#linuxcnc | Logs for 2014-09-10

[00:11:00] -!- syyl_ has quit [Ping timeout: 246 seconds]
[00:13:08] <CaptHindsight> 35.5 × 25.0 × 3.9 mm (1.4 × 1.0 × 0.15 inches) max Edison board https://communities.intel.com/servlet/JiveServlet/previewBody/23139-102-2-27268/edison_PB_331179-001.pdf so small you might forget where you left it
[00:14:21] <CaptHindsight> why is the IO board (with not much IO) so much larger?
[00:14:31] <CaptHindsight> http://www.intel.com/content/www/us/en/do-it-yourself/edison.html
[00:16:36] <CaptHindsight> 12 GPIO and SPI, I wonder how fast they toggle read/write/read
[00:21:48] -!- gambakufu has quit [Read error: Connection reset by peer]
[00:30:32] -!- maZer`- [maZer`-!~info@ip-2-205-81-194.web.vodafone.de] has joined #linuxcnc
[00:30:47] -!- maZer`- has quit [Client Quit]
[00:46:41] -!- Tecan [Tecan!~fasdf@ip-216-234-182-80.wireless.tera-byte.com] has joined #linuxcnc
[00:46:41] -!- Tecan has quit [Changing host]
[00:46:41] -!- Tecan [Tecan!~fasdf@unaffiliated/unit41] has joined #linuxcnc
[00:46:56] -!- WyrM has quit [Ping timeout: 272 seconds]
[00:47:02] -!- PetefromTn_ has quit [Quit: I'm Outta here!!]
[00:47:06] -!- WyrM [WyrM!wyrm@bnc.wyrm.net.br] has joined #linuxcnc
[00:47:06] -!- WyrM has quit [Changing host]
[00:47:06] -!- WyrM [WyrM!wyrm@unaffiliated/wyrm] has joined #linuxcnc
[00:47:10] <PCW> if its anything like the Galileo, not very fast because its I2C
[00:48:12] -!- tom_o_t has quit [Ping timeout: 272 seconds]
[00:48:38] -!- tom_o_t has quit [Changing host]
[00:50:36] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[00:51:04] -!- sirdancealot [sirdancealot!~koo5@98.82.broadband5.iol.cz] has joined #linuxcnc
[01:00:40] -!- FinboySlick has quit [Remote host closed the connection]
[01:03:44] -!- patrickarlt has quit [Remote host closed the connection]
[01:08:45] -!- patrickarlt has quit [Ping timeout: 246 seconds]
[01:08:54] -!- zeitue has quit [Quit: Leaving]
[01:09:24] -!- DrLEcter [DrLEcter!be8a16f0@gateway/web/freenode/ip.] has joined #linuxcnc
[01:13:10] -!- FinboySlick [FinboySlick!~shark@squal.net] has joined #linuxcnc
[01:16:38] -!- tjtr33 [tjtr33!~tjtr33@76-216-190-185.lightspeed.cicril.sbcglobal.net] has joined #linuxcnc
[01:18:21] <tjtr33> i thought singapore had a lot of rules, but in taiwan, they are even tougher http://ibin.co/1ZnhN6ryWLVN http://ibin.co/1ZnhgmORYvpk
[01:18:49] <DrLEcter> hi just for learning purposes somebody already done some cheap project like this youtube/watch?v=Z0vQIaUZGu4
[01:30:22] <tjtr33> its a good cheap intro to a cnc system. just a few old cd drives, maybe linuxcnc and a parport drive with cheap chip drivers.
[01:30:23] <tjtr33> BUT its really wiggly, and that part may be a bad thing to label as 'ok' in any way to a student of a real discipline.
[01:33:37] -!- Thetawaves [Thetawaves!~Thetawave@186-51-178-69.gci.net] has joined #linuxcnc
[01:36:01] <DrLEcter> ok I understand you about label as 'ok' project
[01:37:30] <tjtr33> I've thought about doing that exact for my 5 yr old nephew though.
[01:40:54] <tjtr33> this is simple, and fun too, just 2 steppers, string and a solenoid to control the spray paint http://www.youtube.com/watch?v=V4aXw0Zotzw
[01:45:32] <DrLEcter> jaja graffiti robot rocks
[01:46:11] <tjtr33> make it really big, and put a seat instead of spray can ... wheeeeee!
[01:47:09] <DrLEcter> but don´t let de subway go jaja!!
[01:47:51] <asah> anyone have advice about using mesa 24v io to control a device (vfd) expecting closed contact switching as opposed to 24v input.
[01:47:53] <asah> ?
[01:48:45] <asah> should I use small relays?
[01:52:26] <asah> unless serial control is working well for some people?
[01:54:13] <jdh> My plan is to use serial-485 for speed, but discrete for start/stop
[01:54:15] -!- patrickarlt has quit [Remote host closed the connection]
[01:54:37] <zeeshan> asah what vfd model
[01:54:56] <asah> hy, cheep. =)
[01:55:05] <zeeshan> jdh why not keep them both on serial
[01:55:14] <zeeshan> and just have e-stop run through discrete
[01:55:36] <jdh> 'because'
[01:55:46] <asah> is serial reliable in linuxcnc for vfd control?
[01:55:46] <zeeshan> tell me!
[01:55:47] <zeeshan> thats how i have it
[01:55:57] <zeeshan> asah it's been working for me
[01:56:06] <zeeshan> it's all about how you implement it
[01:56:24] <zeeshan> my vfd has a charge pump kind of deal
[01:56:29] <zeeshan> if linuxcnc crashes, or the computer shuts
[01:56:36] <zeeshan> and the signal is broken
[01:56:44] <zeeshan> the vfd goes in fault mode and everything shuts down
[01:57:01] <zeeshan> most vfds have this type of thing
[01:57:30] <asah> I like the reliability of just going with the logic.
[01:57:39] <asah> ie, digi io.
[01:57:44] <asah> for now.
[01:58:05] <asah> later, when I care more about other features, I can go to serial.
[01:58:21] <asah> but it looks like I can’t just drive the thing with 24v
[01:59:23] -!- patrickarlt has quit [Ping timeout: 240 seconds]
[01:59:57] -!- sirdancealot has quit [Ping timeout: 264 seconds]
[02:01:52] <asah> guess I have to run to radio shack… back in a bit. =)
[02:02:07] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[02:02:32] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[02:02:32] -!- anarchos2 has quit [Read error: Connection reset by peer]
[02:03:01] -!- anarchos2 [anarchos2!~mike@S010600259ce59399.vn.shawcable.net] has joined #linuxcnc
[02:04:14] -!- zeitue [zeitue!~z@] has joined #linuxcnc
[02:08:29] <jdh> what mesa IO?
[02:09:23] <jdh> and c'mon... how many people don't have at least 10 different kinds of relays laying around.
[02:10:04] <jdh> onlinemetals.com 25% off any* online order $100 or more.
[02:14:14] <zeeshan> !
[02:14:17] <zeeshan> interesting...........
[02:14:26] <zeeshan> starrett digital mics are made in china
[02:14:37] <zeeshan> http://i.ebayimg.com/00/s/NjAwWDgwMA==/z/qnMAAOSw-jhUDviZ/$_35.JPG
[02:20:50] <WalterN> hmm
[02:21:07] <WalterN> I cant seem to get envelope deformation to work?
[02:22:32] <WalterN> and, can I do it do a group of objects? like say I'm making a wall with a bunch of stuff on it and want to bend everything so it looks better from a certain perspective
[02:23:30] <WalterN> derp
[02:23:33] <WalterN> wrong chat
[02:23:40] <WalterN> next channel up...
[02:23:51] * WalterN uses the elevator
[02:28:16] <asah> jdh, ha! I have relays on mill I could use, but… you know, I want to have some solid state ones.
[02:28:28] <asah> the SSRs I have will only switch AC.
[02:30:25] <asah> the mesa io is a 7i71
[02:30:36] <asah> I also have a 7i83
[02:32:46] <SpeedEvil> Almost al SSRs wil only switch AC off
[02:33:32] amnesic is now known as amnesic_away
[02:34:51] <asah> the small ones I just bought are 15v max! =(
[02:34:56] <asah> ack.
[02:39:48] -!- DrLEcter has quit [Quit: Page closed]
[02:47:17] -!- syyl_ [syyl_!~sg@p4FD11491.dip0.t-ipconnect.de] has joined #linuxcnc
[02:52:45] -!- jbr has quit [Ping timeout: 264 seconds]
[02:59:49] -!- archivist [archivist!~archivist@host81-149-189-98.in-addr.btopenworld.com] has joined #linuxcnc
[03:01:16] -!- the_wench [the_wench!~the_wench@host81-149-189-98.in-addr.btopenworld.com] has joined #linuxcnc
[03:07:25] <tjtr33> check if you need just a contact closure, ssr is not 'just a contact closure', some things need no power, just a connection between 2 pins
[03:08:15] <tjtr33> urf dc ssr may be a contact closure, the ac are not
[03:13:36] -!- FinboySlick has quit [Quit: Leaving.]
[03:13:56] <pcw_home> the 7I83 enable outputs are DC contact closure equiv
[03:16:00] -!- AR__ has quit [Ping timeout: 272 seconds]
[03:18:12] -!- Valen [Valen!~Valen@c211-30-128-202.blktn6.nsw.optusnet.com.au] has joined #linuxcnc
[03:20:10] -!- sumpfralle has quit [Quit: Leaving.]
[03:26:34] <asah> ok, thanks peter.
[03:26:51] <asah> I am wiring up these relays, they seem to have worked at 24v.
[03:26:59] <asah> we shall see how long they last.
[03:27:31] <asah> I don’t have the 7i83 involved in the mill yet, but I suppose I could add it to the mix, and use the analog outs for spindle speed.
[03:27:45] <asah> spun the spindle today for the first time!
[03:28:18] <asah> big win for the single phase folks.
[03:29:04] <asah> 240V single phase -> transformer to step up to 380V single phase. -> VFD -> three phase.
[03:29:15] <asah> driving the ac induction motor.
[03:29:27] <asah> looked good on paper. nice to see it work out in reality.
[03:30:05] <zeeshan> nice asah
[03:30:18] <zeeshan> theres some vfds that'll step up the voltage also
[03:30:53] <zeeshan> ive been looking at running a 110V single phase in 1HP to 220V 3~ 1HP
[03:31:02] <zeeshan> well its actually not in @ 1HP
[03:31:07] <zeeshan> it draws a lot more current
[03:31:19] <zeeshan> i think sqrt3 times!
[03:32:47] -!- likevinyl [likevinyl!~nerdos@unaffiliated/likevinyl] has joined #linuxcnc
[03:36:24] <asah> sounds like a nice clean solution zeeshan.
[03:36:48] <zeeshan> you avoid the humming transformer!
[03:36:48] <zeeshan> :D
[03:36:53] <asah> prob quite a bit more expensive than my el cheapo solution. we will see how long the HY lasts.
[03:37:06] -!- patrickarlt has quit [Remote host closed the connection]
[03:37:30] <asah> Ill scour ebay in the meantime for a more pro VFD… see what pops out.
[03:39:19] <zeeshan> if you dont mind me asking
[03:39:25] <zeeshan> how much did the 240V to 380V xformer cost
[03:39:29] <zeeshan> and how many watts
[03:39:40] <asah> the transformer came with the mill. =)
[03:39:52] <zeeshan> oh
[03:39:57] <asah> I have another one lying around.
