Back
[01:03:45] <RyanS> Look at what $1 buys you these days
http://www.machines4u.com.au/view/advert/OKUMA-MX-45VAE-CNC-VERTICAL-MACHINING-CENTRE/94455/
[02:01:16] <DJ9DJ> moin
[03:23:10] <WalterN> where is a good place to get light-cured liquid plastic (for rapid prototyping machine)?
[03:44:11] <Tom_itx> http://polymarkinc.com/uv_curables.html
[10:03:56] <skunkworks> pcw_home:
http://imagebin.org/index.php?mode=image&id=265484
[10:25:42] <CaptHindsight> WalterN: how much do you need? For gallons or more
http://bucktownpolymers.com/polymer00.html
[10:40:45] <skunkworks> http://imagebin.org/index.php?mode=image&id=265488
[10:41:54] <archivist> shed extension time again :)
[10:42:07] <skunkworks> no - cleanup...
[10:42:12] <skunkworks> ;)
[10:42:35] <skunkworks> another project
[10:42:56] <archivist> I realise it is another project , that usually means more space needed!
[10:48:39] cradek changed topic of
#linuxcnc to: LinuxCNC is a linux-based opensource CNC control. | Latest release: 2.5.3 |
http://www.linuxcnc.org |
http://wiki.linuxcnc.org
[10:49:27] <skunkworks> I am not understanding this plate
http://imagebin.org/index.php?mode=image&id=265484
[10:49:55] <skunkworks> I get brushless, 60a, and 2000rpm
[10:53:08] <skunkworks> that seems afully high for 10.6ft-lbs (14.5nm)
[10:53:50] <jdh> 3 phase permanent magnet?
[10:54:15] <archivist> bldc/ac servo
[10:54:54] <archivist> so many names !
[10:55:06] <skunkworks> heh
[10:55:32] <CaptHindsight> maybe it's high current low voltage
[10:55:46] <archivist> lets invent a name so only our drive appears to work
[10:55:58] * JT-Shop finally got everyone on the network to speak to each other
[10:56:17] <jdh> maybe it's 6 amps
[10:56:53] <archivist> is that 14.5 A at 60hz
[10:57:16] <skunkworks> archivist - seriously?
[10:57:31] <skunkworks> I hate these type of spec tags
[10:57:41] * archivist out driving miss Daisy
[10:57:43] <jdh> damned germans
[10:58:01] <skunkworks> so it is 10nm?
[10:58:26] <jdh> http://www.ebay.com/itm/161072533907 better version of their plate.
[11:06:21] <CaptHindsight> 150V Y, 14.5A is most likely
[11:16:37] <skunkworks> CaptHindsight: I hope so ;)
[11:17:07] <PetefromTn> morning folks.
[11:19:55] <skunkworks> PetefromTn: moring
[11:19:57] <skunkworks> morning
[11:19:58] <PetefromTn> ;)
[11:21:18] <skunkworks> PetefromTn: - new project
http://imagebin.org/index.php?mode=image&id=265488
[11:21:26] <skunkworks> look - it has coolant control!
[11:21:50] <skunkworks> Yay
[11:25:26] <Jp11> Acraloc?
[11:25:42] <skunkworks> yes
[11:26:02] <skunkworks> another machine with odd tooling.. :) we are used to it.
[11:26:54] <Jp11> They work well easy on feeds and keep holders clean
[11:26:55] <PetefromTn> skunkworks: Cool man those are supposed to be decent machines altho sorta oddball. How far you got with it?
[11:27:13] <skunkworks> I think it is outside the shop door at the moment.. ;)
[11:27:33] <skunkworks> it was delivered yesterday
[11:28:52] <skunkworks> don't know if the oringinal control works yet.
[11:29:31] <jdh> sounds like you should unload some of those lathes
[11:30:42] <skunkworks> jdh: getting there...
[11:31:26] <skunkworks> http://imagebin.org/index.php?mode=image&id=265491
[11:32:18] <jdh> cool buttons
[11:35:04] <cradek> "spindle speed rpm" has 6 digits
[11:35:42] <jdh> it's in metric
[11:36:00] <jdh> RPFN
[11:37:45] <TekniQue> metric revolutions per minute
[11:38:35] <jdh> revolutions per fortnight
[11:38:52] <TekniQue> thumbtwists per tidal period
[11:39:08] <cradek> it could be revolutions per kilosecond
[11:39:49] <cradek> 6000 rpm = 100000 rpks
[11:41:23] <cradek> or 100000 millirevolutions per second
[11:45:29] <skunkworks> heh
[11:49:59] <WalterN> CaptHindsight: oh ok... thanks
[11:51:13] <pcw_home> Neat button hieroglyphics
[11:51:45] <WalterN> CaptHindsight:
http://support.formlabs.com/entries/23813638-Can-I-use-third-party-resins-with-my-Form-1-
[11:54:16] <pcw_home> No you cannot use other resins as it will interfere with our business plan...
[11:54:54] <cradek> 0 people found this useful.
[11:55:36] <WalterN> yeah
[11:55:46] <WalterN> I just need to know what the wavelength is for the laser
[11:56:07] <WalterN> ...which I'm not finding
[11:57:32] <pcw_home> blue laser?
[11:59:57] <WalterN> hmm... this guy is saying that red/orange tends to block UV light
[12:02:08] <WalterN> http://www.lasersafetyindustries.com/35_110_Laser_Safety_Glasses_p/35-110.htm
[12:02:57] <WalterN> and its probably not IR curable
[12:05:10] <PetefromTn> resins and lasers? Whazzat?
[12:12:40] <IchGuckLive> hi all B)
[12:12:59] <WalterN> hmm
[12:13:27] <WalterN> anyone know of good cheap laser engraving machines?
[12:13:42] <IchGuckLive> buld your own one
[12:14:02] <IchGuckLive> otherwise make a good fire enshurens as it will burn ff soon
[12:14:03] <JT-Shop> good and cheap don't fit in the same sentance
[12:14:21] <IchGuckLive> for lasermashines as well
[12:14:23] <WalterN> s/cheap/least expensive/
[12:15:20] <IchGuckLive> what are you going to engrave
[12:15:22] <IchGuckLive> wood
[12:15:25] <WalterN> all the things
[12:15:29] <IchGuckLive> it is a hell of a diference if you need 30W or 60W
[12:15:39] <IchGuckLive> or even 120W
[12:15:49] <WalterN> why not 150?
[12:15:56] <jdh> 221
[12:16:00] <IchGuckLive> so 800USD-80kUSD
[12:16:00] <ReadError_> got some NeverWet to play with :)
[12:16:22] <WalterN> also, used is acceptable too
[12:16:49] <IchGuckLive> WalterN: you are in US or europ
[12:16:53] <jdh> I looked in to the cheap 40w ones recently. Decided against getting one.
[12:16:54] <WalterN> USA
[12:17:09] <jdh> looks like $1800us is a more realistic starting point
[12:19:29] <IchGuckLive> what size do you need 200+300mm = 6"x8"
[12:20:02] <WalterN> hmm
[12:20:04] <IchGuckLive> 40W is good at 930USD at amazon
[12:20:10] <WalterN> all the size?
[12:20:12] <jdh> you can get 40w from keling for $680
[12:20:21] <IchGuckLive> oh thats good
[12:20:29] <jdh> only good in that it is cheap.
[12:21:13] <WalterN> I'd like to be able to stick larger things in it... maybe 15"x15" or bigger?
[12:21:15] <jdh> no Z, moshi draw drivers that burn up
[12:22:04] <jdh> can only use bad chinese software in windows and requires a USB key
[12:22:24] <WalterN> yeah screw that
[12:25:40] <jdh> the 50w ones seem to be much more useful, but cost 3-4x
[12:31:43] <PetefromTn> I'd probably burn the piss out of myself if I built one of those LOL....
[12:31:57] <IchGuckLive> 60W is bst if you can affort it dont forget the loss of power after 50-100Work houres of the tube
[12:32:38] <WalterN> ..?
[12:37:10] <CaptHindsight> WalterN: some resins can get too hot for the Form1
[12:38:27] <CaptHindsight> WalterN: what wavelength for what resin? The BP -v420 resins are sensitive from UV up to 440nm
[12:39:03] <WalterN> CaptHindsight: or 385
[12:39:15] <CaptHindsight> WalterN: just use a Bluray 405nm
[12:39:58] <CaptHindsight> the -UV are sensitive up to ~390nm
[12:40:01] <WalterN> yeah
[12:40:14] <WalterN> CaptHindsight: I'm guessing its UV because of the orange cover
[12:41:04] <CaptHindsight> oraange cover on what?
[12:41:16] <WalterN> the form1
[12:41:31] <CaptHindsight> the Form1 uses it to block light from outside the printer from initiating polymerization and also from the laser hurting anyone
[12:41:55] <WalterN> right, which leads me to believe that its a UV laser of some sort
[12:42:02] <CaptHindsight> they use 405nm
[12:42:41] <WalterN> the form1 uses 405nm?
[12:42:54] <CaptHindsight> yeah, last I checked
[12:43:08] <WalterN> interesting... where did you see that? I couldent find anything
[12:44:14] <CaptHindsight> I formulate photopolymers for BP and others, just use a 405nm with the -v420 and you're done
[12:44:31] <WalterN> oh woah
[12:45:02] <WalterN> no wonder why you know this stuff :P
[12:45:08] <CaptHindsight> the problem with the Form1 and the B9 is the vat
[12:45:19] <WalterN> B9?
[12:45:19] <CaptHindsight> if you use to fast a curing photopolymer it gets soo hot it melts their vat coating
[12:45:28] <WalterN> oh
[12:45:39] <WalterN> I could possibly make a glass vat though?
[12:45:44] <WalterN> hmm
[12:45:59] <CaptHindsight> sure and use the same coating
[12:46:49] <WalterN> what is -v420?
[12:47:00] <WalterN> oh, 420nm plastic goop?
[12:47:03] <CaptHindsight> yeah
[12:47:17] <WalterN> how much would a gallon cost?
[12:47:20] <CaptHindsight> the coating is Dow Sylgard 184
[12:47:27] <WalterN> (or two liters if you prefer)
[12:48:59] <CaptHindsight> it's starts around $180/gal, I can send you a liter to play with
[12:48:59] <WalterN> and it should be fine in the form1?
