#linuxcnc | Logs for 2012-04-09

Back
[00:03:31] <Tecan> http://www.xup.in/dl,14639681/image2gcode.7z/
[00:45:42] <Tecan> sweet i've got the spindle rate being set in gcode from the z depth
[00:48:44] <Tecan> will m2 issue an m5 ?
[01:02:36] <Tecan> self.write("S{}".format(-z*160))
[02:12:57] <DJ9DJ> moin
[05:07:56] <JaggedNZ> Random question of the day: does EMC always send a fixed length pulse to STEP ?
[05:08:44] <mhaberler> man stepgen: stepgen.N.steplen u32 rw
[05:08:44] <mhaberler> The length of the step pulses, in nanoseconds. Measured from rising edge to falling edge.
[05:09:16] <mhaberler> thats a parameter, so fixed at config time
[05:10:53] <mhaberler> no, wrong: it can be changed post config
[05:12:08] <JaggedNZ> hmm cool, thats what I was hoping for. Copied some stepper MCU code that polls rather than using interupts
[05:12:37] <JaggedNZ> couldn't figure out how it would work unless pulse times where near identical
[05:13:15] <mhaberler> yout should be able to scope it with halscope even in sim, and then fiddle with setp stepgen.X.steplen <length>
[06:02:12] <Loetmichel> *baeh* *spit* *pfrrrt* Aluminium swarf EVERYWHERE, in the hair, in the clotes, in the COFFEE... (sh** when one has to mill a workpiece three times the machine... all housing dismantled)
[06:02:17] <Loetmichel> mornin' btw
[06:45:56] <Tecan> awe i fell asleep last night
[06:46:08] <Tecan> luckily i have monday off
[06:59:47] <Loetmichel> sooo, second half of lower connetion bar for alex' mill on the run... (shi* is THAT part big... the maschine can operate in the white area ;-) -> http://www.cyrom.org/pce/connectbar.avi (4MB, H264)
[07:05:11] <Loetmichel> hrhr: the steppers are louder than the spindle milling...
[07:05:37] <Tecan> earplugs
[07:05:58] <Loetmichel> Tecan: the stepüpers are silent
[07:06:49] <Loetmichel> i meant: the whole machine is SO silent i even let the door of the worksop open and my wife 2 rooms away doesent complain!
[07:07:03] <Tecan> neat
[07:07:40] <Loetmichel> yes, and cheap
[07:08:05] <Loetmichel> have payed about 250 eur fpr spindle plus VFD plus set of ER11-chucks
[07:32:21] <Jymmm> Loetmichel: What up with you bouncing for a couple hours?
[07:51:49] <ssi> I need a pendant
[07:52:29] <ReadError> me too ssi
[07:52:33] <ReadError> order 2 and give me one plz
[07:53:08] <ReadError> what kind of tools should i go ahead and order
[07:53:13] <ReadError> in preparation for the machine
[07:53:13] <ssi> for what
[07:53:22] <ssi> the taig mill?
[07:53:22] <ReadError> i need the vactara oil
[07:53:24] <ReadError> yessir
[07:53:42] <ssi> measuring tools, setup stuff
[07:53:47] <ssi> do you have a vise for it?
[07:53:50] <ssi> clamping set?
[07:53:54] <ssi> 123 blocks
[07:54:02] <ssi> parallels for the vise
[07:54:11] <ssi> I don't really have a feel for just how tiny the taig is :)
[07:54:22] <ssi> also, what exactly do you intend to cut
[07:57:00] <ReadError> aluminum mostly
[07:57:03] <ReadError> no steel
[07:57:06] <ssi> what sort of parts
[07:57:14] <ReadError> pcbs
[07:57:26] <ReadError> motor mounts for some quad copters
[07:57:37] <ReadError> i need a nice table to hold larger parts
[07:57:42] <ReadError> not so much a vise
[07:59:09] <ReadError> maybe i can get some of those little hold down clamps
[07:59:15] <ReadError> and use a piece of wood
[07:59:25] <ReadError> i dont want to tear up my board learning
[08:02:36] <ReadError> idealy i would like to make a router to cut my CF on
[08:02:50] <ReadError> but i figured with a mill making a router would be more feasible ;0
[08:03:10] <ssi> it helps
[08:04:29] <ReadError> i was thinking about using some 80/20
[08:04:33] <ReadError> i can mill the mounts
[08:04:48] <ReadError> *sigh*
[08:04:52] <ReadError> so many ideas so little time
[08:04:54] <Loetmichel> Jymmm: telekom Router has a hickup: if its running for a few weeks it fills its ram with logs-> no room for routing rables
[08:05:03] <Loetmichel> reboot router and everyting is fine
[08:05:13] <Loetmichel> i just forgot to reboot it in time ;-)
[08:06:24] <ReadError> i need to decide if i should order the cnc motor kit
[08:06:30] <ReadError> its like 535
[08:06:34] <ReadError> or do it myself ;/
[08:06:43] <ssi> link?
[08:06:47] <Loetmichel> ReadError: better use a perpex/pvc sheet as a sacrificial plate and thenn mount the workpieces with CA or double sided tape
[08:07:04] <ReadError> http://www.deepgroove1.com/gecko/geckodriver.htm
[08:07:06] <Loetmichel> so you can cool with water/oil
[08:07:19] <ReadError> ahh
[08:07:25] <ReadError> i have a bunch of pvc sheet
[08:07:29] <Loetmichel> (from a flower spray can)
[08:07:54] <Loetmichel> perspex was meand
[08:08:03] <Loetmichel> meant
[08:08:06] <ssi> ReadError: that kit comes with power supply too?
[08:08:11] <ReadError> yea
[08:08:12] <ReadError> 400w
[08:08:38] <ssi> yeah that's probably a reasonable deal
[08:09:07] <ssi> here, comparison shop
[08:09:07] <ssi> http://kelinginc.net/CNCNEMA23G540Package.html
[08:09:32] <ssi> that's the setup i use for a couple of my machines
[08:09:39] <ssi> the G540 + that keling 48/7.3 supply
[08:12:44] <Loetmichel> my sacrificial plate is made of a 2 minute polyuretane resin
[08:19:35] <ReadError> 2 minute?
[08:22:59] <SolarNRG> Does anyone in here have a 3d printer?
[08:30:51] <ReadError> 3 pcs NEMA23 KL23H2100-35-4B (1/4” Dual shaft with a flat) 381 oz-in
[08:30:56] <ReadError> those are some MONSTERS huh?
[08:35:21] <kb8wmc> SolarNRG: working on building a 3d printer right now
[08:35:56] <ReadError> why can 3d printers use USB and CNCs not?
[08:40:43] <SolarNRG> hmmm
[08:40:58] <SolarNRG> Well i can drive a stepper motor with a USB arduino....
[08:41:06] <SolarNRG> But CNCs use much more powerful stepper motors
[08:41:30] <SolarNRG> And to switch them very very fast you need like a gecko drive which only uses parallel
[08:42:32] <SolarNRG> I think when switching at very high frequency, USB breaks down because it is a packet switching thing, wheras parallel you are not
[08:43:35] <psha[work]> unbounded latency
[08:43:44] <psha[work]> and slow
[08:52:54] <joe9> the "CNC Programming Handbook" by Peter Smid is awesome. Got it from a local library. a "must read" it seems for cnc folks.
[08:53:16] <ReadError> no pdf? ;(
[08:54:33] <ReadError> nm found it
[08:57:34] <Loetmichel> *ha*, it works... [insert mad scientist laughter here] -> http://www.cyrom.org/palbum/main.php?g2_itemId=13051
[08:58:16] <Loetmichel> ... one part at a time... now i have to drill some, then screw some parts, then make a coupling for the stepper... then the x axis ;-)
[09:02:03] <Loetmichel> ... the wooden y standoffs are just "templates". when the Mill is running the first will be to mill some new ones from 14mm aluminium plate which is already here
[09:04:18] * Thadiusb slaps Thadius around a bit with a large trout
[09:06:52] <mazafaka> Fresh nettle, steamed potato, slightly fried onion for a dinner... Am I simply rich? Yeahhhhhh.
[09:27:30] <CareBear\> mazafaka : do you grow nettles yourself?
[09:27:34] <CareBear\> or did you buy those?
[09:32:53] <ssi> ReadError: 3d printers can use USB because typically people who use 3d printers are hipsters who are using them to make joke glasses frames, and so precision isn't terribly necessary
[09:32:59] <ssi> ReadError: USB isn't realtime
[09:37:41] <CareBear\> ssi : 3d printers have no feedback
[09:37:54] <CareBear\> ssi : that's the difference
[09:38:01] <CareBear\> ssi : it has nothing to do with precision
[09:38:44] <CareBear\> ssi : the 3d printers accept g-code over USB, and do all control in a microcontroller
[09:39:10] <CareBear\> ssi : so feedback can be added, and they can still use USB fine
[09:39:35] <CareBear\> "realtime" is just an illusion
[09:39:46] <CareBear\> a local CPU bus is also not realtime
[09:39:57] <CareBear\> all communication has latency
[09:41:20] <CareBear\> since linuxcnc insists on doing all processing on the CPU the round trip time of USB is much too high
[09:41:25] <CareBear\> it's very simple
[09:41:58] <CareBear\> USB could work perfectly, but it would require some fundamental changes in linuxcnc
[09:44:08] <JT-Shop> LOL I just got an email from bob cad trying to sell me Mach3 now that is funny
[09:44:21] <CareBear\> ReadError : I hope you caught the above
[09:44:46] <ReadError> yup
[09:45:25] <CareBear\> psha[work] : "unbounded latency" is simply not true, as you would know if you looked a little deeper into the spec :) "slow" is also not true for many applications, but yes, round trip time is too long for linuxcnc
[09:45:29] <kb8wmc> CareBear\: thank you, your explanation was a help to my understanding
[09:46:06] <CareBear\> kb8wmc : you're welcome
[09:47:49] <CareBear\> the key for successful USB living is to use a bit more high level communication over the bus, ie. having some bit of control outsourced to a moderately or highly intelligent peripheral, e.g. an axis controller
[09:50:28] <skunkworks> so 'motion' would be in the usb device/
[09:50:30] <skunkworks> ?