[03:40:03] <zeeshan> cause when i was looking for 220V to 575V transformers
[03:40:07] <zeeshan> they're stupid expensive
[03:40:09] <zeeshan> even at 5HP
[03:40:27] <zeeshan> i couldnt find them on the used market either
[03:40:30] <asah> where are you located?
[03:40:37] <zeeshan> near toronto , canada
[03:40:38] <asah> they are super heavy so shipping is rough..
[03:40:41] <asah> =)
[03:40:57] <asah> otherwise I could sell it cheap. =)
[03:41:01] <zeeshan> hehe
[03:41:33] <asah> it came off an industrial robot from detroit.
[03:41:58] <zeeshan> nice
[03:42:10] <zeeshan> i'd like to own an abb robot
[03:42:11] <asah> they auction off old robots (motoman, fanuc) for cheap
[03:42:11] <zeeshan> for home use
[03:42:28] <asah> and those cabs have all kinds of great automation parts in them.
[03:42:35] <zeeshan> cabs?
[03:42:38] <zeeshan> cabinets?
[03:42:43] <asah> and usually have isolation tranformers.
[03:42:45] <asah> cabinets.
[03:42:57] <zeeshan> my last job at eaton
[03:42:57] <asah> colloqual for us robot nerds.
[03:43:07] <zeeshan> my boss hooked me up with a lot of stuff when i was done my coop
[03:43:09] <zeeshan> but no robots
[03:43:27] <zeeshan> only vfds , ssr and some contactors and switches
[03:43:41] <zeeshan> these guys used to throw away anything used for lab testing purposes
[03:43:47] <zeeshan> after doing product evaluation
[03:43:48] <zeeshan> ..!
[03:45:02] <asah> nice to have the money to pull that off.=)
[03:52:19] <SpeedEvil> http://www.ebay.co.uk/itm/321507497743?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[03:52:38] <zeeshan> lol
[03:52:38] <zeeshan> !
[03:52:43] <zeeshan> too bad its in uk
[03:52:49] -!- MC500 [MC500!43352076@gateway/web/freenode/ip.] has joined #linuxcnc
[03:52:53] <zeeshan> that would be so bad ass to own.
[03:53:46] <asah> I have two. =)
[03:54:02] <SpeedEvil> asah: two of similar arms?
[03:54:10] <MC500> I put a negitive number in my P of the PID loop in the tuning app with all other values 0, the axis now moves more or less correctly, somthing must be backwards.. anyone have an idea?
[03:54:11] <SpeedEvil> Or do you mean the bio ones
[03:54:14] <asah> I have a large arm, a fanuc 420
[03:54:22] <zeeshan> how much did you pay
[03:54:31] <asah> and a very small motoman.
[03:54:39] <asah> motoman was $500
[03:54:44] <asah> fanuc was free
[03:54:44] <zeeshan> thats not bad
[03:54:51] <zeeshan> i really want one for welding.
[03:55:11] <asah> I paid more in shipping for the motoman.
[03:55:12] <asah> =)
[03:55:19] <SpeedEvil> how much would the above arm weigh? I'm guessing 300kg?
[03:55:22] <asah> but I got a bunch of really nice cable.
[03:55:41] <asah> my fanuc is around 2500-3000 pounds
[03:55:45] <zeeshan> asah are you in detroit?
[03:55:49] <asah> california
[03:55:52] <zeeshan> oh
[03:55:52] <pcw_home> MC500: if the DRO reads the correct direction the PID output device scale is backwards
[03:56:17] <MC500> pcw_home is that a setting or a cable?
[03:56:25] <asah> the motoman is more like 100 pounds.
[03:57:11] <pcw_home> MC500: what drives and what controller hardware?
[03:58:04] <asah> the motoman used to live its life at the huffy factory making bicycles.
[03:58:14] -!- kb8wmc has quit [Quit: ChatZilla [Firefox 20.0/20130329043827]]
[03:58:17] <MC500> stg isa card, baldor amplifiers, baldor A/C servo motors quad linear scales and one rotary scale hooked directly to the axis not motor
[03:58:45] <pcw_home> does the DRO read correctly?
[03:58:50] <MC500> yes
[03:59:41] <MC500> well... rotary... so + is right - is left but no home so I guess im not sure
[03:59:47] <pcw_home> so you need to reverse the DAC outputs in your HAL file
[03:59:58] <MC500> ok will do thank you
[04:00:33] <pcw_home> not sure how thats done for STG (probably some DAC scale parameter needs to have ist sign changed)
[04:02:00] -!- gambakufu [gambakufu!~ah@bzq-84-110-34-216.red.bezeqint.net] has joined #linuxcnc
[04:04:07] -!- MC500 has quit [Ping timeout: 246 seconds]
[04:04:39] -!- MC500 [MC500!43352076@gateway/web/freenode/ip.] has joined #linuxcnc
[04:06:05] <MC500> sorry got disconnected is it stg_motion.hal
[04:06:30] <MC500> net Aoutput => stg.3.dac-value
[04:08:08] <pcw_home> yeah that the DAC output but what sets the DAC scale?
[04:08:26] <MC500> yea how do I reverse it
[04:08:56] <MC500> oh Im not sure what sets the DAC scale
[04:08:57] -!- `Nerobro has quit [Ping timeout: 245 seconds]
[04:08:59] <MC500> what file is that in
[04:09:26] <pcw_home> probably there or the ini file
[04:10:17] <MC500> dac is not mentioned in the stg.ini file
[04:10:40] <pcw_home> output scale?
[04:12:16] <MC500> INPUT_SCALE = 1000 OUTPUT_SCALE = 1 OUTPUT_OFFSET = 0
[04:12:21] <MC500> thats in the .ini
[04:13:00] <pcw_home> might try
[04:13:02] <pcw_home> OUTPUT_SCALE = -1
[04:13:32] <MC500> k will do thank you
[04:13:54] <pcw_home> beware of runaways!
[04:22:00] -!- The_Ball [The_Ball!~ballen@] has joined #linuxcnc
[04:22:38] -!- `Nerobro [`Nerobro!~wolfeman@] has joined #linuxcnc
[04:26:49] -!- anarchos2 has quit [Read error: Connection reset by peer]
[04:27:22] -!- anarchos2 [anarchos2!~mike@S010600259ce59399.vn.shawcable.net] has joined #linuxcnc
[04:29:52] -!- WyrM has quit [Ping timeout: 272 seconds]
[04:32:09] -!- WyrM [WyrM!wyrm@unaffiliated/wyrm] has joined #linuxcnc
[04:45:27] -!- tjtr33 [tjtr33!~tjtr33@76-216-190-185.lightspeed.cicril.sbcglobal.net] has parted #linuxcnc
[04:56:42] -!- Tecan has quit [Ping timeout: 272 seconds]
[04:57:13] -!- cmorley has quit [Quit: Leaving.]
[04:57:48] <asah> awesome!!! spindle control through axis gui!… finally!
[05:01:02] -!- asheppard_ has quit [Ping timeout: 245 seconds]
[05:01:07] -!- sheppard has quit [Ping timeout: 245 seconds]
[05:02:35] -!- Fox_Muldr has quit [Ping timeout: 276 seconds]
[05:03:11] -!- theorbtwo [theorbtwo!~theorb@cpc8-swin16-2-0-cust252.3-1.cable.virginm.net] has joined #linuxcnc
[05:03:36] -!- ve7it has quit [Remote host closed the connection]
[05:04:08] -!- Fox_Muldr [Fox_Muldr!quakeman@frnk-5f74207f.pool.mediaWays.net] has joined #linuxcnc
[05:17:57] -!- syyl_ has quit [Ping timeout: 246 seconds]
[05:17:57] -!- anarchos2 has quit [Read error: Connection reset by peer]
[05:18:27] -!- anarchos2 [anarchos2!~mike@S010600259ce59399.vn.shawcable.net] has joined #linuxcnc
[05:22:35] -!- sirdancealot [sirdancealot!~koo5@98.82.broadband5.iol.cz] has joined #linuxcnc
[05:25:47] -!- `Nerobro has quit [Ping timeout: 255 seconds]
[05:38:29] -!- GJdan has quit [Quit: WeeChat 1.0-dev]
[06:02:59] Cylly is now known as Loetmichel
[06:03:14] -!- `Nerobro [`Nerobro!~wolfeman@fg300a.skylan.net] has joined #linuxcnc
[06:21:10] -!- zoni [zoni!b6b46bb8@gateway/web/freenode/ip.] has joined #linuxcnc
[06:21:13] <zoni> hi
[06:21:20] <zoni> any one there\
[06:21:26] <archivist> no
[06:21:34] <zoni> fine
[06:21:44] <archivist> in irc ask the real question
[06:22:28] <zoni> I need some help i want to know is it possible to use a microcontroller to change tool with modbus
[06:22:50] -!- jerryitt has quit [Quit: Connection closed for inactivity]
[06:23:09] <zoni> modbus only send tool number and get acknogement via rs232
[06:24:05] <archivist> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?VFD_Modbus
[06:25:46] -!- tmcw has quit []
[06:41:49] -!- cmorley [cmorley!~chris@S0106204e7f8c229b.no.shawcable.net] has joined #linuxcnc
[06:44:18] -!- The_Ball has quit [Remote host closed the connection]
[06:57:50] -!- Thetawaves has quit [Quit: This computer has gone to sleep]
[07:03:11] -!- Deejay [Deejay!~Deejay@unaffiliated/dj9dj] has joined #linuxcnc
[07:03:20] <Deejay> moin
[07:07:10] -!- zoni has quit [Ping timeout: 246 seconds]
[07:07:55] -!- patrickarlt has quit [Remote host closed the connection]
[07:21:30] -!- tjb11 [tjb11!~tjb1@cpe-69-207-197-216.stny.res.rr.com] has joined #linuxcnc
[07:24:52] -!- tjb1 has quit [Ping timeout: 245 seconds]
[07:31:06] -!- The_Ball [The_Ball!~ballen@c114-77-179-73.hillc3.qld.optusnet.com.au] has joined #linuxcnc
[07:31:16] -!- sirdancealot has quit [Ping timeout: 268 seconds]
[07:34:01] -!- automata [automata!~Amit@] has joined #linuxcnc
[07:35:05] -!- md-2 has quit [Remote host closed the connection]
[07:39:47] -!- md-2 has quit [Ping timeout: 245 seconds]
[07:43:26] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[08:07:17] -!- asdfasd [asdfasd!~332332@] has joined #linuxcnc
[08:07:32] -!- ries [ries!~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl] has joined #linuxcnc
[08:08:15] -!- b_b has quit [Changing host]
[08:11:01] -!- rob_h [rob_h!~robh@] has joined #linuxcnc
[08:12:27] -!- anth0ny has quit [Quit: anth0ny]
[08:17:17] -!- Demiurge has quit [Ping timeout: 245 seconds]
[08:19:55] -!- balestrino [balestrino!~balestrin@] has joined #linuxcnc
[08:21:50] -!- balestrino has quit [Client Quit]
[08:25:42] -!- balestrino [balestrino!~balestrin@] has joined #linuxcnc
[08:26:24] -!- alexchally has quit [Ping timeout: 252 seconds]
[08:36:46] -!- MC500 has quit [Ping timeout: 246 seconds]
[08:40:36] -!- toastyde1th has quit [Ping timeout: 246 seconds]
[08:44:31] -!- asah has quit [Quit: asah]
[09:01:52] stt_michaelz is now known as stt_michaelz^afk
[09:01:53] stt_michaelz^afk is now known as stt_michaelz
[09:06:53] -!- asdfasd has quit [Ping timeout: 240 seconds]
[09:08:57] -!- phantoxe [phantoxe!~destroy@acarlosss.broker.freenet6.net] has joined #linuxcnc
[09:14:31] -!- basiclases has quit [Excess Flood]
[09:20:57] -!- amiri has quit [Ping timeout: 264 seconds]
[09:22:50] -!- larryone has quit [Quit: This computer has gone to sleep]
[09:26:21] -!- rosslyoung has quit [Ping timeout: 264 seconds]
[09:27:58] -!- Tomashe has quit [Read error: Connection reset by peer]
[09:27:58] -!- postaL has quit [Read error: Connection reset by peer]
[09:29:16] -!- postaL [postaL!~postaL@c-50-131-18-140.hsd1.ca.comcast.net] has joined #linuxcnc
[09:33:07] -!- larryone has quit [Quit: This computer has gone to sleep]
[09:48:53] -!- Simooon [Simooon!~simon@gw.obelnet.dk] has joined #linuxcnc
[10:03:51] -!- larryone has quit [Quit: This computer has gone to sleep]
[10:11:30] -!- h_maximilian [h_maximilian!~bonsai@dslb-178-007-132-169.178.007.pools.vodafone-ip.de] has joined #linuxcnc
[10:23:10] -!- h_maximilian1 [h_maximilian1!~bonsai@dslb-088-065-148-112.088.065.pools.vodafone-ip.de] has joined #linuxcnc
[10:24:57] -!- h_maximilian has quit [Ping timeout: 260 seconds]
[10:25:52] <automata> hello
[10:26:18] -!- skunkworks_ has quit [Ping timeout: 246 seconds]
[10:26:19] <automata> anyone know how to use UART on the mesa cards?