[12:48:59] <CaptHindsight> yes
[12:49:18] <WalterN> hmm
[12:50:00] <WalterN> I might have to buy one when it starts shipping
[12:50:10] <CaptHindsight> build your own, you already have most of the parts
[12:50:18] <WalterN> not really
[12:50:23] <WalterN> the laser I have is 808nm
[12:50:41] <CaptHindsight> a 405nm Blu ray laser is $25 on ebay
[12:50:43] <WalterN> and I dont have a proper power supply for it either
[12:51:12] <WalterN> I do have a galvo system though
[12:51:36] <CaptHindsight> http://pryntech.com/ improve his design
[12:52:32] <CaptHindsight> prove that a noob can do better than a 20 something know-it-all
[12:52:42] <WalterN> lol
[12:52:47] <CaptHindsight> not that I have anything against 20 somethings :)
[12:53:26] * WalterN hides his 26 years behind the corner
[12:53:44] <CaptHindsight> bbl, new positioners just arrived
[12:54:48] <WalterN> the thing I'm not sure about is how to get those galvos to work with linuxCNC
[13:02:39] <WalterN> oh, thats designed to work with OpenSL?
[13:02:48] <WalterN> meh
[13:02:57] <WalterN> I'd rather it work with linuxCNC
[13:07:44] <WalterN> CaptHindsight:
http://pryntech.com/?product=20kpps-laser-galvanometer-kit is what I have
[13:13:17] <CaptHindsight> WalterN: you can use the Mesa servo cards to drive the galvo's, he gets them from China (or are made there anyway) +- 10V
[13:13:40] <WalterN> ..?
[13:14:52] <WalterN> and then those connect to linuxCNC and the galvo power supply?
[13:15:15] <CaptHindsight> WalterN: the galvo's use either 5 or 10V for their full range of motion
[13:15:16] <WalterN> (hint, I'm not an electronics guy... lol.. sorry)
[13:16:12] <CaptHindsight> the Mesa servo board can supply the analog 10V to the galvo driver board
[13:16:21] <WalterN> which mesa card?
[13:16:24] <CaptHindsight> let me check if they are 5 or 10V, or selectrable
[13:16:50] <CaptHindsight> I forget the Mesa part number
[13:17:23] <CaptHindsight> pcw_home: what's the servo card with 2+ outputs?
[13:17:53] <WalterN> 4I27A 2 axis servo motor controller card (analog out)
[13:17:57] <WalterN> that one?
[13:18:46] <WalterN> or does it need to be PWM?
[13:19:03] <WalterN> I'm rather clueless... lol
[13:19:46] <CaptHindsight> the galvo drives are analog in, not PWN IIRC
[13:20:02] <CaptHindsight> PWN/PWM
[13:20:05] <WalterN> oh ok
[13:20:07] <WalterN> how do the galvo's keep track of their position? (or do they)
[13:20:29] <CaptHindsight> it will be an open loop with those galvos
[13:20:39] <WalterN> which means..?
[13:21:48] <CaptHindsight> http://www.alibaba.com/product-gs/748755952/30Kpps_Speed_Galvo_Scanner_For_Laser.html
[13:24:32] <CaptHindsight> he might use these
http://www.alibaba.com/product-gs/734776255/20Kpps_Speed_Galvo_Scanner_For_Laser.html
[13:25:13] <WalterN> no... the neck of those is plastic
[13:25:23] <CaptHindsight> the high + moves the mirror to one end of its travel and a low to the other end
[13:25:50] <WalterN> lemme go look at mine
[13:26:05] <CaptHindsight> there are a few manufacturers in China that make them, all very similar
[13:27:12] <WalterN> 20k-0012161
[13:27:14] <CaptHindsight> the driver boards use either analog +- 5V or that stage lighting digital protocol ( I forget the name of)
[13:27:35] <WalterN> is the part number printed on... it looks like the two controller boards
[13:29:00] <CaptHindsight> the mirrors are just driven by coils, just like a brushless motor only with <180 deg of travel/rotation
[13:29:13] <WalterN> oh ok
[13:30:25] <WalterN> should I take a picture of what I have?
[13:30:41] <WalterN> or is that necessary... to get the matching stuff for what I need
[13:30:42] <CaptHindsight> the driver board dives the coils, it just tracks the analog voltage that you supply to control position
[13:31:23] <CaptHindsight> the galvos you have are driven by the analog out from a arduino type board, ramps or something
[13:32:22] <WalterN> so... hmm
[13:32:54] <CaptHindsight> ah it is PWM out from his control board
[13:35:57] <CaptHindsight> Mesa 7i33 has 4 PWM servo out
[13:36:28] <WalterN> is that what he is using?
[13:36:28] <CaptHindsight> sorry PWM in
[13:36:39] <WalterN> oh, isnt PWM in what we want though?
[13:36:45] <CaptHindsight> no, he drives them from his arduino like board with an Atmel ARM SOC
[13:37:43] <WalterN> I'm kinda confused now... maybe I should make a flow chart
[13:37:43] <CaptHindsight> yes, then you should be able to use any anything I/O card with PWM outs to control the galvo driver/amp boards
[13:38:41] <CaptHindsight> so Mesa Anthing I/O PWM ----> galvo driver ----> galvo mirrors
[13:39:17] <WalterN> so the galvo driver takes PWM in, and outputs an analog current?
[13:39:23] <CaptHindsight> I'll look over hos wiring later
[13:39:58] <CaptHindsight> did you get any specs with that galvo kit?
[13:40:40] <WalterN> I didnt get diddly
[13:40:43] <CaptHindsight> his schematics are up there somewhere
[13:40:56] <WalterN> "here have the electronics... have fun figuring out what pins do what"
[13:41:11] <CaptHindsight> just have to see if those driver boards he sent you are PWM or +- 5V input
[13:42:45] <WalterN> http://pryntech.com/wiki/index.php?title=Open_SL_Electronics
[13:43:38] <CaptHindsight> " Schematic Todo " I though he had it up before
[13:43:59] <WalterN> mouser parts list... if thats useful at all
[13:44:56] <WalterN> http://pryntech.com/forum/viewtopic.php?id=49
[13:47:10] <WalterN> oh here...
https://github.com/beardface/LAMPS/tree/master/pcb_design/OpenSL_1.0
[13:52:09] <CaptHindsight> it looks like PWM out
[13:54:03] <WalterN> so this is what I want?
http://www.mesanet.com/pdf/motion/4i28ds.pdf
[13:54:59] <CaptHindsight> thats a PC/104 card
[13:55:46] <WalterN> ..?
[13:57:24] <CaptHindsight> http://www.mesanet.com/fpgacardinfo.html anything IO with PWM, several cards have it
[13:59:21] <CaptHindsight> WalterN: do you have a voltmeter and can you solder?
[13:59:52] <WalterN> I can do both
[13:59:59] <CaptHindsight> just checking
[13:59:59] <WalterN> I actually have a nice soldering iron
[14:00:23] <CaptHindsight> some software devs don't have or touch either
[14:00:31] <WalterN> meh
[14:00:43] <WalterN> I'm not a software dev either...
[14:00:55] <WalterN> but I've made a few small programs in C
[14:01:32] <CaptHindsight> can you post a high res pic of the galvo driver board, both front and back?
[14:01:44] <WalterN> hmm
[14:01:47] <WalterN> sure
[14:01:53] <WalterN> should I set up my white box?
[14:02:05] <CaptHindsight> I'll look it over later
[14:02:39] <CaptHindsight> you might be able to drive it with 7i43 or one of the PCI cards
[14:03:38] <jdh> what freq does it want for the PWM?
[14:04:49] <WalterN> well
[14:05:01] <WalterN> if I get a new computer, not too many motherboards have PCI slots, only PCIe
[14:05:28] <CaptHindsight> well the response time is only 20K positions per second, it looks like it takes PWM in and converts that to an analog signal and then amplifies that to supply current to the mirrors coils
[14:06:06] <jdh> 6i25 is pcie
[14:07:43] <CaptHindsight> i just bough a brand new MB
http://www.asus.com/Motherboards/F2A85V_PRO/#specifications 2 x PCIe 2.0 x16 (x16 or dual x8), 1 x PCIe 2.0 x16 (x4 mode), 2 x PCIe 2.0 x1, 2 x PCI
[14:09:45] <CaptHindsight> http://pryntech.com/wp-content/uploads/2013/04/20130426_233436.jpg
[14:10:15] <CaptHindsight> what name is on that board?
[14:10:49] <CaptHindsight> it might have PWM and analog in
[14:11:38] <CaptHindsight> phenixtechnology
[14:12:50] <WalterN> ok, I have some pretty good pictures
[14:13:03] <CaptHindsight> http://www.phenixtechnology.com/uploadfile/file/20120816114123_54373.pdf
[14:17:16] <CaptHindsight> I don't see an DAC on the LAMPS board
[14:18:04] <WalterN> CaptHindsight:
http://tiwake.com/IMG_0402.JPG http://tiwake.com/IMG_0403.JPG http://tiwake.com/IMG_0404.JPG
[14:18:10] <CaptHindsight> http://www.phenixtechnology.com/ they make the galvo board
[14:18:57] <CaptHindsight> heh, it's upside down :)
[14:19:28] <WalterN> it was easier to photograph that way :P
[14:20:42] <CaptHindsight> http://www.phenixtechnology.com/en/product-ari--6.html
[14:20:58] <WalterN> yup, thats it
[14:21:24] <WalterN> ...except I cant download the manual?
[14:23:17] <WalterN> dang, those are some nice pictures
[14:32:23] <WalterN> CaptHindsight: are those pictures good enough?
[14:32:55] <CaptHindsight> looks like the driver board takes PWM and analog
[14:34:02] <CaptHindsight> no good docs
[14:34:19] <CaptHindsight> IN+ IN- G on that 3 pin header
[14:34:52] <CaptHindsight> and possibly FB G on the 2-pin next to it
[14:36:54] <CaptHindsight> I'll call them in 8 hours
[14:37:07] <CaptHindsight> I wanted to talk to them anyway
[14:38:58] <WalterN> call who?
[14:39:14] <CaptHindsight> the manufacturer
[14:39:20] <WalterN> o.0
[14:39:25] <WalterN> you'd do that for me?
[14:39:37] <WalterN> (evidently... heh)
[14:42:23] <WalterN> I'll be getting home from work about 9.5 hours from now
[14:42:34] <WalterN> erm
[14:42:38] <WalterN> 10 hours rather
[14:46:43] <CaptHindsight> that driver might have position feedback
[14:49:52] <CaptHindsight> it's just 3- quad op amps and some regulators on that board
[14:52:01] <CaptHindsight> analog PID
[15:11:47] <spackWrk> w000t
[15:11:54] <spackWrk> 1 parallel port on meth has just arrived via UPS
[15:12:18] * jdh ponders a p-port with rotten teeth and worn out lips
[15:12:19] <ReadError_> ON METH?