[09:52:09] <skunkworks> sounds like a buffered system
[09:52:41] <skunkworks> almost unrealated... http://groups.yahoo.com/group/mach1mach2cnc/message/134478
[09:56:12] <skunkworks> (the video posted is of linuxcnc...) ;)
[10:00:47] <thinpete> Thats the inherent problem with buffered systems with multiple control loci (USB etc)
[10:00:49] <thinpete> To do things right you would end up porting all of LinuxCNC to the remote device (so why not leave it as it is with the PC as the embedded controller)
[10:01:35] <skunkworks> :)
[10:04:01] <mazafaka> CareBear\: I have found one on growing out of the Mother Earth
[10:10:49] <JT-Shop> darn I can't load a yahoo groups page
[10:11:11] <skunkworks> I bet you are heart broken ;)
[10:11:26] <JT-Shop> LOL yea
[10:11:41] <skunkworks> did you get your satellite service yet?
[10:18:39] <JT-Shop> yea, the only problem I have is with my hosting site their log in page has such a fast timeout that I can't log in to my control panel and they keep telling me to tell my ISP
[10:18:56] <JT-Shop> as the ISP can't do anything about satellite latency
[10:19:35] <JT-Shop> s/as/and
[10:19:43] <JT-Shop> so i'
[10:19:50] <JT-Shop> I'm stuck in the middle
[10:29:30] <asdfasd> while you talking for motion, and usb, can I somebody with more knowlage for PC architecture about IDE connector, is it possible for real time control?
[10:30:57] <asdfasd> or may be directly from PCI bus
[10:31:49] <jdhnc> how about a digital IO card, directly on the PCI bus.
[10:32:13] <jdhnc> or even better, a mesa 5i25 card, directly on the PCI bus.
[10:33:11] <asdfasd> actually the question was for DIY , something cheap and minimum hardware, like LPT port but something better
[10:34:01] <asdfasd> eternal motion controller usually cost a bit more
[10:34:10] <asdfasd> external
[10:39:42] <skunkworks> 2nd printer port...
[10:43:07] <JT-Shop> sure you can do it Peter did!
[10:43:32] <JT-Shop> should have been a comma in that broken sentenace I think
[10:49:16] <skunkworks> asdfasd: you can get a 5i25 for like $89
[10:52:59] <skunkworks> JT-Shop: Unless they can increase the speed of light.. You are stuck with your latency.. (26K miles to the satalite.. request to recieve is atleast 100k miles of travel)
[10:53:54] <Thadius> hello all
[10:54:01] <skunkworks> You could probably half that with dialup for the transmit - (that used to be an option) but now you have 2 systems that can go wrong. ;)
[10:54:09] <skunkworks> Thadius: hello
[10:54:50] <Thadius> just started trying out "solidworks"
[10:55:01] <Thadius> not to shabby
[10:56:50] <jdhnc> you can buy 5 5i25's for $71. Anybody want 3?
[10:57:02] <jdhnc> (each)
[10:58:01] <skunkworks> JT-Shop: your going to have a minimum of .5 seconds latency. (when I had hughes net I would get into the second latency)
[11:05:48] <JT-Shop> skunkworks, yea the big problem is getting webhost4life to increase the timeout on the log in page so you can connect via the satellite connection
[11:06:44] <JT-Shop> I think webhost4life wants wildblue to speed up light
[11:17:44] <cradek> I'm talking to you FROM SPACE?
[11:19:59] <JT-Shop> lol
[11:21:24] <Tecan> hmmm
[11:21:48] <Tecan> im one bong hoot away from the moon too
[11:23:23] <Tecan> cradek , theres a copyright on the image to gcode source... if i make changes are we allowed to publish them ?
[11:23:49] <cradek> what file are you talking about?
[11:23:54] <Tecan> i setup the spindle speed gcode for laser
[11:24:08] <thinpete> asdfasd: I've seen some IDE interfaced hardware (in PIO mode 4 you have a pretty simple 16 bit bidirectional bus)
[11:24:10] <Tecan> author.py and image-to-gcode.py
[11:25:14] <asdfasd> thinpete: and how is performing compared to LPT ?
[11:25:27] <thinpete> I would think it much faster
[11:25:30] <cradek> Tecan: both of those files have a full license statement at the top, with a pointer to the full license that you can read to understand your rights and responsiblities.
[11:25:32] <thinpete> its
[11:25:49] <cradek> they are both licensed under the GNU GPL version 2 or later
[11:26:02] <cradek> you can read about those licenses online
[11:26:16] <Tecan> oh oops
[11:26:17] <thinpete> but you need at least some buffers/latches to get I/O
[11:26:19] <Tecan> thanks
[11:27:27] <ReadError> how viable is modeling in corel ?
[11:28:19] <JT-Shop> skunkworks, I got them on the phone and got a tech that sounded like he knew a bit and cared so we will see...
[11:28:47] <Thadius> anyone know of any problems with "Oxford OXPCIe952"chipset for pci-e parallel adapter??
[11:29:20] <Tecan> where do we even submit patches for linuxcnc ?
[11:29:26] <JT-Shop> Thadius, some real good info here http://linuxcnc.org/index.php/english/component/kunena/?func=view&catid=18&id=18917
[11:29:43] <JT-Shop> Tecan, code or docs?
[11:29:47] <Tecan> code
[11:29:59] <JT-Shop> the developers mailing list
[11:30:15] <JT-Shop> you need to make a git patch
[11:32:12] <Thadius> ty JT-shop
[11:32:24] <Thadius> not finding if the card is supported or not, still checking
[11:34:54] <skunkworks> Thadius: you could ask on the mailing list... you might be the first with pci-e though
[11:36:22] <Thadius> remember why i ordered it now, under reviews someone said they used on 10.04 with no problems
[11:36:28] <Thadius> man i'm skatter brained today
[11:37:03] <Tecan> will an m2 command stop everything ? im too lazy to test
[11:37:30] <Tecan> feel like im going blind from my laser
[11:37:38] <Thadius> and there goes my money, ordered my v90
[11:37:58] <jdhnc> Thad: what motors/drives?
[11:38:59] <Tecan> v90 sounds expensive you said it was only 290 ?
[11:39:09] <Thadius> the probostep motors
[11:39:49] <Thadius> well my total order was 1286.03
[11:39:57] <Tecan> ouch
[11:40:47] <Thadius> ya, if i had the tools and a lil more time, i might of gone the DIY route for building my cnc
[11:40:53] <Thadius> seen a lot of cool success stories
[11:41:27] <Tecan> http://www.youtube.com/watch?v=Y58Ab192ggQ& << my first attempt at a gantry
[11:41:56] <Thadius> but i had some luck on a scratch off lottery ticket and decided instead of buy crap, i'd go ahead and start doing what i've wanted to for some time now
[11:42:14] <Tecan> i went overkill on the table ended up buying a z axis 6' x 6' and the table flopped around with that 6' half inch thread rod
[11:42:52] <Thadius> thats cool you have the ability to even get where you're at Tecan
[11:43:13] <Thadius> i'm hoping my adventure into this will lead to a better understanding and eventually a DIY creation of my own
[11:43:45] <JT-Shop> my first gantry CNC http://s47.photobucket.com/albums/f163/johnplctech/Plasma%20Cutter/
[11:43:46] <Tecan> my first one was a flop, the second is pretty rough too but belts made it way easier
[11:44:38] <Thadius> that your creation JT?
[11:44:50] <JT-Shop> yea
[11:44:58] <Thadius> what was lead time on it?
[11:45:33] <JT-Shop> didn't keep track built it over a winter iirc
[11:45:52] <Thadius> very cool
[11:46:06] <JT-Shop> thanks
[11:46:15] <Thadius> and now to order my spindle
[11:46:20] <JT-Shop> I machined all the parts on my BP
[11:47:20] <Thadius> what i'd like to do is get a decent knowledge base of my new machine and over time piece together a creation of my own
[11:47:26] <JT-Shop> seems like I built it 4 years ago http://www.youtube.com/watch?v=e70jdiDm7yc&list=UUt3eCzdbsQeuCifjsTyvvTg&index=10&feature=plcp
[11:47:27] <Tecan> (e70jdiDm7yc) "CNC Plasma first test It's Alive" by "Gnipsel" is "Tech" - Length: 0:00:08
[11:47:31] <Thadius> got lucky on that lottery ticket lol
[11:47:33] <Tecan> thadius i reccommend a driverboard from mechantronics
[11:47:40] <JT-Shop> test run in the beer cave
[11:47:54] <Tecan> they have 2 pots that adjust for awesomeness and the parts are all changable
[11:48:44] <Tecan> indestructable driverboards
[11:49:43] <Tecan> brb im going to take a picture of my plasma cutter lol
[11:49:45] <Thadius> anyone have experience with vacuum hold downs?
