Back
[00:00:36] <pfred1> the concept of putting together a CNC machine is simple enough but getting all of the details to come together can be tricky
[00:02:23] <pfred1> you have to be good at electronics and mechanics to make your own machine
[00:02:44] <Connor> I build robots as a hobby.. so... :)
[00:03:30] <pfred1> so why'd you pick MDF?
[00:03:39] <pfred1> that stuff can't even support its own weight
[00:04:21] <Connor> Because I had no way to build it at the time with anything else.. Alum was too expensive.. and no metal work tools.. no lathe, no mill.. just a drill press..
[00:04:43] <Connor> That machine was built with a table saw, hand drill and the colt router itself.
[00:04:50] <pfred1> believe it or not you can cut aluminum pretty good with woodworking tools
[00:04:56] <Connor> most of it was assembled in my living room.
[00:05:03] <pfred1> like you can rip heavy aluminum plate on a tablesaw
[00:05:26] <pfred1> wear long sleeves thoug hthe chips come off pretty hot
[00:05:38] <pfred1> don't ask me how i know this ...
[00:06:17] <Connor> I know. I use my miter saw with carbide blade to cut 8020 and plate now..
[00:06:33] <pfred1> yup
[00:06:38] <pfred1> rips right through it
[00:06:42] <pfred1> just hang onto it
[00:07:07] <Connor> yea. and stop the saw after your through.. don't bring it back up still going.. or you'll rip off some of the teeth.
[00:08:10] <Connor> The Plexiglas lid came off of a old Dot matrix printer sound proofing box..
[00:08:49] <Connor> It was doubled hinged with that strip on the bottom.. I re did that way so I could open up the bottom and let longer parts tick out if needed.
[00:10:08] <pfred1> Connor did you ever see how I did my Z axis guides?
[00:10:38] <Connor> don't recall if I did or not.
[00:11:02] <r00t4rd3d> i put a old table saw blade on backwards and cut metal/aluminum.
[00:11:53] <pfred1> Connor it cost me $2
http://i.imgur.com/PHZCW.jpg
[00:11:54] <Connor> r00t4rd3d: WHAT? Backwards WTF ?
[00:12:20] <pfred1> I've heard about running the blades backwards I just run them normal
[00:12:45] <pfred1> instead of running backwards just go with a lower tooth rake angle
[00:12:48] <r00t4rd3d> forward sometimes they will bind
[00:12:59] <pfred1> yes that means yo uhave too agressive of a rake
[00:13:22] <r00t4rd3d> ive done that in a skill saw a qazillion times to cut metal roofing
[00:13:42] <Connor> pfred1: Now, That is partial board with laminate on it...
[00:14:06] <pfred1> Connor yes but it is tubed
[00:14:06] <Connor> Larger particls than MDF..
[00:14:17] <Connor> Yea.. interesting setup.
[00:14:31] <pfred1> I think i can even cut it down
[00:14:41] <pfred1> because I was paranoid extended it would wobble
[00:14:55] <pfred1> it doesn't wobble
[00:15:20] <pfred1> I saw a commercial machine with nested steel swuare tubes and I was like hey I could do something like that
[00:15:29] <Connor> My Z has rails on it.. The blocks are mounted and the section with the rails moves..
[00:15:45] <Connor> I have 6" clearance with the gantry and 4.25" Z movement..
[00:15:59] <pfred1> then this is inside it to move the inside tube up and down
http://www.youtube.com/watch?v=fHPKaHLzXes
[00:16:02] <Tecan> (fHPKaHLzXes) "Z_Bearings.avi" by "Paul Frederick" is "Tech" - Length: 0:01:00
[00:16:51] <Connor> pfred1: Good way to keep the screw from getting stuff on it..
[00:16:54] <pfred1> with my crap TB6560 motor drivers i can get it to go 1.2 IP/S
[00:17:08] <pfred1> 72 IPM?
[00:17:14] <pfred1> something like that
[00:17:53] <pfred1> I'm almost done with my Y axis now but I haven't taken any pictures of it yet
[00:18:05] <pfred1> it is ah different too
[00:18:33] <pfred1> I'm thinking by next week I should be done with it
[00:19:17] <pfred1> it is a skate bearing design
[00:20:03] <pfred1> sort of looks like engineered trusses
[00:21:16] <pfred1> LinuxCNC running on Debian
http://i.imgur.com/OjNG078.png
[00:22:06] <pfred1> Connor the fit for my motor clamp and the tube is accurate to 0.002 of an inch
[00:22:36] <r00t4rd3d> ReadError,
http://www.thingiverse.com/thing:65706
[00:22:37] <pfred1> but I'm crediting dumb luck for that
[00:22:45] <r00t4rd3d> print me those so i can fuck with people
[00:23:33] <pfred1> Connor I got my Bosch router at a flea market for $15
[00:23:55] <pfred1> when i saw it I was like oh i know what I can do with that
[00:32:47] <SadieD> Hi
[00:33:42] <SadieD> Would anyone be willing to indulge a few stupid questions?
[00:33:54] <pfred1> that's one
[00:34:10] <pfred1> no one seemed to mind either
[00:34:15] <SadieD> how many do I get?
[00:34:25] <pfred1> as many as you like
[00:34:44] <SadieD> Awesome
[00:35:35] <SadieD> I have a 1992 Haas VF-2 that needs a retrofit, I was thinking of going with Mach Motion but I've heard some good things about Linux CNC
[00:36:18] <pfred1> such as?
[00:37:10] <SadieD> It allows more control over the configuration of your machine
[00:37:12] <Connor> SadieD: You want full closed loop to the controller ?
[00:37:43] <pfred1> I wouldn't say more
[00:37:50] <pfred1> I would use the term total
[00:38:08] <SadieD> Connor: I'm not knowledgeable enough to know if that's necessary
[00:38:20] <SadieD> Connor: But I'm willing to learn
[00:38:54] <Connor> SadieD: Mach doesn't really support Servo's. You have to use a Motion Controller... which handles closed loop control. Mach just issues a Step/Dir command... and hopes the Motion controller can do it.
[00:39:02] <SadieD> pfred1: I'd prefer to use a GNU/Linux solution
[00:39:31] <Connor> Atom baed PC with Mesa 5i25 + 7i77 if your using servo's.
[00:39:33] <pfred1> SadieD you are futureproofed with open source
[00:39:52] <pfred1> with Mach who knows?
[00:40:06] <pfred1> the original guy just dissappeared one day
[00:40:41] <Connor> That's a big machine.
[00:41:02] <Connor> Are you planning on using existing servos ?
[00:41:06] <Connor> and drivers ?
[00:41:08] <SadieD> Would in be less expensive to go with a custom solution?
[00:41:38] <SadieD> Mach Motion is quoting ~$8000 for the control, interpreter, MPG, and PLCs
[00:41:48] <Connor> WOW
[00:42:06] <SadieD> Seems high to me
[00:42:36] <Connor> PC with Mesa cards.. $700.00
[00:42:47] <SadieD> Connor: Yes, my servos and drivers/amps are fine
[00:43:06] <Connor> What kind of input can the drivers/amps take?
[00:43:13] <Connor> and do they output encoder info ?
[00:43:50] <SadieD> I'm still gathering data, I'm waiting on the manufactured for documentation
[00:44:05] <SadieD> I'm kind of diving into all of this
[00:44:28] <Connor> I'm helping a guy who gutted a Cincinati Arrow 500 and replaced the servers / drviers and control system for less.
[00:44:51] <SadieD> Do you do contract work?
[00:45:15] <Connor> No. I do this as a hobby. and he's just a local guy whom I'm helping.
[00:45:50] <SadieD> Quite a hobby
[00:46:05] <pfred1> got to do something for fun
[00:46:05] <Connor> He had a Rf45 that he used for a long time running Mach3.. and then wanted a full VMC.. and sold it.. to fund the Arrow 500
[00:46:12] <SadieD> I haven't found a CNC serivce company that can do this sort of thing
[00:46:31] <Connor> I like doing stuff myself..
[00:46:37] <pfred1> me too
[00:46:59] <pfred1> my motto is screw it up myself and spend
[00:47:05] <Connor> Once you have details on the servos and driver/amps.. we can give you more direction on the control hardware needed.
[00:47:22] <SadieD> I'd like to do this retrofit myself, but there is quite a bit to learn. I haven't even built a RepRap yet
[00:47:42] <Connor> Don't rule out removing the existing servo's and amps and selling them for spare parts and buying new stuff..
[00:47:45] <pfred1> nothing from rep rap would be transferrible
[00:48:11] <Connor> It have a tool changer ?
[00:48:13] <pfred1> servos and steppers are different ballgames
[00:48:52] <Connor> My guy ordered TECO motors and Drivers from Mach Motion.. but they were setup for step/dir for use with Mach3. :(
[00:48:58] <SadieD> The Haas has an umbrella syle tool changer
[00:48:59] <pfred1> SadieD have you signed up at the website?
[00:49:18] <Connor> Tool changers are always the challenge..
[00:49:19] <pfred1> SadieD
http://linuxcnc.org/
[00:49:19] <SadieD> pfred1: Not yet
[00:49:30] <SadieD> Connor: PLC, ladder programming?
[00:49:40] <pfred1> SadieD the forum is the best ah forum for the more involved issues
[00:50:05] <Connor> LinuxCNC does support Ladder.. and I think we now have internal support for doing tool changers without having to use ladder.
[00:53:15] <Connor> SadieD: Question is what's your budget and what's your time frame...
[00:53:55] <SadieD> I'd like to go under $9k
[00:53:56] <pfred1> the thinner the shoestring the further you have to stretch it
[00:54:07] <SadieD> and ~2 months
[00:54:54] <Connor> You should be able to do WELL under that with LinuxCNC if you can use existing servo's and drivers/amps
[00:55:20] <Connor> and maybe a little over half if you have to gut everything and replace the servos and drivers/amps
[00:55:30] <pfred1> SadieD this might be some good bedtime reading
http://linuxcnc.org/docs/EMC2_Integrator_Manual.pdf
[00:56:08] <pfred1> give you an overview of what you are up against
[00:56:45] <Connor> Between the IRC, Forum and Mailing list.. you have allot of support options.