[15:12:24] <ReadError_> bad skin too
[15:12:42] <ReadError_> stupid bugs
[15:18:46] <spackWrk> anyone have a link to the 5i25 hostmot files?
[15:19:49] <spackWrk> 5i25 isn't in the ones included with linuxcnc
[15:20:00] <jdh> does it need them?
[15:21:01] <spackWrk> pretty sure it does
[15:21:03] <DJ9DJ> re
[15:21:24] <spackWrk> pcw_home: you around by any chance?
[15:24:35] <spackWrk> i need to flash the probotix firmware on the card i think
[15:24:50] <spackWrk> or i need them on my computer so inuxcnc can do it
[15:26:53] <JT-Shop> 5i25 does not need any files
[15:27:04] <JT-Shop> to run
[15:27:32] <spackWrk> but does it need files to run my probotix?
[15:27:46] <spackWrk> all i can find are a few threads on it and they all require various files
[15:28:09] <spackWrk> even in pncconf, there is no 5i25 in the list
[15:28:13] <JT-Shop> do you mean directly from the 5i25 ie no daughter card?
[15:28:18] <spackWrk> yes
[15:28:35] <JT-Shop> then you have to flash the 5i25 with the correct firmware
[15:28:43] <spackWrk> which requires... files
[15:29:14] <JT-Shop> does the probix use the parallel port normally?
[15:29:19] <spackWrk> yeah
[15:29:23] <JT-Shop> and the flash utitlty
[15:29:49] <andypugh> What firmware does the 5i25 currently have? (It ought to say on the sticker)
[15:30:26] <spackWrk> just says 5i25
[15:31:04] <spackWrk> unless there is a sticker on the chip side
[15:31:07] <spackWrk> in which case i can't see it
[15:31:24] <andypugh> What is the current pinout?
[15:32:35] <andypugh> OK, so you have had it 15 mins, it might be too early to know that.
[15:32:42] <spackWrk> haha
[15:32:45] <spackWrk> yeah
[15:32:55] <andypugh> open a terminal
[15:32:58] <andypugh> halrun
[15:33:02] <andypugh> loadrt hostmot2
[15:33:09] <andypugh> loadrt hm2_pci
[15:33:12] <andypugh> exit
[15:33:19] <spackWrk> o
[15:33:20] <spackWrk> k
[15:33:22] <spackWrk> oops
[15:33:25] <spackWrk> done
[15:33:28] <andypugh> Then pastebin the result of dmesg
[15:35:05] <spackWrk> http://pastebin.com/jcXkMSGA
[15:35:39] <andypugh> I guess that is the 7i77 firmware
[15:35:52] <andypugh> (7i77 x2 in fact)
[15:35:56] <andypugh> What did you order?
[15:36:01] <spackWrk> just the 5i25
[15:36:26] <andypugh> I think you are meant to specify a firmware at the time you order.
[15:36:28] <andypugh> But, all is not lost.
[15:36:33] <spackWrk> hmm, well nobody asked :|
[15:36:58] <spackWrk> well, i'm down for some reflashing, etc
[15:37:11] <andypugh> You need this set of files, I think.
http://www.mesanet.com/software/parallel/5i25.zip
[15:37:13] <spackWrk> i wasn't expecting this to be easy
[15:37:23] <jdh> I like easy.
[15:38:19] <spackWrk> that's what i was looking for
[15:38:20] <spackWrk> w00t
[15:39:05] <andypugh> In the configs/hostmot2 directory there is a file called prob_rfx2.bit which looks like what you probably want
[15:39:54] <andypugh> And to use that you need to use the utils/linux/mesaflash utility
[15:40:14] <spackWrk> ok
[15:40:39] <andypugh> As the file has been zipped it is likely that mesaflash has lost its execute bit. You may need to make it executable.
[15:40:52] <spackWrk> ok, seems managable
[15:42:16] <andypugh> navigate to the directory where mesaflash downloaded to, then just type ./mesaflash
[15:42:24] <andypugh> I think you should get clues from there.
[15:42:28] <spackWrk> i just did a ./mesaflash --scanpci
[15:42:49] <andypugh> You may need a sudo
[15:43:23] <andypugh> The pin and xml files that match the .bit file are for the benefit of pncconf.
[15:43:56] <andypugh> But you seem pretty quick on the uptake, so might prefer to just read man hostmot2 and configure by hand (based on your existing config)
[15:44:28] <spackWrk> i definitely appreciate the pointers
[15:44:55] <spackWrk> i'll give this stuff a whirl and come back when i get stuck or i get something to move :)
[15:44:55] <spackWrk> thanks
[15:50:42] <JT-Shop> spackWrk: don't forget to cycle power after flashing the 5i25
[15:51:05] <spackWrk> oh, thanks, i would definitely have forgotten
[15:52:23] <spackWrk> brb
[16:01:40] <skunkworks> JT-Shop: you have simens drives in your bridgeport?
[16:02:35] <JT-Shop> on the BP Discovery 308 and the Hardinge CHNC I have siemens drives (different ones)
[16:03:36] <spackWrk> can i stuxnet you?
[16:03:44] <spackWrk> =]
[16:04:35] <andypugh> I suspect I would know what you meant if the answer was yes...
[16:08:23] <skunkworks> JT-Shop: the spindle looks like it uses 3 cards that are hooked togather with a ribbon cable.. I seem to remember seeing a picture like that and I wonder if it was you
[16:09:18] <JT-Shop> my spindle is a big drive on the CHNC and one huge one on the BP
[16:09:52] <skunkworks> ok - then it wasn't you :)
[16:10:13] <JT-Shop> well my axis drives on the CHNC are just cards in a card rack
[16:16:46] <skunkworks> these are brushless
[16:19:18] <skunkworks> pcw_home: did you have a feeling on this
http://imagebin.org/index.php?mode=image&id=265484
[16:22:30] <skunkworks> CaptHindsight: devises - 150V , 14.5A
[16:22:48] <skunkworks> And I would hope that - rather than 60a ;)
[16:23:15] <CaptHindsight> from the size it would seem in the ballpark
[16:23:30] <skunkworks> yes
[16:24:09] <skunkworks> the thing has a lot of movement.
[16:24:17] <CaptHindsight> but they do spread their specs all over the faceplate, sometimes above or below, then left vs to the right
[16:25:15] <skunkworks> heh
[16:26:20] <skunkworks> around 2.2kw - nice
[16:27:24] <skunkworks> hmm - smells like an 8i20 ;)
[16:30:09] <DJ9DJ> gn8
[16:31:11] <spackWrk> ok...
[16:31:25] <spackWrk> so i put the 5i25 .xml in the firmware folder, used pncconf to set up a config
[16:31:41] <spackWrk> started up linuxcnc, chose the new conf
[16:32:41] <spackWrk> and linuxcnc errored out. i can't find the relevant error in the dump. here it is:
http://pastebin.com/wGWkDunX
[16:33:12] <spackWrk> [ 2273.883087] hm2_5i25.0: initialized AnyIO board at 0000:05:02.0
[16:33:14] <spackWrk> [ 2274.375201] hm2_5i25.0: dropping AnyIO board at 0000:05:02.0
[16:33:15] <spackWrk> =[
[16:33:23] <skunkworks> fireball072413.hal:10: parameter or pin 'hm2_5i25.0.pwmgen.pwm_frequency' not found
[16:33:45] <spackWrk> ah, i skimmed over that section
[16:33:49] <spackWrk> didn't see it
[16:34:06] <CaptHindsight> 11nM = ~8 ft. lbs and that would also fit the size of the motor
[16:36:17] <spackWrk> skunkworks: any idea what that means? :)
[16:36:37] <spackWrk> i don't want/need any pwm... dunno where it got that from
[16:36:54] <spackWrk> i guess there are some pwm pins configured in the firmware
[16:38:40] <CaptHindsight> skunkworks: Baldor makes something similar in size and voltage
http://www.baldor.com/support/Literature/Load.ashx/BR1202-F?LitNumber=BR1202-F
[16:39:25] <CaptHindsight> MT-4525-BTYCN page 8
[16:40:45] <CaptHindsight> that's a DC servo
[16:40:52] <spackWrk> the config has setp hm2_5i25.0.pwmgen.pwm_frequency 100000 in it
[16:41:12] <andypugh> spackWrk: I think the problem is the other way round, the config is referring to a pwm that might not exist.
[16:41:24] <spackWrk> hmm
[16:42:27] <andypugh> The firmware looks odd, it actually only seems to contain a single encoder, and nothing else.
[16:42:34] <spackWrk> weird
[16:42:53] <andypugh> Ah, no, wait, perhaps you doidn't configure any stepgens in pncconf?
[16:43:00] <spackWrk> i thought i did
[16:43:22] <spackWrk> going back into pncconf now to see what it sees
[16:43:30] <andypugh> What does the halrun / loadrt hostmot2 / loadrt hm2_pci / exit sequence put in dmesg. That should default to the full set of everything.
[16:44:44] <spackWrk> http://pastebin.com/3jnYNvjh
[16:45:18] <andypugh> That's still the 7i77x2 firmware, I think.
[16:45:21] <spackWrk> hmmf
[16:45:39] <spackWrk> pncconf tells me 2 encoders, 2 pwm generators and 8 step generators
[16:45:53] <spackWrk> pin 1 is listed as unused pwm gen
[16:45:58] <andypugh> You need to power-cycle the card. That might involve shutting down the PC and maybe even unplugging it.
[16:46:26] <andypugh> spackWrk: pncconf is working on what the XML tells it to expect, not what is actually there.
[16:46:29] <spackWrk> i rebooted, but didn't think the card would still be powered
[16:46:34] <spackWrk> ah
[16:47:32] <spackWrk> well, i'll do everything i can to cut power to this board and then try again :)
[16:47:35] <spackWrk> brb
[16:50:24] <spackWrk> now i won't have to rejoin all the time
[16:50:47] <spackWrk> still getting the same error after a shutdown and unplug
[16:51:01] <spackWrk> actually, no
[16:51:03] <spackWrk> it's different
[16:51:28] <spackWrk> insmod: error inserting '/usr/realtime-2.6.32-122-rtai/modules/linuxcnc/lowpass.ko': -1 Invalid parameters
[16:51:48] <andypugh> Interesting.
[16:51:54] <andypugh> dmesg?
[16:51:59] <spackWrk> i see pins and stuff listed now
[16:52:14] <spackWrk> http://pastebin.com/9J4iqQ1a
[16:52:37] <andypugh> Yes, that's got stepgens.