[11:49:53] <JT-Shop> http://www.youtube.com/watch?v=EC1_TTlsOrM
[11:49:53] <Tecan> (EC1_TTlsOrM) "Second Plasma Cuts" by "Gnipsel" is "Tech" - Length: 0:00:33
[11:50:51] <Thadius> JT, what is the maximum thickness you can cut?
[11:50:57] <Thadius> just curious
[11:51:00] <JT-Shop> 1"
[11:51:49] <Thadius> nice
[11:52:05] <JT-Shop> no, 1" is ugly
[11:52:15] <JT-Shop> 1/2" is nice
[11:52:43] <JT-Shop> actually I think it is rated for 1 1/4" thick
[11:55:14] <Thadius> i'm just marveling at the ability to do it, i was thinking more 1/4 or a lil more
[11:55:23] <Thadius> obviously way off base lol
[11:55:28] <Tecan> u guys ready for this ?
[11:55:45] <Tecan> http://i.imgur.com/DkQa8.jpg
[11:56:44] <Tecan> homemade plasma cutter
[11:58:01] <Tecan> im afraid to turn it on
[12:00:33] <Tecan> http://i.imgur.com/TbLgl.jpg << most recent machine got the rails motors and pullys in a kit on ebay for 300
[12:01:20] <ries> Tecan: what about a remote switch?? :D
[12:01:50] <Jymmm> Tecan: whats the heating element for?
[12:02:03] <Thadius> man, I hate electrical work, spent last 5 years working on Wind Turbines, and still have my fear of electricity 0_o
[12:04:33] <Tecan> the truck coil is to jump up the voltage , the cealing pot is my current adjust and the heating element is for dissipation and current dumping
[12:04:54] <Jymmm> Tecan: Love the bare secondary leads to the bridge!
[12:05:02] <ries> Thadius: what about heights??
[12:05:24] <jthornton> since switching to satellite I've been unable to send mail from this computer(Ubuntu 10.04) via Evolution
[12:05:33] <jthornton> I get this error message MAIL FROM command failed: bosauthsmtp02: Host 184.20.140.167: No unauthenticated relaying permitted
[12:05:51] <Jymmm> jthornton: You have to AUTH
[12:05:51] <jthornton> anyone have a clue what that message means?
[12:06:12] <jthornton> Jymmm, what is that?
[12:06:13] <Tecan> its not bare its magnet wire :)
[12:06:16] <Tecan> hehe
[12:06:24] <Jymmm> jthornton: You have to sign in to send mail.
[12:06:55] <cradek> you probably need to change your outgoing mail host to one that your new ISP provides. You are probably pointing to your old ISP.
[12:06:57] <Jymmm> jthornton: You need to update the outgoing smtp authetication method in your email client.
[12:07:10] <jthornton> I'm using my hosting account
[12:07:22] <cradek> I don't know what that means
[12:07:44] <Thadius> ries, i'm fine with heights lol
[12:07:46] <jthornton> my web site is hosted by webhost4life and I have e mail accounts with them
[12:07:59] <jthornton> my ISP is wildblue satellite
[12:08:09] <Jymmm> jthornton: In evolution goto your settings, in OUTGOING/SMTP MAIL there will be an authentication area somewhere
[12:08:12] <jthornton> maybe I need to use wildblue smpt
[12:08:27] <cradek> yes you do
[12:08:43] <Jymmm> cradek: Not if he's using his hosting accoutn he doens't.
[12:08:46] <jthornton> ok let me change that, thanks
[12:08:54] <Thadius> Jthornton, how bad is the latency on wildblue?
[12:08:55] <cradek> http://help.wildblue.net/kb/article/3317
[12:09:04] <Thadius> always been curious
[12:09:18] <jthornton> cradek, thanks
[12:09:53] <cradek> they're too tired to run their own mail host, so you have to give your outgoing mail to google to read
[12:10:13] <Jymmm> cradek: No, he doens't ned to change that
[12:10:29] <Jymmm> jthornton: Just do what I said.
[12:10:34] * cradek shrugs
[12:13:08] <jthornton> well I didn't get to your suggestion cradek as Jymmm suggestion worked
[12:15:14] <jthornton> thanks to both of you for the suggestions
[12:17:31] <jthornton> cradek, will do
[12:18:07] * Thadius away
[12:25:11] <Tecan> whenever i fire my laser my xbox turns on and ejects a disk
[12:25:38] <Jymmm> jthornton: Man, these guys are full of shit. Even though you "lease" the equipment, it only has a two year warranty. WTF is that shit?!
[12:25:56] <Jymmm> jthornton: http://www.viasatresidential.com/wildblue/legal/limited-warranty
[12:27:30] <Jymmm> jthornton: and be sure to check your "daily usage" too http://www.viasatresidential.com/wildblue/legal/data-allowance-policy
[12:27:30] <jthornton> yea you don't want to drop anything on it for sure
[12:27:39] <jthornton> usually it is broken
[12:28:49] <Jymmm> jthornton: Well, after 24 months, I'd make them ship you new equipment with a new 2yr warranty on it.
[12:29:59] <Jymmm> jthornton: Or, if they give you any shit, have them send you in writting that the equipment is under warranty and will be repaired/replaced at no cost to you including all shipping fees.
[12:30:51] <asdfasd> is there any way I can make a box in pyvcp, and pressing a button to send the value in the box as MDI command and then zero that box
[12:31:12] <Jymmm> There is a "durable goods" law in the US that says leased/rent should be working, so that "warranty" policy is BS
[12:33:39] <Jymmm> jthornton: And since it's a "one time accoutn setup fee of $150", never close the account, just have it placed on hold for 6months or indefinantly if you ever decide to change/move and may want it again in the future.
[12:35:58] <alex4nder> yoh
[12:36:01] <alex4nder> machinists
[12:39:37] <jthornton> asdfasd, a spinbox
[12:40:12] <Jymmm> jthornton: FYI... On large file downloads (ISO's as an example) your speed can be reduced by 90%.
[12:40:22] <ssi> jthornton: hey I have a question for you about the toolchange ladder
[12:40:37] <ssi> jthornton: occasionally mine misses a change... it seems like it's popping the stop a little too early
[12:40:45] <ssi> jthornton: is that something I can fix in ladder, or do I need to tweak the encoder magnet
[12:41:04] <jthornton> the encoder magnet might be a little off
[12:41:15] <jthornton> or the stop may be sticky
[12:41:34] <ssi> I can often hear it clack a little as it runs past the last tool before the selected one
[12:41:42] <ssi> and occasionally it'll actually stop at the one before selected
[12:41:44] <jthornton> my stop is sticky after it sits for a day or three and I always miss the first tool but not after that
[12:41:52] <ssi> sometimes it faults the program, sometimes it DOESNT.... which is scary
[12:42:04] <jthornton> yea the magnet timing may be a bit early
[12:42:12] <ssi> how hard is it to tweak the magnet?
[12:43:11] <Jymmm> Tecan: What rifle is that?
[12:43:18] <jthornton> one tiny screw from the top
[12:43:39] <jthornton> after you get the cover off
[12:44:11] <ssi> ah ok
[12:44:32] <Tecan> just a pump action 22 :)
[12:45:22] <Tecan> the gross killer :)
[12:45:36] <Jymmm> Tecan: ah, looks used and abused =)
[12:45:45] <Tecan> grouse
[12:51:21] <jthornton> asdfasd, a pyvcp number input box is a spin box http://linuxcnc.org/docview/html/hal/pyvcp.html#_number_inputs
[12:52:33] <asdfasd> yes I thoght is the same, but how to send the value as MDI
[12:53:34] <jthornton> that is not so simple and I don't know how you would get the number from the spinbox to a variable
[12:58:07] <asdfasd> using gladevcp I can add GLADEVCP= -u ../gladevcp/hitcounter.py ../gladevcp/manual-example.ui
[12:58:55] <asdfasd> probably with that *py file will be easy?
[12:59:10] <asdfasd> can I do same with pyvcp
[13:21:49] <rob__H> spinning the turret too fast can lead to problems also , there is an adjuster to alter the air flow
[13:23:27] <asdfasd> hey rob_h JT-shop told me that you have front-rear tool post on your lathe
[13:24:02] <asdfasd> yesterday I found that mach3 make wrond calculations when you change from front to rear and oposite
[13:24:20] <asdfasd> do you have any problems with emc?
[13:25:21] <rob__H> define problem
[13:25:46] <Jymmm> rob__H: snafu, fubar, etc
[13:25:53] <asdfasd> mach3 mess up with the offsets, changing front to rear and then front again and finally you end with offset on the whole axis
[13:26:14] <rob__H> http://www.youtube.com/watch?v=8S1NIhwCzB0
[13:26:14] <Tecan> (8S1NIhwCzB0) "EMC2 Fitted to Hardinge Superslant Lathe" by "roberth270" is "Tech" - Length: 0:04:01
[13:26:58] <rob__H> i have no problems with tool offsets
[13:28:07] <asdfasd> yesterday I wasted 2 hours to find out where is the problem I though that I wrote wrong G code
[13:28:37] <asdfasd> but later I found that other people have same problem....
[13:29:07] * JT-Shop runs and hides
[13:31:01] <rob__H> its not perfect in EMC , u have to tell axis to change.. so displays right in axis
[13:32:48] <asdfasd> you mean in your ini file something like LATHE = 1
[13:34:07] <rob__H> in the axisrc config file, flip display over for tool post/tool direction
[13:34:22] <rob__H> what lathe is this?