[00:56:55] <SadieD> Well thanks for the info, I'm registering for the forum. I'll get all my information together.
[00:57:17] <pfred1> the manual will be good to start with as soon as possible
[00:57:34] <Connor> almost 2am.. Time for bed.. need to get up tomorrow so I can help him finish up his conversion.
[00:57:58] <SadieD> g'night, thanks
[01:00:39] <Connor> SadieD: One other thing. You can install LinuxCNC Simulator on a existing ubuntu install and play around with it.
[01:01:45] <SadieD> Connor: I'll do that
[01:02:26] <pfred1> yes familiarity is a good thing gives context
[01:03:14] <pfred1> be bold and use a flash crive save a plastic tree
[01:03:19] <pfred1> flash drive even
[01:03:29] <pfred1> plus it'll run better off flash
[01:08:05] <SadieD> Can I just run it from my Kubuntu partition??
[01:11:51] <pfred1> it will run live
[01:11:59] <pfred1> you don't have to install anything
[01:13:19] <pfred1> until you gain some basic familiarity you probably should check out a few live sessions
[01:13:40] <SadieD> pfred1: Coolio
[01:14:04] <pfred1> I mean why just dive in?
[01:14:44] <SadieD> pfred1: Makes sense
[01:15:09] <SadieD> Thanks again for the info, I'm off
[01:15:11] <pfred1> then you can be like me and try to get LinuxCNC to run on Gentoo
[01:15:52] <pfred1> for that extra erg of speed
[01:16:11] <SadieD> pfred1: My brain is overloaded as it is
[01:16:28] <pfred1> yeah now I wouldn't reccommend yo utry to do that right off
[01:17:16] <pfred1> it is good to become immersed though
[01:17:41] <SadieD> pfred1: True
[01:18:01] <pfred1> I can't say a little knowledge never hurt anyone, because it has, but more knowledge usually straightens them out
[01:18:11] <SadieD> pfred1: It's just a bit overwhelming
[01:18:49] <pfred1> the thing is to get exposed to all of the concepts then tackle the details
[01:18:56] <SadieD> pfred1: I'm a machinist/prgogrammer, this is all new to me
[01:19:20] <pfred1> well don't expect it all to click as you go hope that it clicks eventually
[01:19:48] <pfred1> it is like looking at a picture from an inch away you have to scan it a lot for it to become clear
[01:20:13] <SadieD> pfred1: Sound advice. All the information is out there, I just have to absorb it.
[01:20:25] <pfred1> and absorb it over and over
[01:21:15] <pfred1> I guess in a way the process is kind of like building up a hologram
[01:21:27] <SadieD> Well I'll be back on for sure, I'm off to absorb some Rum.
[01:21:29] <pfred1> it starts out fuzzy but the details build up
[01:22:00] <SadieD> Night Fred
[01:22:26] <pfred1> yeah the best advise I can give is just keep on trudging through everything people can clear you up on what really baffles you
[01:23:16] <pfred1> read that manual
[01:37:51] <r00t4rd3d> dog brought to the vet after eating a dozen pot brownies:
[01:37:52] <r00t4rd3d> http://i.imgur.com/vWevas9.jpg
[01:47:01] <Valen> r00t4rd3d: :-,
[01:47:04] <Valen> :-<
[01:59:33] <r00t4rd3d> damn, Austin might be getting Google Fiber.
[01:59:38] <r00t4rd3d> fuck texas
[02:19:24] <DJ9DJ> moin
[02:23:24] <Valen> r00t4rd3d: in theory I'll be getting fiber within a year or so
[02:29:59] <r00t4rd3d> you order some ceral from the US?
[02:32:03] <Loetmichel> mornin#
[05:17:49] <andypugh> Last night my machine stopped. As it was exactly at the end of a toolpath I simply assumed that I had hit the wrong button in PyCAM, and only exported the roughing path.
[05:18:27] <Valen> never a good start andypugh
[05:18:36] <andypugh> Looking more closely the G-code has stopped at an M5, and the machine is dead. I have no idea why this would be.
[05:21:03] <andypugh> It just seems too much like a coincidence. But I have no idea why G-code execution stopped, or why the machine is now dead. It may be that the 24V supply has died, and that there is some tie-in to the machine enable that I have forgotten putting in.
[05:22:04] <andypugh> When I say "dead" the VFD is dead and the steppers don't move, but the 7i43 is fine, and encoder counters are working.
[05:22:36] <andypugh> I guess I need to pull it apart and check stuff.
[05:23:18] <Valen> start with a multimeter
[05:23:43] <Valen> rebooted the machine?
[05:23:59] <Valen> turned the power off to *everything* and started over?
[05:24:07] <andypugh> Yes, though I was _really_ hoping not to lose position as I am in the middle of a job :-(
[05:24:20] <Valen> it should save position
[05:24:23] <Valen> or at least it can
[05:24:29] <andypugh> Steppers....
[05:24:45] <Valen> it won't save position with steppers?
[05:24:53] <andypugh> And I proved that the steppers were dead by jogging.
[05:25:30] <andypugh> Ah well, best get to it.
[05:25:38] <Valen> fingers crossed
[05:26:07] <archivist> what is in common between vfd and steppers
[05:27:16] <Valen> got an I/O card or something?
[05:27:32] <archivist> power supply
[06:00:19] <andypugh> PSUs are good. It is looking like an e-stop thing.
[07:07:29] <jthornton> how do you get updates if you don't have an internet connection?
[07:09:16] <archivist> carrier pigeon
[07:09:52] <jthornton> what does the pigeon carry?
[07:09:57] <archivist> http://tools.ietf.org/html/rfc1149
[07:11:28] <archivist> later updated
http://www.ietf.org/rfc/rfc2549.txt
[07:12:53] <andypugh> Broken wire in the e-stop loop to the button on the milling head...
[07:13:15] <andypugh> Strange failure. It is in a cable chain and everything.
[07:13:34] <archivist> wire type?
[07:14:14] <archivist> single/stranded
[07:15:34] <archivist> we used some stranded for print head ribbon solenoid, it failed rapidly due to the tight bend we had
[07:15:48] <andypugh> For no good reason, 4 x individually-screened 7x0.2
[07:16:04] <andypugh> 100mA current though. Maybe too much?
[07:16:39] <archivist> we replaced with gore flexible, no failures then at all
[07:17:01] <archivist> the way you bend cable really matters
[07:21:23] <archivist> is the cable bundle causing a small radius bend?, there must be no small radius bending
[07:57:02] <andypugh> No idea. It's been replaced and the machine is working again.
[07:57:25] <andypugh> I wonder if I put in some sort of program-stop linked to the e-stop circuit?
[07:59:48] <gabewillen> andy did you ever successfully get you high speed serial hal component working?
[08:00:57] <andypugh> For a while, yes. Then I lost it.
[08:01:31] <gabewillen> maybe you should clean up after your self and you won't loose things so easily... :) kidding
[08:01:45] <andypugh> I lost it when cleaning up files, actually.
[08:01:53] <gabewillen> lol
[08:02:02] <gabewillen> i hate that
[08:02:07] <andypugh> But as it was no longer being used at the time, I didnt worry about it too mush.
[08:02:46] <gabewillen> i had a borrowed copy of windows 7 i use from time to time. I seem to have deleted it and it wasn't backed up on my server. My favorite borrowing website was shut down and now i don't have it
[08:03:55] <andypugh> gabewillen: Here is a sample .comp for using the Mesa UART.
http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=blob;f=src/hal/drivers/mesa_uart.comp;h=be8bf84d4d649f7eb756df6641457ba00fa52631;hb=refs/heads/master
[08:05:00] <gabewillen> im not using the mesa though, i have a alpha rtusb driver working on my 3.6.27 kernel. I can stream 1.5mbs from usb serial on my arm processor. was looking for some more examples
[08:05:30] <andypugh> Ah, well, I don't think my stuff is any help at all.
[08:06:13] <andypugh> The actual serial stuff is all done in the FPGA.
[08:06:28] <gabewillen> ahh, this is strictly from arm to pc
[08:06:49] <gabewillen> its working pretty good for now. Though i don't know if it will work on the 2.6xx kernel
[08:07:54] <gabewillen> i might write a cross platform(microprocessor/arm) C library for interfacing with hal directly similar to that of hostmot2 and mesa
[08:08:01] <gabewillen> over usb
[08:08:21] <gabewillen> i know the all about the redflags when it comes to usb and realtime
[08:08:39] <andypugh> hal_input does a lot of what might be needed.
[08:09:08] <gabewillen> Im using it for reading 7200 cpr encoder.... just got it working at 12:00 AM last night
[08:09:21] <gabewillen> little 90mhz cortex 2
[08:09:26] <gabewillen> i had laying around
[08:11:19] <gabewillen> i did finally succesfully write an fpga program.... last night wow, what a steep learning curve that is.
[08:12:38] <gabewillen> i will admit thought fpga's are pretty bad ass.... the abilitly to whip up an digital circuitry you want on the spot is awesome.
[08:13:35] <gabewillen> i have never been so excited to see an led flash before
[08:31:55] <AR__> lol
[08:32:44] <AR__> gabewillen, verilog or vhdl
[08:39:30] <L84Supper> http://linuxgizmos.com/mantis-hexapod-robot-embedded-linux-computer/
[08:40:59] <L84Supper> http://www.youtube.com/watch?v=1h9Mw-s9mzI&feature=youtu.be 2:30 video with > :30 of credits
[08:41:01] <Tecan> (1h9Mw-s9mzI) "Mantis - Two Tonne Turbo Diesel Hexapod Walking Machine" by "winchymatt" is "Tech" - Length: 0:02:32
[08:43:14] <L84Supper> nice use of CAN bus outside of an auto
[08:46:34] <jdh> CAN on dive computers and marine electronics
[08:47:44] <L84Supper> not sure why they chose this: iCOP PC/104 SBC, based on a DMP Vortex86DX CPU
[08:48:09] <L84Supper> it's powered by a 2.2 liter turbo diesel
[08:48:52] <L84Supper> maybe the RPi was out of stock? :)
[08:49:24] <andypugh> He probably wanted the PC104 format CAN interface.