[10:47:45] -!- Tecan [Tecan!~fasdf@unaffiliated/unit41] has joined #linuxcnc
[11:33:08] amnesic_away is now known as amnesic
[11:37:09] -!- meji3 has quit [Ping timeout: 264 seconds]
[11:43:25] -!- larryone has quit [Quit: Leaving]
[11:44:34] -!- Tecan has quit [Ping timeout: 272 seconds]
[11:48:56] -!- h_maximilian [h_maximilian!~bonsai@dslb-088-065-148-112.088.065.pools.vodafone-ip.de] has joined #linuxcnc
[11:50:59] -!- h_maximilian1 has quit [Ping timeout: 255 seconds]
[11:51:42] <SpeedEvil> http://www.ebay.co.uk/itm/321507497743?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT - last chance! /me resists.
[11:51:58] <SpeedEvil> I'd guess it'd be at least a couple of hundred to get it here.
[11:52:03] <SpeedEvil> (abb arm)
[11:53:38] -!- skunkworks [skunkworks!~skunkwork@68-115-41-210.static.eucl.wi.charter.com] has joined #linuxcnc
[11:54:18] -!- h_maximilian has quit [Quit: Leaving.]
[12:03:23] -!- Tomashe1 has quit [Ping timeout: 240 seconds]
[12:07:42] -!- PetefromTn_ [PetefromTn_!~IceChat9@66-168-88-232.dhcp.kgpt.tn.charter.com] has joined #linuxcnc
[12:16:53] amnesic is now known as amnesic_away
[12:21:34] -!- ramkam2014 [ramkam2014!~ramkam@] has joined #linuxcnc
[12:31:44] -!- sirdancealot [sirdancealot!~koo5@98.82.broadband5.iol.cz] has joined #linuxcnc
[12:35:46] -!- sylphiae [sylphiae!~sylphiae@99-7-4-250.lightspeed.cicril.sbcglobal.net] has joined #linuxcnc
[13:05:22] <automata> hi pcw_home
[13:07:01] -!- ramkam2014 has quit [Quit: Colloquy for iPad - http://colloquy.mobi]
[13:08:32] -!- FinboySlick [FinboySlick!~shark@] has joined #linuxcnc
[13:09:18] <PetefromTn_> automata Pete is on California time so is probably just getting up right now. Maybe try him a little later on.
[13:10:09] <automata> I'll try that... I am about 12 hrs ahead of peter... so I thought I'd take a chance and see if he is up yet...
[13:10:54] <PetefromTn_> sounds good just letting you know if you were not already aware. Good luck
[13:18:41] -!- syyl_ [syyl_!~sg@p4FD10F9F.dip0.t-ipconnect.de] has joined #linuxcnc
[13:18:51] -!- Tecan [Tecan!~fasdf@unaffiliated/unit41] has joined #linuxcnc
[13:23:56] -!- zeitue has quit [Quit: Leaving]
[13:24:59] <automata> thanks petefromTn_
[13:25:14] <PetefromTn_> NP man have a great day.
[13:26:26] -!- ramkam2014 [ramkam2014!~ramkam@] has joined #linuxcnc
[13:28:35] <zeeshan> i need some advice
[13:28:38] <zeeshan> my friend 3d printed me something
[13:28:43] <zeeshan> i need to know what something like this goes for:
[13:29:11] <zeeshan> damn it
[13:29:12] <zeeshan> lost the pics
[13:29:55] -!- jbr has quit [Client Quit]
[13:30:47] <zeeshan> http://i.imgur.com/UldT2g8.png
[13:30:49] <zeeshan> he printed that
[13:31:02] <zeeshan> its 1.5" tall, 2" wide, 3" long
[13:31:19] <zeeshan> whats a fair price to pay for that
[13:31:23] <zeeshan> he wont charge me money
[13:31:26] <zeeshan> but im gonna send him some anyway
[13:31:43] <zeeshan> ABS material
[13:33:04] <TekniQue> send him 20 bucks, and send me 50 for the advice
[13:33:19] <zeeshan> lol
[13:33:19] * archivist wants the tax
[13:35:06] <jdh> restaraunt gift card
[13:35:40] <jdh> you should have just machined that out of 6061
[13:35:42] <archivist> a grannygram
[13:35:47] <jdh> yah!
[13:35:47] <zeeshan> why
[13:35:52] <zeeshan> it needed to be non conductive
[13:36:32] <jdh> because it would be prettier and that's what really counts.
[13:36:33] <PetefromTn_> It is hard to know what CNC work is worth sometimes. I am struggling with a bid on a cool project right now for a local customer. always worried about quoting too high a price.
[13:36:55] <PetefromTn_> or too low
[13:36:58] <archivist> it is a bugger if you underprice too
[13:38:09] <archivist> factor in machine replacement cost and your time and loads of profit (well a bit
[13:38:31] <PetefromTn_> yeah right... I wish I was bidding on jobs that big LOL
[13:38:36] <Rab> zeeshan, did you send him a ready-to-print STL/etc, or did he do any processing of the file?
[13:38:50] <zeeshan> i sent him a solidworks file
[13:38:58] <zeeshan> he did some conversion on his end
[13:39:04] <Rab> That's the real time sink.
[13:39:37] <zeeshan> whats that rab
[13:39:49] <Rab> Tweaking the job.
[13:39:55] <PetefromTn_> Why don't you ask your friend how many hours it took to make it from start to finish and then give him a number you are comfortable with for that much time.
[13:40:17] <zeeshan> im gonna give him $20
[13:40:19] -!- arturia [arturia!~sylphiae@99-7-4-250.lightspeed.cicril.sbcglobal.net] has joined #linuxcnc
[13:40:24] <zeeshan> ill be doing some machine work for him in the future :P
[13:40:40] <Rab> $20 plus beers next time you see him.
[13:40:50] <zeeshan> rab it's [cube] from the other chan
[13:41:11] <zeeshan> he's funny he printed it and sent it on friday
[13:41:15] <zeeshan> and then bounched to china for 2 weeks
[13:42:30] -!- asdfasd [asdfasd!~332332@] has joined #linuxcnc
[13:43:00] <zeeshan> rab make anything else cool on your cnc?
[13:43:23] -!- sylphiae has quit [Ping timeout: 240 seconds]
[13:43:41] <jdh> if I had a 3d printer I would make a new light mount for my handlebar
[13:43:48] -!- paideia [paideia!~paideia@] has joined #linuxcnc
[13:43:59] -!- ramkam2014 has quit [Quit: Colloquy for iPad - http://colloquy.mobi]
[13:44:01] <zeeshan> build one jdh
[13:44:03] arturia is now known as sylphiae
[13:44:21] <zeeshan> CaptHindsight: did you ever use those diesel injectors for 3d printing
[13:44:38] <Rab> zeeshan, I did some engraving and tried cutting some stuff out of aluminum sheet. Cutting through in one pass worked great, but light passes softened the aluminum and almost grabbed the tool. I think the friction from the end of the mill was too much.
[13:44:53] <Rab> No coolant.
[13:45:11] <zeeshan> coolant makes a huge different on aluminum
[13:45:18] <zeeshan> *difference
[13:45:39] <zeeshan> like when im say milling .125" deep you'll start seeing a rolled up edge on the workpiece itself cause aluminum cuts so damn gummy
[13:45:49] <zeeshan> just by spraying water
[13:45:54] <zeeshan> it's enough to stop that from happening
[13:46:04] <Rab> Last night my friend and I hooked up a logic analyzer to tap the serial comms between the FlashCut software and the control box, to try to r/e the protocol. It looks pretty obtuse though.
[13:46:32] <zeeshan> hackers
[13:46:43] <Rab> We may just ditch everything and run the stepper drivers straight from the BeagleBone.
[13:47:20] <zeeshan> have others run stepper drivers from bb before?
[13:47:42] <Rab> I was hoping the protocol was fairly high-level, maybe even sending g-code directly...but it's very verbose, lots and lots of numbers for every movement.
[13:48:09] <Rab> zeeshan, yeah, works great apparently. I have Machinekit installed for that purpose.
[13:48:37] <zeeshan> that must be cool not needing a huge ass motherboard
[13:48:51] <jdh> relatively huge
[13:48:59] <jdh> and it does lots of things for you.
[13:49:12] -!- ramkam2014 [ramkam2014!~ramkam@] has joined #linuxcnc
[13:49:14] <zeeshan> i wanted to make a car computer out of the BB
[13:49:25] <zeeshan> all it needs is usb output, wireless in
[13:49:36] <mozmck> Rab: how is the FlashCut software? I presume you don't like something about it if you're looking to change?
[13:49:36] <zeeshan> and the ability to process an audio signal :P
[13:50:55] <Rab> mozmck, I have v1.4 which is very ancient and primitive. Win3.1-era stuff. The next version introduced some things like toolpath optimization, but this is what came with my controller (on three floppies).
[13:51:04] <mozmck> Ah, I see.
[13:51:10] <zeeshan> haha floppies
[13:51:16] <mozmck> I've heard of flashcut but never have seen it.
[13:51:29] <Rab> I'm sure modern FlashCut is nicer. Having said that, it's pretty well thought out and works great for what it does.
[13:52:36] <Rab> It's tied to their proprietary hardware, so unless you want to pay for a Windows-only hardware solution I don't think you would encounter it.
[13:52:50] <zeeshan> die mach 3!!!
[13:53:23] -!- ramkam2014 has quit [Client Quit]
[13:54:02] <Rab> It also has a decent "just works" quotient...LinuxCNC seems to be a lot more complex to set up and tune.
[13:54:29] <Rab> But that is my aim, which is why I'm here.
[13:54:41] -!- sumpfralle [sumpfralle!~lars@on1.informatik.uni-rostock.de] has joined #linuxcnc
[13:54:52] <zeeshan> you know my favorite thing about linuxcnc?