[13:34:50] <asdfasd> Im not sure it was something to enable lathe screen
[13:35:03] <asdfasd> in [display] section
[13:35:09] <rob__H> lathe=1 put linuxcnc into lathe mode yes
[13:36:21] <asdfasd> I didnt tried yet, I will keep banging my head with mach3 just a little bit more, if more problems comming then I will change on the lathe too
[13:38:50] <asdfasd> http://linuxcnc.org/docs/html/common_machining_center.html#cap:Tool-Orientations
[13:39:04] <asdfasd> that link is broken, it was working few days ago
[13:39:15] <rob__H> John probly broke it
[13:39:48] <JT-Shop> I thought I read somewhere that Mark3 would not work with a lathe
[13:41:07] <JT-Shop> http://linuxcnc.org/docview/html/lathe/lathe-user.html
[13:41:36] <JT-Shop> the 2.4.x docs have been replaced with the 2.5.0 docs so old links might not work anymore
[13:41:57] <JT-Shop> rob__H, did you mean this http://linuxcnc.org/docview/html/lathe/lathe-user.html#_lathe_tool_orientation_a_id_lathe_tool_orientation_a
[13:53:22] <ssi> rob__H: does the collet closer pressure adjustment also affect the turret's air pressure?
[13:53:45] <rob__H> no
[13:53:46] <rob__H> well
[13:53:53] <rob__H> maybe if ucan not supply the volume of air
[13:54:02] <rob__H> as it loves to eat all you can give it
[13:54:11] <ssi> air volume isn't an issue
[13:54:14] <ssi> have 19cfm shop air
[13:54:41] <rob__H> most time we would swithc the hydrovane compressor on to keep it happy and all other stuff in shop too
[13:54:47] <joe9> what is a good brush to apply "way oil" often? any art brush that you would recommend?
[13:55:05] <rob__H> how about an oil can
[13:56:03] <rob__H> ssi, i took sometimes have problem of turret points missing
[14:27:00] <Tecan> dam
[14:27:04] <Tecan> anyone here good with gcode ?
[14:27:17] <Tecan> how do i set spindle speed without setting surface speed ?
[14:28:33] <cradek> please elaborate, it's not clear what you mean
[14:28:46] <Tecan> Defines speed, either spindle speed or surface speed depending on mode
[14:28:50] <Tecan> s
[14:29:06] <archivist> what are you reading
[14:29:24] <Tecan> wiki on gcode
[14:29:57] <archivist> we cant see the page unless you are more explicit
[14:30:05] <Tecan> http://en.wikipedia.org/wiki/G-code
[14:30:11] <Tecan> im going to try and set g97 mode
[14:30:30] <cradek> maybe you are asking about http://linuxcnc.org/docview/html/gcode/gcode.html#_g96_g97_spindle_control_mode_a_id_sec_g96_g97_spindle_a
[14:31:13] <cradek> g97 is the default mode and is the only one that makes sense on a mill. g96 only makes sense on a lathe.
[14:31:33] <asdfasd> cradek do you read personal messages? is there any problem with me or my question?
[14:32:51] <cradek> no problems, I prefer to answer questions in public so they benefit everyone, sorry if it seemed like I was ignoring you
[14:33:21] <cradek> also on irc I think untargeted questions are best, because anyone can answer them if they have time and knowledge
[14:35:39] <asdfasd> yeah, but some time the question is for a specific person and also Im not very conviniet to ask anything in public specially if I know how stupid may sound for someone else with much more experience
[14:37:21] <asdfasd> I thought that something gone wront with our last conversation about acceleration, Im sorry if I did anything wrong
[14:38:30] <cradek> no, I did not even remember. it's hard for me to remember all the people I help in irc, sorry. only after I see someone around for a long time do they become a real individual person and I start remembering conversations...
[14:38:52] <cradek> I wish I was better at that.
[14:41:05] <asdfasd> anyway I will find another solution, may be another day
[14:44:57] <Tecan> how do i edit gcode in the manual control box ?
[14:45:27] <cradek> Tecan: file/edit will pull up your choice of editor, with the loaded file in it
[14:45:48] <Tecan> pimping aint easy :)
[14:45:56] <cradek> ?
[14:46:14] <Tecan> thats a nice method
[14:47:19] <Tecan> gedit wont let me edit though
[14:47:43] <cradek> you can't edit the splash screen gcode.
[14:47:51] <cradek> you can save a copy of it and edit that, also under the file menu
[14:47:52] <Tecan> aha
[14:50:13] <Loetmichel> soo.... little by little i am making progress... no i have to make a motor mount and two gears... ant then y axis will be nearly finished... -> http://www.cyrom.org/palbum/main.php?g2_itemId=13054
[14:56:38] <asdfasd> Im not that patient to make all so precise and beautifull
[14:57:06] <Loetmichel> i am not patient at all
[14:57:19] <asdfasd> I just rush to make it working....
[14:57:23] <Loetmichel> but i can make it right the first time or rebuild it 5 times.
[14:57:42] <Loetmichel> making it right the first time is MUCH faster ;-)
[14:57:57] <asdfasd> I mean right but not looking nice
[14:58:23] <asdfasd> I know nice looking is much better but....
[14:58:58] <asdfasd> I just finished this one
[14:59:11] <asdfasd> http://imagebin.org/207345
[14:59:37] <Loetmichel> hrhr, multitool for a lathe ;-)
[15:01:14] <asdfasd> actually it is working really nice, but it looks like will fall apart every moment
[15:02:05] <archivist> looks ok to me
[15:02:29] <asdfasd> 10x
[15:02:38] <Loetmichel> your lathe looks three numbers bigger than mine ;-)
[15:02:45] <archivist> add a servo mode to the spindle and some live tooling next
[15:02:58] <asdfasd> the spindle is servo
[15:03:01] <asdfasd> 2kw
[15:03:26] <asdfasd> step/dir control
[15:03:28] <cradek> gang tooling is nice. fastest tool changes possible.
[15:03:38] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=13027 <- just made me 2 new chuck keys
[15:04:03] <asdfasd> yes, that was the reason, each tool will be used for a few seconds only, so tool changing will take most of the time
[15:06:34] <Loetmichel> anyone knows how to convert tooth spacing in european Module in gears?
[15:06:58] <djdelorie> today's lesson: if you're using tabs/tags to hold your workpiece in while you profile it's outline, make sure your Z is touched off the BOTTOM of the piece. Tags don't work if they're not there because your Z is wrong :-P
[15:09:18] <Loetmichel> hrhr
[15:12:58] <archivist> asdfasd, have you been attempting to tap with mach3?
[15:15:21] <asdfasd> Loetmichel: Im not sure if that will help you - copy from wiki
[15:15:23] <asdfasd> Countries which have adopted the metric system generally use the module system. As a result, the term module is usually understood to mean the pitch diameter in millimeters divided by the number of teeth. When the module is based upon inch measurements, it is known as the English module to avoid confusion with the metric module. Module is a direct dimension, whereas diametral pitch is an inverse dimension (like "threads per inch").
[15:16:22] <archivist> Loetmichel, see HPC catalogue technical section
[15:17:04] <archivist> downloadable pdf has all sorts of useful gear info
[15:17:38] <asdfasd> archivist: tapping not tested yet, still writing G code, but I think will be fine
[15:17:39] <skunkworks> if the spindle can be run as a step/dir 'axis' then tapping with mach3 will work just fine.. (for some definition of)
[15:17:46] <djdelorie> http://www.delorie.com/photos/cnc/bldc-bracket-1.html
[15:18:11] <archivist> asdfasd, will be ok on emc
[15:18:33] <djdelorie> ignore the inside, I stopped it when I realized Z was way too low
[15:19:25] <asdfasd> I already setup mach3 before I tried emc
[15:19:51] <asdfasd> may be in future
[15:20:34] <archivist> Loetmichel, http://www.hpcgears.com/technical.htm
[15:21:33] <archivist> asdfasd, you already found one problem, now google mach3 and screw cutting
[15:22:34] <asdfasd> yes I read that
[15:22:48] <asdfasd> I hope I can make it work somehow
[15:23:38] <archivist> you probably need an old style sliding tap holder, not rigid for mach3
[15:24:18] <asdfasd> I can control the spindle like rotating axis, so it should be fine
[15:25:25] <JT-Shop> I thought mack couldn't tap on a lathe
[15:26:13] <Thadius> ok, anyone ever had "bad character 'a'used" ?
[15:26:38] <asdfasd> G1 Z-12.5 A3600 - this will make 10 turns on the spindle and move the tap if the tap is 1.25 pitch
[15:27:09] <archivist> Thadius, yes when you try to run gcode with an a axis and machine does not have an A
[15:27:20] <asdfasd> it should be syncronized move
[15:27:30] <Thadius> ahh, guess i need to check my cam setup?
[15:27:52] <Tecan> dont eat yellow pee
[15:27:56] <archivist> possibly
[15:37:10] <Thadius> got it, .con file for Meshcam had [R] in two places
[15:37:25] <Thadius> or rather the EMC.con file
[15:37:45] <ssi> so what kind of servo amplifiers are out there for small nema23 or 34 sized servo motors?
[15:37:56] <ssi> all the drivers I've seen emulate step/dir
[15:44:06] <jdhnc> +/- 10VDC
[15:44:24] <ssi> I mean who makes one
[15:44:24] <jdhnc> step/dir seems so lame for a servo
[15:44:32] <ssi> all I can find is stuff like the gecko 320
[15:44:36] <ssi> which I agree seems lame
[15:44:52] <cradek> you're just looking at hobbyist stuff then?