[08:50:57] <andypugh> I guess it isn't "for" anything?
[08:51:03] <L84Supper> pcw_home:
http://linuxgizmos.com/99-sbc-runs-embedded-linux-on-1ghz-dual-core-arm-soc/
[08:51:34] <Felix29> hi there
[08:51:41] <andypugh> Hi
[08:52:19] <Tom_itx> andypugh you didn't fatigue your estop wire fixing the PSU the other day did you?
[08:52:31] <andypugh> Different machine.
[08:52:35] <Tom_itx> oh
[08:53:22] <andypugh> It's actually quite annoying. There are several jobs I could be doing, all of which need the lathe, which is doing a long CNC milling job...
[08:56:05] <Felix29> L84Supper: I am the coors you met a few days ago
[08:56:24] <Felix29> I guess people think it's a stupid nickname, so I changed it :)
[08:58:27] <IchGuckLive> hi all B)
[08:58:45] <Tom_itx> Felix29, i just figured you lived in colorado
[08:59:39] <L84Supper> Felix29: I thought coors was fine, unfortunately IRC is the playground of many petty tyrants and the self absorbed
[08:59:42] <IchGuckLive> colorado is a nice state
[09:00:16] <IchGuckLive> felix is a european name
[09:00:20] <Felix29> Well, I'm in indiana
[09:00:34] <Felix29> and from germany, so that explains the name i suppose :P
[09:00:36] <Tom_itx> i noticed that the other day
[09:00:50] <IchGuckLive> Felix29: im in germany B)
[09:02:12] <Felix29> you don't see it in writing, but people always notice when i speak
[10:16:32] <IchGuckLive> im off by
[10:30:15] <skunkworks> JT-Shop: how are you physically interfacing your vfd using serial?
[10:40:34] <DJ9DJ> i assume with wires ;)
[10:41:45] <pfred1> maybe they went with an optical interface?
[10:41:53] <pfred1> then it would be with fibers
[10:42:17] <pfred1> would be nice isolation that's for sure
[10:42:44] <DJ9DJ> yeah, or by radio ;)
[10:42:59] <pfred1> that might be difficult to accomplish in a noisy environment
[10:43:24] <pfred1> it'd be impressive if you could manage to do it though
[10:52:12] <JT-Shop> skunkworks: yes via modbus
[11:46:05] <mhaberler> anybody using hardy on an amd64 platform ? please get in touch with me
[11:52:21] <pfred1> mhaberler hardy heron?
[11:52:27] <archivist> mhaberler, I have that one this pc I am using
[11:53:13] <mhaberler> yes, ubuntu 8.04 with an RTAI kernel on an amd64 platform, I am trying to find the people who actually use that
[11:53:35] <pfred1> I can never keep Ubuntu's names straight with their versions
[11:53:55] <pfred1> I just setup LinuxCNC on Debian Squeeze
[11:54:08] <pfred1> it was a pain but I finally managed to do it
[11:54:30] <pfred1> I need to try Gentoo again
[11:55:18] <pfred1> so I'm pretty up on the kernel config for an AMD64 now because that is the system I did it on
[11:55:34] <pfred1> although I set it up for 32 bits
[11:56:01] <pfred1> still being multicore that does make it a little different than a single core
[11:56:36] <pfred1> you have to pick choose sparse IRQs for it to work
[11:56:41] <mhaberler> thats fine - I am specifically looking for people who run hardy and want something newer than v2.5_branch
[11:56:59] <pfred1> I used kernel 2.6.38.8
[11:57:16] <pfred1> it is the latest the tarball of RTAI will patch against
[11:57:22] <pfred1> I couldn't get their cvs to build at all
[11:57:48] <pfred1> well I didn't try too hard either it failed and I was like back to the tarball
[11:57:50] <mhaberler> well that's what we get for using university fallout. Anyway:
[11:57:58] <archivist> this is not running rtai though
[11:58:05] <mhaberler> the issue is as follows:
[11:58:08] <pfred1> xenomai?
[11:58:50] <carper64_lb> gong to setup on 12.04.2 in the next cpl of days using the xenomai kernel
[11:58:58] <mhaberler> the new RTOS branch will not build on RTAI in sim mode, and it needs some absurd workarounds for rt mode
[11:59:20] <mhaberler> this is why I suggest dropping support for hardy/amd64 after v2.5_branch
[11:59:42] <pfred1> ah I think i see what you mean
[11:59:47] <mhaberler> which mean that master (into which the RTOS work likely will be merged) will not work on hardy
[12:00:04] <mhaberler> but then I still have to find a scenario why one would ever want to do that
[12:00:07] <pfred1> hardy is pretty old now
[12:01:34] <mhaberler> the issue on the table therefore is: I propose to declare hardy end-of-life after v2.5_branch, and want to hear from folks which are a) affected b) cannot upgrade for good reasons to a postwar base platform) c) need to run master for some reason
[12:01:54] <pfred1> isn't hardy EOL as far as Ubuntu is concerned?
[12:02:08] <mhaberler> "Ubuntu 8.04's support ended on 12 May 2011 for desktops"
http://en.wikipedia.org/wiki/List_of_Ubuntu_releases#Ubuntu_8.04_LTS_.28Hardy_Heron.29
[12:02:19] <pfred1> like are there even any repos that host it now?
[12:02:20] <mhaberler> but this project has a fad with collector item releases
[12:02:28] <archivist> I still have hardy on the machines too but 32 bit
[12:03:06] <pfred1> archivist yes but if you wanted to install any software you couldn't just apt-get it today well not without some fooling around
[12:03:11] <mhaberler> it is installable on vintage amd64 machines - it fails for instance on my ASUS E45 - the kernel doesnt come up properly any more (hardware built 2011)
[12:03:36] <pfred1> mhaberler you can put almost any kernel you like into almost any Linux distro
[12:04:16] <mhaberler> sure, but this is a management of expectations question, not really a technical one
[12:04:21] <pfred1> there are some compatibility issues with other system utilities that limit how far you can advance
[12:05:08] <pfred1> like Debian Squeeze ships with 2.6.32 I think but I bumped it to 2.6.38
[12:05:35] <archivist> LTS should mean LTS as once a machine is working you dont want to be futsing with it too much
[12:05:42] <mhaberler> this question is really directed at the folks who arent that familiar with upgrading themselves
[12:06:01] <pfred1> archivist some knuckleheads worry about security updates more than I think they should
[12:06:06] <mhaberler> amen - that includes installing master and expecting it to work 2 years after platform EOL
[12:06:36] <pfred1> I mean I ran SUSE 8.0 for about 8.0 years before that system self immolated in the dead of the night on me
[12:06:47] <mhaberler> well if one thinks an RT kernel is a viable firewall she might be in for surprises
[12:06:55] <pfred1> had nothing to do with Linux was a total hardware meltdown
[12:07:39] <pfred1> usually after a year or so I stop installing new software on systems
[12:07:48] <pfred1> by then I have about all I want
[12:08:31] <archivist> boost seems to break anything it is involved with, that normally stops me upgrading software
[12:08:52] <pfred1> really?
[12:09:30] <mhaberler> where do you get that from? I still have to see a problem with it
[12:09:47] <pfred1> yeah my CNC PC I am not interested in anything other than LinuxCNC running on it
[12:10:01] <pfred1> heck I want to kill cron and the loggers
[12:10:36] <mhaberler> archivist: next time you find 'boost breaks anything' I do expect you to file a bug report
[12:10:54] <mhaberler> that's the way to do it, sorry
[12:11:05] <archivist> has been in other projects, I hate it
[12:11:12] <pfred1> bah who wants to bug folks with bug reports?
[12:11:40] <mhaberler> if such statements are the alternative, I do prefer a bug report
[12:11:49] <archivist> I dont run the latest like others do
[12:12:05] <pfred1> yeah when stuff breaks I figure I'm doing something wrong
[12:12:17] <mhaberler> did you ever have a unsresolved problem with boost _and_ linuxcnc master?
[12:12:33] <archivist> I am fed up with the "must upgrade a pile" that use of libraries like that cause
[12:13:11] <pfred1> archivist my opinion varies on upgrades once a system runs how I like it I am usually less inclined to change anything on it
[12:13:44] <pfred1> I mean when I'm good I'm good
[12:13:53] <archivist> pfred1, I agre if not broken dont change
[12:13:57] <pfred1> why tempt the fates?
[12:14:33] <pfred1> some folks like the gentoo crowd they're upgrade fanatics
[12:14:45] <pfred1> new stuff came out 5 minutes ago I have to install it!
[12:16:55] <pfred1> in order to get RTAI to build i had to downgrade a bunch of stuff and they thought I was nuts
[12:17:59] <pfred1> well the biggest problem was python for LinuxCNC I guess
[12:19:39] <L84Supper> sounds more like Arch linux than Gentoo
[12:19:44] <IchGuckLive> will there be a new liveCD including 2.5.2
[12:21:08] <pfred1> heh I built and ran 2.6.0~pre
[12:21:08] <IchGuckLive> hi chopper79
[12:21:14] <pfred1> http://i.imgur.com/FUL7PwG.png
[12:21:19] <IchGuckLive> and did the hack of thc work
[12:21:38] <IchGuckLive> pfred1: how does not
[12:21:59] <IchGuckLive> pfred1: whar r are yoiu on
[12:22:27] <pfred1> IchGuckLive I did it in Debian Squeeze but I may give Gentoo another crack now that I've worked out all of the bugs
[12:22:48] <IchGuckLive> why gentoo
[12:23:02] <IchGuckLive> all the big CAD CAM are on ubuntu
[12:23:02] <pfred1> because i can custom build the whole system for the arch
[12:23:17] <pfred1> actually the kernel that ships with LinuxCNC is a mess
[12:23:28] <pfred1> how it is configured
[12:23:36] <IchGuckLive> ah so this is the goal for the manufactures to go linuxcnc on there mashines
[12:23:37] <pfred1> not too optimal for anyone really
[12:23:57] <pfred1> strange i couldn't even get the kernel source for LinuxCNC either
[12:24:48] <pfred1> there must be some trick to it that I don't know
[12:24:50] <IchGuckLive> there are lots of manufactures reliing on linuxcnc and get CProggers on his to tranfer the code to there system
[12:25:29] <pfred1> because i can usually manage to get kernel sources for systems that are dpkg maintained
[12:25:49] <pfred1> well always up til when i tried for LinuxCNC
[12:25:54] <IchGuckLive> pfred1: im not on your thinking stay clear and go open as mutch as needed
[12:26:22] <pfred1> but I did have the config file
[12:27:51] <pfred1> diffing that against one of my custom configs was somewhat illuminating
[12:41:18] <IchGuckLive> Felix29: ?