[13:55:10] <zeeshan> other than it being realtime
[13:55:11] <jdh> quit wasting time on the flashcut snip the wires, hook up the drivers.
[13:55:24] <zeeshan> you have the ability to write whatever code you feel like to make whatever you want to work
[13:56:17] <Rab> jdh, I think that's where we're at. ;) The r/e effort is mainly for fun.
[13:57:21] <zeeshan> oh and i forgot to mention another important thing.. linuxcnc has a community full of knowledgeable people ;p
[13:57:25] <zeeshan> EXCEPT JDH!!!
[13:57:28] * zeeshan hides
[13:57:30] * jdh nods
[13:57:38] <zeeshan> i kid
[13:57:47] <jdh> not really
[13:58:00] <zeeshan> you helped me out major in the mechanical mounting for the lathe
[13:58:10] <Rab> zeeshan, yeah, I just made a USB joystick for the FlashCut software. It spits out keyboard shortcuts for the GUI because there's no other way.
[13:58:21] <Rab> Presumably easier with LinuxCNC.
[13:58:59] <zeeshan> dont you have a spare computer
[13:59:02] <Rab> http://i.imgur.com/gA2jqPO.jpg
[13:59:11] <zeeshan> you've always got 29038321098123 electrical related stuff
[13:59:30] <Rab> Spare computer for what?
[13:59:50] <zeeshan> to use a standard parallel port with for a cheapo BoB
[14:00:45] <Rab> Man, I don't want to deal legacy ports and latency and all that crap. That's why I'm focusing on BBB.
[14:01:01] <jdh> get a 5i25
[14:01:08] <zeeshan> realistically what ipm are you planning to run on your machine?
[14:01:20] <zeeshan> cause i can hit 200 ipm over a parallel port
[14:02:08] <zeeshan> but i will be getting a 5i25 for my cnc mil project
[14:02:10] <Rab> zeeshan, I have rapids capped at 120 ipm and that's plenty fast for me.
[14:02:12] <zeeshan> cause i need more outputs
[14:02:21] <zeeshan> rab then a parallel port should work good for you
[14:02:43] <zeeshan> its these hardcore guys running crazy servos that need the speed =D
[14:02:48] <zeeshan> cough cough... SSI
[14:02:51] <Rab> I don't want to run a PC. They're big, power-hungry, they die.
[14:02:52] <zeeshan> with his 1200 ipm laser
[14:03:08] <Rab> The shop has no AC.
[14:03:21] <SpeedEvil> Rab: Modern PCs use ~10W
[14:03:33] <SpeedEvil> http://www.ebuyer.com/630828-gigabyte-ga-j1900n-d3v-intel-celeron-j1900-2ghz-vga-dvi-hd-audio-mini-itx-ga-j1900n-d3v
[14:03:44] <SpeedEvil> Fanless CPU, ...
[14:04:03] <SpeedEvil> Though there are aso ports for the beaglebone black
[14:04:06] <Rab> SpeedEvil, is there a parallel port on that board?
[14:04:41] <zeeshan> fail @ parallel port lacking
[14:04:44] <SpeedEvil> yes
[14:04:55] <zeeshan> oh you need a header for it
[14:05:01] <SpeedEvil> 1 x parallel port header
[14:05:11] <SpeedEvil> $2 on ebay
[14:05:48] <zeeshan> i kind of like the beagle bone idea though
[14:06:04] <zeeshan> i could have run a smaller enclosure with it
[14:06:54] <jdh> except that it is only a bb
[14:07:23] <Rab> The objective of the project is to spend as little money as possible...since I got the FlashCut stuff for free, and the BB for free, those are the obvious options.
[14:07:47] <zeeshan> im sure you have friends that have old computers
[14:07:48] <zeeshan> lol
[14:07:54] <zeeshan> my buddy hooked me up with his pentium 4
[14:08:09] <zeeshan> this thing hits 7000 ms on the jitter test
[14:08:10] <Rab> BBB seems to work great, but graphics performance is very poor so AXIS backplot is annoying.
[14:08:11] <zeeshan> which is nothing
[14:08:56] * jdh ponders 7000 ms
[14:09:04] <Rab> zeeshan, I have money. The financial constraint is just to make it interesting.
[14:09:15] <zeeshan> rab i know you have money!
[14:09:16] <zeeshan> :P
[14:09:25] <zeeshan> jdh i meant ns
[14:09:33] <zeeshan> not ms!
[14:10:36] <zeeshan> http://i41.tinypic.com/2ykaqo1.jpg
[14:11:24] amnesic_away is now known as amnesic
[14:16:38] -!- rosslyoung has quit [Quit: Leaving]
[14:20:04] -!- damien__ has quit [Quit: Quitte]
[14:22:54] -!- anth0ny has quit [Quit: anth0ny]
[14:25:22] -!- sumpfralle has quit [Ping timeout: 252 seconds]
[14:25:39] -!- damien__ has quit [Quit: Quitte]
[14:31:10] -!- patrickarlt has quit [Remote host closed the connection]
[14:32:48] -!- mozmck has quit [Read error: Connection reset by peer]
[14:33:02] -!- syyl_ws [syyl_ws!~sg@p4FD10F9F.dip0.t-ipconnect.de] has joined #linuxcnc
[14:36:07] -!- patrickarlt has quit [Ping timeout: 245 seconds]
[14:39:00] <PetefromTn_> Well I just put in my bid for the prototype project from a BIG local company and thankfully they did not even bat an eye about it. I actually priced it a bit higher than I felt I would do okay with it just to be sure. The designer said the price was fine with him but he has to run it up the ladder. He did not seem to think it was going to be a problem tho. So crossing fingers and waiting patiently for a confirmation email.
[14:40:39] -!- kb8wmc [kb8wmc!~chatzilla@] has joined #linuxcnc
[14:42:01] -!- sumpfralle [sumpfralle!~lars@on1.informatik.uni-rostock.de] has joined #linuxcnc
[14:43:47] <kb8wmc> I just finished downloading the 2.6.3 LinuxCNC Debian .iso file by the instructions, but cannot find where it is located, any help is appreciated...
[14:43:49] -!- almccon has quit [Client Quit]
[14:44:13] <jdh> what did you use to d/l it?
[14:44:49] <kb8wmc> zsync
[14:45:01] <kb8wmc> that was recommended
[14:45:38] -!- Valen has quit [Quit: Leaving.]
[14:45:39] <jdh> no clue.
[14:46:09] <cradek> you ran the zsync command at the shell, right?
[14:46:26] <kb8wmc> yes I did cradek
[14:46:34] <cradek> what directory were you in when you ran it?
[14:46:52] <kb8wmc> will look
[14:47:05] <kb8wmc> home directory
[14:47:12] <jdh> I'd look there!
[14:47:24] <kb8wmc> rgr...will look
[14:49:17] -!- mozmck [mozmck!~moses@] has joined #linuxcnc
[14:49:27] <kb8wmc> found it....duh....sorry for not looking beforehand...
[14:49:45] <cradek> :-)
[14:50:54] -!- MrSunshine [MrSunshine!~mrsun@c-403de253.4551536--62697410.cust.bredbandsbolaget.se] has joined #linuxcnc
[14:52:31] <kb8wmc> thanks
[14:53:16] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[14:55:26] -!- sumpfralle has quit [Ping timeout: 252 seconds]
[14:55:30] <CaptHindsight> zeeshan: yes
[15:02:52] <archivist> cradek, any known bug causing axis to stop listening to OS messages (kb/mouse etc) ?
[15:02:59] <archivist> ver 2.5.3
[15:03:42] <cradek> no. check dmesg and stdout/stderr
[15:04:25] <archivist> I am updating the box at the moment, not been used for some weeks, will try again
[15:05:23] <archivist> I was mid making some thing mdi mode
[15:07:04] <skunkworks> us gmane down?
[15:07:05] -!- sylphiae has quit [Ping timeout: 276 seconds]
[15:07:11] <skunkworks> is
[15:13:32] -!- jduhls has quit [Ping timeout: 245 seconds]
[15:16:50] -!- mle has quit [Ping timeout: 276 seconds]
[15:26:14] -!- skunkworks has quit [Ping timeout: 252 seconds]
[15:27:33] <CaptHindsight> no pads routed for the UART, no problem http://postimg.org/image/erzb4a0fz/ just cut through the package :)
[15:28:00] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[15:28:23] <archivist> best quality
[15:28:30] <pjm> nice - love it
[15:33:01] <ssi> zeeshan: you hush
[15:33:14] -!- anth0ny has quit [Quit: anth0ny]
[15:35:17] <ssi> so I have a friend who wants to build a 24x48x6" cnc router
[15:35:24] <ssi> and I'm trying to help him build it right
[15:35:29] <ssi> but he keeps wanting to cheap out on everything
[15:35:40] -!- patrickarlt has quit [Remote host closed the connection]
[15:35:59] -!- skunkworks [skunkworks!~skunkwork@68-115-41-210.static.eucl.wi.charter.com] has joined #linuxcnc
[15:36:03] <ssi> he wants to buy some cheap all in one breakout/drive kit from probotix
[15:39:35] -!- fomox_ [fomox_!~chatzilla@] has joined #linuxcnc
[15:39:45] <PetefromTn_> teach him to follow the path Grasshopper...
[15:39:50] <ssi> tryin
[15:39:57] <PetefromTn_> Preferrably the MESANET path LOL
[15:40:05] <ssi> yeah I'm trying to do that too
[15:40:18] <ssi> on that aspect I think the website scared him
[15:40:40] <PetefromTn_> yeah I agree the website is techie based at best
[15:41:50] <ssi> he was also thinking acme screw drive on this machine
[15:41:55] <ssi> but I feel like he'd be better off with belt
[15:43:23] <PetefromTn_> tell him to get a set of tinkertoys
[15:43:33] <ssi> heh
[15:45:26] -!- patricka_ has quit [Ping timeout: 255 seconds]
[15:46:04] <PetefromTn_> whatre you planning to build the frame from?
[15:47:56] <roycroft> "build it right" and "cheap out" are mutually exclusive, and the latter almost surely doomed to failure
[15:48:57] <roycroft> does anyone have a sense of whether the 80/20 extrusions are straighter than cold rolled steel generally is?
[15:49:49] <roycroft> doing a cost comparison, there's not a heck of a lot of difference between the two, especially when the cost of labor to weld the steel is added in
[15:51:48] <PetefromTn_> I would build it from steel simply because I think it IS cheaper and I enjoy tig welding LOL
[15:52:45] <ssi> he's gonna build the frame from steel rect tube I think
[15:52:50] <ssi> he can get all that stuff from work for nothing
[15:53:06] <ssi> it's mostly the drive mechanism, stepper drivers, and interface electronics I'm concerned about
[15:53:20] <ssi> I told him to budget about $1000 for motors/drivese/electronics and he about had a heart attack
[15:53:28] <PetefromTn_> hell tell him to get me some and I will weld two up LOL
[15:53:59] <ssi> he works for delta, and has an aerospace welding cert
[15:54:02] <ssi> I think he's got that under control
[15:54:32] <PetefromTn_> just a thought man..