[15:45:04] <ssi> well, yeah I suppose
[15:45:09] <ssi> thinking about making my g0704 a servo machine
[15:45:14] <ssi> since I have a spare 7i33
[15:45:28] <ssi> want to do 23 frame x/y motors and a 34 frame Z motor
[15:46:00] <jdhnc> the cncfusion g0704 kit has a 23 Z mount
[15:46:01] <cradek> pico makes an inexpensive pwm 20A servo drive
[15:46:17] <ssi> if I use the '33, though, it's analog out
[15:46:26] <ssi> so I figure I'd be better off with an analog input drive
[15:49:45] <jdhnc> http://pico-systems.com/pwmservo.html
[15:50:56] <skunkworks> it is pwm in though..
[15:51:05] <jdhnc> I'm planning on using 483oz x/y and some bigger 34 for Z
[15:51:28] <skunkworks> you can get analog amc drives off of ebay pretty cheap usually
[15:51:58] <ssi> http://www.ebay.com/itm/AMC-Brush-Type-PWM-Servo-Amplifier-25A20T-LF3-/250896539331?pt=LH_DefaultDomain_0&hash=item3a6a995ec3#ht_1330wt_1378
[15:52:01] <ssi> something like that? :D
[15:52:30] <skunkworks> yes - that is 25A peak - 200v
[15:52:36] <skunkworks> 12.5 cont
[15:53:21] <skunkworks> we have had good luck with them.
[15:54:38] <ssi> unfortunately there's not more than one of any given drive on ebay :P
[15:54:48] <skunkworks> we are running 40A ones on the K&T
[15:55:15] <skunkworks> heh - I thought you where only looking for one...
[15:55:43] <ssi> just trying to make plans for how I want to run the g0704
[15:55:50] <ssi> I always figured that you could just sorta buy the stuff
[15:55:54] <ssi> but apparently not :)
[15:56:40] <ssi> I can get little 23 or 34 frame servo motors from keling, but the only low-cost amps that seem to be available are either something like a gecko that's step/dir drive (I guess you call that position mode?) or the pico pwm-input amps
[15:56:54] <ssi> I'd think that something like a gecko 320 but without the step/dir drive and encoder feedback would actually be CHEAPER
[15:57:07] <skunkworks> there are some amps that take step/dir or analog..
[15:57:11] <ssi> stuff like those amc amps retail price is insanely high
[15:57:17] <skunkworks> yes
[15:57:34] <djdelorie> define "insanely high"
[15:58:07] <ssi> $1k
[15:58:13] <DJ9DJ> more than 3 ;)
[15:58:16] <asdfasd> it is not such a problem to build one
[15:58:21] <djdelorie> just to run a brushless motor at a random speed?
[15:58:26] <cradek> the only place you find cheap servo amps is attached to very large pieces of cast iron
[15:58:29] <asdfasd> it is much easy than step/dir
[15:58:30] <ssi> brushed motor more like
[15:58:46] <ssi> yeah there's not that much to an analog servo amp
[15:58:47] <djdelorie> even so, a brushed motor is just a brushless motor with one less wire :-)
[15:59:00] <djdelorie> (as far as the driver electronics is concerned :)
[15:59:03] <ssi> but I don't really feel like adding that project onto the stack :)
[15:59:42] <skunkworks> http://www.granitedevices.fi/index.php?q=servo-drive-vsd-e
[15:59:56] <djdelorie> if you can drive the board with something other than analog, you could just use mine :-)
[16:00:07] <skunkworks> VSD-E features industry standard pulse/direction, quadrature, PWM and analog control inputs
[16:00:21] <ssi> skunkworks: what do those drives cost?
[16:00:39] <cradek> djdelorie: an incremental software-only improvement for your amps would be to let them accept quadrature input instead of step/dir
[16:00:51] <cradek> that lets you "step" faster
[16:00:54] <djdelorie> that's two bits in a control register, the hardware already supports it
[16:01:16] <djdelorie> the limiting factor at the moment is *not* the step rate, it's the mechanicals in the machine itself
[16:01:22] <cradek> ah
[16:01:29] <djdelorie> the board will accept step pulses up to 1 MHz :-)
[16:01:42] <djdelorie> (and probably faster, but the step generator stopped there)
[16:01:44] <cradek> I mean you can generate them faster with software stepping
[16:01:52] <djdelorie> I understood
[16:01:54] <cradek> ok
[16:02:12] <cradek> no finicky setup/hold timings either
[16:02:14] <djdelorie> the hardware has step/dir, fwdstep/revstep, and quadrature built in
[16:02:19] <JT-Shop> the 7i29 does 22.5A at 165V
[16:02:42] <DJ9DJ> gn8
[16:02:46] <djdelorie> I have three more ADC inputs left on the RX62T chip too, I could break one of those out to a connector.
[16:02:51] <JT-Shop> I assume that is for a DC servo motor
[16:03:43] <JT-Shop> ssi what size amp are you looking for?
[16:03:45] <djdelorie> Mine will do 15A at 450V
[16:04:04] <JT-Shop> did you make them?
[16:04:08] <djdelorie> yes
[16:04:12] <JT-Shop> nice
[16:04:22] <djdelorie> http://www.delorie.com/electronics/bldc/
[16:04:44] <djdelorie> I didn't think to put analog control inputs on, though. Maybe if I respin it I'll add those
[16:04:56] <djdelorie> -10 to +10 is standard?
[16:05:00] <skunkworks> I like analog.
[16:05:04] <skunkworks> djdelorie: yes
[16:05:13] <JT-Shop> yea
[16:05:30] <djdelorie> digital output would still be faster than the motor control loop ;-)
[16:06:20] <djdelorie> hmmm... in sergvo drive, the PC wants the encoder signals, yes?
[16:06:26] <djdelorie> servo drive, not sergio drive :-)
[16:06:31] <JT-Shop> yea
[16:06:32] <cradek> yes
[16:06:50] <skunkworks> well.. with linuxcnc (and most controls) - you are limited to the trajectory/motion of 1khz to 10khz anyway...
[16:07:07] <djdelorie> if the controller board *also* wants them, it would have to include a "copy out" connector for it, if the controller also wants the encoder signals
[16:07:40] <cradek> yes it's pretty normal for the amp to give you encoder outputs to use for the position loop
[16:07:55] <djdelorie> does the pc normally want the usual ttl/cmos, or some EMI-resistant differential pair?
[16:08:15] <cradek> good encoder input boards like mesa and pico take differential in
[16:08:18] <djdelorie> in my case, the motor's hall and encoder signals are on the same connector, and you'd want to use the encoder for vector drive anyway, smoother operation
[16:08:42] <djdelorie> next question: do I *really* want to make this more complicated than it already is? ;-)
[16:08:50] <skunkworks> heh
[16:08:53] <asdfasd> what I noticed is that PID controller need few pulses error on the encoder to make control, so the actual accuracy is a bit lower
[16:09:09] <skunkworks> asdfasd: compared to?
[16:09:11] <djdelorie> I think the 62T has PWM input on the step/dir pins too, I'll have to check
[16:09:43] <djdelorie> skunkworks: one thing the encoder signal is used for is a more accurate RPM reading
[16:09:45] <asdfasd> compared to the encoder accuracy
[16:10:17] <cradek> the mesa does a good sub-period velocity reading too, now
[16:10:32] <cradek> by timestamping the encoder edges
[16:11:00] <JT-Shop> http://linuxcnc.org/index.php/english/component/kunena/?func=view&catid=38&id=19055#19055
[16:12:14] <cradek> not really. you'd have to run two linuxcncs on two computers, and coordinate them somehow (hook some IOs together and use M codes for sync?)
[16:12:36] <cradek> but my understanding is this is what commercial controls did too - just use two of them
[16:16:14] <ssi> JT-Shop: honestly not sure... something that'd drive a servo that'll run that little mill
[16:16:16] <asdfasd> ssi asked for something cheap :)
[16:16:38] <ssi> JT-Shop: the keling 23 frame servos are listed as 36VDC, 4A cont, 19A peak
[16:16:44] <ssi> and 230oz peak torque
[16:17:00] <ssi> 350oz version is 60VDC, 4/20A
[16:17:01] <asdfasd> then better go to step/dir servo
[16:17:04] <asdfasd> is not so bad
[16:19:45] <asdfasd> most of the motors are 3000 rpm - 50 rev per second, 1000 counts encoder give you 4000 pulses or total 200khz at maximum speed
[16:20:19] <asdfasd> perfect tuned PID can control with 5 counts error
[16:20:22] <alex4nder> hey
[16:20:33] <asdfasd> ops 5 pulses
[16:21:12] <asdfasd> that mean actual frequency with guaranteed acuracy will be 40 khz
[16:22:42] <Tecan> arc radius too small... gaaaa
[16:23:55] <JT-Shop> ssi, the Mesa one does 22.5A and 165V it says H bridge so I assume that is for a brushed DC servo motor
[16:23:59] <Tecan> im going to loose my pickles
[16:24:18] <JT-Shop> ssi, do you need the motors too?
[16:24:44] <JT-Shop> cradek, thanks for the comment
[16:28:18] <skunkworks> asdfasd: analog allows you to do cool stuff like.. http://www.electronicsam.com/images/KandT/conversion/testing/tuning.png
[16:30:00] <ssi> JT-Shop: yeah I do
[16:30:12] <ssi> JT-Shop: if you have a suggestion on a good combination, I'm definitely interested :)
[16:30:20] <ssi> I can always punt and go stepper instead
[16:30:43] <ssi> but I would really prefer to do servos controlled via that '33
[16:31:15] <asdfasd> skunkworks: no questions, much better, but some time not needed that much
[16:31:23] <skunkworks> if it where me - I would keep an eye on ebay and buy a few amc drives as they come up...