[12:50:46] <andypugh> mhaberler: I am running a machine made in 2003 with Dual Xeons (the original, first, Xeon) and even I am not running Hardy any more.
[12:52:07] <mhaberler> well I didnt have you in mind for that question ;)
[12:53:01] <mhaberler> the whole point of the new RTOS work is to support _newer_ platforms, and that gets at odds with supporting stuff which is EOL for years now
[12:53:36] <archivist> the poor in society have to use what they can get their hands on
[12:54:06] <mhaberler> are you saying one cannot upgrade from hardy to say lucid for monetary reasons?
[12:55:39] <archivist> I use older hardware for monetary reasons, I did try lucid for a a few minutes for a job(server) but went back to hardy for that
[12:56:38] <archivist> investing time in learning an os and its foibles is a cost too
[12:57:37] <bedah> <-- is using linux mint now, it's debian as i am used to it, now blinky and effects like in original debian/ubuntu
[12:57:48] <bedah> -now+no
[12:58:58] <bedah> hm.. maybe it would be worth to make linux mint the basis for linuxcnc
[12:59:16] <IchGuckLive> im on 10.04 its amazing all telescop printer mashines router plasmas robots even the refrituater on EIB takes it
[12:59:52] <IchGuckLive> the hole house system managment
[13:06:11] <andypugh> archivist: Though there is the other question of whether your old machines need any features that are not in 2.5.2
[13:07:27] <archivist> not at the moment no
[13:09:01] <jdh> http://raleigh.craigslist.org/tls/3725596266.html
[13:09:17] <jdh> any good?
[13:10:43] <archivist> looks usable, has screw cutting gearbox
[13:11:25] <archivist> possibly has self acting facing
[13:13:03] <jdh> never heard that term
[13:13:47] <IchGuckLive> jdh: lots of work on this
[13:13:57] <IchGuckLive> does not look to be that stable
[13:14:03] <archivist> look at the middle lever on the apron looks like it has 3 positions
[13:14:37] <jdh> like a power feed for teh cross slide?
[13:14:49] <archivist> yes exactly that
[13:15:11] <archivist> my southbend has it
[13:16:41] <jdh> In 1967 a minor change was made to the apron-mounted mechanism that engaged the (standard-fit) power cross feed with the provision of a simple, ball-ended toggle arm to slide the selector button in and out.
[13:16:55] <archivist> http://www.lathes.co.uk/atlas/
[13:17:35] <jdh> has no belt, so no way to tell if it really works.
[13:18:14] <IchGuckLive> jdh: 1967 way older then the most here
[13:18:48] <IchGuckLive> im off by for today ! B)
[13:20:03] <andypugh> That's about the perfect age for a machine tool.
[13:21:48] <jdh> I could put it next to my Jet 9x20 that has been sitting in the corner under a tarp since I bought it.
[13:21:59] <pfred1> sweet
[13:22:16] <pfred1> does metal get harder as it ages?
[13:22:34] <pfred1> or is that just a delusion I suffer from?
[13:23:26] <archivist> cast iron should be aged before machining for best results
[13:23:30] <pfred1> I can even swear I've noticed it in some metal I've owned for a very long time in my life
[13:23:56] <pfred1> like i have these really crappy punches I bought whe n Iwas a little kid
[13:24:03] <pfred1> they're not nearly as crappy today
[13:24:44] <pfred1> I swear they are harder than they used to be
[13:26:22] <pfred1> I have some really old woodworking tools too and the steel of them can be pretty hard I can imagine it might not have been as hard a long time ago
[13:27:13] <pfred1> I know concrete never stops getting harder as it ages
[13:27:47] <pcw_home> I know glass get harder to the point that its almost impossible to cut after a number of years
[13:27:55] <pfred1> well glass flows
[13:28:05] <SWPadnos> pfred1, if you actually use the punches, they may be getting hardened by the pounding
[13:28:10] <SWPadnos> a little like railroad steel
[13:28:12] <pfred1> and it gets more brittle
[13:28:37] <pfred1> SWPadnos I thought about that and they were so bad for so many years I didn't really use them
[13:28:47] <SWPadnos> heh
[13:29:02] <SWPadnos> well, they could be air-hardening steel or something funny
[13:29:19] <pfred1> I thought that meant after being heated up?
[13:29:27] <pfred1> not just sitting in the air
[13:29:31] <archivist> glass flowing is supposed to be an old wives tail, think about toughened glass
[13:29:49] <pfred1> yes I've heard old windows were just made badly
[13:30:06] <SWPadnos> no, some machine steels (like for tools or jigs or whatever) are air-hardening or oil-hardening
[13:30:11] <pfred1> but i have seen old windows that were all kinds of distorted and i can't imagine anyone put them in like that
[13:30:15] <SWPadnos> never used any, but I've seen them at McMaster
[13:30:36] <archivist> there are other effect in cutting too, make sure you clean the glass and score it properly
[13:30:53] <pfred1> nah old glass doesn't cut right
[13:31:03] <pfred1> I've heard it from glaziers that knew their trade
[13:31:28] <archivist> pfred1, old glass was flattened from a blown tube
[13:31:36] <pfred1> might be due to some strange pitting and getting dirt in it though
[13:32:03] <archivist> I have cut curves in old glass, was fun
[13:32:10] <pfred1> another trick to cutting glass is using kerosene to lubricate the cutting wheel
[13:32:25] <pfred1> pros don't cut glass dry
[13:32:43] <archivist> seen that and its true
[13:32:51] <pfred1> if yo uwatch them they have a little tin of crap they dip the cutter into
[13:32:55] <andypugh> <puts on materials scientist hat>. Oil-hardening or Air-hardneing in steels only decribes how they need to be quenched from red-hot.
[13:33:14] <pfred1> that crap is kerosene
[13:33:35] <archivist> kerosene/paraffin
[13:33:52] <andypugh> Old glass really is very dificult to gut. This is due to tiny amounts of water being absorbed, just enough to put the outer skin into a slight compression, which toughens it.
[13:34:12] <pfred1> yeah the you can't cut old glass thing isn't a myth
[13:34:23] <pfred1> I mean yo ucan cut it but it isn't nearly as good as fresh glass
[13:34:29] <pcw_home> Thats been my experience
[13:34:47] <pfred1> I'm like 50::50 on oldglass and 100% on fresh
[13:35:18] <sadara> Is a M40x1.5 thread threaded onto a 40mm od rod?
[13:36:23] <pfred1> I never understood metric threading
[13:36:45] <andypugh> So, I was stood there looking at my work on the mill, wondering how I could drill holes in the right places to flip it and get the right alignment. Then I remembered it is sat on a CNC milling machine :-)
[13:37:00] <andypugh> sadara: Yes.
[13:37:06] <pfred1> I mean I've read about it and right after I've read about it I can suss out what is going on but whatever is going on never sticks with me
[13:37:20] <andypugh> Mecause 40mm pitch on a 1.5mm diameter would be very difficult to cut :-)
[13:37:27] <archivist> sadara, cut or rolled
[13:37:56] <pfred1> I made some not half bad acme taps out of acme threaded rod
[13:38:14] <archivist> either way it fits in a 40mm hole
[13:38:19] <pfred1> only trouble is the threat they cut is too tight because the tap isn't clearance sized
[13:38:41] <pfred1> but I've cut shallow threads into mild steel with them
[13:39:10] <pfred1> I did that for a really snug fit on my lead screw stop collars
[13:39:25] <pfred1> I still have grub screws in them though
[13:39:38] <andypugh> Actually, I do keep intending to make a table of threading depths. 40mm is the nominal diameter that it would be if the threads came to a sharp point, but they have a flat so are a bit smaller. Then there is the fact that a nut has clearance (never try to wind a tap into a die-nut!)
[13:40:50] <sadara> Thanks!
[13:40:51] <pfred1> I've seen people making some pretty gnarly homemade taps on the net
[13:41:11] <andypugh> The only really easy thing to calculate with metric threads is the tapping drill size. Just subtract the pitch from the diameter.
[13:41:14] <pfred1> so when I made mine I wrote an article about how I did it
http://www.instructables.com/id/Make-an-Acme-Tap/
[13:42:21] <andypugh> So, turning to 40mm with a formed insert then threading to the expected depth with that insert will give you a very much oversize thread. Though this is true of any thread, not just metric.
[13:43:05] <pfred1> I still need to find the right hole to drill for tapping my lead nuts
[13:43:14] <pfred1> so far I'm on the tight side
[13:44:04] <pfred1> plastic is pretty stretchy though
[13:44:29] <pfred1> I'm making my lead nuts out of HDPE
[14:48:56] <Nick001-Shop> I have a compound slide that has steppers which I have running. The other end of the ball screws have a brake to hold it in position. Is there a way of operating them automatically when I call a move command or do I have to go the M function route to clamp and unclamp them?
[15:03:33] <Connor_Arrow500> Heya
[15:04:59] <Connor_Arrow500> Working on Pete's Arrow 500... We're getting some realtime delay mesgs..
[15:05:37] <cradek> usually a stepper has plenty of holding power. I've never heard of a stepper being powerful enough to move an axis but not hold it in place.
[15:05:44] <pcw_home> what servo thread rate?
[15:05:55] <Connor_Arrow500> 500,000
[15:06:11] <pcw_home> that will be marginal on a Atom MB
[15:06:25] <Connor_Arrow500> okay.. back it down to 1Millon ?