[15:54:35] <ssi> :)
[15:54:40] <ssi> I sold my clausing lathe yesterday
[15:54:41] <ssi> to dan
[15:54:45] <ssi> so I still have access to it :P
[15:57:24] -!- fomox_ has quit [Ping timeout: 252 seconds]
[16:01:53] -!- patrickarlt has quit [Remote host closed the connection]
[16:02:11] <roycroft> $1k is a low estimate for that stuff
[16:02:22] <ssi> agreed
[16:03:16] <roycroft> i'm building a 3'x2'x8" router
[16:03:24] <roycroft> i got a good deal on some ball screws for it
[16:03:32] <roycroft> they still cost me $250
[16:03:40] <ssi> yeah ballscrews are expensive
[16:03:45] <ssi> I think there's little danger of him doing ballscrew
[16:03:47] <roycroft> the rails and pillow blocks will be more than that
[16:03:54] <ssi> but he was planning to do acme
[16:04:07] <ssi> but had no idea that he'd have to machine ends and make bearing mounts
[16:04:07] <roycroft> so i'm probably at $600-$700 just for that stuff
[16:04:12] <ssi> thought someone could sell him that stuff
[16:04:18] <ssi> which may be possible, but if so it's not gonna be cheap
[16:04:24] <roycroft> and we're not even talking motors/power supplies/drivers yet
[16:04:47] <roycroft> my ball screws shipped today btw
[16:04:50] <roycroft> from china
[16:05:00] <roycroft> and they're due to arrive tuesday next
[16:05:10] <roycroft> i was expecting shipping to take much longer
[16:05:28] <ssi> not bad
[16:05:51] <ssi> most of the ballscrew work I've done has been with roton screws
[16:05:52] <roycroft> i should be able to do some design work by next weekend
[16:05:55] <ssi> and they're pretty reasonably priced
[16:05:59] <roycroft> i'm designing around my ball screws
[16:06:08] <ssi> getting them pre-machined?
[16:06:11] <roycroft> yes
[16:06:17] <ssi> yeah that saves a lot of effort
[16:06:18] <roycroft> in a stock size
[16:06:28] <roycroft> stock sizes, rather
[16:06:32] <jdh> you can get them from lmb2008 any size you want dirt cheap
[16:06:42] <ssi> is that an ebay seller?
[16:06:52] <jdh> linearmotionbearings2008
[16:06:55] <roycroft> mine are 350mm, 950mm, and 1150mm
[16:07:02] <roycroft> $250 including shipping
[16:07:10] <roycroft> and that's who i got them from
[16:07:21] <roycroft> had i ordered custom sizes it would have cost me about $100 more
[16:07:25] <roycroft> which is still dirt cheap
[16:07:29] -!- patrickarlt has quit [Ping timeout: 260 seconds]
[16:07:31] <roycroft> but mine were dirt cheaper :)
[16:07:48] <PetefromTn_> I would think some well setup timing belts tensioned would work pretty well no? and be relatively cheap
[16:07:54] <ssi> http://www.ebay.com/itm/3-ballscrews-ballscrew-3set-SBR-rails-3sets-BK-BF12-3-couplings-Promotions-/251449603748?pt=BI_Heavy_Equipment_Parts&hash=item3a8b9076a4
[16:07:59] <jdh> just teh screws or bearings/etc also?
[16:08:01] <ssi> that would be good if the shipping wasn't $130
[16:08:10] <roycroft> i have bearings too
[16:08:12] <roycroft> for that price
[16:08:24] <roycroft> the double bearings for the head end and single bearings for the tail
[16:08:37] <roycroft> it really was a good deal
[16:08:42] <ssi> PetefromTn_: yeah that's my thinking... in terms of cheap and easy to build, timing belt would be fine
[16:08:49] -!- samcfc has quit [Quit: Page closed]
[16:08:57] <roycroft> they claim to be "zero backlash"
[16:09:03] <roycroft> that's what i have to determine right away
[16:09:04] <jdh> ssi: $400ish for that shipped is cheap
[16:09:11] <roycroft> if they are, then bob's my uncle
[16:09:31] <roycroft> if not, then i need to order more nuts, and the second sets of nuts will limit travel
[16:09:41] <zeeshan> roycroft: what kind of cold roll steel are you looking at
[16:09:48] <zeeshan> tubing or solid
[16:09:48] <ssi> I'm sure the backlash will be fine
[16:09:49] <PetefromTn_> belt drive would also be smoother and faster
[16:09:52] <roycroft> i could try oversize bearings if there's backlash
[16:09:54] <roycroft> tubing, zeeshan
[16:09:56] <ssi> I've had good results with the roton screws
[16:10:11] <roycroft> i don't know if oversize bearings can be used in rolled screws
[16:10:12] <zeeshan> im assuming round?
[16:10:27] <Connor> the C&'s from lm2008 are better than the roton as far as backlash.
[16:10:28] <ssi> roycroft: they can
[16:10:32] <zeeshan> i dont think ive ever seen cold rolled rectangular tuving
[16:10:33] <CaptHindsight> beware of crunchy ballscrews and linear bearings, you might have to spend time cleaning and relubing
[16:10:55] <Connor> lm2008 is a well known seller and makes things right.
[16:11:12] <CaptHindsight> look for ebay feedback with comments about getting bad or dirty sets
[16:11:14] <ssi> maybe I should just build myself a router
[16:11:14] <ssi> heh
[16:11:16] <roycroft> yeah, i'll probably dump the bearings out of the nuts and look for rust first thing
[16:11:19] -!- jasen [jasen!4e5330fa@gateway/web/freenode/ip.] has joined #linuxcnc
[16:11:24] <zeeshan> most cold roll is within +/- 0.003
[16:11:29] <ssi> roycroft: have you ever repacked one?
[16:11:29] <zeeshan> depending on section type and size
[16:11:30] <PetefromTn_> never had any trouble from my ROTON screws on the RF45 build. Double nut zero backlash...worked well
[16:11:32] <roycroft> no, ssi
[16:11:36] <jdh> there won't be rust. might be missing a ball or two
[16:11:36] <ssi> it's a huge pain :)
[16:11:37] <roycroft> it will be an experience :)
[16:11:42] <zeeshan> ssi
[16:11:42] <zeeshan> its not!
[16:11:43] <ssi> do it over a tray
[16:11:47] <roycroft> yes, of course
[16:11:48] <zeeshan> vaseline
[16:11:52] <zeeshan> and the ball chill
[16:11:55] <zeeshan> *balls
[16:11:58] <ssi> zeeshan: I use a dab of klueber
[16:12:05] * roycroft doesn't like the sound of ball chilling
[16:12:07] <Connor> PetefromTn_: Was referring to single nuts.. Not sure on the roton double nuts.
[16:12:16] <jdh> I didn't use anythign.
[16:12:28] <jdh> why vaseline?
[16:12:33] <ssi> especially if you put a dab of grease in the recirc tube
[16:12:35] <roycroft> ssi: i'm an engineer
[16:12:43] <ssi> so when you put it upsidedown back on the nut the balls stay put
[16:12:48] <roycroft> it's deeply ingrained in my psyche to immediately disassemble anything new
[16:12:48] <ssi> roycroft: we're all engineers :)
[16:12:49] <jdh> no recirc tube on these
[16:12:54] <Connor> jdh: To keep the balls in the race to make it easier to repack. ?
[16:12:55] <roycroft> i can't help it
[16:13:07] <zeeshan> the grease makes the balls stick
[16:13:11] <zeeshan> so they dont fall out
[16:13:14] <ssi> roycroft: hey on the bright side, you'll be repacking it on purpose
[16:13:18] <roycroft> yes
[16:13:22] <ssi> roycroft: instead of because you jogged it off the end of the screw
[16:13:25] <ssi> like I did on my lathe :)
[16:13:27] <jdh> Connor: did you repack yours? I can't see grease helping anything
[16:13:33] <roycroft> and i'll be unpacking over a tray
[16:13:43] <roycroft> unlike unpacking over a lathe bed
[16:14:00] <ssi> yep
[16:14:16] <ssi> but yea I repacked some of my roton single nuts with 0.0001" oversize balls and it helped a fail bit
[16:14:25] <zeeshan> vaseline is so sticky that you dip a screw driver tip in it
[16:14:30] <zeeshan> and touch the ball and it sticks :D
[16:14:33] <Connor> jdh Yes.. I've "repacked" mine for my router.. only I was using RM1610 no RM1605's
[16:14:46] <ssi> s/fail/fair/ haha
[16:14:50] <Connor> which means larger balls.. because it's 1 turn for 10mm vs 1 turn for 5mm
[16:14:55] <Connor> Was not fun.
[16:14:58] * likevinyl is away: chanserv is watching ;)
[16:15:13] <Connor> and was WAY back before I knew of any of the tricks..
[16:15:37] <Connor> and depending on how your end of the screw is machined can be almost impossible.
[16:15:39] <jdh> I jsut did it like the youtube videos
[16:15:44] -!- syyl_ has quit [Ping timeout: 252 seconds]
[16:15:52] <roycroft> it must have been fun picking ball bearings out of a sea of swarth
[16:16:12] <roycroft> anyway, can one use oversize balls with a rolled screw?
[16:16:20] <jdh> sure
[16:16:30] <Connor> Yes.. you need to get a assortment
[16:16:31] <zeeshan> roycroft: how do you think they are preloaded
[16:16:32] <roycroft> so if there's backlash that might be an option
[16:16:41] <ssi> yeah
[16:16:44] <ssi> I think you're overthinking this a bit :)
[16:16:49] <Connor> It's not un heard of to do something like this.. OooOooO in a race..
[16:16:56] <roycroft> well i have nothing better to do until they arrive ;)
[16:17:13] <Connor> Put a large ball every 2 or 3 smaller balls..
[16:17:17] <zeeshan> the only thing thats a big piece of shit on the chinese ball screws
[16:17:22] <zeeshan> from linmotionbearings2008
[16:17:25] <jdh> you could drool over pictures of cheap chinese c7 sces
[16:17:29] <zeeshan> is the way the nut secures at the end
[16:17:35] * roycroft reminds himself that this is a router, not a mill, and his goal is to hold 0.005", not 0.0001"
[16:17:47] <jdh> yours were secured?
[16:17:58] <zeeshan> for the actual screw
[16:18:01] <zeeshan> with that square nut
[16:18:14] <zeeshan> it has those allen key things that distort the the thread
[16:18:14] <CaptHindsight> some ballscrews come without any means of attaching nuts to the end :)
[16:18:15] <Connor> Yea.. single nut.. with a brass set screw..
[16:18:16] <zeeshan> its a piece of shit
[16:18:26] <Connor> doesn't work.
[16:18:33] <CaptHindsight> saves you the hassle of having to set preload
[16:18:35] <Connor> double nuts are really needed
[16:18:48] <CaptHindsight> shim into place
[16:18:59] <zeeshan> connor there isnt enough thread on the standard ones
[16:19:01] <zeeshan> for a double nut
[16:19:06] <ssi> that sucks
[16:19:06] <skunkworks> that is how the big boys do it...
[16:19:08] <ssi> what about thin jam nuts?
[16:19:12] <roycroft> i'm hoping to avoid double nuts
[16:19:13] <Connor> zeeshan: I know.
[16:19:16] <zeeshan> i just ended up using loctite
[16:19:17] <roycroft> that will eat about 50mm of travel
[16:19:40] <Connor> ssi: That's what I'm saying.. jam nut is needed..but like zeeshan said.. not enough threads on their standard threaded versions.
[16:20:02] <ssi> yeah but what I'm saying
[16:20:02] <ssi> http://www.grainger.com/product/Hex-Jam-Nut-6CB60
[16:20:08] <ssi> they make thin nuts specifically for jam nuts
[16:20:14] <ssi> maybe able to put two thin nuts in place of one regular
[16:20:24] <ssi> have a pic of the fixed end of one of these screws?
