#emc | Logs for 2010-10-26

[00:04:03] <theorb> theorb is now known as theorbtwo
[00:14:32] <andypugh> Unused cnc milling machine, has to be a bargain at that price: http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=270654491900#ht_500wt_1110
[00:15:39] <SWPadnos> Cranbrook. I know that town
[00:16:10] <MattyMatt> bloody students. all those toys to play with :)
[00:16:48] <SWPadnos> heh. the student I knew there was the 16 year old daughter of a friend of mine
[00:16:59] <SWPadnos> I didn't notice any machine shops at that school :)
[00:17:23] <SWPadnos> I did notice that the school was built in 1536 or something, which I thought was very cool
[00:17:54] <andypugh> And probably hadn't been repainted since.
[00:18:24] <MattyMatt> and still had the original torture chamber for wayward pupils
[00:18:51] <SWPadnos> it didn't seem to need paint when I visited, but that was a few years ago
[00:18:53] <MattyMatt> * MattyMatt had a brief internment in a 'posh' school
[00:19:17] <andypugh> Shouldn't you be running the country then?
[00:19:50] <SWPadnos> no. I didn't attend the school (except for a field hockey or cricket game)
[00:19:58] <SWPadnos> oh, him
[00:20:03] <MattyMatt> I was a scholarship boy. I wasn't actually a member of the ruling class
[00:21:01] <MattyMatt> I rubbed shoulders with a the offspring of the rich and powerful. and I bit their ears. I was tight head prop
[00:28:19] <andypugh> Ah, the Other Code
[00:28:38] <MattyMatt> The day I got bored with rugby was probably also me quitting politics :)
[00:30:19] <MattyMatt> far too many scrums for comfort in the 15 a side game
[00:30:47] <andypugh> At least they have scrums. League scrums are a travesty
[00:31:31] <MattyMatt> all those dudes in wheelchairs in the front row at twickenham. they were all props :)
[00:32:28] <andypugh> Clearly the solution is helmets and enormous body armour.
[00:32:42] <MattyMatt> how ingenious
[00:33:04] <andypugh> Perhaps those American chaps can try it and tell us how it works out?
[00:33:43] <MattyMatt> and you could have the team swap over for a breather between pushes
[00:34:43] <andypugh> You could have several full teams, so that nobody ever plays for more than 5 minutes, and the game stops every minute. I am sure that would be a marvelous game to watch
[00:35:10] <MattyMatt> it'd be even more entaining with weapons, for the viewer
[00:35:39] <andypugh> My feeling is that I would far rather watch Union and fa rather play League.
[00:46:20] <andypugh> Right chaps. Goodnight all.
[02:14:07] <pcncx> hi honey I'm home!
[02:16:20] <pcncx> anyone here know what CONFIG_PREEMPT_RCU does in the kernel?
[02:17:59] <pcncx> or does anyone here know why it says no user servicable parts on PC power supplies?
[02:31:50] <flyback> HOLY MOTHER OF CANUCK
[02:31:57] <flyback> I am seriously going to cry like a baby
[02:32:07] <flyback> all my 36 yrs I struggled to solder much anything past thru hole
[02:32:25] <flyback> and turns out half the problem was shit irons, shit flux and shit wire
[02:32:27] <flyback> wow
[02:32:31] <flyback> kynar wire is freaking awesome
[02:32:41] <flyback> pcncx, yeah to scare off the user
[02:32:46] <flyback> because it's a high voltage chopper
[02:32:48] <flyback> about 300v
[02:33:22] <flyback> next time any of you have to jumper wire patch a circuit board etc
[02:33:24] <pcncx> well this machine was howling last night and giving me such a headache
[02:33:26] <flyback> look into kynar wire
[02:33:29] <flyback> freaking awesome
[02:33:33] <flyback> you mean cause of the fan?
[02:33:37] <pcncx> yes
[02:33:39] <flyback> you can open it and replace the fan
[02:33:51] <flyback> just be careful the caps can store a nasty wallop for hrs/days etc
[02:33:54] <pcncx> so I took it all apart got the fan out peeled off the label and under it was a rubber plug
[02:34:02] <flyback> I usually just replace them
[02:34:09] <flyback> because they don't usually last even after oiling
[02:34:11] <pcncx> popped that out and got sone drops of chainsaw bar oil in there
[02:34:13] <flyback> once the bearings are scored
[02:34:24] <pcncx> man whisper quiet
[02:34:27] <flyback> cool :)
[02:34:32] <pcncx> hell yeah!
[02:34:48] <pcncx> I never seen one of these fans you could even get to the bearings on before
[02:35:10] <flyback> well that's probably a cheap sleeve bearing fan
[02:35:15] <pcncx> trick is chainsaw bar oil awesome stuff
[02:35:20] <The_Ball> pcncx, getting rid of large collections of dust will also help it cool down properly so it doesn't need to speed up
[02:35:22] <flyback> ball bearing you can't do unless it's 1 ball 1 sleeve
[02:35:35] <flyback> fluid bearing are perm sealed
[02:35:37] <pcncx> The_Ball this is one speed cools all
[02:35:42] <flyback> because they need to be to build up the pressure
[02:36:12] <flyback> * flyback is in absolute shock how much better this wire is
[02:36:24] <pcncx> but of course while i had it apart i hit it with an acid brush and some compressed air
[02:36:32] <flyback> FAIL
[02:36:45] <flyback> you don't mean literal acid on there right?
[02:36:49] <flyback> or had any before
[02:36:52] <pcncx> I don't know as long as wire is clean I've no problems soldering any of it
[02:37:05] <flyback> also a brush might generate 1000's of volts of static
[02:37:06] <pcncx> no just that is what i call flux brushes
[02:37:09] <flyback> and damage chips
[02:37:23] <pcncx> nah we used them all the time when i made PCBs
[02:37:34] <flyback> kynar wire is floropolymer coated wire
[02:37:37] <pcncx> how else are you supposed to clean flux off a board?
[02:37:46] <flyback> wet
[02:37:57] <flyback> and those brushes are usually for pipes I thought
[02:38:02] <pcncx> well we'd brush til the remover dried
[02:38:15] <pcncx> then hit it up again
[02:38:28] <pcncx> be brushing them freaking boards al lday long
[02:38:31] <flyback> man I never knew about kapton tape
[02:38:34] <flyback> kynar wire
[02:38:48] <flyback> and mix of acetone/isopropyl is wicked on flux
[02:38:52] <pcncx> I use steel wool on old wire sometimes
[02:38:56] <flyback> uh
[02:38:58] <flyback> bad idea
[02:39:01] <pcncx> why?