[15:06:36] <pcw_home> I would
[15:07:20] <pcw_home> you wont gain much with a 2 KHz thread rate unless you are running torque mode drives
[15:07:34] <Connor_Arrow500> okay. nope. just Speed mode.
[15:07:53] <Connor_Arrow500> Pete's standing by me.. He says Hi!
[15:08:37] <pcw_home> yeah the velocity loop on the drive should 'keep on the road' for 1 ms
[15:09:21] <pcw_home> Hi pete from pete
[15:09:29] <Connor_Arrow500> ROFL
[15:14:56] <andypugh> Nick001-Shop: Sorry to take so long to reply, but someone did once come up with a scheme to lock-out his Z axis when not moving.
[15:16:03] <Nick001-Shop> Any write-ups somewhere?
[15:16:23] <JT-Shop> my vmc has a lock on the Z axis but it is off when the drive is in control and goes on when the drive goes to off IIRC
[15:16:25] <andypugh> I am trying to find it.
[15:16:43] <andypugh> But as Cradek said, it is probably unnecessary
[15:17:36] <JT-Shop> unless your trying to hold the axis when power is off...
[15:17:43] <pcw_home> it does seem odd with step motors unless it there to keep it from moving when powered down
[15:17:50] <pcw_home> Ha
[15:18:03] <JT-Shop> ^^^ what Peter said :)
[15:18:07] <Nick001-Shop> need to keep the 1 axis from vibrating when the other axis is drilling a hole
[15:18:09] <archivist> how coarse is the leadscrew
[15:18:53] <archivist> and steppers on odd microsteps will not hold that well
[15:19:04] <Nick001-Shop> about 0.125 lead - 8 turns to the inch
[15:20:09] <Nick001-Shop> not using microstepping
[15:21:43] <Nick001-Shop> slide assembly came with them installed and seems like a good ides for what I'm doing
[15:22:16] <andypugh> Nick001-Shop: This is the thread I was thinking of:
http://thread.gmane.org/gmane.linux.distributions.emc.user/26042/focus=26053
[15:22:57] <andypugh> But you can probably do it with just HAL, a timedelay that resets on the output of a ddt.
[15:25:09] <pcw_home> I think peak stepmotor torque would be 90 electrical degrees from null or 1/1600"
[15:25:11] <pcw_home> so unless that's too much, I don't imagine you will gain much by locking
[15:26:01] <pcw_home> unless the stepmotors cant hold against the forces
[15:29:13] <Nick001-Shop> Don't know if they'll hold or not. Still have to build a 3 spindle block to mount on top of the slides and then try it out. guess I'll just leave then energized to test it out. Then see if I have to ask again.
[15:42:14] <archivist> you should always keep steppers energised for the holding torque
[15:44:58] <sadara> Is there currently a method of connecting a "single ended/unbalanced" 5v quadrature encoder directly to a parport/breakout? ( this is for a gantry stepper, however lets not get into that for the moment)
[15:45:06] <Nick001-Shop> the stepper will stay energized - the brake is on the opposite end of the ballscrew and is used to make sure the slide stays in position while holes are being drilled and tapped
[15:45:50] <pcw_home> sadara: yes this is supported by linuxcnc
[15:46:17] <pcw_home> (by the encoder component)
[15:46:32] <sadara> What sort of speed will it be able to accept? (parport hardware wise, not rpm)
[15:46:57] <sadara> 100hz?
[15:47:01] <pcw_home> perhaps 1/2 the base thread rate (if you include margin)
[15:47:07] <sadara> ok
[15:47:16] <sadara> great, thanks
[15:49:33] <sadara> the next question, and I would bet that this has been covered a lot, is is there any benifit to including the encoders in the loop? I only included the encoders to generate a fault if the steppers miss a step
[15:50:20] <sadara> From my research a cpl of years ago, there was no advantage, and several disadvantages
[15:50:55] <sadara> There now seem to be step/dir drivers with encoder input, so I was wondering if anything has changed?
[15:51:35] <pcw_home> There are use cases where its is a advantage (say good linear scale and high ustep ratio step drive)
[15:52:30] <pcw_home> this is not the same as commercial close loop stepdrives (which use the stepmotors as servos)
[15:52:49] <pcw_home> closed loop I should say
[15:53:43] <sadara> I'm trying to avoid ustepping for the moment... however adding linear encoders is feasible
[15:54:33] <sadara> also, I read of a system that, when detecting a missed ustep, would subtract a ustep from the path of the other axis, and pulse a couple more volts into the step drivers. I can see how this could work with emc2, but would it be worth it?
[15:54:35] <pcw_home> if you use feedback you want a high ustep ration
[15:55:17] <pcw_home> ratio
[15:55:29] <sadara> ^^^^ could you expand on that? I'm not following
[15:55:55] <pcw_home> if you have a feedback system you want small steps in the corrections
[15:57:21] <sadara> I have 200 step motors, connected to a 5mm lead screw, so this system will be limited by my step rate as it is. The encoder is only 400pulse
[15:57:57] <sadara> should I be using 1/4 stepping to match my step size to the encoder ?
[15:58:09] <pcw_home> in that case I would just use the encoder as a check
[15:58:31] <sadara> yeah, thats what I thought
[15:58:56] <sadara> because this is a double stepper gantry, I don't want it to rack
[15:59:24] <sadara> is there a cheap step generator compatible with EMC yet?
[15:59:39] <pcw_home> if you had a 100 nm linear scale and 256 ustep drives you would be better with a closed loop system
[16:00:02] <pcw_home> hardware stepgen?
[16:00:25] <sadara> yeah
[16:00:32] <sadara> MESA?
[16:00:50] <pcw_home> if you are no ustepping theres not much point
[16:02:00] <sadara> this is a well designed (engineered) gantry, 3m x, 2m y, 600mm z. The gantry is limited by step rate already
[16:02:33] <pcw_home> really depends on your drives. if you have high performance drives then a hardware stepgen will make a big difference
[16:03:06] <sadara> Not sure about "high performance" but they were cheap -> wantai motor
[16:03:28] <pcw_home> what voltage?
[16:03:46] <sadara> one sec...
[16:03:50] * sadara getting drives
[16:05:04] <sadara> the drives are 80v max ( i'm not sure I would be comfortable running them at that though
[16:05:06] <pcw_home> (you can sort of draw a dividing line between the 24V-35V drives that use a single chip and better 50 to 80V drives)
[16:05:20] <sadara> ahhh
[16:05:39] <pcw_home> so a hardware stepgen would probably help
[16:05:44] <sadara> no, proper drives, not TDA ones
[16:06:49] <sadara> I'm running 1700 oz nema 34 steppers, with 6 amp drives and 48v 8 amp supplies each
[16:07:45] <sadara> the drives support 10'000 step pulses per rev (that seems ... a lot)
[16:08:16] <sadara> I'm not sure on the max frequency
[16:08:41] <pcw_home> in my experience there is not much to gain above about 32 usteps (6400)
[16:09:40] <pcw_home> (6400/rev)
[16:09:45] <sadara> yeah, unless your AC driving ( i think it's called wave driving?) but I don't think these drives have a +/- 10v input
[16:09:53] * sadara googles for max useful torque and rpm
[16:10:37] <sadara> I havn't worked through this with anyone else, so could you let me know if this all sounds reasonable?
[16:11:51] <pcw_home> I may have odd standards for "reasonable"
[16:12:17] <sadara> don't we all
[16:12:37] <sadara> there is no torque curve / RPM on the website
[16:13:22] <pcw_home> thats rather important for a stepmotor/drive combo
[16:13:46] <sadara> yeah, there are torque curves, just not for this exacp motor
[16:13:53] <sadara> s/exacp/exact
[16:14:11] <sadara> what would you guess the max RPM would be for the steppers? 1200? 1600?
[16:14:27] <sadara> (at 0 torque)
[16:15:15] <pcw_home> Hard to guess
[16:16:26] <sadara> I was reading the wrong drive btw, this one has 51200 Pulses per rev, and is 8amp peak, 5.6 amps continious
[16:16:54] <pcw_home> thats common = 256 ustep max
[16:17:18] <sadara> ok
[16:17:59] <pcw_home> note that a single 1/256 step may not even make the rotor move
[16:18:43] <sadara> lol, the windings arn't that accurate anyway, the motor could move backwards
[16:18:56] <sadara> for serveral 1/256 steps
[16:19:32] <sadara> ok, assuming 1500 rpm 0 torque, that is a step rate of 5000 (this may be a little high, but would be a good worst case)
[16:19:32] <pcw_home> yes and the torque vs displacement is very flat at small dispacements
[16:19:43] <R2E4> hi all
[16:20:01] <sadara> hi
[16:20:14] <sadara> the quadrature encoder would be outputing 10Khz
[16:20:24] <R2E4> pcw: the R2E4 is working 100%
[16:20:32] <pcw_home> so thats easily doable with software stepping (and counting)
[16:20:48] <R2E4> Got the iron squared away and it powered right up.
[16:20:57] <sadara> ok, thats where I am now
[16:21:03] <andypugh> sadara: 1500rpm is very optimistic for steppers
[16:21:06] <pcw_home> R2E4: Great!
[16:21:33] <andypugh> You are running high voltage, which helps, but NEMA34, which doesn't.
[16:21:57] <pcw_home> maybe not so optimistic at 0 torque
[16:22:34] <andypugh> No, there is that, but who cares about zero torque? That means you can't stop from your 1500rpm cruise.
[16:23:04] <pcw_home> you can stop , you just dont know where :-)
[16:23:09] <sadara> I'm rying to work to worst case
[16:23:33] <sadara> lol, never concidered that... (should have got steppers with a break :) )
[16:23:50] <pcw_home> but hardly anyone would do full steps
[16:23:56] <sadara> if I go to 64 usteps (again, worst case) and I want to still be able to get to max rpm, is it just 64 * 5000 as my step rate?
[16:24:06] <andypugh> Yes.
[16:24:12] <pcw_home> yes 320 KHz
[16:24:18] <DJ9DJ> gn8
[16:24:21] <sadara> 0.3Mhz
[16:24:47] <sadara> a standard drives even capable of accepting that?