[16:20:25] <roycroft> i thought we had lathes to make nuts thin
[16:20:27] <Connor> on my mill, I even had to get someone make me slightly longer spacers because the nut rubbed on the seal of the bearing block.
[16:20:30] <jdh> he will put however much thread on there you want
[16:20:36] <zeeshan> ssi its not a standard thread
[16:20:41] <ssi> of course it's not :(
[16:20:44] <zeeshan> i will make custom jam nuts
[16:20:51] <zeeshan> its like m11x1 or something like that
[16:20:55] <zeeshan> depending on the ball screw
[16:21:01] <roycroft> it's m12x1 for the 16mm ball screws
[16:21:03] <ssi> can you get him to cut it in a standard size?
[16:21:13] <roycroft> which is pretty standard
[16:21:15] <jdh> I used mcmaster jam nuts for my lmb2008 screws
[16:21:19] <zeeshan> i actually ordered 4 nuts
[16:21:22] <zeeshan> with the plans to face them down
[16:21:28] <zeeshan> and make them smaller like jam nuts
[16:21:33] <zeeshan> just havent gotten around to doing it
[16:21:36] -!- syyl_ [syyl_!~sg@p4FD10F9F.dip0.t-ipconnect.de] has joined #linuxcnc
[16:21:40] <ssi> 12x1 doesn't seem to be standard
[16:21:51] <ssi> I see 12x1.25 and 12x1.5 and 12x1.75
[16:21:56] <zeeshan> yea its not
[16:21:59] <jdh> Connor: did you get ballscrews for your mill?
[16:22:15] <Connor> Not yet.. I was stil saving up $$$
[16:22:31] <zeeshan> the other question is
[16:22:37] <Connor> I have enough now for the screws, the mesa kit, the shims, and the bearings.. just haven't pulled the trigger.
[16:22:47] <zeeshan> how did you guys actually stop the screw from turning when locking your jam nut?
[16:22:51] <ssi> I need to find one of these lmb2008 kits that's long enough
[16:23:17] <ssi> zeeshan: if the machine is together, you could run the axis up to the end of its travel and that'll stop it
[16:23:26] <jdh> Connor: I probably have spare hardware if you need some (shims/nuts/etc)
[16:24:04] <zeeshan> what i ende dup doing was making a solid coupler that went at the end of the ball screw. tightened it down
[16:24:09] <Connor> Looks like somewhere around $551.45 for everything..
[16:24:11] <zeeshan> put a big ass vise grip on the od of the coupler
[16:24:17] <zeeshan> and thats how i stopped it from spinning lol
[16:25:17] <ssi> roycroft: are you doing one screw or two for the gantry?
[16:26:24] <roycroft> one
[16:26:39] <ssi> I would prefer two
[16:26:40] <roycroft> the gantry will extend below the bed
[16:26:45] <roycroft> and there will be a ball screw in the center
[16:27:02] <Connor> how wide ?
[16:27:14] <Connor> is the gantry ?
[16:27:16] <roycroft> machinery's handbook lists m12x1, m12x1.25, and m12x1.5 as metric fine threads
[16:27:20] <roycroft> ~2'
[16:27:28] <Connor> Okay. you should be good then.
[16:27:37] <CaptHindsight> http://machinedesign.com/linear-motion/importance-ballscrew-end-fixity answers the how's and why's of ballscrew end fixity
[16:27:39] <roycroft> yeah, if i were doing a 4'x8' router i'd use two
[16:27:42] <Connor> more than that, you could have issues with racking.
[16:27:59] <roycroft> the linear rails come with two pillow blocks each
[16:28:07] <roycroft> i'm planning on spacing them fairly far apart
[16:28:16] <roycroft> that should help with racking issues, shouldn't it?
[16:28:27] <Connor> It does yes.
[16:28:34] <Connor> but.. you can still have issues with it.
[16:28:38] <ssi> even more so for twist about X
[16:28:39] <roycroft> sure
[16:28:41] <Connor> but @ 2' it's not going to be.
[16:28:46] <ssi> er Y in your case I guess :)
[16:28:50] <roycroft> and again, my goal is to hold 0.005"
[16:29:03] <ssi> the router is going to want to make the gantry nod forward/aft
[16:29:15] <ssi> and you want a wide bearing footprint at the gantry ends to combat that
[16:29:22] <roycroft> right
[16:29:31] <ssi> probably 12" apart
[16:29:41] <roycroft> that's pretty wide
[16:29:52] <roycroft> but doable
[16:29:57] -!- maximilian_h [maximilian_h!~bonsai@dslb-088-065-148-112.088.065.pools.vodafone-ip.de] has joined #linuxcnc
[16:29:58] <ssi> especially since you want 8" of travel
[16:30:02] <ssi> Z travel
[16:30:06] <ssi> that's gonna mean a tall gantry
[16:30:10] <roycroft> yes
[16:30:12] <ssi> which'll exaggerate that moment
[16:30:15] <Connor> if using split rails, you can rotate the rails 90 degrees so that the open side of the bearings face the opposite set.. This helps combat the nodding issue.
[16:30:56] <roycroft> i haven't worked out how i'm going to mount the rails yet
[16:31:00] <ssi> Connor: you mean supported round rail?
[16:31:05] <ssi> like, mounting it on the outside faces of the frame
[16:31:08] <roycroft> it's been suggested that i mount them underneath the frame
[16:31:11] <ssi> so the split in the bearing is facing inboard rather than down?
[16:31:21] -!- patrickarlt has quit [Remote host closed the connection]
[16:31:35] <ssi> roycroft: yeah I was wondering how well those bearings can handle all their force pulling against the split
[16:31:38] <Connor> ssi yes
[16:31:39] <CaptHindsight> http://machinedesign.com/archive/precise-moves-depend-ball-screw-threads
[16:33:36] <Connor> I mounted mine to the bottom instead of 90 degrees.. and had issues with the gantry pulling up on the back.
[16:33:42] -!- patrickarlt has quit [Read error: Connection reset by peer]
[16:33:51] <Connor> the ball screw helped combat it.. but.. it's not ideal.
[16:34:02] <ssi> Connor: which machine is this?
[16:34:10] <Connor> My little MDF router.
[16:34:27] <Connor> http://www.ivdc.com/cnc/cnc1.jpg
[16:34:37] <Connor> http://www.ivdc.com/cnc/cnc8.jpg
[16:34:40] <ssi> ah
[16:34:49] <ssi> ooh I haven't tried lasercutting mdf yet
[16:34:57] <ssi> I bet it smells terrible
[16:35:06] <ssi> how well does that machine work?
[16:35:14] <Connor> MDF is a b!tch to laser cut.. You'll have to use very thin MDF
[16:35:19] <ssi> is it?
[16:35:29] <CaptHindsight> Connor: don't those blocks also have less bearing rows then others?
[16:35:32] <ssi> more so than hardwood ply?
[16:35:41] -!- md-2 has quit [Quit: Leaving...]
[16:36:08] <Connor> CaptHindsight: Open blocks I think only have 4 races.. closed ones have 5.
[16:36:17] <Connor> and the position of them might not be ideal either..
[16:36:31] <Connor> the split tends to flex if you pull UP on the block..
[16:36:46] <Connor> and you may only end up with 1 race in contact with the underside of the rail.
[16:36:56] <ssi> Connor: look at this
[16:36:57] <ssi> http://www.cnccookbook.com/CCPlasmaCNCIdeaGallery.htm
[16:37:00] <ssi> that top machine
[16:37:00] <Connor> ssi: MDF has tons more glue.
[16:37:09] <ssi> it looks like that huge ass gantry is hung by split bearings
[16:37:11] <ssi> with the split up
[16:37:15] -!- maximilian_h [maximilian_h!~bonsai@dslb-088-065-148-112.088.065.pools.vodafone-ip.de] has parted #linuxcnc
[16:37:18] <ssi> and I can't imagine that working well
[16:37:46] <Connor> much large diameter bearings.. who knows.
[16:37:52] <Connor> can really tell.
[16:39:18] <mozmck> with that large of gantry it is bound to be quite heavy as well, requiring *much* larger motors to get the acceleration you really need for plasma
[16:39:37] -!- diepchess [diepchess!~diep@a213-84-245-71.adsl.xs4all.nl] has joined #linuxcnc
[16:40:01] <Connor> http://www.igus.com/wpck/8107/5_linear_bearing_mistakes_you_do_not_know_you_are_making?C=US
[16:40:35] <diepchess> hi Good Morning!
[16:40:46] <diepchess> can smoothieboard work somehow with linuxCNC?
[16:40:55] <diepchess> or what controllerboard would you advice for linuxcnc?
[16:41:13] <Connor> If it's USB No.
[16:41:17] -!- maximilian_h [maximilian_h!~bonsai@dslb-088-065-148-112.088.065.pools.vodafone-ip.de] has joined #linuxcnc
[16:41:24] <diepchess> there is also ethernet version
[16:41:28] <Connor> probably not if it's ethernet either.
[16:41:32] <diepchess> linuxcnc doesn't connect to usb?
[16:41:35] <Connor> MESA is your best bet.
[16:41:55] <Connor> diepchess: for Mouse / Keyboard.. not for controllers and drivers.
[16:41:56] <diepchess> what's the reason to not choose for connectivity to usb nor ethernet?
[16:42:14] <diepchess> i looking for controllerboard for building simple CNC machine (milling metals)
[16:42:20] <diepchess> aluminium
[16:42:44] <diepchess> and there is so much software out there and hardware that seems not communicatingwith each other
[16:42:47] <Connor> LinuxCNC IS the controller and therefor needs to be realtime, USB isn't real time and they're working on Ethernet support for some mesa products.
[16:42:53] <diepchess> that i am thinking: "let's first figure out the software stack"
[16:43:20] <Rab> Is there a standard method for holding workpieces on a router? I'm thinking of making an aluminum plate with 10-32 holes on a 1" grid and using random hardware to bracket work down...what do people usually use?
[16:43:21] <Connor> You can do simple Parallel port to a break out board (BOB) to drive steppers..
[16:43:31] <diepchess> what is 'realtime' type of connection?
[16:43:43] <Connor> Parallel port, or using mesa card
[16:43:59] <diepchess> parallel port has slower latency than if we use ethernet card
[16:44:30] <diepchess> but i guess bandwidth through parallel port more than sufficient for some gcode
[16:44:47] <Connor> pport is used to send step/dir commands to stepper drivers.
[16:44:59] <CaptHindsight> diepchess: that board was specifically designed to avoid using Linuxcnc
[16:45:05] <diepchess> linuxcnc is not shipping gcode commands?
[16:45:09] <Connor> which linuxcnc generates.
[16:45:12] <mozmck> diepchess: smoothieboard is a stand-alone motion/CNC controller. Linuxcnc does all the motion control in the PC
[16:45:32] <ssi> linuxcnc is not a cut-rate 3d printer control software, so no it doesn't ship gcode commands
[16:45:38] <jdh> heh
[16:45:38] <ssi> it's a motion controller
[16:45:44] <Connor> diepchess: No, it processes the gcode and generates the signals itself.