[02:39:09] <flyback> you get any of that metal dust on your board etc
[02:39:11] <flyback> bzzzzz
[02:39:20] <pcncx> oh i don't do it over a board
[02:39:26] <flyback> well even on the wire
[02:39:32] <flyback> you need to wipe it good with damp cloth with alochol
[02:39:44] <flyback> so you don't leave metal sprag all over the wire to come off on something
[02:40:04] <pcncx> haven't had a problem
[02:40:11] <flyback> nice
[02:40:21] <pcncx> cept once i was cleaning out an analog meter there do not use steel wool
[02:40:21] <flyback> I soldered up a old 486 laptop this weekend
[02:40:30] <flyback> soldered together a cf adaptor for it
[02:40:33] <flyback> get rid of the hd
[02:40:40] <flyback> it's just to run a serial debug program
[02:40:50] <flyback> and I managed to solder all 44 wires/88 joints without a lot of problem
[02:40:51] <flyback> using kynar
[02:40:53] <flyback> :)
[02:41:06] <flyback> oh you might like this
[02:41:30] <flyback> http://www.bgmicro.com/2gbindustrialgradecompactflash.aspx
[02:41:44] <flyback> if you rather run a machine without a hd to worry about
[02:42:22] <flyback> I would assume linux cnc can be tweaked to minimize the write cycles or do them elsewhere
[02:42:50] <pcncx> would really have to put together a small system or use a few
[02:43:21] <flyback> also I was telling someone on this network
[02:43:47] <flyback> at least pata/eide drives have ATA cut command you can force the drive to only be a certain size
[02:43:59] <flyback> so you don't have to find a used 500 meg drive on ebay for a old cnc machine etc
[02:44:06] <flyback> you can use a modern disk and just restrict the size
[02:44:38] <pcncx> I'm up to 1.9 BG on my root already and I'm not done builsing this system though I'm sure i could take out a lot too
[02:44:43] <pcncx> GB even
[02:44:59] <flyback> those aren't normal cf's they are industral
[02:45:08] <flyback> they have smart you can see the write cycles and wear level reporting
[02:45:19] <pcncx> does that mean a GB is bigger?
[02:45:19] <flyback> they have a software write lock feature and a 5 second emergency erase
[02:45:22] <flyback> no
[02:45:24] <flyback> heheeh :)
[02:45:37] <pcncx> with industrial sized gigabytes!
[02:45:40] <flyback> just means if you optimize your sw to avoid needless writes it should last for yrs and yrs
[02:45:57] <flyback> i'm going to have to figure out how to do this to win98se :P
[02:46:02] <pcncx> well once a system is setup I don't see a whole lot of writing going on
[02:46:27] <pcncx> just mount /tmp and /var in a ramdisk
[02:47:23] <pcncx> heck live distros work put a live distro on it
[02:47:29] <flyback> I haven't done any cnc stuff, I wanted to build something but never got chance, I am just here helping a friend find a home for a large milling machine
[02:47:33] <flyback> yeah
[02:47:45] <pcncx> there ain't no writing going on with those
[02:47:56] <flyback> I know this one doesn't even know what a hd is
[02:48:00] <flyback> let alone flash etc
[02:48:00] <flyback> :P
[02:48:02] <flyback> 80's
[02:48:14] <flyback> mazlak cnc and forget the control system
[02:48:25] <flyback> but you can run it in dumb mode and pipe the data from a pc to it's serial port
[02:51:40] <pcncx> ha ha ha I know what circuit this is
[02:51:46] <pcncx> http://www.bgmicro.com/weeblinky.aspx
[02:52:35] <pcncx> I made one of those about 30 years ago using the hanging garden method
[02:53:23] <flyback> wow
[02:53:29] <flyback> you are pretty old then :P
[02:53:41] <pcncx> I started pretty young
[02:54:17] <flyback> well I was taking things apart at age 6
[02:54:22] <flyback> and then got into IT and still my hob
[02:54:23] <flyback> job
[02:54:24] <pcncx> http://www.reprise.com/host/circuits/transistor_flasher.asp
[02:54:36] <flyback> but got into electronics in the last 7 yrs just finally actually trying to do hands on now
[02:54:38] <flyback> 36
[02:54:54] <flyback> mostly cause the last 7 yrs has been nothing but depression and health issues
[02:55:07] <pcncx> I always loved electronics
[02:55:26] <flyback> you should idle in #electronics then
[02:55:33] <pcncx> years ago people called me Mr. Wizard because i was always fooling around with the stuff
[02:55:58] <flyback> hehehe
[02:56:04] <flyback> he just died like 3-4 yrs ago
[02:56:29] <flyback> http://www.bgmicro.com/monocrystalline10wattsolarpanel.aspx
[02:56:31] <flyback> DAMN
[02:56:33] <flyback> * flyback cries
[02:56:44] <pcncx> I wish I knew if this stuff was good or bad for RTAI and EMC2 CONFIG_PREEMPT_RCU
[02:56:54] <flyback> they might know
[02:57:00] <flyback> they play with steppers
[02:57:11] <pcncx> I tried to google it came up empty
[03:03:58] <pcncx> this is interesting: http://lwn.net/Articles/105571/
[03:45:46] <pcncx> RTAI is so fast they have mail archives from July 2017! https://mail.rtai.org/pipermail/rtai/?&MMN_position=21:21
[04:16:21] <MattyMatt> flyback GREENPOWERSCIENCE!!! http://www.youtube.com/watch?v=5Zy3ELxwdtE&feature=channel
[05:02:39] <Connor> Hi guys, question, how do you all normally set up your work so that you know where X and Y is on it.. What I mean is.. I have a part that is already cut to size.. and I now need to router a area out of it.. what's the best way to get it setup in the machine so that my cuts are in the right place?
[05:08:13] <skinnypup> An edgefinder or jog the tool slowly to the part edge until the tool just pinches a thin piece of paper b/t the tool and the work and touch off that axis the numeric value of the tool radius on the offset you intend to use . i.e. g54 g55 etc
[05:11:34] <Connor> Not sure I understand G54, G55 yet.. and I was wanting something less sloppy than to jog slowly to the part edge.. especially with fluted bits.. I always end up with the wrong side against the part.
[05:14:17] <pcncx> edge finders have flutes?
[05:14:36] <Connor> No. I use the tool bit normally...
[05:14:40] <Connor> Just starting out..