[16:24:57] <pcw_home> sure
[16:25:01] <andypugh> A secondary issue is step granularity. A 10kHz base-thread can do 10kHz steps. Or 5kHz steps, Or 3.33kHz steps. Those are big speed jumps for a motor.
[16:26:38] <andypugh> That was why I went for a Mesa card on my machine, but as my steppers are topping out at about 300rpm I didn't actually see any speed improvement. I did see a "suddenly I have plenty of IO" improvement.
[16:26:56] <sadara> aside from mesa, are there any other step gens out there?
[16:27:45] <andypugh> sadara:
http://pico-systems.com/osc2.5/catalog/product_info.php?products_id=30
[16:27:45] <sadara> emc compatible?
[16:28:06] <sadara> anything ... cheaper?
[16:28:13] <pcw_home> Picosystems, the RT-8P8C
[16:28:37] <andypugh> Cheaper than the $79 that a Mesa 5i25 costs?
[16:29:02] <sadara> I'm not going to buy a Mesa 5i25
[16:29:45] <andypugh> 7i43? About the same price.
[16:29:58] <sadara> no, I'm getting a lot more mesa stuff
[16:30:09] <sadara> I'm the guy with the weird servos...
[16:30:15] <sadara> 320v resolvers?
[16:30:31] <andypugh> Are the resolvers 320V, or just the motors?
[16:30:53] <sadara> if they don't co-operate, I will see
[16:31:29] <sadara> I wanted to try something else
[16:31:45] <andypugh> This is my machine. 320V and resolvers on the servos:
http://youtu.be/haO5ZJq-Y3c
[16:32:09] <sadara> I figure if I'm going to spend lot's of $$$ on a mesa system, I should try something else for the steppers
[16:32:12] <pcw_home> 320V sounds like a motor not a resolver (unless you like expensive smoke)
[16:32:50] <gene77> Hi all
[16:33:22] <sadara> yes, the moters are 320v, but I don't have the drives, I was asking you about the mesa 8i20, but it wond do that little current
[16:33:59] <gene77> I need a math lesson, see figure 15.4 in the gcode reference
[16:34:00] <sadara> ~100 watt, 320v brushless dc motors
[16:34:35] <sadara> pcw_home: have you concidered knock out a low current HV servo drive?
[16:34:47] <pcw_home> thats tough unless you find the original drives...
[16:35:11] <sadara> driving them isn't that hard
[16:35:19] <andypugh> 100W 300V is a _tiny_ current.
[16:35:29] <sadara> yeah
[16:35:41] <gene77> I am doing the ends of an = sign, but the top bar is at a plus value, the bottom bar is at a much less negative value. How do I calc a working g2 J value
[16:35:48] <pcw_home> yes maybe 1.5 A peak
[16:36:05] <sadara> there are a lot of servos out there that are 320v and a lot are <400w
[16:36:07] <andypugh> Hmm, put 500w resistors in parallel with each motor winding and use a 1500W drive?
[16:36:32] <sadara> I can make my own FOC drives for them
[16:36:50] <sadara> but then I can't communicate with the MESA system for the resolvers
[16:37:15] <sadara> and the mesa -> hostmot -> emc
[16:38:09] <andypugh> If you make your drives accept the Mesa three-phase PWM then you can hook up a resolver card to the other connectors. The problem here is current feedback.
[16:38:55] <r00t4rd3d> scored a small slab of 100+ year old Cherry today :D
[16:39:49] <gene77> I hope its been inside ALL that 100 years, cherry, on the ground and in the weather, does not do well
[16:40:09] <r00t4rd3d> It was inside. The interior of a gold club house bar.
[16:40:14] <r00t4rd3d> golf*
[16:40:25] <gene77> Ahh, nice
[16:40:33] <sadara> there are single chip drives for servo this size
[16:40:45] <r00t4rd3d> they had tables made from most of it for the dining room
[16:40:46] <sadara> I think ST make single chip drives up to 1200v 40A
[16:41:34] <sadara> (just the drive part, not the controller)
[16:41:38] <pcw_home> well there are Hybrid Hbridges (none of the HVICS are single chip)
[16:42:07] <sadara> http://www.st.com/web/catalog/sense_power/FM148/SC1276/PF253316
[16:42:32] <gene77> Outside on the ground, you have maybe 6 monhs to salvage it, theres about 20 grand worth on the hill across the street, rotten all the way thru in 3 years.
[16:42:59] <andypugh> sadara:
http://uk.rs-online.com/web/p/mosfet-power-drivers/6880811/
[16:44:10] <andypugh> Ah, sorry, I miss-read the boundary, that needs external mosfets.
[16:45:04] <sadara> yeah, 200mz might be a little low, even for tine servos
[16:45:12] <sadara> 200ma rather
[16:47:01] <sadara> these seem ideal :
http://www.st.com/web/catalog/sense_power/FM148/SC1276/PF251661
[16:47:09] <gene77> G2 y# i# j# where J is an offset, but from the starty or the endy?
[16:47:59] <andypugh> sadara: 500V 2A (there is a 1A version too)
http://uk.rs-online.com/web/p/motor-driver-ics/0415213/
[16:50:11] <sadara> andypugh: so your recomendation is to cut up the 8i40 ?
[16:50:16] <andypugh> But that's still a 1kW drive, for a 100W motor.
[16:50:30] <andypugh> 8i40?
[16:50:43] <sadara> the mesa 3 phase BLDC drive
[16:50:56] <andypugh> No, not really.
[16:51:11] <andypugh> I would start from scratch, I think.
[16:51:27] <sadara> but then how do you communicate with the mesa?
[16:51:43] <andypugh> I wonder if the motors would be fast enough at 50V
[16:52:43] <sadara> I can re-organise there phases, and run them at 80v (not a rewinds, cjust go from wye to delta)
[16:53:11] <andypugh> Hmm, PyCAM still reckons it is going to takr 3 hours to finish, so I think I will turn the machine off.
[16:53:16] <sadara> 50v is a little low
[16:53:33] <andypugh> All it will cost you is speed, not torque.
[16:54:00] <sadara> it defeates the purpose of having servos
[16:54:11] <sadara> I might as well keep the steppers
[16:54:49] <andypugh> Yes. I would suggest selling these servos and finding some that are either lower voltage, or higher power.
[16:55:18] <sadara> they are the right size thoungh, and they were the right price
[16:55:21] <r00t4rd3d> wow you really use pycam?
[16:55:36] <andypugh> Or, keep an eye out on eBay for a suitable analogue-controlled drive.
[16:55:56] <andypugh> r00t4rd3d: I am trying it, it seems to work.
[16:56:29] <sadara> It's unlikely I will find a 320v drive with resolver input
[16:56:52] <sadara> I am kinda in hacking the mesa 8i40 though
[16:57:54] <pcw_home> The 8I20 could be built with a 5A Hall sensors (though it would not save any money)
[16:58:22] <sadara> I'm not really concerned with the price
[16:58:42] <sadara> I wouldn't mind if it cost a bit extra
[16:59:08] <sadara> all the mesa stuff is really well priced
[16:59:25] <sadara> I also have 8 of these, so it's not like a 1 off
[16:59:31] <andypugh> pcw_home: Can the 7i39 be up-voltaged?
[16:59:50] <pcw_home> Nah, its all overpriced
[17:00:46] <pcw_home> Its not isolated so I would not suggest it (I think it could go to 200V though)
[17:02:16] <sadara> I only _need_ 80v
[17:02:36] <emc> hi and hello from germany, hurray, my first access with unbuntu-linuxcnc
[17:06:52] <sadara> pcw_home: would the 7i39 go to 80v?
[17:07:30] <andypugh> sadara: Hang on, I have just realised that you can use _any_ Hall-input brushless drive.
[17:08:03] <andypugh> the HAL bldc component can convert resolver angles into umulated Hall signals.
[17:08:12] <andypugh> (hi emc)
[17:10:22] <andypugh> sadara:
http://www.ebay.com.au/itm/121084418877 (though that is another 15A device)
[17:12:52] <sadara> are you sure?
[17:13:22] <andypugh> Sure about what? About that drive, no. About the idea, yes.
[17:14:05] <pcw_home> the 7I39 could go to 80V but if you started with 320V drives and they were wye connected delta would only get you down to 184V
[17:14:31] <sadara> yes, but I can live with 80v
[17:14:58] <sadara> below that and I would be better off with hobby rc motors
[17:14:58] <andypugh> You would end up commutating trapezoidally, but that isn't a disaster.
[17:15:51] <sadara> andypugh: I'm not sure I'm following
[17:16:06] <sadara> how is the drive tracking shaft angle?
[17:16:13] <andypugh> http://youtu.be/47y6RgAK--8 this movie compares sinusoidal and trapezoidal commutation:
http://youtu.be/47y6RgAK--8
[17:16:14] <Tecan> (47y6RgAK--8) "BLDC.mov" by "andy pugh" is "Tech" - Length: 0:00:38
[17:16:34] <andypugh> Sinusoidal is quieter, but other than that there isn't much in it.
[17:18:11] <sadara> Are you talking about sensorless feedback?
[17:19:57] <andypugh> No. I am talking about using the Resolver feedback then computing an equivalent hall-sensor signal for the measured angle.
[17:21:20] <andypugh> The HAL bldc component can convert pretty much any motor feedback signal into any other.
[17:25:22] <pcw_home> if you search for AMC drives on ebay they may be a solution as well (though low current one are less common)
[17:26:52] <andypugh> That one I linked was an AMC. Hall-sensor and encoder input. But if we have Resolver feedback it's easy to emulate such a lower-quality signal. :-)
[17:30:17] <pcw_home> this one is 25A peak but settable down to 3.75A peak
[17:30:18] <pcw_home> http://www.ebay.com/itm/ADVANCED-MOTION-CONTROL-BRUSHLESS-SERVO-AMPLIFIER-BD25A20C-/181106394878?pt=LH_DefaultDomain_0&hash=item2a2ac846fe
[17:30:53] <Tom_itx> what's the _new_ atom board of choice nowdays?
[17:32:10] <sadara> pcw_home: is that a better option than a modded 8i20? (regardless of price, and assuming a MESA resolver input + other mesa equiptment)
[17:39:42] <andypugh> If the only problem with the 8i20 is the lack of current sensitivity, and fitting different sensors fixes that, then it sounds like a good, if expensive, solution.