[16:45:53] -!- kb8wmc has quit [Ping timeout: 240 seconds]
[16:46:08] <diepchess> well price interests me of course :)
[16:46:21] <diepchess> i'm building my own spindle here now
[16:46:26] <jdh> linuxcnc is free. run it on your dumpstered computer
[16:46:31] <Connor> the mesa card offloads the step/dir generation from the CPU on the computer.. but, still leaves linuxcnc as the motion control is still under linuxcnc's control
[16:46:31] <diepchess> and looking what steppers to order and which controllercard
[16:46:38] <PetefromTn_> plus smoothieboard just sounds kinda gay LOL
[16:46:48] <diepchess> yeah well i realize it's free. and a GUI frontend is nice as well to have
[16:47:16] <diepchess> so it has its own protocol over the parallel port
[16:47:21] <Connor> What sort of machine you looking to build
[16:47:26] <diepchess> as a result i hope there is real dirt cheap controller cards?
[16:47:44] <diepchess> well this will be a simple small cnc milling machine
[16:47:46] <diepchess> just for aluminium
[16:47:48] <Connor> diepchess: No. It generates a step / direction pulse train that a stepper motor controller uses
[16:47:57] <PetefromTn_> Mesanet cards are reasonably cheap and excellent quality. Can run about any machine.
[16:47:58] <diepchess> similar like 3d printer, just with ball screws and nema23
[16:48:13] <diepchess> so it can only steer steppers?
[16:48:26] <diepchess> the spindle i'm bilding from outrunner motor, yet i have brushless controller with it
[16:49:00] <diepchess> so it's simple 3 axis vertical milling machine
[16:49:04] <diepchess> just intended for alu
[16:49:05] <Connor> No. You can do encoders / servos.. but it's a bit limited using pport.. most who use encoders/servos use MESA
[16:49:24] <diepchess> i want to just use steppers
[16:49:44] <diepchess> with good microsteping
[16:49:49] -!- GJdan has quit [Quit: WeeChat 1.0-dev]
[16:49:54] <ssi> Connor: 1/4" mdf cuts beautifully at 50ipm
[16:49:56] <ssi> about the same as oak ply
[16:49:59] <ssi> about to try 1/2"
[16:50:10] <PetefromTn_> LOL saw that coming..
[16:50:15] -!- patricka_ has quit [Remote host closed the connection]
[16:50:22] <Connor> ssi Cool, the hackerspace guys were using 40W laser..
[16:50:27] <diepchess> [gonna look at mesacards]
[16:50:28] <CaptHindsight> diepchess: http://www.micromo.com/microstepping-myths-and-realities know what it really means
[16:50:54] <diepchess> yes i know that.
[16:50:57] -!- Flipp-_ has quit [Ping timeout: 264 seconds]
[16:51:11] <diepchess> however cheap ballscrews c7 have error easily 0.08 mm
[16:51:23] <diepchess> so the microstepping with stepper accuracy together is more than sufficient then
[16:51:43] <mozmck> diepchess: you don't even need a mesa card if you are using steppers and want to go cheap.
[16:51:59] <diepchess> i'm listening mozmck
[16:52:01] <diepchess> continue :)
[16:52:03] <mozmck> the microstepping is usually done in the motor drive itself
[16:52:12] <diepchess> yes it is
[16:52:35] <mozmck> The parallel port built into the computer will send step and dir signals
[16:52:41] <Connor> mozmck: That's just not good advice.. he can pick up the 5i25 for $99 and use it as a pport with standard BOB and still be way better off than straight pport.
[16:52:42] <diepchess> there is no interface card to parallel port with built in motordrivers?
[16:52:53] -!- paideia has quit [Ping timeout: 240 seconds]
[16:53:06] <mozmck> all you probably need is a simple buffer chip on a bob
[16:53:10] <diepchess> i'm not gonna do that much effort if it's 99 dollar a card
[16:53:10] <Connor> diepchess: No. Those you find on ebay are supper supper cheap and die.
[16:53:47] <Connor> $99 be too much $$$ ?
[16:53:50] <diepchess> what mesacard you'd be advicing connor?
[16:53:59] <Connor> 5i25
[16:54:07] <diepchess> i'm saying that 99 dollar for a card means i'm not gonna fiddle myself
[16:54:17] <mozmck> Connor: that's not true at all. You can skip the 5i25 and just use a bob off the parport
[16:54:27] <CaptHindsight> LPT breakout boards are ~$15 and up
[16:54:33] -!- patrickarlt has quit [Read error: Connection reset by peer]
[16:54:38] <mozmck> no fiddling needed with parport.
[16:54:48] <mozmck> hook up a BOB and go
[16:54:53] <diepchess> link?
[16:55:02] <Connor> mozmck: I know.. I'm doing it.. I'ms aying.. mesa 5i25 gives you better performance.. no need to run the base thread.
[16:55:11] <Connor> http://www.mesanet.com/
[16:55:13] <CaptHindsight> getting clear specs on the BOB is another story
[16:55:35] <Connor> http://store.mesanet.com/index.php?route=product/product&path=83_84&product_id=215
[16:55:53] <Connor> that's a 5i25 and BOB combo.. which is what I will be upgrading too.
[16:55:55] <Connor> BUT...
[16:56:03] <PetefromTn_> http://www.cnccookbook.com/img/PlasmaTable/Ideas/sideview.jpg I like that design..
[16:56:08] <mozmck> Connor: better performance is relative. I run my 4' x 4' router that way, and the linuxcnc computer will work faster than the motors can keep up
[16:56:13] <PetefromTn_> Looks cool and well built.
[16:56:35] <Connor> http://www.geckodrive.com/geckodrive-step-motor-drives/g540.html
[16:57:02] <ssi> Connor: 1/2" is a little iffier, but it cuts at 10ipm
[16:57:10] <ssi> my 2" lens is really too short for that
[16:57:23] <Connor> OR http://www.automationtechnologiesinc.com/products-page/digital-stepper-motor-driver/3-axis-dsp-based-digital-stepper-drive-max-60-vdc-6-0a
[16:57:23] <ssi> I have a 3" lens coming which I'm hoping will make 1/2" and 3/4" stuff easier
[16:57:32] <Connor> combined with a 5i25, no need for a break out board.
[16:57:47] <Connor> both have internal BOB's
[16:58:05] <ssi> whoa I didn't know they made a 3 axis all in one
[16:58:06] <ssi> that's fancy
[16:58:10] <ssi> is there a 4 axis one?
[16:58:31] <ssi> doesn't look like it
[16:58:32] <Connor> Not that I know of.. the MX3660 is nice.. I like it better than the G540 stepper driver..
[16:58:49] <ssi> yeah but the g540 is four drives :/
[16:58:57] <ssi> I prefer those keling digital drives these days
[16:59:03] <ssi> I'm using three KL5056D's on the laser
[16:59:07] <diepchess> ah ja the gecko 540 i saw on youtube
[16:59:18] <ssi> laser is only 2 axis, and dual gantry drive
[16:59:43] <ssi> maybe if I build a router I'll get a MX3660, swap it into the laser, and get a fourth 5056D and use those for the router
[16:59:55] <PetefromTn_> you can use another axis to point the laser at your neighbors house and light that baby up!
[17:00:34] <ssi> PetefromTn_: I honestly think if I shot this thing across the street it probably would catch a house on fire
[17:00:39] <ssi> the 100W tube is effin serious
[17:01:01] <PetefromTn_> no doubt
[17:01:30] <ssi> I have a piece of plywood with holes cut in it below the 1/2" aluminum honeycomb that is my bed
[17:01:46] <ssi> and when I'm cutting slowly, it'll catch that bottom plywood on fire and it has melted the aluminum in spots
[17:01:55] <CaptHindsight> diepchess: example BOB's http://www.ebay.com/itm/310702269807 http://www.ebay.com/itm/111149514180
[17:01:56] <ssi> I need to replace the plywood with expanded metal
[17:03:12] -!- rob_h has quit [Ping timeout: 245 seconds]
[17:03:26] -!- balestrino has quit [Ping timeout: 276 seconds]
[17:03:35] -!- maximilian_h [maximilian_h!~bonsai@dslb-088-065-148-112.088.065.pools.vodafone-ip.de] has parted #linuxcnc
[17:03:51] <Jymmm> ssi: Uh, what speed/power are you setting when you say "cutting slowly"
[17:04:34] <CaptHindsight> diepchess: http://www.cnc4pc.com/Store/osc/product_info.php?cPath=33&products_id=425
[17:05:25] <ssi> I dunno, sub-10ipm at 100W
[17:05:44] <diepchess> wow 22 dollar
[17:05:45] <Jymmm> ssi: Then droop the power DRAMATICALY
[17:05:59] <diepchess> but it still needs stepper drivers right?
[17:06:00] <ssi> the problem is my optics are too short for a lot of stuff
[17:06:32] <CaptHindsight> diepchess: that one has no pwm or relay for spindle
[17:06:38] <Jymmm> ssi: doesn't matter, There are cases where 100% won't cut shit, yet drop to 65% and lower speeds will.
[17:06:41] <CaptHindsight> http://www.cnc4pc.com/BB.htm has more info
[17:06:44] <ssi> hm
[17:07:04] <Jymmm> ssi: and VERY cleanly too. Can't think like mechanical cutters here.
[17:07:29] amnesic is now known as amnesic_away
[17:07:36] <Jymmm> ssi: Can even have different settings for summer/winter too.
[17:07:50] <ssi> Jymmm: you have any thoughts on cutting carbon fiber?
[17:07:51] <CaptHindsight> diepchess: yes, that just buffers and isolates the LPT port from the machine, you connect the BOB to the stepper drivers and limit/home switches
[17:08:02] <ssi> I didn't have much luck with it
[17:08:15] <Jymmm> ssi: And you won't unless you have a Yag
[17:08:19] <ssi> dang
[17:08:21] <Jymmm> YaG
[17:08:52] <CaptHindsight> diepchess: this combines the BOB, 3 stepper drivers and home/limit switch IO all into one http://www.automationtechnologiesinc.com/products-page/digital-stepper-motor-driver/3-axis-dsp-based-digital-stepper-drive-max-60-vdc-6-0a
[17:09:34] <ssi> hm there's a guy on cnczone who claims to have cut some 1mm carbon sheet at 50% power with a 120W co2 laser
[17:10:08] <CaptHindsight> with the right optics
[17:10:25] <ssi> I have short optics
[17:10:31] <ssi> shorter would be better in this case
[17:10:43] <ssi> pretty sure I tried the 1.5" lens on it
[17:10:46] <Connor> CaptHindsight: Your a little late.. I already showed him that. :)
[17:10:46] <ssi> which is about as short as it gets
[17:10:54] <Jymmm> ssi: focal point size would be the major factor
[17:11:07] <ssi> 1.5" lens on my system calcs to like a 67um spot
[17:11:21] <Jymmm> Ok, fair enough.
[17:11:22] -!- kfoltman [kfoltman!~kfoltman@] has joined #linuxcnc
[17:11:30] <ssi> 64.6um
[17:11:39] <ssi> and 30kw/mm^2 at 100W
[17:11:43] <ssi> which is absurd power
[17:12:02] <Jymmm> it's not always about power
[17:12:06] <ssi> I've read that it takes 10kw/mm^2 to cut steel
[17:12:08] <ssi> but I wasn't able to get through it
[17:12:18] <ssi> tried 26ga sheet, and it'll mark it enough you can feel it with a fingrenail, but that's it
[17:12:38] -!- Tecan has quit [Ping timeout: 272 seconds]
[17:12:48] <jdh> how wide is the mark?
[17:12:53] <ssi> pretty narrow
[17:13:55] <ssi> hang on, pic incoming
[17:14:00] -!- anth0ny_ has quit [Quit: anth0ny_]
[17:14:49] <ssi> https://pbs.twimg.com/media/BxMAiYAIUAAfUAM.jpg
[17:15:10] <ssi> might be doable with a pierce delay, slower speed, higher pressure
[17:15:11] <ssi> I dunno
[17:15:15] <ssi> there's so many damn variables
[17:16:11] <Jymmm> ssi: do you need to cut a LOT of carbon fiber or just playing?
