[18:30:45] <jmkasunich> yeah
[18:31:02] <archivist> cnc boring for less boring
[18:31:23] <SWPadnos> milling about while milling is almost as boring as boring
[18:31:32] <toastatwork> this is madness
[18:31:42] <SWPadnos> yes. time to make breakfast
[18:31:44] <jmkasunich> http://www.youtube.com/watch?v=zJauvs2FDI4
[18:35:41] <skunkworks_> yikes - Normally you don't bore without pre-drilling..
[18:36:57] <archivist> boring bar == one sided drill :)
[18:37:11] <jmkasunich> he's sort of trepanning
[18:37:13] <toastatwork> i tried to drill with a boring bar once
[18:37:19] <toastatwork> it got a full inch into the part
[18:37:50] <toastatwork> then went all crazy.
[18:38:27] <archivist> I have done small depths
[18:39:15] <archivist> I need to reboot firefox to see youtube :(
[18:39:19] <toastatwork> =(
[18:39:50] <anonimasu> lab
[18:40:14] <anonimasu> archivist: I that's pretty shitty
[18:41:01] <archivist> it loses it after a week or two without a restart
[18:41:06] <anonimasu> err iab
[18:41:17] <anonimasu> archivist: oh, I mean the one hour in contact
[18:41:28] <anonimasu> 20 minutes and the cutter looks worn -_-
[18:42:13] <archivist> tool lifetime is a lot shorter than often thought
[18:42:24] <anonimasu> but comeon 20 minutes -_-
[18:42:32] <anonimasu> that's pretty extreme
[18:42:36] <anonimasu> and this are dormer tooling
[18:42:39] <archivist> boss needs training in tool lifetime
[18:42:46] <anonimasu> your?
[18:42:52] <archivist> yes
[18:42:54] <anonimasu> hehe
[18:42:57] <anonimasu> * anonimasu has no idea
[18:43:06] <anonimasu> I've never ran alot of jobs with alot of contact before
[18:43:28] <anonimasu> but well, 20 minutes in cut is a shitty tool life..
[18:43:30] <anonimasu> for a 12mm cutter..
[18:43:48] <anonimasu> with the tool manufacturers data.. -_-
[18:43:54] <archivist> read Sandvik site about contact time for the better carbides, its still not long
[18:44:16] <anonimasu> weird..
[18:44:28] <anonimasu> hm, the carbide I'm using sounds alot more nice
[18:44:35] <anonimasu> even after 4 hours of cutting
[18:45:12] <jmkasunich> that means you aren't pushing it
[18:45:33] <anonimasu> following the same source for the tooling
[18:45:36] <jmkasunich> most of what I've read says that carbide tooling running at high-production speeds and feeds lasts 20 mins or so
[18:45:52] <anonimasu> I can hear that the spindle is getting noticeably loaded..
[18:46:13] <anonimasu> hehe
[18:46:17] <anonimasu> I have neither ;)
[18:46:30] <anonimasu> jmkasunich: the thing im wondering is if it's worth buying hss tooling
[18:46:59] <anonimasu> at all that is..
[18:47:14] <archivist> hand taps yes
[18:47:24] <anonimasu> hehe milling cutters
[18:47:57] <jmkasunich> I don't know much about the economics of cutters
[18:48:03] <archivist> depends on the cut (intermittent or not)
[18:48:17] <anonimasu> I'm not doing production but I dont like to change tooling after 10 minutes of cutting
[18:48:37] <anonimasu> also, I dont like tweaking the feed override in mid work because the cutter is dulling
[18:48:40] <anonimasu> :)
[18:49:06] <anonimasu> if the case is that cutters wear this fast, maybe I should take a 12*12mm cut instead
[18:49:15] <archivist> poor cooling can kill cutters, best left with none rather than intermittent
[18:49:28] <anonimasu> and use more of the endmill
[18:50:01] <anonimasu> using 4mm if a 26mm long endmill dosent make sense
[18:52:46] <anonimasu> err of..
[18:52:48] <anonimasu> right?
[18:54:25] <archivist> if the cutter gets hot its ok, sudden cooling creates micro cracks and bits fall off
[18:55:11] <anonimasu> yeah, I know
[18:55:21] <anonimasu> but that's with hss isnt it?
[18:55:24] <anonimasu> err carbide..
[18:55:27] <anonimasu> not so much with hss
[18:55:56] <anonimasu> hss shouldnt get hot ever :p
[18:56:10] <archivist> I tend to run dry unless its sticky and needs something
[18:56:27] <anonimasu> :)
[18:56:35] <anonimasu> well, the stuff I run the most is ferrous stuff
[18:56:45] <archivist> heh I blued a gear parting off with HSS the other day
[19:00:46] <anonimasu> :)
[19:00:57] <archivist> the talk I went to by the new Mach owner was about tooling for cnc, on that subject he seems to know his stuff
[19:01:22] <anonimasu> :)=
[19:01:24] <anonimasu> cool
[19:56:33] <Vortex> good evening
[20:01:01] <toastatwork> hai
[21:47:02] <Vortex2> http://www.youtube.com/watch?v=MNPTofQksJw
[21:54:01] <skunkworks_> Vortex2: very nice!!!