[17:41:33] <sadara> that is something I would be willing to do myself, as long as the firmware doesn't need to change as well
[17:44:19] <pcw_home> Tom_itx:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813128561
[17:45:12] <Tom_itx> does that take the same ram as the 525?
[17:45:22] <pcw_home> No idea
[17:46:37] <pcw_home> Only problem if the firmware was not changed is that the current scaling would be off (1/6 setpoint)
[17:47:09] <sadara> I figured as much
[17:48:26] <pcw_home> (assuming 5A Hall effect devices)
[17:50:52] <pcw_home> I think the Atom uses DDR3 SODIMMs but the Brazos CPU uses standard 240 pin DDR dimms
[17:50:54] <pcw_home> You can still get non Intel D525 based MBs:
[17:50:55] <pcw_home> http://www.newegg.com/Product/Product.aspx?Item=N82E16813153242
[17:51:13] <andypugh> I have a dead 8i20 you could practice on, but it is on the wrong side of the world.
[17:51:31] <sadara> what part died?
[17:54:24] <andypugh> The capacitors burst into flames, but only after I jumpered a burned-out internal track: This is _before_ the capacitors burst into flames. (it turns out that the damaged ones (which checked out OC) weren't OC at 320V.
https://plus.google.com/photos/108164504656404380542/albums/5747722155741347649/5857937110515956322?banner=pwa
[17:55:09] <andypugh> U1 and U2 are, I think, the sensors you would need to change.
[17:56:39] <andypugh> It just seems _wrong_ to use something at 1/10 it's rated capacity. There ought to be something 1/10 the price for the jo.
[17:56:41] <andypugh> (b)
[17:56:54] <sadara> Look at it this way
[17:57:04] <sadara> I got the servoes for $10 each
[17:57:04] <andypugh> Tomorrow I might try SMT in the oven.
[17:57:23] <sadara> The list price is about $800
[17:57:48] <sadara> so if I pay $250 for the drives I'm well ahead
[17:59:41] <andypugh> Ah, yes. I got three servos (750W) for £50. They cost me £600 for a machine to use them on, £120 for a PC, £120 for a monitor, about a year of spare time, £200 for castings (and counting) probably £100 in parts, connectors (the connectors for each motor cost more than the full set of motos cost) and sundries.
[18:00:21] <andypugh> Odd how bargains can be so expensive :-)
[18:00:40] <pcw_home> You are ahead the final system is useful, what system has such small servos?
[18:01:01] <pcw_home> ahead if
[18:01:12] <sadara> Foam cutter
[18:01:30] <sadara> foam grinder to be more precise
[18:01:42] <andypugh> 100W servos are quite possibly no better than steppers.
[18:02:03] <sadara> and an smt pick and place
[18:02:27] <pcw_home> Yeah it may be cheaper to just use step motors as decent (leadshine 542 or so)
[18:02:36] <pcw_home> step drive
[18:02:52] <sadara> that is what I am going with atm
[18:03:05] <sadara> but the gantry is limited by stepper RPM
[18:03:50] <andypugh> It is likely to be limited by servo rpm too, as the servos will have less torque at 0 rpm, so will need to be geared down.
[18:04:17] <sadara> these servos go to 14000RPM
[18:04:51] <andypugh> The video I linked earlier was a 100W servo on my milling machine Z axis. It had no real performance advantage over the steppers.
[18:05:30] <sadara> most of the time the long axis (3m) will be doing constant acceleration or deceleration, as the machine goes back and forth machining
[18:05:34] <andypugh> And I don't think you will get 14,000 rpm with an 8i20 and software commutation. (I am glad you mentioned that)
[18:05:49] <sadara> but that was a short heavy axis
[18:07:18] <sadara> the steppers can accelerate to max speed very quickly
[18:07:36] <sadara> (ie more torque = faster acceleration)
[18:08:03] <sadara> the steppers should reach max speed in 5cm or so
[18:08:41] <L84Supper> http://www.soliforum.com/topic/1954/fullfledged-lcd-touchscreen/ this approach would be nice if tablets had a real time interface or usable SDIO
[18:09:03] <sadara> the servos can't reach max V, as they are not powerful enough, and have to start decelerating at the half way point
[18:09:23] <andypugh> The 8i20 can only commutate at servo-thread rate. If that is 1kHz, with a 4-pole motor and wanting to commutate 6x per electrical cycle, that is about 2500rpm.
[18:09:33] <sadara> but they reach a much faster speed after a about 200-250 mm
[18:10:03] <andypugh> Which is faster than a stepper, by a fair margin.
[18:11:22] <sadara> yeah, I can just under acceleration and deceleration, and still get optimal results, unlike metal and woodwork
[18:12:05] <pcw_home> a 8I20 will work with a 14000 RPM motor but you would need a 4-5 KHz servo thread
[18:12:31] <sadara> on PC?
[18:12:44] <pcw_home> Yes
[18:12:52] <sadara> I'm not hopeing to reach 14000 rpm btw
[18:13:26] <sadara> that would be terrifing
[18:13:46] <pcw_home> but the fact that its 100W at 14000 RPM means you have a lot of gearing down to do to get that 100W
[18:14:15] <andypugh> These are nice little 7i39-compatible motors, I have a couple, but I did most emphatically not pay that price:
http://www.ebay.com.au/itm/250799732438
[18:15:42] <sadara> I can do the math, and even with-out gearing down, the 100w servos still outperform the steppers
[18:17:37] <sadara> (on the long axis, one shorter axis y and z, the steppers are a better choise, actually, it would be marginal on y for bigger jobs, but y travel isn't as smooth , ie, more accel/decel)
[18:20:32] <andypugh> http://granitedevices.com/index.php?q=servo-drive-vsd-e#.UWCqF6XgKH8 might be worth a look. They claim to work down to 5W, and LinuxCNC can definitely synthesise Hall signals (and quite possibly encoder signals too).
[18:21:21] <sadara> and LinuxCNC can definitely synthesise Hall signals <- ahhhh.... that make WAY more sense now
[18:22:29] <andypugh> Hmm, the Granites are not any cheaper than the 8i20, though.
[18:22:57] <andypugh> Ah, had I failed to clearly explain what I was saying?
[18:23:40] <sadara> yes, which was why I was asking how the drives tracked shaft angle
[18:24:54] <sadara> if EMC can generate hall sensor information, can it generate 3 phase PWM ?
[18:25:17] <sadara> then I just need a three phase power stage in a chip
[18:25:37] <andypugh> Yes. I thought I already explined that too.
[18:26:16] <sadara> If you did, I didn't get it then
[18:26:55] <sadara> I wonder if there is a hostmot module
[18:27:52] <pcw_home> Yep
[18:28:32] <andypugh> With the right firmware a Mesa 5i20 + 7i49 can read the resolvers and output the correct 3-phase drive signals for the Inverter bridge. _However_ you would need to do current-control on your external board.
[18:28:50] <sadara> that is easy
[18:29:06] <pcw_home> so for a simple voltage mode drive its very easy (+ - blowing up the bridge with wrong drive signals)
[18:29:14] <andypugh> That is basically all that a 7i39 is. It's a dumb drive that responds to a 3-phase PWM signal from the PCI card.
[18:32:00] <sadara> I assume it's the same for a 7I40?
[18:32:18] <andypugh> pcw_home: Ah, but, a Mesa card would surely never give wrong drive signals?
[18:32:51] <andypugh> 7i40 is for brushed motors, I think.
[18:33:19] <sadara> sorry, I meant the 7i29
[18:33:24] <andypugh> (and, by the way, a set of brushed motors would almost certainly work brilliantly for you)
[18:33:40] <sadara> I know
[18:34:00] <sadara> but after spending this much time talking about them, I have to use them now
[18:34:09] <andypugh> 7i29 is for brushed motors too.
[18:34:29] <sadara> 7i39?
[18:34:32] <sadara> googles
[18:34:42] <andypugh> That is basically all that a 7i39 is. It's a dumb drive that responds to a 3-phase PWM signal from the PCI card.
[18:34:44] <sadara> yes, the 7i39
[18:34:54] <pcw_home> At startup all high/low side signals are high
[18:35:01] <andypugh> The one I mentioned earlier?
[18:35:06] <pcw_home> so you need to be careful
[18:35:22] <sadara> is there a serperate "enable" ?
[18:36:37] <andypugh> Some of the bridge-on-a-chip devices have built-in shoot-through disable.
[18:38:29] <sadara> 3ppwmgen
[18:38:37] <sadara> found it
[18:38:38] <andypugh> yes
[18:38:44] <andypugh> man hostmot2
[18:39:05] <sadara> google, I don't have an emc computer nearby
[18:39:49] <Connor> pcw_home: Got Pete's MPG working on the Field I/O without issue.
[18:40:06] <andypugh> sadara:
http://www.linuxcnc.org/docs/html/man/man9/hostmot2.9.html
[18:40:07] <Connor> looks to work just fine.. even at the distance..
[18:43:18] <andypugh> Ok chaps, (and chapesses, if there are any). It is time for me to log off.
[18:43:41] <sadara> same, It 7 am
[18:43:47] <sadara> need sleep
[18:44:02] <sadara> may also explain why I'm a little slow
[18:44:18] <andypugh> Ah, I normally assume that when I see an Aussie here they are up early, not down late.
[19:07:39] <pcw_home> Connor Good. its a better match for a MPG since it does 1x mode
[19:47:03] <R2E4_> Hi all
[20:56:59] <Connor> pcw_home: Still working out the PID and a few other things with the machine... After we home the machine... the Z would show something like 0.003 or so... not sure what that's about..
[21:26:18] <pfred1> tools are toys
[22:20:43] <pcw_home> Connor: if Z has a static load it may show a static error unless you have some I term
[22:21:22] <Connor> you mean I as in the I in PID ?
[22:22:10] <Connor> I think we had the Z around 11 on P, .0001 on I, and i don't recall what D was..
[22:23:18] <Connor> I know when we first fired it up.. the servo buzzed.. I think we had P at 12 at that point..