[16:31:38] <skunkworks> (to use the 33)
[16:33:36] <skunkworks> (or your analog drive of choice)
[16:34:02] <asdfasd> I made serial output on my controller so I can see the error in the same way
[16:36:11] <JT-Shop> rob__H, got some AC servos and drives that he used to upgrade his CHNC he should know the place
[16:36:58] <skunkworks> ssi: http://www.ebay.com/itm/Advanced-Motion-Controls-Brush-Type-PWM-Servo-Amplifier-/170821953256?pt=LH_DefaultDomain_0&hash=item27c5c826e8
[16:37:09] <skunkworks> how many volts do you need?
[16:37:15] <skunkworks> that is 80 max
[16:40:19] <asdfasd> why in USA all technical stuff is much cheaper...
[16:43:11] <skunkworks> That + mesa + encoders + servos makes a cheap true closed loop system.
[16:46:27] * skunkworks is a closed loop snob\
[16:48:52] <asdfasd> yeah... different people different mean of cheap
[16:49:14] <skunkworks> heh
[16:50:02] <skunkworks> bbl
[16:50:06] <asdfasd> if you are using encoders with less than 2000 cpr
[16:50:25] <asdfasd> then all that is not worh
[16:51:43] <Loetmichel> asdfasd / archivist: thx
[16:55:17] <asdfasd> Loetmichel I dont understand inches, why.... why it should be that confusing...
[16:56:16] <stevegt_1> Hey all -- I'm trying to figure out how jepler and John K. did that povray animation of the sphere being cut a few years ago, the one on youtube. Did either of them (or anyone else) ever publish any code for that?
[16:56:30] * stevegt_1 goes looking for the youtube url in case anyone wants it
[16:57:58] <stevegt_1> http://youtu.be/_tA-W7nTNSc
[17:10:57] <ssi> hrm those that skunk linked look pretty good
[17:12:53] <asdfasd> it looks like skunk think only proffesional :)
[17:14:28] <asdfasd> there is sin/cos encoders, who can provide over 100k ppr t
[17:14:48] <asdfasd> I think then it is worth to make it with all accesories
[17:17:09] <asdfasd> otherwise it sounds like C7 ballscrew on woodworking mill
[17:22:11] <asdfasd> ssi : what kind of machine you are planning to drive?
[17:33:51] <Tecan> was image 2 gcode removed in linuxcnc 2.5 ?
[17:46:52] <rob__H> ssi, how about some servos like mitsubishi 200W i don tknow mounting sizes u are asking for so hard to know servo size you want
[17:49:06] <SolarNRG> Does anyone in here know how to convert a .stl Openscad file into a .dwg autocad file?
[17:50:27] <Tom_itx> solidworks might
[17:50:54] <SolarNRG> I don't have solidworks, if I upload you my file, would you be willing to convert it for me?
[17:51:40] <rob__H> ud have to get it into a DXF other wise be a soldworks DWG and ud still be stuck
[17:52:36] <SolarNRG> I tried uploading my stl file to a company and they can't open it they want me to send it to them as a dwg
[17:53:24] <Tom_itx> rob__H you might be able to save it as a dwg
[17:53:26] <Tom_itx> ?
[17:53:37] <Tom_itx> mmm
[17:53:49] <SolarNRG> Is there a pastebin of files?
[17:53:50] <Tom_itx> what about iges or dxf export?
[17:54:09] <rob__H> is it a 3d part?
[17:54:14] <SolarNRG> The openscad export won't let me export my model as a dxf because it is not a 2d part
[17:54:15] <SolarNRG> Yes it is a 3d part
[17:54:23] <Tom_itx> if it's 3d you may be out of luck
[17:54:39] <rob__H> stls are quite universal
[17:54:40] <rob__H> and iges
[17:54:46] <rob__H> open them all the time
[17:55:05] <rob__H> just a pain in but when you have to repatch surfaces becasue the other system could not save in stl properly etc
[17:55:21] <Tom_itx> yep we ran into that all the time with catia
[17:55:29] <SolarNRG> Am I correct in saying the Industry ONLY accepts autocad files and if you're not rich enough to be able to BUY autocad, you can't design and order custom parts with open source files, they won't open them and there's no way of cracking autocad so you MUST pay thousands in order to even get a 3d model from one PC to another these days?
[17:55:50] <Tom_itx> no
[17:56:00] <Tom_itx> iges is quite common
[17:56:44] <djdelorie> STEP is talked about a lot, but mostly in the "everyone else supports it, we should too" way
[17:56:54] <Tom_itx> ask them what formats they support
[17:57:08] <SolarNRG> This company supports DWG only
[17:57:11] <rob__H> think depends on what kinda programs tho people you are talking to have.. i find there is quite afew out there still that dont have all the latest stuff
[17:57:16] <Tom_itx> find a new company
[17:57:25] <SolarNRG> Tom I like your thinking
[17:57:25] <Tom_itx> they are living in the dark ages
[17:57:31] <rob__H> too true
[17:57:39] <rob__H> the complex shapes and parts get now days, u cant afford not to have 3D
[17:57:52] <SolarNRG> How about this, if I send you a stl file, will someone please print that part for me and post it to me and I'll send them twenty quid in return?
[17:57:55] <Tom_itx> and my software is probably 10 yrs old or more
[17:58:19] <Tom_itx> wtf am i gonna do with a squid???
[17:58:21] <Tom_itx> :)
[17:58:48] <Tom_itx> i'll look at it i guess. unless i get interrupted here
[17:59:55] <rob__H> need upload it
[17:59:58] <rob__H> or something
[18:00:03] <Tom_itx> although my solidworks is rather old
[18:00:26] <rob__H> only odd one i know of is, Sketchup u dont do anything with that
[18:00:29] <Tom_itx> as is my acad and smartcam
[18:00:33] <rob__H> i got 2012 here
[18:00:42] <Tom_itx> you're ahead of me then
[18:00:50] <SolarNRG> Is there a pastebin for files?
[18:00:50] <rob__H> my wallet dont think so
[18:00:51] <Tom_itx> was that a volunteer i heard??
[18:01:11] <Tom_itx> i'm sure there is
[18:01:38] <rob__H> SolarNRG, PM u maybe email better?
[18:02:03] <SolarNRG> yeah email would work
[18:39:19] <ssi> back
[18:39:55] <rob__H> if any one wants a good free STl converting tool, check this out - http://meshlab.sourceforge.net/
[18:40:19] <Tom_itx> heh
[18:40:26] <Tom_itx> did you just find it?
[18:41:07] <Tom_itx> oh wow, i posted a link to that the other day
[18:41:26] <Tom_itx> i forget what it was in reference to
[18:42:04] <rob__H> yea try convert SolarNRG file as its not true geometry
[18:42:06] <Tom_itx> oh yeah.. a point file like from a cmm etc to cad
[18:42:43] <Tom_itx> somebody was asking about probing a part iirc
[18:42:55] <rob__H> nice
[18:43:32] <Tom_itx> might come in handy some day
[18:43:51] <rob__H> when finaly afford that cmm
[18:44:24] <Tom_itx> no no...
[18:44:27] <Tom_itx> http://www.vinland.com/Touch-Probe.html
[18:45:26] <rob__H> all my mills are too bussy making a mess all day long eating metal
[18:45:45] <ssi> rob__H: I ran 40 production parts this weekend on my HNC :)
[18:46:03] <Tom_itx> i don't make a living at it
[18:46:07] <Tom_itx> just mess around nowdays
[18:48:30] <rob__H> prob best way
[18:58:25] <SolarNRG> Beddy byes for Snoozy pies
[18:58:34] <ssi> go melt metal
[19:13:38] <JT-Shop> what's the syntax to wait for an input in a M1xx bash file?
[19:18:41] <cradek> consider using M66 instead, but if you must use M1xx here are some clues: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?CustomMcodesToHal
[19:21:59] <JT-Shop> cradek, thanks
[19:35:09] <Tecan> wholy man cradek you really modified the crap outa image-to-gcode lol
[19:35:28] <Tecan> i just found the orig copy from the site with a skull on it
[19:35:58] <Tecan> http://www.junktech.de/en/software/15-modified-heightmap-script.html
[19:47:21] <cradek> the one that comes with linuxcnc is a zillion times better than what I originally wrote
[19:48:09] <Tom_itx> that's alot
[19:49:19] <cradek> only the basic idea and algorithm is the same - jepler rewrote the whole thing
[19:49:24] <JT-Shop> that is a bunch of numbers
[19:50:06] <kb8wmc> cradek: I have been running the 2.5~pre and keeping up with the upgrades, should I install version 2.5?
[19:52:58] <cradek> are you using the buildbot 2.5 packages?
[19:53:21] <kb8wmc> I have been
[19:53:45] <kb8wmc> have not had any upgrades in several weeks though
[19:53:48] <cradek> then you already have 2.5.0 and perhaps a change or two afterward (I think there were only doc fixes so far)
[19:53:56] <cradek> odd - you should have
[19:54:17] <kb8wmc> ok
[19:54:28] <Tecan> cradek in 2.5 you cant open image-to-gcode in the open menu of linuxcnc
[19:54:54] <kb8wmc> my last was .626
[19:55:05] <cradek> if you have the filter options set properly, you can just open the image and it will run image-to-gcode for you.
[19:55:54] <Tecan> thats what i mean its not in there to pick
[19:55:57] <cradek> kb8wmc: the latest on 2.5 branch should be dc10e7130a
[19:56:09] <Tecan> ngc is the only thing there
[19:56:10] <kb8wmc> will check
[19:56:17] <cradek> file/open, select an image file like a jpg?