[05:15:33] <skinnypup> then chuck up a piece of 1/4 or 1/8 roundstock and touch off with that ... adjust to the radius of the tool you plan to use
[05:16:06] <pcncx> math is your friend!
[05:16:31] <skinnypup> yeah he's sitting next to me as we speak :P
[05:16:41] <skinnypup> j/k of course
[05:18:04] <pcncx> me i try to never make anything where tolerances are so critical I need to worry about edge finding
[05:18:54] <Connor> If I was doing the whole part from scratch, that would be easy..
[05:18:56] <pcncx> but I do have a wobbler
[05:19:10] <Connor> but, my part is already made, and I have to modify it..
[05:19:14] <skinnypup> wobblers handy things too
[05:19:46] <pcncx> Connor in most shops they strategically place these cylindrical objects around the floor often with fluted sides for those parts
[05:19:47] <Connor> Wobbler?
[05:19:56] <pcncx> Connor its called the trash can!
[05:20:09] <skinnypup> so are a few dowels to line up said part so you can return without that problem
[05:21:19] <pcncx> a wobbler has a floating head on a shank you set it off so it wobbles then you bring it into the part when it stops wobbling you're there
[05:22:12] <Connor> okay, saw a demo of that... Not sure how well that works with a highspeed router. I would be afraid it would sling the wobble apart.
[05:22:19] <skinnypup> or if the lower left corner is 0,0 chuck up whatever you have pointy and locate to the corner best you can
[05:22:27] <pcncx> yeah no you don't run wobblers fast
[05:22:33] <skinnypup> nope
[05:22:35] <pcncx> my mill goes down to 80 RPM
[05:24:45] <Connor> I'm trying to figure out what bits I need... I need a few different ones.. one for PCB milling, 1 for plastics and/or wood. and for aluminum (2d cutout mostly)
[05:25:06] <Connor> All of which would be in my Bosch Colt router..
[05:25:15] <pcncx> oh I don't want to be there when you route aluminum!
[05:25:33] <Connor> yea, I've heard that can be harsh...
[05:25:41] <pcncx> a little bit
[05:25:50] <Connor> I've heard very SHALLOW passes..
[05:25:55] <pcncx> a word to the wise get ear plugs
[05:26:01] <Connor> and vacuum a plenty.
[05:26:11] <pcncx> oh don't worry about that
[05:26:14] <Connor> Well.. My router is in a enclosure..
[05:26:16] <pcncx> the chips will go 20 feet
[05:26:54] <Connor> Dry Milling BTW.. I don't even know if it's possible..
[05:27:07] <pcncx> sometimes i cut aluminum on a table saw and thats harsh
[05:27:19] <Connor> yea. I do that too..
[05:27:28] <pcncx> router bout the same
[05:27:37] <skinnypup> wd40 helps al from sticking the bit
[05:27:43] <Connor> Okay, well.. the enclosure should help with that.
[05:27:49] <Connor> Yea.. but machine is made of MDF.
[05:28:24] <pcncx> well let us know how you make out
[05:28:55] <Connor> You all probably laughing you a$$ of at me right now... thinking what the hell is this guy thinking... :)
[05:29:13] <skinnypup> oh we all started somewhere
[05:29:20] <pcncx> actually i was thiking about the knights that say Neek scene from Holy Grail
[05:29:35] <pcncx> cut down the largest tree in the forest with a herring!
[05:32:20] <pcncx> there are lots of different kinds of aluminum too
[05:32:29] <pcncx> some are softer than others
[05:32:35] <Connor> yea. That's true.
[05:32:53] <pcncx> ever tried to cast the stuff?
[05:33:00] <Connor> Mostly would be using stuff that could be easily bent on a break
[05:33:00] <pcncx> its dead easy
[05:33:23] <pcncx> if you make your own you can alloy it with say lead and make it really machinable
[05:33:50] <Connor> Mostly using this to make parts for Robots and such.. not sure how casting would work with that...
[05:34:05] <pcncx> I mean just make billets then machine those
[05:34:29] <pcncx> you can melt aluminum on a barbeque if you have a blast running on the fire
[05:34:43] <pcncx> like say a hair drier
[05:35:04] <pcncx> stuff melts at about 1250 F
[05:35:18] <pcncx> which is less than double lead
[05:35:43] <Connor> I've done lead before making cast bullets.
[05:35:45] <pcncx> I seen people use old paint cans for crucibles
[05:36:00] <pcncx> you don't get many sessions out of one but it works
[05:36:38] <pcncx> yeah aluminum isn't much further up the heat scale than lead
[05:37:06] <pcncx> when i did it I was suprised how easy it really is to do
[05:37:23] <pcncx> then you can machine beercanium!
[05:37:40] <pcncx> cans BTW are pure aluminum
[05:38:07] <pcncx> molten aluminum is cool too it looks like quick silver
[05:38:39] <pcncx> it looks like the T-2000 terminator
[05:39:22] <Connor> interesting... http://www.youtube.com/watch?v=4BXAPaqpa7I&feature=related
[05:39:56] <pcncx> You need to upgrade your Adobe Flash Player to watch this video.
[05:40:15] <Connor> Well, upgrade then.. :) It's youtube. :)
[05:40:30] <pcncx> this is my CNC machine
[05:40:52] <Connor> So? You can install the adobe plugin.
[05:43:52] <pcncx> I wonder if i have this? Enable support for the Intel(R) I/OAT DMA engine present in recent Intel Xeon chipsets.
[05:44:01] <pcncx> how can I tell?
[05:47:52] <Connor> What do you need that for?
[05:48:17] <pcncx> so if i was transferring data on discs my CPU wouldn't be tied up
[05:48:26] <pcncx> but thats OK I don't think I have it
[05:49:09] <Connor> would have to check the CPU Number...
[05:49:28] <pcncx> intel says it'd have to be in the system bios and enabled
[05:49:28] <Connor> in terminal do a cat /proc/cpuinfo
[05:53:38] <waprat> waprat is now known as packrat
[05:55:12] <pcncx> well it's probably not the best kernel ever configured but I can always do it over
[05:58:37] <pcncx> I don't understand why this CONFIG_ACPI_DOCK needs to be selected for RTAI
[06:02:41] <alex_joni> pcncx: the most cautious RTAI config disables ACPI all together
[06:03:07] <pcncx> alex_joni I'm following the directions on linuxcnc.org
[06:03:46] <pcncx> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Debian_Lenny_Compile_RTAI#Configuring_the_kernel
[06:04:24] <pcncx> I'm enabling it this time around but it if works I'm going back to shut it off
[06:04:44] <pcncx> just to see whats what
[06:07:34] <pcncx> the more I read about all this realtime stuff the more I'm drawing the conclusion that nobody really knows what is going on!