[22:25:28] <Connor> And I think we found that FF2 made it not "bang" as hard.. I have no idea what FF0, FF1 FF2 any BIAS are for..
[22:25:47] <pcw_home> Did you look at JTs tuning tutorial?
[22:25:59] <Connor> I was just looking for the link..
[22:26:04] <Connor> Not sure what I did with it.
[22:26:20] <pcw_home> FF1 is critical for velocity mode tuning
[22:29:52] <pcw_home> but you will only see its effect during motion
[22:29:56] <pcw_home> http://gnipsel.com/linuxcnc/tuning/servo.html
[22:30:26] <Connor> Yea... that's the one that helped with the "banging" which was a jarring the machine when we changed directions..
[22:31:34] <Connor> Wow. This uses HAL scope.. We didn't even bother with that yet..
[22:31:40] <Connor> Just trail and error.
[22:31:47] <pcw_home> You need to use halscope to see whats happening dynamically
[22:33:24] <Connor> okay. Well.. Next time I go over there to help him.. I'll bust this link out and use it to help fine tune the PID the.
[22:35:49] <Connor> So, he's not even doing anything with I or D
[22:35:53] <Connor> *boggle*
[22:37:01] <pcw_home> D is not normally needed for velocity mode drives
[22:37:22] <pcw_home> (the drive handles 'D')
[22:37:40] <Connor> This needs to be done under load right? Not the motor simply on the bench..
[22:38:10] <pcw_home> Yes tuning will be different under load
[22:41:03] <Connor> We also got a alarm on his Z from the driver a few times.. not sure I understand it..
[22:41:24] <Connor> Error: 1. The main circuit voltage is exceeded maximum allowable value. (410V) 2. Regeneration voltage is too high.
[22:42:06] <Connor> Corrective Actions: 1. Use multi-meter to check whether the input voltage is within the specified limit. 2. Check the Parameter Cn012 if it is setting correctly. 3. If this alarm appears during operation. Extend ac/deceleration time or reduce load ratio in the permitted range. Otherwise, an external regeneration resistor is needed. (Please contact yo
[22:42:30] <Connor> We no the main circuit voltage was good.. and this was during operation..
[22:44:28] <pcw_home> if you stop too fast and dont have a braking resistor, you will get this
[22:45:01] <pfred1> is it caused by back EMF generated by the motor coils?
[22:45:20] <Connor> Okay, That's what I was thinking.. So, He might need a breaking resistor on the Z.. Didn't have it happen on the X.. but, it wasn't under the load the Z was..
[22:45:38] <Connor> or, I need to decrease the Acc/Dec in the INI file for the Z
[22:45:54] <Connor> He DOES have a break on the Z though.
[22:46:28] <pcw_home> its caused by the mechanical energy being dumped back into the power supply when deccelerating
[22:47:07] <Connor> Which means, it'll happen on the Z because of the weight of the Head when it goes from a Down move to a up move rapidly..
[22:47:11] <pfred1> the servo acts like a generator?
[22:47:19] <Connor> pfred1: Yes.
[22:47:23] <pcw_home> Yes when stopping
[22:48:04] <pcw_home> very fancy/large drives have a inverter to dump the energy back into the power lines
[22:48:34] <pcw_home> simpler drives must do with a braking resistor
[22:48:50] <pfred1> no one uses filter capacitors to store it?
[22:48:54] <Connor> I think this is more than enough info to send to Pete..
[22:50:11] <pcw_home> The mechanical energy is much larger that practical to store in the filter caps
[22:50:23] <pcw_home> s/that/than/
[22:52:06] <pfred1> must be a big resistor to dump it as heat then
[22:52:12] <pcw_home> (well to large to store with a practical delta V)
[23:23:35] <r00t4rd3d> blaa, midnight and I wanna go cut stuff
[23:25:04] <WalterN> what kind of power density do I want if I'm going to make a laser sintering machine?
[23:26:37] <WalterN> it would depend on the material type I guess... but were would I find this information for something like plastic or nylon or something
[23:28:46] <r00t4rd3d> you just wanna know how big a laser to get?
[23:29:12] <WalterN> mostly
[23:29:31] <WalterN> but that depends on the focal lense stuff I get too
[23:30:37] <WalterN> read through this...
http://www.parallax-tech.com/faq.htm
[23:31:05] <WalterN> I guess if its powerful enough, a lense would not be needed XD
[23:32:42] <WalterN> at the moment I'm trying to figure out what the cheap plastic stuff is thats commonly used for laser sintering
[23:33:04] <r00t4rd3d> http://www.cnczone.com/forums/forum.php#laser_engraving_cutting_machines
[23:33:39] <r00t4rd3d> the people here that probably know are probably in bed
[23:35:00] <r00t4rd3d> - 1st probably :)
[23:35:13] <WalterN> well
[23:35:33] <WalterN> after fixing up my car so I can get to work and back, I have a bunch of money leftover
[23:35:47] <r00t4rd3d> can i have some ?
[23:35:48] <WalterN> I caught myself looking at new computer stuffs
[23:36:04] <WalterN> "what am I doing? I need to make a CNC machine first"
[23:36:17] <r00t4rd3d> i wouldnt bother with a laser
[23:36:27] <WalterN> why?
[23:36:37] <r00t4rd3d> what do you want to make?
[23:36:44] <r00t4rd3d> whats your purpose for cnc?
[23:37:01] <r00t4rd3d> lets start with that
[23:37:12] <r00t4rd3d> lasers are expensive for one
[23:37:14] <WalterN> well, I want to either make a laser sintering machine, or a light weight CNC mill machine/engraving
[23:37:16] <r00t4rd3d> good ones
[23:37:38] <r00t4rd3d> wth is sintering
[23:37:49] <FinboySlick> Is there any success with home-made laser sintering?
[23:38:10] <WalterN> its sinter with ing added to it :P
[23:38:29] <r00t4rd3d> creating objects with powder?
[23:38:32] <r00t4rd3d> lol
[23:38:57] <r00t4rd3d> wait is that how they made that cresent wrench?
[23:39:24] <FinboySlick> r00t4rd3d: That wasn't sintering, it was plastic.
[23:39:36] <WalterN> FinboySlick: donno... there isnt much about it online.. most of the prototyping machines made seem to be the inferior extruded plastic type method
[23:39:38] <FinboySlick> Laser sintering is typically for metals.
[23:39:43] <r00t4rd3d> oh i remember when they took it out is was all powdery
[23:40:16] <WalterN> extruded plastic gets terrible resolution
[23:40:18] <FinboySlick> r00t4rd3d: Yeah. That's pretty much filling up an inkjet with glue and printing your layers on plastic powder.
[23:40:41] <FinboySlick> WalterN: If you know what you're doing, they can still be pretty good.
[23:40:51] <WalterN> I dont like pretty good :P
[23:41:11] <FinboySlick> You're not going to get anywhere near pretty good on home made metal sintering.
[23:41:30] <WalterN> I'm not talking about metal though
[23:42:02] <WalterN> though, the only difference would be in the laser I think
[23:42:12] <WalterN> everything else should stay the same
[23:42:38] <FinboySlick> I think you'd be better off looking into stl.
[23:42:44] <r00t4rd3d> doesnt seem like many people are DIY laser sintering
[23:43:52] <WalterN> I already considered it mostly...
[23:44:09] <WalterN> what is the smallest extrusion size?
[23:44:18] <WalterN> 1mm or something?
[23:44:36] <FinboySlick> For extrusion, I've seen 0.25mm.
[23:44:54] <WalterN> where are those?
[23:45:10] <FinboySlick> you make the extruder yourself.
[23:45:15] <WalterN> thats getting into my tollerance range
[23:45:15] <FinboySlick> With a lot of care.
[23:45:31] <pfred1> the plastic extrusion 3D printing market is so much hype now
[23:46:36] <pfred1> Make put up a post on G+ and people thought an extruder did it turned out it was a media fuser that did the stuff they posted
[23:46:55] <WalterN> what is G+?
[23:46:57] <pfred1> idiots were saying oh a printer that costs $1,000 can make those
[23:46:59] <FinboySlick> pfred1: Maybe it'll be like the internal combustion engine... Clunky, clumsy thing that we've ended up over-refining to something pretty awesome.
[23:47:15] <pfred1> WalterN Google's social network
[23:47:19] <WalterN> oh that
[23:47:21] <WalterN> heh
[23:47:55] <WalterN> FinboySlick: back in the day, it was a lot more amazing compared to the other stuff avaliable (steam power)
[23:48:16] <pfred1> I don't know there are some pretty amazing steam engines
[23:49:42] <WalterN> FinboySlick: can a .25mm extruder thing be bought?
[23:50:13] <FinboySlick> WalterN: I guess. The one I've seen was a brass cap-nut with some careful dremel work on it.
[23:50:23] <pfred1> that filament stuff is pretty expensive too
[23:50:28] <WalterN> dremel? o.0
[23:50:43] <WalterN> sigh
[23:50:51] <FinboySlick> It worked.
[23:51:15] <pfred1> are Dremels being made any better today than they used to be?
[23:51:21] <pfred1> because they used to be real garbage
[23:51:35] <WalterN> I'm going to be using ball screws and stuff... nothing dremel is going to touch what I make... lol
[23:51:36] <pfred1> I think Bosch owns them now though
[23:51:50] <FinboySlick> He just ground the nut into the right shape, put a flat at the end.
[23:52:19] <FinboySlick> Then used a lathe-like 3d printed rig to drill with a 0.25mm bit without breaking it.
[23:52:27] <FinboySlick> It works.
[23:52:31] <WalterN> hmm
[23:52:41] <WalterN> is melting point the thing that I need?
[23:52:44] <WalterN> bah
[23:53:02] <WalterN> there seems to be no information on this
[23:57:41] <r00t4rd3d> ive never heard anyone talking about sintering here
[23:57:49] <r00t4rd3d> either*
[23:58:20] <FinboySlick> r00t4rd3d: you can make a candy-printer with powdered-sugar sintering ;)
[23:58:59] <r00t4rd3d> ive wanted to put a frosting pump on my cnc machine and do cakes.