[19:56:31] <Tecan> its not working here
[19:56:50] <cradek> then maybe your config doesn't have image-to-gcode set up as a filter
[19:57:06] <Tecan> oh that might be :) i just used a sample config
[19:57:06] <cradek> you could try the sim/axis config to see how it all works, I'm pretty sure it's set up there.
[19:57:17] <ssi> I bought three of those AMC drives
[19:57:39] <kb8wmc> my last was 626-g402b6cd
[19:58:32] <cradek> that's way way old
[19:58:44] <kb8wmc> oooohhh
[19:59:30] <djdelorie> cradek: there's nothing magic about the python-as-gcode stuff in linuxcnc, is there? With a config line, any language would be supported?
[19:59:33] <cradek> from January some time, just the start of the rebranding
[20:00:00] <cradek> djdelorie: nothing is magic. anything that spits out gcode on stdout makes a fine filter.
[20:00:06] <djdelorie> sweet
[20:00:30] <kb8wmc> I am going to check my software sources
[20:00:32] <cradek> I've used sed as a filter to read a very slightly different dialect of gcode
[20:01:42] <kb8wmc> hmmm....I believe it is correct as follows: "ht--://buildbot.linuxcnc.org/ lucid v2.5_branch-rt"
[20:02:34] <cradek> did you censor your http?
[20:02:51] <kb8wmc> yes
[20:02:54] <cradek> but yeah I think that's what it should look like. what happens when you do apt-get update, apt-get upgrade?
[20:03:08] <kb8wmc> will do so right now....
[20:04:20] <pfred1> maybe that is why my system never broke when I updated it I always used aptitude update;aptitude safe-upgrade
[20:04:42] <kb8wmc> it does the update, and nothing to upgrade
[20:10:48] <cradek> http://buildbot.linuxcnc.org/dists/lucid/v2.5_branch-rt/binary-i386/?C=M;O=D
[20:11:04] <cradek> you should get 2.5.0.9.gdc10e71
[20:11:30] <cradek> I bet you've got emc2 installed, not linuxcnc
[20:11:46] <kb8wmc> ah, yes sir, I do
[20:11:56] <cradek> aha
[20:12:18] <cradek> just apt-get install linuxcnc
[20:12:43] <cradek> or switch to the releases if that toots your horn
[20:12:56] <kb8wmc> rgr, will do....tnx for the help
[20:13:00] <cradek> welcome
[20:14:27] <kb8wmc> doing the install now....
[20:20:31] <kb8wmc> cradek: installed ok, all is well so far
[20:23:39] <kb8wmc> now I will move some files around
[20:24:37] <kb8wmc> cradek: oh, what will be the new resource url?
[20:25:07] <cradek> I don't understand what you mean by resource url
[20:25:43] <kb8wmc> the software source location for updates
[20:26:15] <cradek> for buildbot packages it's the same as what you have now. for releases it's different
[20:26:18] <cradek> which are you wanting?
[20:26:42] <cradek> releases are deb http://linuxcnc.org lucid linuxcnc2.5
[20:27:20] <kb8wmc> I best leave it as is....
[20:27:27] <kb8wmc> thanks again
[20:27:32] <cradek> welcome
[20:50:54] <skunkworks__> logger[psha]:
[20:50:54] <logger[psha]> skunkworks__: Log stored at http://psha.org.ru/irc/%23linuxcnc/2012-04-10.html
[21:01:24] <asdfasd> the closed loop is back :)
[21:03:59] <skunkworks__> ;)
[21:09:43] <skunkworks__> I do have a couple of stepper machines....
[21:11:13] <skunkworks__> seb!
[21:11:35] <Tecan> how do i speed up my gcode ?
[21:11:47] <asdfasd> I hate myself that I used steppers for the last project
[21:11:49] <Tecan> my table is slow as air drying
[21:13:20] <djdelorie> Tecan: for my table, I reduced the steps/rot but that's not something you can do with a stepper ;-)
[21:13:40] <Tecan> no its something in gcode im going to try g97
[21:13:47] <djdelorie> make sure the "max speed" is correct in your ini file for each axis
[21:14:13] <djdelorie> make sure there isn't some confusion between IPM and IPS ?
[21:15:05] <Tecan> whats the fastest feed speed i can set ?
[21:15:17] <Tecan> i have g01 f200
[21:16:29] <djdelorie> the fastest you can set depends on your machine's limits, linuxcnc won't exceed those limits regardless of the gcode
[21:17:13] <skunkworks__> depending on what the moves are - your acceleration has a lot to do with it also..
[21:18:56] <Tecan> its not my limits its the code
[21:19:16] <asdfasd> change F
[21:20:34] <skunkworks__> have you tried http://linuxcnc.org/docs/2.5/html/gcode/gcode.html#sec:G61-G61.1-G64
[21:20:45] <djdelorie> does the runtime in File->Properties change if you load the gcode up in the simulator?
[21:20:59] <skunkworks__> g64px.xxxqx.xxxx
[21:22:06] <asdfasd> g64 do not affect straight lines, is it moving fast on straigt lines?
[21:27:07] <OhSnapp> not moving fast in straight lines
[21:27:29] <OhSnapp> its probably because im setting my S50 value for spindle speed
[21:27:31] <asdfasd> then g64 will not change
[21:27:39] <joe9> is it a bad idea to connect the -VDC (48Volts) to the taig frame?
[21:27:54] <OhSnapp> is it suposed to be S or F for spindle speed setting ?
[21:28:03] <asdfasd> s
[21:28:29] <asdfasd> it is a bit tifferent if you are using g95
[21:28:41] <OhSnapp> i have g96 on
[21:28:53] <asdfasd> but in normal conditions S is RPM and F is units per min
[21:28:57] <OhSnapp> and 97
[21:29:21] <OhSnapp> hmm brb i try some more things
[21:30:23] <asdfasd> ops
[21:30:29] <asdfasd> I mean G96
[21:30:35] <asdfasd> I just checked
[21:31:24] <asdfasd> you set constant surface speed and constant spindle, that mean it will stick to the spindle RPMs
[21:31:25] <OhSnapp> if i move my feed rate override it gets a little better
[21:31:53] <asdfasd> you can try G94
[21:32:02] <asdfasd> feen per minute
[21:32:05] <asdfasd> feed
[21:32:10] <joe9> alex4nder: is it a bad idea to connect the -VDC (48Volts) to the taig frame?
[21:32:20] <asdfasd> then F200 will give you 200 IPM
[21:32:47] <alex4nder> joe9: you mean as a common ground?
[21:32:50] <asdfasd> this is too fast actually
[21:33:00] <asdfasd> be carefull
[21:33:57] <asdfasd> joe9: ground loops are tricky
[21:34:02] <joe9> alex4nder: yes, but not the GND pin of the ac connection.
[21:34:18] <alex4nder> joe9: I don't ever depend on shared grounds for signals
[21:34:29] <joe9> this is the -ve pin of the PSU.
[21:34:33] <alex4nder> I think it's because I've done a lot of race car wiring, and that stuff will drive you insane if it goes wrong.
[21:34:56] <alex4nder> what's your goal?
[21:35:16] <joe9> i am trying to come up with a crude home switch
[21:35:26] <alex4nder> oh, like a wire running into the chassis?
[21:35:30] <alex4nder> "chassis"
[21:35:32] <joe9> yes.
[21:35:52] <alex4nder> where are you going to contact that isn't painted/annodized?
[21:36:00] <alex4nder> -n
[21:36:19] <joe9> oh, the anodized stuff does not conduct electriccity?
[21:36:31] <alex4nder> it will, but it's an oxide layer
[21:37:11] <alex4nder> so you're going to have trouble keeping it conductive
[21:37:41] <joe9> i was thinking of having the chasis to the -ve pin of the psu. and having a small copper plate mounted on a double sided scotch tape that is the +ve pin (with a resistor and all that).
[21:38:16] <joe9> so, that when the small copper sheet touches the chasis, it is home.
[21:38:17] <alex4nder> it's going to depend on where the ground is attached
[21:38:36] <OhSnapp> lol
[21:38:49] <OhSnapp> f is the spindle speed and s is feed
[21:39:01] <alex4nder> joe9: I think you'll find that approach to be flakey
[21:39:29] <alex4nder> and going to be heavily dependent on where you expect current to flow, and where you're forced to remove paint
[21:39:29] <asdfasd> G95 G96 G97 make a lots of changes
[21:39:41] <joe9> oh, so I am better off with a tactile switch like this: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&itemSeq=113121133&uq=634695990904919904
[21:40:09] <asdfasd> better try in G94 mode first, so you can setup all correctly
[21:40:30] <alex4nder> joe9: that didn't load for me, but yah.. a real switch makes sense.
[21:40:30] <joe9> stick it with double sided tape to the axis plate which comes in contact with the chasis at the end of travel.
[21:40:43] <alex4nder> I'd hard mount it
[21:41:07] <alex4nder> unless it's PSA with specified compression characteristics, it's going to be flakey
[21:41:54] <djdelorie> or four switches, in corners of a plate that the bit can hit, wired so that two switches in opposite corners have to close to "activate" ?
[21:42:02] * djdelorie is overthinking again
[21:42:15] <joe9> http://search.digikey.com/us/en/products/TL3310AF200QG/EG4589CT-ND/1794143 alex4nder try this link.
[21:42:56] <alex4nder> joe9: do you think that switch will survive a mill environment?
[21:44:02] <joe9> alex4nder: it will be on the underside of the axis, protected by the waycovers for the x and y axis
[21:44:02] <djdelorie> the four-switch option lets you use a rubber gasket around the perimiter to seal it
[21:44:30] <alex4nder> joe9: well if you trust it
[21:44:55] <joe9> alex4nder: it is just 87cents. should not be a big deal to replace it, if it does not.