[06:08:14] <alex_joni> pcncx: that is probably close to the truth
[06:08:46] <alex_joni> the combinations of BIOS, motherboards, chipsets, processors and periferials are virtually unlimited
[06:09:00] <pcncx> I'm telling you nobody comes right out and says it but that seems to be the state of things
[06:09:01] <alex_joni> so something that truly works is close to a wonder :)
[06:09:25] <pcncx> yeah I'm over here trying to grab ass in the dark to make my system a little faster
[06:09:57] <pcncx> so far i can't beat high latency of 20,000 but it is so rare!
[06:10:15] <pcncx> but i still have ot go with it because i hit it every now and again
[06:11:01] <pcncx> this is maddening!
[06:11:38] <pcncx> its like trying to play draw straws and the guy is holding out 1,600 straws in front of you
[06:12:01] <alex_joni> if you really want good latency you might try 2 things: either disable ACPI completely, or enable only a very small subset of it, until SMP works for you
[06:12:29] <pcncx> well that is why I'm thinking they pick dock it is something nobody has
[06:13:08] <pcncx> so you're sort of fooling the computer into looking for something it'll never see
[06:14:39] <alex_joni> well, maybe whoever wrote that page was planning to use it for a laptop?
[06:14:58] <pcncx> well they got it off another document on RTAI's website
[06:15:07] <pcncx> the list is identical along with a lot of the instructions
[06:17:12] <alex_joni> pcncx: whatever :)
[06:17:22] <alex_joni> I'm quite happy with the 10.04 config mozmck used
[06:17:28] <pcncx> yeah
[06:17:35] <alex_joni> and with the 8.04 I did a long time ago ;)
[06:17:39] <pcncx> I don't care much for Ubuntu
[06:18:20] <pcncx> though i guess i could steal the config file out of it
[06:18:54] <pcncx> what kernel version is it?
[06:19:18] <pcncx> I'm using v2.6.30.5 now
[06:20:38] <alex_joni> http://lmgtfy.com/?q=ubuntu+10.04+kernel+version&l=1
[06:21:17] <alex_joni> 2.6.32
[06:21:46] <pcncx> config would probably be similar
[06:22:01] <pcncx> I just grep them anyways
[06:22:51] <pcncx> like here is one that drives me nuts: Timer frequency should i pick a low or a high speed?
[06:24:36] <pcncx> CONFIG_HZ_100 or CONFIG_HZ_1000
[06:28:42] <alex_joni> I think I used 100 the last time
[06:28:59] <alex_joni> it shouldn't matter much to RTAI
[06:29:12] <pcncx> that is what I went with
[06:31:38] <pcncx> I'm also thinking when I really do the CNC for real to leave out all networking junk I can always have a different kernel I boot if i need to network the machine
[06:32:13] <skinnypup> CONFIG_HZ_250=y from config-2.6.24-16-rtai
[06:32:33] <pcncx> so the stock shipped value
[06:33:05] <pcncx> last time I went with 300 this time 100 I'll let you know if the Earth moves :)
[06:33:08] <skinnypup> think so i just looked at the 10.04/emc file
[06:33:34] <pcncx> like you said probably doesn't matter
[06:33:36] <skinnypup> i tried them all on a gaming box with no change ... wasnt rt kernel either though
[06:34:40] <pcncx> I'm getting too fixated trying to get this cast off box running as good as it can fact that it runs should be enough for me
[06:35:13] <skinnypup> want me to grep any other values b4 i head off ?
[06:35:32] <pcncx> no I'm in make now thanks maybe another time?
[06:35:39] <skinnypup> k-sure
[06:36:07] <pcncx> build it boot it see whats what then I'm off
[06:36:37] <pcncx> probbaly won't even boot
[06:41:21] <alex_joni> fwiw the 2.6.24-16 config is in git
[06:41:23] <alex_joni> http://git.linuxcnc.org/gitweb?p=infrastructure.git;a=tree;f=livecd/kernels/hardy;h=630b19a6b1f3b7d43ba1dbc46fa988beadbca8d5;hb=HEAD
[06:42:23] <pcncx> sweet
[06:42:42] <pcncx> thanks
[08:51:07] <lilalinux> is it possible to get a grid in the preview?
[08:51:30] <lilalinux> When using G41/G42 it's hard to tell, if the paths are correct
[08:52:03] <lilalinux> It would be nice to get some coordinates printed when clicking a path segment
[08:57:58] <MattyMatt> or just a running display of mouse 'x' & 'y' values in the axis aligned views
[09:00:06] <MattyMatt> was that a feature request or are you hoping it's there already?
[09:16:24] <lilalinux> MattyMatt: if it's not there yet, then it's a feature request
[09:38:46] <boot|sleep> boot|sleep is now known as bootnecklad
[10:38:31] <lilalinux> http://pastebin.com/ismy5d2Q
[10:38:35] <lilalinux> what's wrong with that code?
[10:38:55] <lilalinux> the result should be a square rectangle from -3,-3 to 3,3
[10:39:00] <lilalinux> the tool is 1.6mm
[10:40:51] <lilalinux> the entry move doesn't go to the lower right corner of the square
[10:41:11] <lilalinux> is that an emc bug or a cam bug?
[10:41:44] <lilalinux> I thought the coordinates in the gcode are independant from the tool size
[10:42:49] <lilalinux> so my entry move should stop at -3,-3 in the g-code, like it does?
[10:55:22] <SWPadnos> lilalinux, yes. you program the coordinates of the edge of the tool when using cutter comp
[10:55:48] <lilalinux> SWPadnos: can you see what's wrong with my code?
[10:55:58] <SWPadnos> and emc puts the center of the tool on the left or right side of the programmed path, depending on G41 vs G42
[10:56:13] <SWPadnos> no, I'm just looking at it rather than running it
[10:56:24] <SWPadnos> what error do you get?
[10:56:36] <SWPadnos> or what do you see that tells you it's wrong?
[10:56:42] <lilalinux> I don't get an error, but the cut doesn't start the the right position
[10:56:58] <lilalinux> I expected the cut to start at the compensated corner or the square
[10:57:21] <lilalinux> it shoult start at (-2.2, -2.2)
[10:57:28] <SWPadnos> what's the tool diameter for T1?