[16:52:52] <spackWrk> HAL: ERROR: duplicate variable '.in'
[16:53:19] <andypugh> Aha, that was a wierd bug caused by using a blank name somewhere.
[16:53:26] <spackWrk> heh
[16:55:06] <andypugh> Does the pinout in the dmesg match what the Probotix expects? (eg, Pin 2 is X-Step, Pin 3 X-Dir, etc?)
[16:56:35] <spackWrk> i'm not really sure how to read this
[16:56:49] <spackWrk> pin 2 should be x step, pin 3 should be x-dir
[16:57:37] <spackWrk> pin 5 should be y-dir, but dmesg shows it as PWMGen #0?
[16:57:57] <spackWrk> or should i be looking at the pin numbers in parentheses?
[16:59:04] <andypugh> Yes, look in the Parentheses under P3-NN
[16:59:06] <spackWrk> i guess the parentheses, since there aren't 33 pins in one port :)
[16:59:20] <andypugh> Indeed, the 5i25 has 2 ports
[17:00:46] <spackWrk> so the probotix expects 6 pins and those are matched in dmesg
[17:00:59] <andypugh> That's going to help a lot :-)
[17:01:02] <spackWrk> :)
[17:01:12] <spackWrk> i think the configuration splits them over 2 ports though
[17:01:28] <spackWrk> some of the pins are in the io connector 2 tab and some are in io connector 3
[17:01:41] <andypugh> That seems strange.
[17:01:55] <spackWrk> only pins 1,4,5,6,7 are shown in connector 2
[17:02:07] <spackWrk> and 0,1,2,3 are shown in connector 3
[17:02:12] <spackWrk> 0 is shown 3 times
[17:02:33] <spackWrk> and then there are a bunch of greyed out areas with no numbers next to them
[17:02:45] <spackWrk> (still in pncconf here)
[17:02:51] <andypugh> The greyed-out ones may change as you select options
[17:03:22] <andypugh> P3 is the one that appears on the back panel, P2 is the on-board header (I think)
[17:04:00] <andypugh> Yes, that seems to be the casse.
[17:04:03] <andypugh> You have
http://www.mesanet.com/pdf/parallel/5i25man.pdf ?
[17:04:18] <andypugh> You need to be sure about 5V tolerance at the very least
[17:05:52] <spackWrk> got it now :)
[17:06:17] <spackWrk> i mean the pdf
[17:06:21] <spackWrk> not working :/
[17:07:16] <andypugh> You need to configure P3, then choose X for stepgen-0, Y for Stepgen-1, Z for stepgen-2 (I think, I don't use Pncconf)
[17:08:02] <spackWrk> it's weird, i can't change the pin types
[17:08:10] <spackWrk> they're all grayed out except the pwm one that isn't even used
[17:08:18] <andypugh> (Not because there is anything wrong with Pncconf, just that when I am configuring things it is generally brand-new hardware that pncconf doesn't know anything about)
[17:08:37] <andypugh> You can't change the pin types because those are set by the .bit file.
[17:08:59] <spackWrk> for example they're all set to step gen A
[17:09:00] <andypugh> What you need to do is choose which axis the pins are associated with.
[17:09:09] <andypugh> Oh.
[17:09:15] <spackWrk> so i can't set one axis to step gen a and one to b
[17:09:32] <spackWrk> so how about i go into this .xml file
[17:09:40] <andypugh> Which LinuxCNC version?
[17:09:54] <andypugh> I doubt that editing the XML is the way to go...
[17:10:34] <spackWrk> not sure which linuxcnc versino
[17:10:38] <spackWrk> *version
[17:10:42] <spackWrk> ubuntu 10.04
[17:10:48] <spackWrk> if that's the deciding factor
[17:11:10] <spackWrk> downloaded within the last 2 weeks
[17:11:29] <spackWrk> let's see...
[17:11:38] <spackWrk> LINUXCNC - 2.5.2
[17:12:25] <spackWrk> i'm not sure what this .xml is even saying...
[17:12:40] <andypugh> I have started a VM, bear with me
[17:12:51] <spackWrk> ok...
[17:12:58] <spackWrk> what is secondaryinstance?
[17:14:18] <andypugh> Right, you need to click the "accept component changes" then select tab 3
[17:14:57] <spackWrk> hmm, maybe i need to do that again
[17:15:01] <andypugh> Where it says "Unused StepGen" click that and select "X-Axis StepGen"
[17:15:59] <spackWrk> i still only have 0,1,2,3 pins in connector 3
[17:16:05] <spackWrk> with 3 instances of pin 0
[17:16:10] <andypugh> That is correct
[17:16:49] <andypugh> Those 0: mean Stepgen-00, Encoder-00 and PWMGen-00
[17:17:16] <andypugh> So, set Unused StgepGen:0 to "X-Axis StepGen"
[17:17:21] <spackWrk> ok
[17:17:37] <skunkworks> http://imagebin.org/index.php?mode=image&id=265521
[17:17:38] <andypugh> Those are not pin numbers, they are Hostmot2 function instance numbers.
[17:17:43] <spackWrk> ah
[17:17:46] <spackWrk> of course
[17:17:47] <spackWrk> :S
[17:18:16] <andypugh> skunkworks: Serious wire clamps. What are those on?
[17:18:24] <spackWrk> andypugh: i just jogged x
[17:18:32] <andypugh> W00t!
[17:18:44] <spackWrk> ^5
[17:18:53] <skunkworks> andypugh: the middle 3 cards? spindle drive
[17:18:54] <spackWrk> thanks man
[17:19:12] <andypugh> skunkworks: I mean, which machine? The Emco?
[17:19:51] <skunkworks> heh - no. dad got a acroloc...
[17:20:29] <spackWrk> got some errors jogging y and z though
[17:20:32] <spackWrk> and now x errors out too
[17:20:43] <andypugh> What error?
[17:21:01] <spackWrk> starting linuxcnc gives the error:
[17:21:02] <skunkworks> andypugh:
http://imagebin.org/index.php?mode=image&id=265488
[17:21:05] <spackWrk> insmod: error inserting '/usr/realtime-2.6.32-122-rtai/modules/rtai_hal.ko': -1 File exists
[17:21:28] <andypugh> This is on the "test axis" tab?
[17:21:36] <andypugh> I would probably not worry about that.
[17:21:39] <spackWrk> no, i closed that
[17:21:43] <spackWrk> started up linuxcnc
[17:21:45] <spackWrk> and got that error
[17:22:06] <andypugh> Hmm, that sounds like Pncconf didn't tidy up after itself properly.
[17:22:10] <spackWrk> i was jogging x ok, then i continued on to jog y
[17:22:24] <spackWrk> jogging y gave some kind of error and then i went back to x and got the same error
[17:22:26] <andypugh> The only way I know to clear that error is to reboot.
[17:22:31] <spackWrk> (in the test axis tab)
[17:22:44] <spackWrk> so much rebooting...
[17:22:55] <spackWrk> for a second i thought i was using linux here
[17:22:56] <spackWrk> ;)
[17:23:10] <spackWrk> i'm gonna get the test axis tab error before i reboot
[17:23:34] <andypugh> Yes, but you are delving into the deep part of the kernel where errors are properly erroneous
[17:23:34] <spackWrk> ok, i can jog x in the test axis tab again
[17:23:51] <spackWrk> proceeded forward to y axis
[17:24:02] <spackWrk> and now it says realtime thread not running when i click the test axis tab
[17:24:50] <spackWrk> if i close that window and click enable, i get "/usr/bin/pncconf" line 9204, in update_tune_axis_params halrun.flush()
[17:24:54] <spackWrk> IOerror:[errno 32] broken pipe
[17:25:35] <spackWrk> reboot?
[17:25:42] <andypugh> I am afraid pncconf is a bit of a mystery to me.
[17:26:06] <andypugh> I would run through the config, ignoring the axis test, get out the other side, and try the config it creates.
[17:26:28] <spackWrk> ok
[17:26:31] <andypugh> Only reboot if you get the "module exists" error.
[17:27:33] <spackWrk> got the module error again
[17:27:34] <spackWrk> rebooting
[17:28:49] <andypugh> If you are interested, the reason that you can't clear that module error is that there are two module who both think they are being used by the other, and insmod / modprobe won't remove an in-use module.
[17:29:33] <spackWrk> oh
[17:29:49] <spackWrk> can't even force it...?
[17:29:50] <spackWrk> well i get an error now
[17:29:51] <spackWrk> but it's at least different
[17:30:01] <spackWrk> insmod: error inserting '/usr/realtime-2.6.32-122-rtai/modules/linuxcnc/lowpass.ko': -1 Invalid parameters
[17:30:21] <spackWrk> and still the: HAL: ERROR: duplicate variable '.in'
[17:30:45] <andypugh> Can you pastebin the INI?
[17:30:50] <spackWrk> http://pastebin.com/VXMNATSP
[17:30:56] <spackWrk> ^^that's the last error dump
[17:31:27] <andypugh> We need to figure out what signal is being called blank.in
[17:32:00] <spackWrk> http://pastebin.com/k2H0DSmn
[17:32:01] <spackWrk> .ini
[17:32:07] <andypugh> What is happening is that the clever part of pncconf that creates signal names is trying to call lots of things "" (ie, blank)
[17:32:29] <andypugh> Can i see the HAL too?
[17:32:57] <spackWrk> http://pastebin.com/FP3JJJtw
[17:33:01] <spackWrk> ^^hal
[17:34:06] <andypugh> The problem is lines 13 and 14.
[17:34:11] <spackWrk> in hal?
[17:34:14] <andypugh> Yes
[17:34:20] <spackWrk> aha
[17:34:21] <spackWrk> i see that
[17:34:25] <andypugh> names= {nothing}
[17:34:48] <andypugh> I guess that comes from the spindle config
[17:34:54] <spackWrk> so i can comment it out?
[17:35:00] <spackWrk> i have no spindle
[17:35:30] <andypugh> I think that is the problem, and how you have confused pncconf.
[17:35:49] <andypugh> just comment-out lines 13 and 14 in the HAL (as a test).
[17:36:01] <andypugh> #loadrt abs names=
[17:36:08] <spackWrk> yeah
[17:36:22] <spackWrk> that got rid of that error
[17:36:36] <spackWrk> now i get
[17:36:38] <spackWrk> fireball072413.hal:67: parameter or pin 'hm2_5i25.0.gpio.032.invert_output' not found
[17:36:44] <spackWrk> when i start linuxcnc
[17:37:11] <spackWrk> http://pastebin.com/2Sx4bKyc
[17:37:14] <spackWrk> ^^error dump
[17:37:18] <andypugh> That's line 67. Not sure what that is about.