[21:45:20] <alex4nder> right, if your time is free, and the annoyance of a failed switch costs nothing.
[21:45:35] <Tom_itx> might be a big deal to replace what it breaks when it fails
[21:46:10] <alex4nder> joe9: what's the goal of the switch?
[21:46:22] <joe9> Tom_itx: probably not, it is a home'ing switch.
[21:46:29] <asdfasd> joe9: better take something more durable, and better dust resistant
[21:46:39] <joe9> alex4nder: to identify the end of travel for that axis.
[21:46:54] <alex4nder> so an emergency limit?
[21:47:01] <joe9> and I plan on using limits in the .ini file that correspond to the actual travel per the taig spec.
[21:47:04] <Tom_itx> it looks like a smt rubber bumper switch
[21:47:08] <joe9> alex4nder: no, a home switch.
[21:47:26] <alex4nder> that happens to be where a limit switch is. :D
[21:47:35] <joe9> yes.
[21:47:45] <joe9> only on one side.
[21:47:47] <djdelorie> you probably don't want a home switch that the carriage could damage if it smashes into it. Better to be one that the carriage can go past
[21:48:20] <alex4nder> if you fixture a lot, I could see the reason to make home switches a priority.. but I've never had a need for home switches.
[21:48:37] <alex4nder> I spend all my time with touchoff.
[21:49:25] <djdelorie> I used MS0850505F045S1A switches for mine
[21:50:36] <alex4nder> djdelorie: yah exactly
[21:50:49] <alex4nder> I think the roller and lugs are the ticket
[21:51:38] <djdelorie> the X has a bump-out that swipes past the switch so it won't hurt it; the other two have the switches next to the actual ends so the machine takes the impact
[21:51:44] <alex4nder> joe9: this will all work, the question is 'how long' and 'how well'..
[21:52:02] <alex4nder> I don't use my mill in a professional environment, and my steppers are 166 oz/in.. so ..
[21:52:24] <djdelorie> most reliable is probably a hall switch, uses magnetics, neither dust nor coolant will stop it from working
[21:52:26] <joe9> djdelorie: how did you attach them to the machine?
[21:52:36] <joe9> double-sided scotch mounting tape?
[21:52:43] <joe9> is what I am thinking of.
[21:52:45] <alex4nder> djdelorie: yah, and the sensors are relatively cheap
[21:53:13] <djdelorie> #2 screws through the sides http://www.delorie.com/photos/cnc/img_2571.html
[21:53:36] <joe9> i do not want to drill my machine to attach the screws.
[21:53:36] <djdelorie> http://www.delorie.com/photos/cnc/img_2561.html shows the X with the bump-out
[21:54:27] <alex4nder> joe9: there are quite a few mounting locations that could be used
[21:54:55] <alex4nder> I was thinking of mounting a bracket under the way cover mount point
[21:55:02] <alex4nder> and then the switches would screw into that
[21:55:24] <djdelorie> if you rely on tape and it fails, your next home damages the machine
[21:55:35] <joe9> alex4nder: good idea.
[21:56:08] <joe9> djdelorie: i noticed that when I get the axis past the end, it just rotates and nothing happens. not sure if it is a good idea to let it happen many times.
[21:56:35] <alex4nder> do that at 60 IPM
[21:56:37] <alex4nder> it's fun
[21:57:01] <djdelorie> I managed to bend a 3/8-16 screw rod on my test jig when the motor didn't stop at the end
[21:57:15] <alex4nder> you're also using frankenservos
[21:57:31] <djdelorie> 100 oz-in continuous
[21:59:41] <alex4nder> joe9: the cool thing is, now that you have a mill you can make all the brackets yourself
[21:59:45] <alex4nder> the world is your oyster
[21:59:46] <OhSnapp> where do i set default speed mode
[22:00:39] <OhSnapp> G97 is using it on me
[22:01:07] <joe9> alex4nder: ok, thanks.
[22:01:38] <alex4nder> joe9: do you have a giant e-stop switch yet?
[22:01:44] <alex4nder> one that you can reach out and smack?
[22:01:45] <djdelorie> alex4nder: the problem with restoring an antique machine, is that you *have* to make all the parts yourself
[22:02:03] <joe9> alex4nder: no, I am just using a light switch as an e-stop still.
[22:02:06] * djdelorie still needs to add a *power* switch
[22:02:13] <alex4nder> joe9: the light switch? you mean for the spindle?
[22:02:21] <alex4nder> or some other light switch
[22:02:24] <joe9> no, as the e-stop switch too.
[22:02:32] <alex4nder> ah
[22:02:41] <alex4nder> well if you ask me, a giant e-stop is way more important than limit switches
[22:02:43] <alex4nder> or home
[22:02:55] <joe9> ok, will get one from digikey.
[22:03:05] <alex4nder> djdelorie: one is probably a machining badass before the antique machine purchase.
[22:03:13] <alex4nder> djdelorie: plus, machines beget more machines. ;)
[22:03:15] <djdelorie> not in this case
[22:03:24] <djdelorie> it was my father-in-law's lathe before he passed away
[22:03:25] <alex4nder> well the best of luck then
[22:04:10] <joe9> e-stop switches are pretty expensive.
[22:04:44] <alex4nder> you just bought a $1200 mill
[22:04:52] <joe9> yeah, i know.
[22:05:16] <alex4nder> I personally like Omron's giant e-stop switches
[22:05:41] <djdelorie> perhaps I should use standard big-tool power switches, the type you can just smack to shut off
[22:06:01] <Valen> estop stuff should be bashable
[22:06:17] <Valen> because if you are stopping in the case of an E you will bash it ;->
[22:06:25] <djdelorie> most of my woodworking tools have a "hit without looking" off switch
[22:06:33] <Valen> we ran without limit switches for ages
[22:06:36] <alex4nder> djdelorie: that's what Omron's e-stop switches are.
[22:06:41] <Valen> actually we still are running without limits
[22:06:44] <alex4nder> smash to kill, rotate to reset.
[22:06:53] <djdelorie> except these are the main power switches, not the big e-stop switches
[22:07:03] <Valen> but now we have the index pulses on our slides connected as home switches
[22:07:08] <alex4nder> nice
[22:07:13] <Valen> so it doesn't hit the limits anyway
[22:07:37] <Valen> we had fuses in the motor lines so running into the end made a bad noise and blew a fuse
[22:07:43] <djdelorie> like this one: http://www.grizzly.com/products/Magnetic-On-Off-Switch-110V-6-Amps/G8992
[22:08:17] <Valen> i wouldn't use one of those unless it had one of those covers that means bashing it will only turn it off
[22:08:30] <Valen> http://www.grizzly.com/products/110V-Paddle-On-Off-Switch/H8241 like this kinda
[22:08:43] <alex4nder> yah, that you could throw a baseball at
[22:08:44] <Valen> but those are a pain, the twist ones are easy to use
[22:09:06] <Valen> problem with the two button one is you could still hit the wrong one
[22:09:18] <djdelorie> most of mine have a recessed ON and a protruding OFF. Except the delta, which just puts a collar around ON so you can't "just" smack it
[22:09:53] <djdelorie> like this, which I hate: http://www.grizzly.com/products/110-220V-On-Off-Switch/H8238
[22:10:16] <alex4nder> I'm going to replace my Omron with a Hello Kitty lightswitch
[22:10:27] <alex4nder> it's illuminated
[22:10:37] <djdelorie> and you enjoy smacking it ? ;-)
[22:11:05] <alex4nder> yes; I always thought Hello Kitty sucked
[22:18:16] <OhSnapp> cradek theres something wrong with g97, it wont allow itself to be set without adjusting the feed rate
[22:25:37] <OhSnapp> no wonder... feed rate is the feed per revolution... i should still be able to make it go faster wth
[22:47:49] <OhSnapp> its strange if i set g01 f6 or f7 it gets its fastest
[22:48:01] <OhSnapp> any higher or lower is bad
[22:52:16] <OhSnapp> awe sweet its getting faster
[22:53:44] <OhSnapp> tis working afterall sorry
[22:55:17] <OhSnapp> it seems to be capped though
[22:56:42] <OhSnapp> i have to many lines of gcode im thinking
[23:15:48] <Jymmm> SWPadnos: Pack Size : 22" H x 17" W x 11" D
[23:25:03] <OhSnapp> uploading video
[23:28:41] <Tecan> http://youtu.be/18_fyBeGRSk
[23:28:49] <Tecan> look how slow its going
[23:36:37] <ssi> http://www.ebay.com/itm/HARDINGE-TURRET-TOP-PLATE-3-8-WITH-TOOLING-/290695609249?pt=LH_DefaultDomain_0&hash=item43aecefba1#ht_500wt_1413
[23:36:44] <ssi> if that was about half that money, I might snatch it :P
[23:52:39] <Tecan> http://i.imgur.com/CBPI1.jpg
[23:52:44] <Tecan> lol sooo slooow
[23:55:57] <ssi> lol
[23:56:11] <ssi> the more I think about it, the more I want a little laser
[23:56:17] <ssi> how well does that 40w tube do marking steel?
[23:58:19] <Tecan> hmmm one sec i think you need 60 for that
[23:58:49] <ssi> 60 may be doable
[23:58:53] <ssi> above that it gets real spendy real fast
[23:59:12] <Tecan> 40 had no effect
[23:59:20] <Tecan> i was just using the test button though
[23:59:40] <ssi> well I've also heard you need some special coatings to make engraving work well with light power lasers