[10:58:07] <lilalinux> 1.6mm
[10:58:08] <SWPadnos> oh, you said 1.6mm. what's in your tool table?
[10:58:12] <lilalinux> 1.6mm
[10:58:18] <SWPadnos> ok, that's good :)
[10:58:20] <lilalinux> :-)
[10:58:28] <lilalinux> the rest of the square is ok
[10:58:35] <lilalinux> but the entry move is wrong
[10:58:39] <SWPadnos> where are you expecting to see 2.2 (vs. 3)?
[10:58:55] <lilalinux> line 19
[10:59:03] <lilalinux> it says end point is -3,-3
[10:59:05] <SWPadnos> yes, I see that there's a linear move just after the G42, then an arc move which looks like an entry move
[10:59:16] <lilalinux> but with compensation it should really be -2.2,-2.2
[10:59:29] <SWPadnos> the programmed point is still -3, -3
[10:59:41] <SWPadnos> but emc will move to -2.2, -2.2
[10:59:48] <lilalinux> but it doesn't
[11:00:02] <lilalinux> wait, i'll givbe you a screenshot
[11:00:14] <SWPadnos> I think the DRO still shows -3, -3 (but I'm not sure)
[11:00:55] <lilalinux> http://bilder.sayegh.de/paste/Bild+5.png.html
[11:01:31] <lilalinux> I expected that it dives in at the junctions
[11:01:33] <SWPadnos> one thing that cradek suggests is to make the compensated code into a subroutine (so lines 17-29 maybe), and call the subroutine once with the tool you want and once with a zero diameter tool (just different TxM3 lines before the call)
[11:01:58] <SWPadnos> dives in?
[11:02:23] <lilalinux> i.e. cutting depth
[11:03:01] <lilalinux> did you see the image?
[11:03:07] <SWPadnos> yes
[11:03:38] <lilalinux> I want to cut inside the square
[11:04:10] <SWPadnos> it looks like a large arc "up high", then a plunge, then a not-quite-aquare, then maybe a G0 Z up move
[11:04:21] <SWPadnos> s/aquare/square/
[11:04:57] <lilalinux> actually it is a perfect square
[11:05:02] <lilalinux> that's just the perspective
[11:05:07] <lilalinux> i just verified
[11:05:19] <lilalinux> it's from -2.2/-2.2 to 2.2,2.2
[11:05:33] <lilalinux> but the cut starts too far in -y
[11:05:46] <lilalinux> and stops too far in -x
[11:06:26] <SWPadnos> you have to be careful with the first and last lines
[11:06:37] <lilalinux> unfortunately I know nothing about subroutines
[11:07:20] <SWPadnos> http://pastebin.com/W2WUUbZg
[11:07:26] <SWPadnos> I think that might work
[11:07:51] <SWPadnos> it should give you two passes, unless you turn off the "conditional execution" button
[11:08:21] <SWPadnos> you may have to define a zero diameter tool and use that instead of T0, I'm not sure
[11:08:41] <lilalinux> still wrong
[11:08:50] <SWPadnos> yes
[11:09:07] <SWPadnos> usually, you do an entry move on a side of the square
[11:09:45] <SWPadnos> if you enter on a corner, then you have to do an extra move at the end
[11:10:00] <SWPadnos> you basically have to cut the first side (or part of it) twice
[11:10:23] <SWPadnos> this is because four line segments only describe three corners. you need a fifth line segment to describe the last corner
[11:10:52] <lilalinux> ok
[11:10:59] <lilalinux> but that's a different problem?
[11:11:03] <SWPadnos> the compensation code can only know that it needs to stop before the end of the programmed line if there's another path "in the way"
[11:12:03] <lilalinux> i'll try a rotated square
[11:12:12] <SWPadnos> add this line between 25 and 26: G1 Y0
[11:12:19] <lilalinux> in your code or mine
[11:12:20] <lilalinux> ?
[11:12:25] <SWPadnos> err, mine
[11:12:52] <SWPadnos> actually, it's the same :)
[11:13:13] <SWPadnos> just before the G1 Z 3 line
[11:13:39] <lilalinux> ok
[11:13:44] <lilalinux> that fixed the end
[11:14:00] <SWPadnos> yep. it made a corner that the compensation code could "see"
[11:15:19] <lilalinux> and for the entry move?
[11:15:47] <SWPadnos> well, normally there would be a CAM option to do the entry/exit moves in the middle of an edge
[11:16:03] <SWPadnos> rather than at a oorner
[11:16:06] <SWPadnos> corner
[11:16:10] <lilalinux> I'll try to figure out where to fix that in dxf2gcode
[11:16:26] <SWPadnos> if dxf2gcode is dumb enough, you might just change the drawing to be 5 lines instead of 4
[11:16:57] <SWPadnos> ie, split one edge in the middle, and use that as the starting point for dxf2gcode
[11:17:23] <lilalinux> good idea
[11:17:33] <SWPadnos> let's hope it works :)
[11:17:37] <lilalinux> but I fear it's smart enough to optimize that away :-)
[11:18:04] <lilalinux> there are messages like "... reduced to X contours"
[11:18:37] <SWPadnos> oh. in that case it should be smart enough to put entry/exit moves in the middle of an edge ;)
[14:12:56] <cradek> http://www.quad-i.com/products-vises.html
[14:13:11] <cradek> looks like some nice ideas, but "call for price"? get real.
[14:13:28] <elmo42> that drives me nuts
[14:13:37] <cradek> that drives me permanently away
[14:13:46] <elmo42> same with machine resellers. always call for price
[14:14:10] <JT-Work> I've been getting my vises from Glacern
[14:14:16] <JT-Work> like the price and quility
[14:14:36] <cradek> my newish kurt is really good, but I like the idea of putting the fixed jaw in front
[14:14:50] <cradek> I guess I could mount it that way, but having the screw on the back would be a minor pain
[14:15:14] <JT-Work> http://www.glacern.com/gpv_615r
[14:15:36] <cradek> hmmm
[14:15:41] <elmo42> cradek: did you see the vice our shop uses? They are lovely! http://www.lang-technik.de/images/qp_04.jpg http://www.lang-technik.de/images/qp_08.jpg
[14:16:27] <cradek> neato (to both of you)
[14:16:48] <elmo42> they did a trial a few years back. now we have over 20 of them.
[14:16:54] <elmo42> perfect for 4 and 5-axis machines
[14:17:55] <elmo42> we were also able to reduce production costs by not requiring such a large stock size due to the clamping style.