[21:58:33] <Vortex2> jep finaly mastes / slave servo setup
[21:58:38] <Vortex2> master
[21:59:32] <Vortex2> works
[22:00:06] <Vortex2> only proplem was PID loop tuning: D was too low
[22:00:55] <Vortex2> and because of that two servos started to resonate
[22:01:24] <Vortex2> increasing D they stapilised
[22:03:52] <skunkworks_> * skunkworks_ is no expert at pid loops
[22:21:44] <skunkworks_> a bit different... old http://www.electronicsam.com/images/KandT/servostart/boardagain.png
[22:21:51] <skunkworks_> new http://imagebin.ca/img/BaUwDpQ.png
[22:22:10] <archivist> looks the same to me
[22:23:01] <archivist> are you going to redesign it to death or make one
[22:23:07] <archivist> :)
[22:23:38] <skunkworks_> heh - I think I am done for now. I need to get some machine time.
[22:32:48] <jmkasunich> one threaded, three to go
[22:32:50] <jmkasunich> what a PITA
[23:32:15] <maddash> what's the simplest way to generate V(t) = sin(t)?
[23:32:32] <maddash> whoops, I meant, V(t) = abs(sin(t))
[23:33:13] <maddash> I tried running the output of a signal generator through a rectifier, but I get weird truncated curves instead
[23:33:16] <fenn> #include <math.h>
[23:34:08] <fenn> rectifier has a forward voltage drop, about 0.35V for a schottky diode, is that what you're seeing?
[23:35:15] <fenn> also diodes have a non-zero switching time
[23:35:26] <fenn> what frequency is this at
[23:35:27] <maddash> I'm seeing something like V(t) = { sin(t) when sin(t) >0, and 0 when sin(t) < 0 }
[23:35:37] <maddash> 52 Hz
[23:35:43] <fenn> you're only half-wave rectifying
[23:35:53] <fenn> use a full bridge instead (4 diodes)
[23:36:18] <archivist> whats the purpose
[23:36:55] <fenn> http://en.wikipedia.org/wiki/Rectifier#Half-wave_rectification <- read this
[23:37:18] <maddash> this (http://www.talonix.com/shop/item.aspx?itemid=5464) is not a half-wave rectifier
[23:37:30] <maddash> I'm using that rectifier chip
[23:37:36] <archivist> http://sound.westhost.com/appnotes/an001.htm
[23:38:03] <fenn> ugh
[23:38:10] <fenn> just get some diodes so you know wtf is going on
[23:38:16] <maddash> okey
[23:38:17] <archivist> again what is your purpose
[23:38:38] <maddash> archivist: to make a "flipped-up" sine curve?
[23:39:02] <archivist> if its the transfer function then see that page I linked
[23:44:51] <jmkasunich> maddash: what voltage is the sine wave you are rectifying?
[23:46:05] <jmkasunich> there will be 0.6 or so volt drop across the diode, so don't expect anything remotely resembling accuracy if you are rectifying a 3 or 4 V sine wave
[23:46:47] <jmkasunich> also, if you are trying to rectify a "signal" (a wire referenced to ground) instead of some power source like a transformer winding, a bridge will NOT give you full wave rectification
[23:48:21] <archivist> and also see http://www.analog.com/static/imported-files/application_notes/130445851AB109.pdf
[23:49:20] <jmkasunich> for rectifying a signal (absolute value), you probably want something like this: http://www.play-hookey.com/analog/full-wave_rectifier.html
[23:49:51] <jmkasunich> except whoever drew that is an idiot
[23:49:57] <jmkasunich> op-amps have two input pins, not one
[23:50:27] <jmkasunich> better - third circuit down on this page: #1011 = [#1010 * -1.0 * TAN[#1005]]
[23:50:30] <jmkasunich> oops
[23:50:30] <archivist> hehe see the pdf I linked at least its correct
[23:50:35] <jmkasunich> http://sound.westhost.com/appnotes/an001.htm
[23:51:06] <jmkasunich> archivist: that is some weird-ass circuit with a fet in it
[23:51:14] <jmkasunich> gimme opamps and diodes any day
[23:51:29] <jmkasunich> ;-)
[23:51:46] <archivist> :)
[23:51:53] <jmkasunich> oh, look, your ckt is the next one down on the 2nd page I linked
[23:52:52] <archivist> I love good app notes from the suppliers