[17:37:49] <spackWrk> oops
[17:37:52] <spackWrk> not error dump
[17:38:17] <andypugh> My guess is that pncconf is coping badly with the 5i25.
[17:38:26] <spackWrk> http://pastebin.com/0CFqtHch
[17:39:07] <andypugh> Each axis section in the HAL has an invert_output that does not appear to match the associated pin in the config.
[17:39:23] <andypugh> Did you select any "invert" options in pncconf?
[17:39:40] <spackWrk> yeah
[17:39:42] <spackWrk> all of my axes are inverte
[17:39:42] <spackWrk> d
[17:40:55] <andypugh> OK, so you need to change line 67 to refer to gpio-010 and 94 to gpio-012
[17:41:31] <spackWrk> dunno why it doesn't have any invert stuff in the x axis section
[17:41:57] <andypugh> (This is going to end badly, the only thing to drink in the house is water or whisky. I am out of milk, juice and squash)
[17:42:09] <spackWrk> haha
[17:42:56] <andypugh> And it is stupidly hot (for a brit). 26.7C in my house (that's 80F)
[17:43:18] <spackWrk> i made the changes you suggested
[17:43:24] <andypugh> And?
[17:43:24] <spackWrk> that got me into linuxcnc without an error
[17:43:29] <spackWrk> i can jog x
[17:43:29] <andypugh> Wohoo
[17:43:34] <spackWrk> but i can't jog y or z
[17:43:42] <spackWrk> it says they are moving but they don't move
[17:44:09] <andypugh> Curious.
[17:44:15] <spackWrk> =/
[17:45:04] <spackWrk> wait, you drink squash? :O
[17:45:13] <andypugh> Because, for some reason, Y and Z are on stepgen 05 and 07 ?
[17:45:23] <andypugh> "Squash" refers to fruit jiuce concentrate that you dilute to make into a drink.
[17:45:26] <spackWrk> ah
[17:45:28] <spackWrk> heh
[17:45:31] <syyl> 27°C and whiskey? sounds not like the best idea ;)
[17:45:32] <andypugh> I am speaking English, you see :-)
[17:45:36] <spackWrk> i thought you were going to juice a pumpkin
[17:45:40] <spackWrk> :)
[17:45:53] <kwallace> Hello. I have a 6i25 system that normally uses three stepgens, but I need to test the system using only stepgen.02. It seems I can't invoke stepgen.02 without invoking 00 and 01 which need to be GPIO. Is it best to get a custom firmware that maps the stepgen.02 pins to stepgen.00, or is there another way?
[17:45:58] <spackWrk> should i put them on step gen 1 and 2?
[17:46:16] <andypugh> spackWrk: Yes
[17:46:29] <andypugh> spackWrk: 5 and 7 are on the other connector!
[17:46:38] <spackWrk> :S
[17:47:04] <spackWrk> retrying
[17:47:19] <spackWrk> yay
[17:47:22] <spackWrk> x,y, and z movement
[17:47:30] * spackWrk slaps pncconf
[17:47:36] <spackWrk> with a dead trout or whatever
[17:47:51] <andypugh> kwallace: It is a choice made by the Hostmot2 driver to enable the functuons in order. You could look at how it does it, and change that.
[17:48:48] <andypugh> I suspect that you can convert stepgens 0 and 1 to gpio by toggling enable bits with the "enable_raw" option. That allows you to write directly to the registers.
[17:48:50] <Jp11> Anyone here playing with the beaglebone
[17:49:26] <andypugh> Jp11: I know some guys are, but I don't know if they are here.
[17:50:02] <andypugh> kwallace: Do you have the "regmap" file?
[17:50:42] <andypugh> (Thinking about it, that is too late, as by the time you get to use the raw writes, Hostmot2 has created the pins)
[17:51:22] <andypugh> kwallace: time to cross-connect the wires in a short length of DB25 ribbon?
[17:52:46] <andypugh> spackWrk: You are now at the point that you won't be using Pncconf again, as if you do it will over-write all the changes.
[17:53:59] <spackWrk> yeah
[17:54:04] <spackWrk> i realized that
[17:54:07] <spackWrk> probably for the better
[17:54:32] <spackWrk> it'll jump start my linuxcnc abilities a bit
[17:54:35] <spackWrk> i guess
[17:54:59] <kwallace> andypugh: I discounted that solution because the board under test has a fixed pin-out, but thinking more, the cable isn't being tested so could be changed. I'll have to think more on this.
[17:56:21] <andypugh> Pncconf looked simple when Chris started it, there were about 3 cards and 6 daughter-boards. Then it got complicated, and the pre-flashed boards like 5i25 appeared, then the smart-serial boards that choose_their_own_pin_names! and then the job became overly complex and doomed to suffer many failure modes.
[17:58:52] <kwallace> andypugh: Remapping pins with a custom firmware file is an option isn't it?
[17:59:01] <andypugh> kwallace: It could even be a very short pin-swapper pigtail
[18:00:03] <andypugh> Yes, it is an option. But I don't think that the LinuxCNC Makefile-based Hostmot2 firmware repo can make 6i25 firmwares.
[18:01:36] <andypugh> No 6i25 or 5i25 UCF files here:
http://git.linuxcnc.org/gitweb?p=hostmot2-firmware.git;a=tree;h=d44ca84cd1c74e4c643fa660c71a48191b4d1a64;hb=d44ca84cd1c74e4c643fa660c71a48191b4d1a64
[18:02:24] <andypugh> Obviously PCW can do it.
[18:04:09] <andypugh> <aside> I am not liking the new eBay front page, full of suggestions. I search a fair bit for "30 INT" tooling, and my front page is filled with "30 inch waist, gay int" items. I guess I should be grateful I don't have a 50 INT machine.
[18:06:20] <kwallace> Okay, I can change the cable and HAL config file, and be running in short order. Thank you for your help.
[18:14:11] <kwallace> andypugh: Now you'll get those same kind of ads while browsing other sites too. You might be helping your mother pick out gardening supplies on-line and all these 30 Int ads will pop up in the margins.
[18:15:24] <andypugh> I think she has given up on grandchildren anyway.
[18:25:51] <spackWrk> andypugh: how do i invert an axis in the hal file?
[18:26:15] <andypugh> As in, make it move in the opposite direction?
[18:26:15] <spackWrk> i commented out the gpio.invert_output true lines because they weren't doing anything for me
[18:26:18] <spackWrk> yerah
[18:26:19] <spackWrk> -r
[18:26:48] <andypugh> There are two ways. You can give the stepgen a negative scale, or invert the direction pin.
[18:27:13] <andypugh> (There are hardware ways too)
[18:28:18] <spackWrk> hmm, ok
[18:28:18] <jdh> swap the a,a/ wires to the motor (after powering off the drive)
[18:28:27] <spackWrk> all the wires are hard lines
[18:28:39] <andypugh> HAL ought to be easiest.
[18:28:46] <spackWrk> how do i invert the direction pins?
[18:29:08] <jdh> make the scale - in the ini
[18:29:12] <andypugh> You had the almost-correct HAL lines, but they were pointing at the wrong GPIO pins)
[18:29:51] <andypugh> I prefer a negative stepgen scale (as suggested by jdh)
[18:29:59] <spackWrk> oh
[18:30:22] <spackWrk> how do i do this? :)
[18:30:24] <jdh> SCALE = -16000.0
[18:30:28] <spackWrk> ok
[18:30:32] <spackWrk> .ini, not hal then...
[18:30:36] <spackWrk> got it
[18:31:13] <spackWrk> what units are the max velocity and max acceleration in
[18:31:21] <spackWrk> and why are they in both hal and ini files?
[18:31:29] <andypugh> Yeah. In my hand-built configs it is all in the HAL, but auto-built configs put the numbers in the INI
[18:31:31] <spackWrk> i guess everything is inches
[18:31:35] <spackWrk> now that i'm looking at .ini
[18:31:54] <andypugh> If your machine was created in metric then it should all be in mm.
[18:32:01] <spackWrk> hmm, hal and ini have different numbers for max velocity and max acceleration
[18:32:25] <jdh> I have about 20 tape measures and none of them even have metric
[18:32:30] <andypugh> The axis max accel and the stepgen max accel _need_ to be different.
[18:32:50] <spackWrk> what units is stepgen max accel in?
[18:33:04] <andypugh> machine units per second per second
[18:33:10] <spackWrk> ok
[18:33:23] <spackWrk> so a machine unit is a step/dir pulse?
[18:33:51] <andypugh> No, it is either an inch or a mm.
[18:34:03] <spackWrk> so why should it be different?
[18:34:16] <andypugh> Why should what be different from what?
[18:34:31] <spackWrk> why should axis max accel be different from stepgen max accel
[18:34:54] <andypugh> The stepgen needs a bit of overhead to keep up to the axis.
[18:35:12] <spackWrk> oh ok
[18:35:25] <spackWrk> are the velocities in in/s or in/min?
[18:35:32] <spackWrk> pncconf was using in/min
[18:35:45] <spackWrk> but stepconf used in/s
[18:35:47] <andypugh> In the INI / HAL it is all /sec
[18:35:53] <spackWrk> ok, cool
[18:36:45] <spackWrk> just installed a new rubber shaft coupling and switched to 16th steps
[18:37:02] <spackWrk> time to experience smoothness
[18:38:05] <spackWrk> hmm, no movement
[18:38:26] <andypugh> That's sad
[18:38:44] <andypugh> Though no doubt extremely smooth?
[18:38:48] <spackWrk> haha
[18:38:54] <spackWrk> yep
[18:39:14] <spackWrk> oh i see
[18:39:20] <spackWrk> it was moving
[18:39:29] <andypugh> But 1/16 the speed?
[18:39:33] <spackWrk> less than that
[18:39:51] <spackWrk> i expected slower speed because of that
[18:39:57] <spackWrk> but also forgot to bump up the jog speed
[18:40:04] <spackWrk> heh
[18:47:19] <spackWrk> ahhhh, so smooth
[18:47:37] <andypugh> Not compared to linear motors :-)
[18:48:48] <spackWrk> i have no experience with those
[18:49:00] <spackWrk> but i might need one
[18:49:01] <andypugh> Nor do I, actually :-)
[18:49:02] <spackWrk> at some point
[18:49:06] <spackWrk> if they're the epitome of smoothness :)
[18:50:04] <spackWrk> my next step, if needed, was going to be the servo route that we were talking about
[18:50:18] <spackWrk> but at this point, i think the screw drive system is more important
[18:50:22] <andypugh> Brushless motors can be pretty good:
http://youtu.be/47y6RgAK--8
[18:50:23] <spackWrk> giving more noise than the motor
[18:52:17] <spackWrk> well, i think i'll head home
[18:52:28] <spackWrk> gonna leave this running overnight
[18:52:43] <spackWrk> because, after 5 minutes, i totally trust it
[18:52:43] <spackWrk> :)
[18:53:06] <spackWrk> actually, i'll just pause it
[18:53:17] <spackWrk> i can let it run for 30 mins in the am while i do some other stuff
[18:54:11] <andypugh> If it's working, I would tend to trust it to work. From it's point of view it does 1000 things every second, it is already into the millions of iterations.