[14:18:04] <elmo42> no need for parallels, either
[14:18:08] <JT-Work> elmo42: that is a cool vise how big is the opening between the jaws?
[14:18:26] <elmo42> whatever you want. they have all sizes. just go to the main URL
[14:19:03] <elmo42> they have center holding (the pics), fixed jaw, double jaw... many differnet types
[14:19:23] <elmo42> and you can reverse the jaws to accommodate larger parts.
[14:19:53] <elmo42> it practically makes a 4-asix machine a 5-axis due to the accuracy in rotating on the subplate 90deg.
[14:23:37] <elmo42> the nice thing about the subplate is you can use the 'pins' and make your own fixture. now it can be rotated 90deg accurately as well as the vice.
[14:30:18] <JT-Work> http://www.chickworkholding.com/4dcgi/public/index.shtml
[14:32:52] <elmo42> nice vice.
[14:34:30] <elmo42> no prices?
[14:35:41] <JT-Work> heh, request a quote lol
[14:35:55] <elmo42> prob because they use distributors.
[14:37:06] <cradek> the jaw attachment looks nice
[14:37:18] <cradek> plain flat bed too
[14:41:10] <elmo42> flat is nice. no chips hitting your face when you clean it.
[14:54:57] <pcw_home> Well that was disappointing, no chick holding the work...
[14:55:21] <elmo42> lol
[15:04:10] <elmo42> elmo42 is now known as elmo40
[15:24:10] <Connor> anyone have any ideas on how to route out a piece of wood to mate up with the rounded inset of this computer case? http://tinyurl.com/34pzkdz
[16:28:43] <lilalinux> * lilalinux doesn't click on tinyurls
[16:29:16] <elmo40> I do all the time. I have no worries... flash block, script block, Linux... what harm could happen?
[16:29:23] <skinnypup> its really a comp case if thats worth anything
[16:29:36] <elmo40> bland case at that.
[16:30:51] <skinnypup> ive gotten into these itx boxes with atom procs...
[16:31:52] <skinnypup> man small is nice , too bad via thinks so much $$ of their pico factor
[16:32:21] <psha> skinnypup: cheap ones are not rare
[16:32:58] <psha> i've bought mine for ~30$
[16:34:24] <Connor> Trying to be nice.. here is the long URL http://www.newegg.com/Product/ImageGallery.aspx?CurImage=11-154-091-S03&SCList=11-154-091-S01,11-154-091-S02,11-154-091-S03,11-154-091-S04,11-154-091-S05,11-154-091-S06,11-154-091-S07,11-154-091-S08,11-154-091-S09,11-154-091-S10,11-154-091-S11,11-154-091-S12,11-154-091-S13,11-154-091-S14,11-154-091-S15&S7ImageFlag=2&Item=N82E16811154091&Depa=0&WaterMark=1&Description=APEX%20MI-008%20Black%20Steel%20Mini-ITX%20Tower%20Compu
[16:35:18] <skinnypup> psha 30 bux , wow
[16:36:37] <psha> it's not too bad for computer cases here at moscow
[16:42:11] <lilalinux> SWPadnos: regarding my compensation problems. Couldn't that be fixed in emc?
[16:42:20] <SWPadnos> no
[16:42:23] <lilalinux> actually I don't think it's a CAM error
[16:43:12] <SWPadnos> cutter comp means that emc will keep the tool from going "outside the box" (in very loose terms)
[16:43:47] <SWPadnos> at the beginning and end of the compensated path, you have no intersecting path to let emc know that it should stop short of the end of the programmed line
[16:44:19] <SWPadnos> since you don't tell emc that there is an "obstruction", it proceeds to the end of the programmed line
[16:44:57] <cradek> you are expecting it to assume your hole is square. program a pentagon and you'll see that the assumption you want it to make is wrong.
[16:45:24] <lilalinux> i see
[16:45:25] <cradek> if you have four corners you must program five fully compensated lines to make the four corners
[16:45:54] <cradek> counting entry and exit moves, that makes 7 moves you must control to cut an inside square
[16:46:00] <cradek> there is an example of cutting an inside square in the docs
[16:46:43] <cradek> it uses tangential entry and exit arcs along a side of the square
[16:46:49] <SWPadnos> cradek, do you know if dxf2gcode has an option to do entry/exit moves in the middle of one of the lines (instead of a corner)?
[16:47:01] <cradek> no idea
[16:47:14] <cradek> I counted once: there are 72 programs on the internet named dxf2gcode
[16:47:17] <SWPadnos> ok. me either (that's what lilalinux is using to generate the code)
[16:47:19] <SWPadnos> heh
[16:47:22] <lilalinux> :-)
[16:47:24] <SWPadnos> I mean this one
[16:47:27] <SWPadnos> or was it that one
[16:47:41] <lilalinux> http://code.google.com/p/dxf2gcode/
[16:47:58] <skinnypup> i used that one too
[16:48:24] <lilalinux> I filed a bug report today: http://code.google.com/p/dxf2gcode/issues/detail?id=11
[16:48:27] <skinnypup> get sheetcam it lets you adjust entry and exit arcs and their lead-in-out radius
[16:48:38] <lilalinux> skinnypup: is sheetcam for linux?
[16:48:43] <skinnypup> yup
[16:49:01] <lilalinux> payware
[16:49:07] <skinnypup> not for lin
[16:49:17] <lilalinux> IC
[16:49:47] <SWPadnos> there is a Linux version, I think
[16:49:53] <SWPadnos> doesn't Les Newell sell that?
[16:49:59] <skinnypup> the linux version isnt updated as often as the other but its WAY more mature than other dxf to gcode
[16:50:05] <skinnypup> yeah Les
[16:50:46] <JT-Work> I use sheetcam on Linux for my plasma
[17:03:15] <elmo40> how much is it?
[17:03:35] <skinnypup> sheetcam for linux is free
[17:03:42] <elmo40> I need to make cases out of sheetmetal and my plasma. it would be nice to have something like that.
[17:03:48] <elmo40> oh?
[17:03:53] <elmo40> link?
[17:04:01] <skinnypup> sheetcam.com iirc
[17:04:06] <elmo40> works with 10.04LTS?
[17:04:16] <skinnypup> idk im on lenny
[17:04:23] <skinnypup> worked on 804
[17:04:53] <cradek> I think he currently says you have to use the windows version under wine on ubuntu 10
[17:05:06] <cradek> I think his linux version only works on ubuntu 8
[17:05:50] <JT-Work> you can try a limited output version for free, but the license is 110 pounds english money
[17:06:33] <JT-Work> but I see he doesn't say that for the linux versions
[17:07:02] <skinnypup> or install virtualbox and run sheetcam on 804 virtual machine
[17:08:03] <JT-Work> it does allow you to pick the entrance and exit moves but only at the ends of arcs or lines
[17:09:05] <elmo40> 2.0.2-1 Linux version. 3.1.5 Windows version.
[17:11:18] <elmo40> have you tried the plugin MyNesting?
[17:12:28] <elmo40> ah, windows only
[17:12:32] <skinnypup> i havent. won't touch m$ with a 10' pole
[17:13:19] <skinnypup> looks neat though
[17:20:34] <skunkworks> http://www.youtube.com/watch?v=19RM35yKGEc
[17:25:39] <elmo40> skunkworks: :)
[17:26:41] <elmo40> need more light... but I love it! :) The temperature display is slick.
[17:26:45] <IchGuckLive> Hi all from the cold Germany
[17:27:01] <elmo40> how is the Gulf leak?
[17:27:29] <IchGuckLive> seeld since a wihle
[17:28:03] <elmo40> so they say. is there no more live feed from the bottom?
[17:28:47] <IchGuckLive> i go to get some for you
[17:28:54] <elmo40> skunkworks: not a heated table, eh?
[17:29:14] <skunkworks> no clue. (just found the video)
[17:30:51] <IchGuckLive> elmo40: all feeds closed
[17:32:18] <elmo40> you can make a blob of plastic and do some engraving after it cools. I want to make my machine like that :)
[17:33:07] <IchGuckLive> elmo40: how mutch woudt you like to spend on
[17:36:00] <elmo40> I have the cnc part. just need the melting head
[17:39:22] <IchGuckLive> melting head what are you gone do wirer cut on styropor
[17:40:40] <elmo40> I want an extrusion (3D printing)
[17:40:49] <elmo40> same as that video http://www.youtube.com/watch?v=19RM35yKGEc
[17:50:58] <alex_joni> http://www.speedtest.net/result/1005470640.png
[17:51:07] <alex_joni> upgraded my internet connection today
[17:51:34] <alex_joni> should be 25/4 Mbit
[17:52:31] <micges> cool
[17:52:38] <skunkworks> nice!
[17:52:43] <micges> you have upc :|
[17:52:55] <micges> it is unaviable in my town
[18:01:14] <alex_joni> it's ~30$ / month (including 120 TV channels and phone)
[18:02:48] <micges> great
[18:03:15] <psha> andypugh: thanks for wiki text edits
[18:04:13] <andypugh> No problem. Hope you agree with what it now says?
[18:07:07] <psha> yes, it's correct
[18:09:06] <IchGuckLive> elmo40: the kress on the cnc woudt mill the part 100times faster
[18:18:35] <KimK_> !later Connor (sent via !later) Re: routing the rounded wood piece, couldn't you use a ball-nose cutter and rastering, as in the "Tux" example at nc_files/examples/3D_Chips.ngc ?
[18:19:05] <Connor> huh?
[18:19:39] <KimK_> Oh, you're on, sorry, I looked a bit ago and you were off.
[18:19:53] <Connor> no clue.. question is.. how to get the profile in gcode.
[18:20:18] <motioncontrol> excuse a simple question . what is the canned cicle fuction on lathe in emc ?
[18:21:59] <JT-Work> are you asking if there is one or about a G gode?
[18:22:01] <KimK_> Connor: Just measure it with a straightedge and maybe a radius gauge if you have a set, and do some judicious guesstimating?
[18:22:22] <motioncontrol> Yes in gcode
[18:22:35] <KimK_> Connor: Assuming you're not set up for probing
[18:22:39] <JT-Work> which g code?
[18:23:27] <motioncontrol> type g76 ecc...emc which canned cycle have on lathe ?
[18:23:27] <Connor> no. I'm not.. although.. I might be able to come up with a probe setup... I have a few servos laying around that I could turn into a 3d probe.
[18:24:15] <KimK_> motioncontrol: G76 is a threading cycle
[18:24:46] <Connor> Is there a way to do it with pictures maybe?
[18:25:28] <andypugh> motioncontrol: There are no roughing cycles, but I think you can use the drilling canned cycles
[18:25:29] <motioncontrol> ok g86 fuction on lathe ?
[18:25:49] <JT-Work> yes, you just have to set X to 0.000
[18:26:06] <JT-Work> and set the plane to G18 (I think)
[18:26:39] <motioncontrol> and g33.1 ?
[18:27:24] <JT-Work> yes
[18:28:05] <JT-Work> I might be wrong about the G18 the more I think about it you need to be in G17 if I recall correctly
[18:28:06] <KimK_> Connor: Maybe, but I would think you'd get more accurate results by just specifying what you want in g-code from the measurements. Besides, you could also just "freehand sculpt" a piece of wood like that with a belt sander.
[18:28:27] <motioncontrol> i thing g18
[18:28:51] <KimK_> JT-Work: You are correct, it's G18 for lathes
[18:30:05] <cradek> most canned cycles will not work in G18 on a lathe, because that plane means the cycle is along Y, which the lathe doesn't have
[18:30:23] <cradek> for example, if you want to use G81 to drill in Z, you have to switch to G17
[18:30:29] <JT-Work> that is what I was thinking about
[18:30:30] <andypugh> motioncontrol: If you want roughing cycles, then both JT and I have similar but different ways of doing it.
[18:30:47] <cradek> if you want to "peck part" (or whatever) with G83 along X, you have to switch to G19
[18:30:49] <JT-Work> it hung me up the other day can cradek came to the rescue
[18:31:04] <cradek> JT-Work: I need theme music
[18:31:40] <andypugh> http://www.bodgesoc.org/lathe/lathe.html is my way
[18:31:48] <JT-Work> I'm listening to Fats Domino singing Jambalaya with Jerry Lee Lewis atm
[18:31:58] <andypugh> For JTs way look on the forum under "subroutines"
[18:32:03] <motioncontrol> More thenks at all for info
[18:32:10] <KimK_> motioncontrol: G86 includes a spindle stop and withdraw in the middle. There are other choices available also, if that's not what you want.
[18:32:46] <KimK_> cradek: Thanks for the lathe tip, good to know.
[18:32:52] <motioncontrol> my question is only for have a idea the can the emc on lathe
[18:33:47] <motioncontrol> i don't know the g81 ,g82 ecc.have some image the work with associaton the gcode caned cicle ?