[18:56:11] <spackWrk> yeah, i do trust it, but i was having the occasional skipped steps while using software stepping
[18:56:23] <spackWrk> was leading to things running out of bounds
[18:56:47] <spackWrk> i doubt that will happen now, but at least i can give it another few minutes before risking it :)
[18:56:59] <WalterN> CaptHindsight: let me know when you call those guys :3
[19:11:07] <CaptHindsight> WalterN: I'll probably have a clear answer in a couple of days
[19:11:50] <CaptHindsight> if you ask what the input should be the answer will be: What are you trying to do? vs just the answer.
[19:20:11] <WalterN> no rush I guess
[19:21:40] <WalterN> CaptHindsight: is one of those mesa controller cards all I need though? nothing else is needed in between?
[19:23:29] <andypugh> WalterN: Care to elaborate on the question?
[19:27:45] <WalterN> andypugh: using linuxCNC to control these
http://pryntech.com/?product=20kpps-laser-galvanometer-kit
[19:29:49] <andypugh> There seems to be no indication at all what sort of motor they are
[19:30:11] <WalterN> its a coil evidently
[19:30:27] <WalterN> or something
[19:30:29] <andypugh> Ah, OK. So what is the input signal?
[19:30:42] <WalterN> yeah, thats the problem... no documentation
[19:31:17] <WalterN> CaptHindsight was going to call the manufacturer about it
[19:32:35] <andypugh> And what is included? The picture in the $150 kit (cheap enough to buy and worry about later) includes a separate item listed at $550 (not cheap enough to buy on a whim)
[19:32:52] <WalterN> if I called I'd ask all the wrong questions
[19:34:02] <andypugh> At the price I am tempted to buy it and see what arrives :-)
[19:34:58] <WalterN> oh, that $500 kit or whatever includes a stepper and a bunch of other garbage
[19:34:58] <WalterN> or something
[19:35:16] <NickParker> has anybody here done work with 4 phase stepper drivers? They take in PHA1 PHAC PHA2 PHB1 PHBC PHB2 signal to control them. plus some other stuff about limits, not sure what that's about.
[19:35:24] <WalterN> the galvo kit I got (from there) includes power supply, and a driver board for each galvo
[19:35:54] <WalterN> its just nothing is documented
[19:36:16] <NickParker> I've got an old PC so linuxcnc seems like my cheapest option to adapt this mill to something more modern than bandit.
[19:36:54] <NickParker> I think I understand how the input signals work, so worst case i'm using an attiny to interpret step/dir for each axis.
[19:39:25] <andypugh> That sounds like 6-wire unipolar steppers. You can run those bipolar 2-phase if you simply ignore the C wires
[19:39:39] <NickParker> yeah they are 6 wire unipolar
[19:39:58] <NickParker> they're also 8 amps though, so i would be spending an extra $300 or so to buy modern drivers, and the existing ones drive current just fine.
[19:40:02] <andypugh> You can run them on any bipolar driver. That should expand your horizons.
[19:41:23] <spackWrk> NickParker:
http://www.linengineering.com/resources/wiring_connections.aspx
[19:41:33] <spackWrk> if that helps at all...
[19:42:41] <spackWrk> looks like the 6 lead unipolar diagram matches your case
[19:42:46] <andypugh> Ah, well, in that case,
http://www.linuxcnc.org/docs/html/man/man9/stepgen.9.html step types 5 to 8
[19:42:50] <NickParker> well at the very least I think I know which wires are which now, which is great. I was checking with a ohmmeter like I usually do with smaller steppers, but these have like 1/4 ohm of resistance, so that was kind of tricky
[19:42:51] <spackWrk> if you have any of those color codes, you should be gtg
[19:43:04] <NickParker> i believe i have code 1
[19:43:46] <andypugh> The way to diagnose phase ends and polarity with steppers is to connect wires together and see if you can still rotate the shaft :-)
[19:44:05] <spackWrk> looks like C on that chart is your A2, and D is your B2
[19:44:21] <NickParker> yep
[19:44:33] <spackWrk> wee
[19:44:47] <NickParker> can you do microstepping with unipolar motors?
[19:44:59] <NickParker> seems like you could..
[19:45:09] <NickParker> eh it doesn't matter i need an oscilloscope..
[19:45:24] <andypugh> I would never trust wire colurs after getting some 8-wire motors that had some phases reversed relative to the documentation in the _box_.
[19:45:50] <NickParker> I *think* the stock drivers in this mill (created 1979 btw) are full step drive
[19:45:57] <andypugh> Types 9 and 10 are unipolar half-step
[19:46:17] <NickParker> so half-step would be great for me, because the mechanics probably aren't good enough to do much better
[19:46:20] <NickParker> stiction and all that
[19:46:54] <andypugh> Basically, if LinuxCNC can't drive your hardware, then you just have to make some developer take offence at that fact :-)
[19:47:35] <andypugh> Sorry, reading the words, Type 9 is the unipolar half-step.
[19:48:55] <WalterN> andypugh: what would be a good servo (would I want a stepper instead?) for the small Z-movements for a rapid prototyping machine?
[19:50:31] <andypugh> There is a much better documentation page for stepgen somewhere, with wave diagrams etc, but I don't seem to be able to find it.
[19:51:37] <andypugh> Ah, here it is:
http://www.linuxcnc.org/docs/html/hal/rtcomps.html#_stepgen_a_id_sec_stepgen_a
[19:53:07] <andypugh> WalterN: For Z I can see that a stepper might be easier. But if the Z increments are not integer multiples of steps then you might end up with layer-height aliasing. (a phrase I just invented, it will only make sense if you are aware of aliasing as a concept)
[19:53:10] <NickParker> ok i'll read through this
[19:54:10] <WalterN> andypugh: yeah, that would take into account screw pitch and rotation of a single step
[19:54:32] <WalterN> andypugh: such a thing wouldent be an issue with a servo?
[19:54:53] <andypugh> It would depend on the encoder resolution.
[19:55:39] <andypugh> It definitely wouldn't be a problem with a resolver-feedback servo, but I doubt they are within budget.
[19:56:00] <WalterN> ok well... how much is a good servo setup going to cost? lol
[19:56:18] <WalterN> and the same question for a stepper
[19:56:26] <andypugh> I have no idea, all my stuff is bought from eBay when the prices look good :-)
[19:58:25] <andypugh> I recently bought this:
http://www.ebay.co.uk/itm/290941411815 and as yet have no idea what I even bought :-)
[19:58:42] <andypugh> (A friend colllected it)
[19:59:06] <WalterN> all caps description
[19:59:08] <WalterN> ugh
[19:59:18] <andypugh> But have been watching this set:
http://www.ebay.co.uk/itm/261144227546?ssPageName=STRK:MEWAX:IT&_trksid=p3984.m1438.l2649
[20:00:12] <Tom_itx> so are alot of others
[20:00:26] <Tom_itx> probably not a good sign for you
[20:00:28] <WalterN> how much do good encoders cost?
[20:00:54] <andypugh> Too much, best to get them already on a motor
[20:01:27] <WalterN> so what... $500?
[20:01:32] <WalterN> $50?
[20:01:45] <WalterN> $5?
[20:02:46] <andypugh> About $50 for an encoder and $400 for a resolver. But I got 3 x 750W servos with resolvers for £50 the set.
[20:04:00] <WalterN> what is a resolver?
[20:04:19] <andypugh> A friend (actually the same one who currently has the other stuff) recently bought a pair of these from the same vendor for $200 (the pair).
http://www.ebay.co.uk/itm/MAVILOR-AC-Servo-Motor-BT0559-99-099-A2-MSP-0307-Drive-/120624827694
[20:04:55] <WalterN> oh... I dont think the servo/stepper for what I want has to be beefy at all
[20:05:04] <spackWrk> hmm, with drivers?
[20:05:12] <andypugh> Those are surprisingly small.
[20:05:31] <andypugh> If you can buy a motor driver that's always a big saving in hassle.
[20:05:45] <spack> i want that
[20:05:55] <spack> i also want this
[20:06:01] <spack> http://chicago.craigslist.org/nwc/tls/3956168077.html
[20:06:32] <andypugh> I am not sure you do, that's quite expensive for that vintage
[20:07:22] <andypugh> Whereas this:
http://london.craigslist.co.uk/bfs/3908763411.html
[20:08:22] <NickParker> I think I'm going to end up using attiny's on each axis. What I don't see in all this stepgen stuff is any reference to driving the common sides of coils.
[20:08:53] <NickParker> the stock electronics in this mill actually put the common high and the phase low when they want to end a pulse, to get sharper wave forms
[20:09:30] <RyanS> I can see with this is going. tormach machines are going to be a ripoff in Australia
[20:09:33] <andypugh> That's a job for the drive, but on a unipolar setup the commons either go to ground or power. (and which is purely a function of if the power drivers in the drive are NPN or PNP)
[20:09:51] <spack> i don't see how anything could be cheaper in the UK, but that's a low price for a machine like that around here
[20:10:15] <spack> to see any running 4000lb mill for that price, even a manual one, isn't that common
[20:10:23] <andypugh> spack: Quite, everything costs are last dollar-equivalent in the UK.
[20:10:30] <NickParker> Oh, I suppose I could build a circuit to do that for me, because attiny probably has a pretty low max frequency
[20:10:55] <Tom_itx> 20Mhz
[20:11:04] <NickParker> really? nevermind then..
[20:11:29] <Tom_itx> pretty sure some are
[20:11:32] <andypugh> NickParker: What input signals do the existing drives take? I am pretty sure that LinuxCNC can provide that signal. The rest of the high-current stuff is the job of the drive.
[20:12:10] <NickParker> I think, and i need to get the existing logic center working while i have an oscilloscope to prove this, but i think all the driver boards do is take 6 logic voltage inputs, and amplify them to 6 motor voltage outputs
[20:12:24] <NickParker> so if PHA1 input is high, motor voltage goes to phase A1 on the motor
[20:12:33] <NickParker> and the same applies to PHAC and PHBC
[20:12:47] <andypugh> Unipolar only takes 4 inputs.