[18:34:16] <cradek> motioncontrol: start with the gcode quick ref, and you can click on them to get to the full documentation
[18:34:39] <cradek> start here: http://www.linuxcnc.org/docs/html/gcode.html
[18:35:13] <motioncontrol> good more thanks
[18:41:52] <KimK_> andypugh: That's a very interesting lathe panel you've got going there, I'll have to think about that awhile. (P.S.: s/boting/boring and s/defaut/default . Sorry, it's why I thought I could help on docs, lol)
[18:43:35] <andypugh> I am hopeless at reading my own text, because I know what I wrote.
[18:45:06] <KimK_> No problem. A friend of mine (who is a notoriously bad speller) says, "It's a small mind that can only spell a word one way."
[18:45:32] <andypugh> Fixed
[18:47:37] <KimK_> Yes indeed. Thanks. Any thoughts on your Arduino resolver situation? Or is that fixed already too?
[18:49:23] <andypugh> Seems to be working now, it was the uint that needed to count negatively...
[18:50:06] <KimK_> OK, congratulations
[18:50:37] <andypugh> Time will tell if it works. There is a fair bit of dither on it.
[18:51:54] <andypugh> I am wondering if i should have some sort of filter on the resolver outputs.
[18:53:00] <atmega> nice looking stuff, did you also post it to the arduino playground?
[18:53:33] <andypugh> I didn't. I might when I have it actually doing something.
[18:53:57] <andypugh> I did build a complete BLDC drive based on the same basic idea and a power driver.
[19:03:29] <KimK_> cradek: Back to your lathe tips for a moment, are those tips documented anywhere? I don't see them in the "Lathe Specifics" , or "G17, G18, G19" , or the "Canned Cycles" intro. There should probably be something in the "Lathe Specifics" section, especially since all the emphasis is on G18. I'll try to write something up. I'd appreciate hearing from you any that you've used and know for sure work.
[19:04:32] <cradek> KimK_: perhaps a rewrite of the canned cycle section would include information about how to use them on a lathe? We've talked about how the CC docs are not very good currently...
[19:05:06] <cradek> "if you want Z to peck, you have to be in G17" isn't really specific to lathes at all
[19:05:44] <KimK_> OK, maybe a canned cycle rewrite would be good too.
[19:06:33] <KimK_> I'll no doubt need your help on a few "mysteries" there, though.
[19:07:36] <cradek> if your changes are for trunk - not 2.4 - there is new behavior for the retract/initial plane (g98/g99)
[19:08:29] <cradek> you can find the relevant commits (and my idea of documentation) with git log --grep=g98
[19:09:02] <KimK_> OK, there's something else that I'm not up on. Yes, thanks, I'll have to catch up on it first.
[19:09:12] <JT-Work> oh, that is how you find that stuff, I assume you need the proper branch checked out
[19:09:40] <cradek> JT-Work: you could add --all to search other branches too, I think
[19:09:53] <JT-Work> cradek: thanks
[19:10:46] <KimK_> cradek: Yes, thanks. You're just full of good tips today.
[19:11:36] <JT-Work> * JT-Work e-mails the tip to myself so I can remember it by the time I get home
[19:27:20] <andypugh> Anyone know what G70 is intended to do?
[19:27:43] <cradek> I think on some machines (not emc) it means inch mode
[19:27:55] <andypugh> Aha!
[19:27:55] <ds2> isn't that stock removal cycle?
[19:29:14] <andypugh> I don't think so, in this context: http://www.pastebin.ca/1973850
[19:29:23] <andypugh> (problem G-code from the forum)
[19:30:09] <cradek> yep looks like it means 'inch mode'
[19:30:13] <ds2> G70 FINISHING CYCLE (P,Q)
[19:30:15] <ds2> G71 O.D. / I.D. STOCK REMOVAL CYCLE (P,Q,U,W,I,K,D,S,T,R1,F) (SETTING 72, 73)
[19:30:22] <cradek> but that's just a guess without knowing what system it was meant for
[19:30:29] <andypugh> Is the : meant to be a : ?
[19:30:37] <cradek> ds2: it's clearly not that, in this gcode
[19:30:59] <cradek> andypugh: I don't know what the : is supposed to do
[19:31:04] <ds2> ah yes, the context suggests it is not a lathe g-code
[19:31:06] <cradek> andypugh: clearly this isn't emc gcode
[19:31:25] <andypugh> No, from a Ungraphics postprocessor
[19:31:39] <JT-Work> : might mean a comment
[19:32:10] <cradek> in context, the : looks odd
[19:32:24] <cradek> maybe it's block-deletable?
[19:32:33] <JT-Work> or a post processor hiccup
[19:32:35] <cradek> I can see doing that for a tool change
[19:32:44] <cradek> brb
[19:34:00] <andypugh> Crikey, an impeller. He doesn't lack ambition.
[19:35:16] <JT-Work> andypugh: did you get it to run?
[19:35:23] <andypugh> Aye
[19:35:38] <andypugh> Swapped G70 to G20 and changed the : to an N
[19:37:04] <andypugh> It seem unnaturally big.
[19:37:26] <KimK_> I notice that the N numbers repeat, so apparently that's not a problem on a Unigraphics postprocessor?
[19:37:47] <andypugh> No idea. But I think it is meant to be in mm, 28" is fairly big.
[19:38:25] <KimK_> G21 then?
[19:42:53] <cradek> but 28mm is awfully small
[19:43:25] <skunkworks> screen shot?
[19:43:49] <andypugh> You would fabricate a 28" one, you would buy a 100mm one. Perhaps you would have to make a 28mm one?
[19:44:17] <cradek> 47.2mm/min is not much feed either
[19:45:13] <cradek> especially at 3500 rpm
[19:45:19] <cradek> I think it's an inch program
[19:46:52] <andypugh> www.imagebin.ca/view/3s122kab.html
[19:47:24] <andypugh> Or maybe not...
[19:47:58] <cradek> nope
[19:49:13] <andypugh> www.imagebin.ca/view/KldGZZ.html
[19:49:51] <cradek> heh, can't tell the scale from that...
[19:51:43] <JT-Work> secret submarine propeller...
[22:28:52] <flyback> * flyback hates life
[22:38:23] <flyback> * flyback is not suicidal, just very tired of a lot of bullshit
[22:51:36] <flyback> * flyback bbl
[23:37:10] <Connor2> Connor2 is now known as Connor