[20:13:04] <andypugh> Unless you have crazy steppers (of course)
[20:13:22] <NickParker> i have crazy drivers. the steppers are very normal 6 wire 4 phase motors
[20:13:27] <NickParker> http://pastebin.com/LegASPG2
[20:13:40] <NickParker> I found somebody online who has retrofitted a similar mill
[20:14:01] <NickParker> http://www.wallacecompany.com/machine_shop/Shizuoka/Shizuoka_Step_Drv-1a.png
[20:14:09] <NickParker> this is the waveform he used
[20:14:15] <andypugh> kwallace logged off a while ago.
[20:14:32] <NickParker> oh he's in here on occasion? that'll be nice.
[20:14:54] <andypugh> He is on here rather a lot of the time.
[20:15:22] <andypugh> Did you see type 15 here:
http://www.linuxcnc.org/docs/html/man/man9/stepgen.9.html
[20:15:57] <NickParker> i did not. That looks more or less like what I want.
[20:16:10] <andypugh> If any of your phases are redundant (ie, always match each other) then that will work.
[20:17:08] <andypugh> If not, then it may not be all that hard to change the limits. (give a good reason, and I see no reason it shouldn't be 256 steps and 16 phases)
[20:18:23] <RyanS> I would gather 1600rpm is useless for CNC, but what about 3000?
[20:18:36] <WalterN> meh
[20:18:51] <WalterN> where should I even look for these things anyway?
[20:18:59] <WalterN> (excluding ebay)
[20:19:20] <andypugh> I have just noticed that user_step_type appears to have basically no documentation :-(
[20:20:02] <NickParker> so type 15 will essentially have 10 (ordered, i guess) output states and step up or down along the list right?
[20:20:08] <NickParker> and i define those 10 states
[20:20:21] <andypugh> RyanS: Not understanding? Given that most stepper systems struggle above 300rpm 1600 is _lots_
[20:20:36] <NickParker> andypugh: RyanS may be talking spindle speeds?
[20:21:33] <andypugh> Then 1600 is plenty for a lathe, and 600 more than my milling machine. Not that I am pleased by that last fact. But it will spin a big cutter.
[20:21:34] <RyanS> Yes sorry I was talking about the a spindle
[20:22:02] <andypugh> (My milling machine has gears from 45 rpm to 1000 rom)
[20:22:18] <NickParker> still talking spindle?
[20:22:22] <NickParker> wtf is 45 rpm good for there?
[20:22:31] <NickParker> wax or something fancy like that?
[20:23:29] <andypugh> NickParker: It has uses, like 1 min into this video:
http://www.youtube.com/watch?v=i4fTythQj5s
[20:23:43] <RyanS> This is for a small machine, desktop. I think 1600 is too slow even for manual?
[20:24:31] <andypugh> (single-point threading a 60mm external thread)
[20:24:49] <WalterN> http://oceancontrols.com.au/CNC-113.html maybe something like that?
[20:25:58] <andypugh> That's pretty spendy for a 70W brushed kit. Have you seen:
http://www.dmm-tech.com
[20:26:03] <kwallace> I usually have to dig up old documentation to find the step types. maybe page 6 here:
http://www.linuxcnc.org/docs/2.2/EMC2_Integrator_Manual.pdf
[20:26:17] <kwallace> Oops page 66
[20:26:41] <andypugh> Ah, kwallace is still here, who was looking for him?
[20:27:24] <andypugh> NickParker please be introduced to kwallace
[20:27:28] <WalterN> andypugh: I dont see any DC servos
[20:27:45] <andypugh> WalterN: And why would you want DC servos?
[20:28:06] <andypugh> Ghastly things, with Victorian Brushes and Edwardian controls.
[20:28:38] <WalterN> because... AC is beefier than what I want?
[20:28:53] <andypugh> Category error
[20:29:45] <andypugh> Those tiny model aircraft use brushless motors, and AC servos are identical to brushless motors.
[20:29:56] <kwallace> My HNC brushed motors work just fine. I can still get any parts I may need.
[20:30:09] <NickParker> hey kwallace am I understanding the function of these driver boards correctly? They juft put motor voltage to a given lead based on whether the corresponding logic pin is high or low?
[20:30:11] <andypugh> I have some lovely little 100W brushless servos
[20:31:08] <andypugh> The smallest DMM-tech motor is 220W
[20:31:50] <andypugh> (I will point out that the matching drive is probably dangerous in voltage control mode and I can't recommend it)
[20:32:01] <kwallace> Well sort of. There is a common supply wire and four wires that get grounded through PWM for current control.
[20:32:30] <andypugh> (The input pin floats high ("Ramming Speed!") when open-circuit
[20:33:56] <kwallace> NickParker: The commons are AA and BB on the schematic.
[20:36:18] <kwallace> NickParker: Each phase is A1, A2, B1, B2, these four inputs pretty directly control the respective outputs.
[20:36:50] <kwallace> Oops again "the four inputs..."
[20:37:16] <andypugh> WalterN: Postage is mad for me but:
http://www.aliexpress.com/item/3000rpm-high-torque-ac-servo-motor/792459749.html
[20:37:19] <kwallace> http://www.wallacecompany.com/machine_shop/Shizuoka/Driver%20Schematic1.jpg
[20:38:00] <andypugh> WalterN: This looks good for you:
http://www.aliexpress.com/item/200W-0-637N-M-AC-SERVO-MOTOR-AND-DRIVER/1046667334.html
[20:40:05] <NickParker> kwallace: so in your waveform A inhibitor and B inhibitor are AA and BB right? And you're calling them inhibitors because you're basically pulsing them to get negative voltage over the phases to turn them off quicker?
[20:41:10] <andypugh> An possibility of doing that logic in HAL?
[20:42:45] <kwallace> NickParker: No, the inhibit affects the inputs only and is used to switch from half to fill stepping. I only us half stepping so the inhibit is tied low.
[20:43:30] <kwallace> The inhibit pulse time changes with step speed.
[20:44:22] <andypugh> Goodnight chaps.
[20:44:47] <kwallace> Sleep tight.
[20:44:52] <NickParker> I don't follow..
[20:45:01] <WalterN> what is the difference between price and total price?
[20:45:56] <NickParker> so when you put inhib high it's the equivalent of putting 1 and 2 low?
[20:46:08] <kwallace> Yes.
[20:46:28] <NickParker> ohhhh I see how that works in the circuit now
[20:47:00] <NickParker> so the inhib signal never gets across the optos. the inhib signal is how the opto's get grounded, so if it's high voltage across the opto inputs becomes 0
[20:47:25] <kwallace> Yep.
[20:49:02] <NickParker> what does that do for us though? I mean, why not just keep it grounded forever and make the pulses to the A1 A2 etc shorter?
[20:51:19] <NickParker> oh, that half/full stepping you mentioned earlier makes sense now too
[20:52:03] <kwallace> I didn't bother to figure out the timing of the inhibit. I think it is a fixed Off pulse width but triggered at the step rate. At slow rates the pulse has more On time and half steps. As the rate goes up the pulse begins to disappear.
[20:52:45] <NickParker> wait is this waveform what the bandit outputs? I thought it was what you were outputting with linuxcnc
[20:56:10] <kwallace> The diagram was the Bandit output that I made in order to figure out what I needed. For LInuxCNC, I didn't bother to implement it. I still don't really understand it. I just tied inhibit to ground.
[20:56:26] <NickParker> Ah I see.
[20:58:23] <NickParker> Oh one other thing, did your mill have 3d moves before you switched to linuxcnc?
[20:58:37] <NickParker> mine doesn't. The guy who gave it to me said it can only move Z while XY is still.
[20:58:50] <NickParker> So I'm a little curious where the bottleneck is that causes that.
[20:59:29] <NickParker> actually nevermind. steppers take the most power when they're not moving. I was worried that i wouldn't have the power to drive 3 axes at once
[20:59:43] <NickParker> So that's probably just a quirk of the bandit.
[21:00:13] <NickParker> thanks kwallace, you've been a huge help.
[21:00:21] <NickParker> my mill is moving friday, can't wait to start playing
[21:01:13] <kwallace> No, as far as I could tell mine did not do 3D moves.
[21:02:43] <kwallace> That and trying to get g-code files into the RS232 port made up my mind to do the conversion.
[21:04:19] <kwallace> NickParker: How is your machine getting moved? I moved mine myself and it was a day long nightmare.
[21:10:59] <NickParker> I hired Dunkel logistics to do it for $725
[21:11:13] <NickParker> parents are paying it to dissuade me from trying to move it ourselves to save money
[21:11:59] <NickParker> what was so awful about your move? We rolled mine to the door of the shop the other day and it was surprisingly easy
[21:12:41] <NickParker> the only part i was really worried about was rolling it off the truck.
[21:13:01] <NickParker> well, and my paver driveway, but i think big rollers would have made that ok
[21:13:27] <cradek> gotta rent a forklift to get it off the truck.
[21:15:03] <NickParker> the guy i was talking to claimed you could roll it safely, can't remember his technique. in any case, that sounded super sketchy and once i added in the forklift rental pros were only ~$200 extra, so why risk it yknow
[21:31:46] <PCW> Wee! Fanuc absolute encoder HM2 firmware module works! (+- a 1/2 a day chasing a stupid bug)
[21:33:28] <CaptHindsight> NickParker: how heavy is the mill?
[21:36:57] <jp_> PCW: congrats
[21:37:38] <jp_> what if any absolutes are supported now
[21:38:00] <PCW> Someone brought a motor/encoder over otherwise I never would have gotten around to it
[21:38:13] <jp_> lol
[21:38:51] <jp_> guess just fanuc then eh
[21:39:18] <PCW> The HM2 resolver interface is absolute but pretends to be increments and has a simulated index
[21:39:33] <PCW> incremental
[21:39:50] <jp_> oh and biss too right?
[21:40:15] <PCW> there are also SSI and BISS interfaces in the firmware but not tested unil i get an encoder
[21:40:54] <CaptHindsight> PCW: are the PWM outputs from the xi25 boards all 5V?
[21:41:05] <PCW> 3.3V
[21:41:13] <CaptHindsight> thanks
[21:41:34] <PCW> 5V from the daughtecards
[21:42:19] <CaptHindsight> looking at the possible boards to drive galvo driver boards with an analog PID
[21:43:23] <CaptHindsight> the input specs are missing from the docs, it's either 5V PWM or +- 5V analog
[21:44:54] <CaptHindsight> www.phenixtechnology.com/uploadfile/file/20120816114123_54373.pdf
[21:46:14] <skunkworks> Machine Specifications:
[21:46:16] <skunkworks> Year: 1989
[21:46:17] <skunkworks>
[21:46:19] <skunkworks> Specifications
[21:46:20] <skunkworks> Control: Fanuc
[21:46:22] <skunkworks> X Axis Travel: 40"
[21:46:23] <skunkworks> Y Axis Travel: 24"
[21:46:25] <skunkworks> Z Axis Travel: 30"
[21:46:26] <skunkworks> Quill Travel: 9.6"
[21:46:28] <skunkworks> Table Size: 26" x 45"
[21:46:29] <skunkworks> Max Table Load: 2,500 LBS
[21:46:31] <skunkworks> Spindle Motor: 15 HP
[21:46:33] <skunkworks> Tool Capacity: 15 ATC
[21:46:34] <skunkworks> 4 Axis
[21:47:29] <spack> max table load 2500lbs...
[21:47:35] <spack> machine must weight... 10,000lbs?
[21:47:44] <skunkworks> around that
[21:47:49] <spack> 30" z trvel...
[21:47:59] <spack> nice beast
[21:48:42] <spack> i wish i had a machine like that
[21:50:23] <CaptHindsight> skunkworks: the specs for that Acroloc? ^^
[21:51:16] <WalterN> CaptHindsight: also, what kind of servo or stepper would I want for the Z-movement in the rapid prototyping machine?
[21:51:48] <PCW> CaptHindsight: PWM frequency?
[21:51:59] <CaptHindsight> WalterN: either, but lets see what input that driver board actually uses first
[21:52:43] <WalterN> hmm
[21:52:46] <WalterN> alright
[21:52:53] <CaptHindsight> PCW: not sure yet, but the analog PID only works to 20-30k positions per second
[21:53:18] <CaptHindsight> WalterN: stepper is easy and cheap for that size
[21:54:36] <WalterN> yeah
[21:54:53] <PCW> Why Galvos instead of a scanning mirror? speed because of maximum laser duty cycle?
[21:55:00] <CaptHindsight> WalterN: maybe it will just need a 6i25
[21:55:08] <skunkworks> CaptHindsight: yes
[21:55:38] <skunkworks> pcw: did you see -
http://imagebin.org/index.php?mode=image&id=265484
[21:55:43] <skunkworks> 7i80?
[21:55:47] <CaptHindsight> PCW: $90 for 2 galvo mirrors + driver card setup
[21:56:04] <PCW> Spindle motor?
[21:56:11] <skunkworks> axis
[21:56:31] <PCW> (I was thinking 0 for a old laser printer)
[21:57:00] <CaptHindsight> PCW:
http://www.phenixtechnology.com/en/product-ari--6.html
[21:57:05] <PCW> Thats a pretty hefty axis drive
[21:58:17] <WalterN> hmm
[21:58:25] <WalterN> I need to get a different laser too... lol
[21:58:46] <CaptHindsight> http://pryntech.com/?product=opensl-board only has PWM out, unless I'm missing something
[22:00:00] <CaptHindsight> http://pryntech.com/?product=20kpps-laser-galvanometer-kit is the
http://www.phenixtechnology.com/en/product-ari--6.html
[22:00:25] <skunkworks> pcw: we are hoping that the specs are 150v 14.5a... not 60...
[22:00:31] <CaptHindsight> WalterN: nah, you can cure and sinter
[22:00:43] <skunkworks> 2.2kw...
[22:01:02] <WalterN> CaptHindsight: with 808nm?
[22:01:16] <WalterN> cause thats the wavelength of my laser
[22:01:38] <WalterN> for the plastic goop, I thought it was 400nm and under
[22:01:40] <PCW> ~3 HP seem more reasonable than 12
[22:01:43] <CaptHindsight> WalterN: sinter with 808nm and cure with a $25 Blu ray 405nm
[22:01:54] <CaptHindsight> two lasers and 1 galvo
[22:01:58] <skunkworks> pcw: my thoughts tooo
[22:02:01] <PCW> so its likely 14.5A
[22:02:04] <WalterN> oh, and swap out lasers?
[22:02:07] <WalterN> I can dig that
[22:02:14] <WalterN> XD
[22:02:22] <WalterN> I still need a new laser though
[22:02:29] <CaptHindsight> have the lasers on a slider
[22:02:33] <spack> what are you sintering?
[22:02:58] <PCW> eyeballs if not careful
[22:03:08] <spack> heh
[22:03:20] <spack> sintered cornea
[22:04:01] <CaptHindsight> i think he already bought some laser safety goggles
[22:04:18] <WalterN> for 808nm, yeah
[22:04:26] <WalterN> they were bloody expensive
[22:04:39] <spack> 'bout $200?
[22:05:42] <WalterN> something like that
[22:05:48] <WalterN> $180... something
[22:06:09] <CaptHindsight> the 808nm can also cure the resin but it won't be as controlled as with the 405nm
[22:06:15] <WalterN> CaptHindsight: oh, does the plastic goop change volume when it cures?
[22:07:11] <CaptHindsight> WalterN: it ranges from no to shrinks by 1-20%
[22:07:28] <CaptHindsight> depends on the formulation
[22:07:53] <CaptHindsight> they can also be made to expand
[22:07:54] <WalterN> hmm
[22:08:04] <WalterN> the cheap stuff
[22:08:08] <WalterN> lol
[22:08:15] <spack> is there such a thing?
[22:08:18] <spack> lol
[22:08:50] <CaptHindsight> the $45 per liter polyester might shrink by 1-2%
[22:09:26] <CaptHindsight> depending on the viscosity and how it's cut
[22:09:44] <WalterN> hmm
[22:09:45] <CaptHindsight> cut/reduced
[22:10:54] <skunkworks> is that the spindle belt?
[22:11:12] <skunkworks> is the motor on the big or small side?
[22:11:29] <skunkworks> still in process
[22:11:35] <CaptHindsight> if you order a tanker the price drops <$20
[22:11:36] <skunkworks> don't hear anything
[22:11:55] <WalterN> because if I have the laser above the vat of liquid shining down, with the base slowly submerging... would have to change the focal length slightly... unless I have some other thing that keeps the liquid level the same
[22:12:16] <skunkworks> pcw: I meant 8i20... ;)
[22:13:25] <CaptHindsight> WalterN: it's insignificant, you won't have to change the focus
[22:13:39] <WalterN> if its 20%... lol
[22:14:02] <WalterN> if 2% is typical, then yeah
[22:15:33] <WalterN> CaptHindsight: so the galvos are PWM?
[22:15:49] <CaptHindsight> WalterN: don't know yet for sure
[22:17:15] <WalterN> alright
[22:18:58] <WalterN> actually
[22:19:13] <CaptHindsight> https://raw.github.com/beardface/LAMPS/master/pcb_design/OpenSL_1.0/board.jpg
[22:19:20] <WalterN> CaptHindsight: I could use the laser in my bluray burner... it recently had some stuff inside break
[22:20:20] <CaptHindsight> the ARM soc only outputs PWM, they might convert it to analog with the OPA4171
[22:20:36] <CaptHindsight> free 405nm laser!!
[22:21:07] <WalterN> "free" heh... that was an expensive burner
[22:21:32] <WalterN> I bought it over a year ago
[22:27:19] <CaptHindsight> http://www.aliexpress.com/item/Ag-20k-laser-mirror-high-speed-mirror-laser-galvanometer-pattern/894190646.html
[23:35:13] <RyanS> A lot of the people on practical machinist seen unrealistic. Some new person asks about a desktop CNC mill & they immediately get shot down with comments like "get a real VMC" as though hobbyists have $50k+ to spend
[23:35:46] <CaptHindsight> http://search.newport.com/?q=*&x2=sku&q2=RV240PP I was a bit surprised to see a gates belt between the motor and the worm gear
[23:36:14] <CaptHindsight> RyanS: or room for a VMC
[23:43:10] <RyanS> I think the members of PM like to talk about the machine their workplace owns as 'their' machine .. No problem, mortgage your house, buy one
[23:43:47] <WalterN> start small
[23:45:13] <sbbrtn> what is a VMC?
[23:45:32] <WalterN> vertical milling center
[23:45:50] <WalterN> (pretty sure anyway)
[23:46:00] <CaptHindsight> Vertical Machining Center
[23:46:05] <sbbrtn> ahh ok
[23:46:07] <sbbrtn> thx
[23:46:23] <sbbrtn> What kind of CNCs you guys running?
[23:46:30] <RyanS> Even $5k is a ton for a hobbyist
[23:47:28] <CaptHindsight> sbbrtn: all different types, from the very small bench tops to some pretty big machines
[23:48:03] <sbbrtn> here is mine
http://www.youtube.com/watch?v=MbZ4h3zCk2U
[23:48:16] <WalterN> sbbrtn: work? home? or personal machines?
[23:48:17] <Tecan> (MbZ4h3zCk2U) "Ryno CNC Machine Test #3" by "Ryan Hamilton" is "Education" - Length: 0:03:56
[23:48:41] <sbbrtn> just for fun
[23:48:45] <sbbrtn> personal
[23:48:52] <sbbrtn> love linuxcnc
[23:49:06] <CaptHindsight> lots of Bridgeport conversions
[23:49:33] <sbbrtn> mine is the tiag x3 I think?
[23:49:40] <sbbrtn> I want a bridgeport soo bad
[23:50:12] <sbbrtn> I wish my machine had more neck rigidity
[23:51:26] <CaptHindsight> RyanS: there are deals to be found for some good sized VMC's
[23:52:13] <WalterN> I dont have any machines that I own :-/
[23:52:14] <WalterN> I have absolutely no place to put anything
[23:52:33] <Tecan> http://en.wikipedia.org/wiki/Coffman_engine_starter
[23:53:51] <CaptHindsight> RyanS:
http://imagebin.org/index.php?mode=image&id=265488 this was on Craigslist for cheap
[23:54:12] <RyanS> not in Australia
[23:54:12] <sbbrtn> Where in LinuxCNC is the "Main" loop?
[23:54:19] <sbbrtn> in the code that is
[23:55:10] <CaptHindsight> RyanS: I found this for a bargain
http://imagebin.org/265317
[23:56:49] <RyanS> People also forget your buying an old machine for a few $k which is still a lot of money if the whole thing breaks down imagine the cost of parts and tooling
[23:58:27] <CaptHindsight> http://chicago.craigslist.org/nwc/tls/3956168077.html
[23:59:13] <CaptHindsight> I notice more deals in late winter - early spring