[17:16:31] <ssi> I'd like to be able to cut a lot
[17:16:50] <Jymmm> The flodd the compartment with nitrogen and try it that way.
[17:16:53] <Jymmm> flood*
[17:17:01] <ssi> ehh I'm not in a good position to do that just yet
[17:17:06] <ssi> what's the "just playing" option?
[17:17:23] <Jymmm> There wans't one.
[17:17:52] <ssi> hahaha
[17:19:48] <Jymmm> There's also PPI and DPI settings that can affect things as well.
[17:20:53] <ssi> I've been running 5000PPI mostly
[17:20:55] <ssi> what's DPI?
[17:21:13] <Jymmm> dots per inch
[17:21:18] <ssi> wtf is the difference? :P
[17:21:20] <Jymmm> why 5000 ppi?
[17:21:29] <ssi> I dunno, that's what I settled on back when I had the 40w tube in
[17:21:34] <ssi> I need to reevaluate everything with the new tube
[17:21:45] -!- PetefromTn_ has quit [Quit: I'm Outta here!!]
[17:21:46] <ssi> I need to realign too
[17:22:10] <ssi> I got a power meter, and I took a power shot just out of the output coupler and got 102W at 28ma, but only like 86W at where the lens would be
[17:22:26] <ssi> not sure if that's poor alignment, or just inevitable falloff
[17:22:35] <Jymmm> I think you need to set everythign close to lowest settings and stop playing Tim the toolman taylor.
[17:23:26] <ssi> this might shock you, but it's possible to be helpful and not condescending at the same time
[17:24:07] <Jymmm> you have everything set to the highest, what else do you expect?
[17:24:29] -!- automata has quit [Ping timeout: 260 seconds]
[17:24:34] <Jymmm> MORE POWER UG UG UG
[17:24:49] <ssi> I don't, actually
[17:24:56] -!- BellinganRoy [BellinganRoy!~Bellingan@host243-229-dynamic.43-79-r.retail.telecomitalia.it] has joined #linuxcnc
[17:24:56] <ssi> I'm running the tube at the "best life" current
[17:24:59] -!- phantoxe has quit [Remote host closed the connection]
[17:25:01] <Jymmm> 200ppi and 500 dpi covers most materials.
[17:25:20] <ssi> i'd be very surprised if that's true
[17:25:27] <ssi> and again, what the hell is the difference between ppi and dpi
[17:25:35] <Jymmm> you tell me
[17:25:39] <ssi> how can they be different?
[17:26:17] -!- Guest84981 has quit [Quit: Ex-Chat]
[17:26:40] <ssi> every pulse makes a dot
[17:26:48] <ssi> how do I pulse 200 times per inch, but somehow end up with 500 dots?
[17:26:55] <ssi> I don't have a DPI control
[17:27:01] <ssi> so I'm asking a genuine question here
[17:27:07] <Jymmm> what makes you think that every pulse makes a dot?
[17:27:15] <ssi> I guess I'm just a simpleton
[17:28:19] <Jymmm> Well, good luck.
[17:28:29] <ssi> why are you being antagonistic about this?
[17:28:36] <ssi> I geniunely don't understand what you're talking about
[17:28:44] <ssi> remember, I don't have a dsp
[17:28:48] <ssi> I'm running this with linuxcnc
[17:28:59] <ssi> I have a pulser module which fires the laser some number of times per inch
[17:29:02] <ssi> for 2.5ms at a time
[17:29:12] <ssi> so I have a ppi control
[17:29:13] -!- f1oat [f1oat!~f1oat@AMontsouris-553-1-22-30.w92-151.abo.wanadoo.fr] has joined #linuxcnc
[17:29:22] -!- kfoltman has quit [Quit: Ex-Chat]
[17:29:26] <ssi> I don't understand what you mean by dpi as something that's independent of ppi
[17:29:29] <ssi> I don't have a dpi control
[17:29:36] <roycroft> every pulse is sacred
[17:29:40] <roycroft> every pulse is great
[17:29:43] <roycroft> if a pulse gets wasted
[17:29:49] <roycroft> god gets quite irate
[17:30:46] <Connor> ssi: sounds like ppi and dpi are the same thing ?
[17:30:51] <diepchess> CaptHindsight: my spindle is an outrunner motor with its own brushless controller that is programmable for speed
[17:31:02] <diepchess> so i'm afraid it's not easy to have linuxcnc control the spindle speed anyway
[17:31:13] <diepchess> so that 22 dollar card is tempting
[17:31:15] <ssi> Connor: not according to Jymmm, but I'm too stupid to understand why so apparently he's decided to withhold his bounty of information
[17:31:39] <ssi> so I guess I'll just go back to doing it the way i have been
[17:32:08] <Connor> well.. technically.. PPI is pixels per inch.
[17:32:14] <ssi> pulses per inch in this case
[17:32:41] -!- pcw_home has quit [Ping timeout: 276 seconds]
[17:32:45] <diepchess> 24 volt is what the $22 card steers right?
[17:33:08] <ssi> I think dpi is relevant to raster engraving only, which I don't care about at all right now
[17:33:55] -!- Flipp- [Flipp-!Flipp@nat/google/x-kvrcsvpnuiwompls] has joined #linuxcnc
[17:34:08] <ssi> I think the disconnect in language here is because Jymmm runs his laser from a print driver
[17:34:15] -!- pcw_home [pcw_home!~chatzilla@c-50-143-148-115.hsd1.ca.comcast.net] has joined #linuxcnc
[17:34:20] <ssi> whereas mine is run like a real machine
[17:34:33] <ssi> my vector paths have infinite DPI
[17:34:50] -!- Jymmm [Jymmm!~jymmm@unaffiliated/jymmm] has parted #linuxcnc
[17:34:53] <Connor> DPI is a graphics / printer term..
[17:35:09] <ssi> right
[17:35:15] <ssi> and his "laser" is just a printer :P
[17:35:40] <jdh> diepchess: how do you controll the speed on the outrunner drive?
[17:37:17] -!- anth0ny has quit [Ping timeout: 245 seconds]
[17:37:24] <diepchess> it can be programmed jdh
[17:37:37] <diepchess> according to the chinese website :)
[17:37:48] <diepchess> i paid 11.25 dollar for a programming box for it :)
[17:37:50] <ssi> diepchess: is it like an RC ESC?
[17:38:00] <diepchess> SBEC
[17:38:09] <ssi> got a link?
[17:38:32] <diepchess> yes but not sure i like the store where i bought it, if i see the flood of emails i got later on
[17:38:43] <ssi> I just want to see what it is
[17:38:46] <jdh> can it use an external analog for control?
[17:39:10] <ssi> if it's an RC ESC, it takes a particular kind of pwm
[17:39:17] <ssi> which it would be possible to set up linuxcnc to run
[17:39:23] <diepchess> here : 1x #TGY_DEL_PB/17619 Turnigy dlux Programming Box w/Data Logging
[17:39:30] <ssi> I've been meaning to do that myself, since I have a brushless spindle with an ESC
[17:39:34] <diepchess> and 1x #TGY_DEL100A/16367 Turnigy dlux 100A SBEC Brushless Speed Controller
[17:39:50] <diepchess> and outrunner: 1x #SK3-6364-245/18128 Turnigy Aerodrive SK3 - 6364-245kv Brushless
[17:39:50] <ssi> yeah that's what I'm talking about
[17:39:56] <ssi> the speed is set by an RC servo signal
[17:40:14] <ssi> which is pwm with a 20ms period, and the duty varies from 1ms on to 2ms on
[17:40:25] <ssi> 1ms on gives you 0% speed, 2ms gives you 100% speed
[17:40:42] <Connor> http://protofusion.org/wordpress/2010/06/victor-speed-controllers-and-emc2/
[17:41:20] <ssi> beautiful
[17:41:20] <diepchess> sounds pretty fast
[17:41:28] <diepchess> but it's a stand alone controller
[17:42:04] <jdh> so
[17:42:08] <ssi> I wonder if a mesa pwm can be set up to run servo pwms easily
[17:42:23] <Connor> I bet pcw could tell yea. :)
[17:42:28] <ssi> I'm sure he could
[17:43:52] <diepchess> that brushless speedcontroller 100A is supposed to be able to set the speed of that outrunner
[17:43:56] <diepchess> gonna work you guess?
[17:44:15] <jdh> sounds likely.
[17:44:17] <diepchess> as a few days from now that package should be here
[17:44:28] <diepchess> my job is then to build an accurate spindle on the lathe here :)
[17:44:46] <diepchess> i won't be putting an ER collet through the outrnner
[17:44:51] <ssi> so my 2" lens has a 93um spot
[17:44:53] <diepchess> yet will connect the outrunner to a small spindle
[17:44:56] <ssi> which is about 3.6 mil
[17:45:17] <ssi> so ppi needs to be above 277 in order to overlap
[17:45:18] -!- patricka_ has quit [Remote host closed the connection]
[17:45:30] <ssi> lemme see what 300ppi looks like on 1/4" mdf
[17:47:40] -!- sumpfralle [sumpfralle!~lars@on1.informatik.uni-rostock.de] has joined #linuxcnc
[17:49:04] <ssi> not nearly enough
[17:49:09] <ssi> even at 10ipm won't cut through
[17:50:23] <Connor> ssi: Is there a way you can simulate a 40w or 60w laser ?
[17:50:38] -!- patrickarlt has quit [Ping timeout: 272 seconds]
[17:51:30] <ssi> yeah I can turn the current down
[17:51:36] <ssi> 500ipm and 20ipm cuts 1/4" mdf pretty well
[17:51:40] <ssi> er, 500ppi
[17:51:50] <Connor> I would like to know what a 40w and 60w would do with MDF
[17:52:06] <ssi> I'm about to turn the current down and see where it stops cutting
[17:52:09] <Connor> no way am I going to have enough room for a 100w
[17:52:51] -!- jerryitt has quit [Quit: Connection closed for inactivity]
[18:00:00] -!- asah [asah!~asah@c-69-181-176-3.hsd1.ca.comcast.net] has joined #linuxcnc
[18:00:01] <ssi> Connor: I have it cutting 1/4" mdf pretty nicely at 14mA, and I stuck the power meter in the beam at the OC at 14mA and it read 56W
[18:01:09] -!- patrickarlt has quit [Ping timeout: 264 seconds]
[18:05:21] -!- f1oat has quit [Ping timeout: 264 seconds]
[18:06:26] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[18:10:48] -!- automata [automata!~Amit@triband-mum-] has joined #linuxcnc
[18:12:34] <automata> hi pcw
[18:13:08] <automata> can you guide me on how to use the uart from hostmot2
[18:14:51] -!- Nick001-shop [Nick001-shop!~chatzilla@50-32-101-83.adr01.dlls.pa.frontiernet.net] has joined #linuxcnc
[18:15:51] <diepchess> that C50 board only supports 5 volt it says here http://cnc4pc.com/Tech_Docs/C50R1_user_manual.pdf
[18:16:08] <diepchess> how to give tad more juice to steppers than 5 volt?
[18:18:47] <pcw_home> I think you will need Andys help with the UART, it does n ot expose any HAL pins directly
[18:18:48] <pcw_home> and relies on a helper component tha created the actual hal pins (like the BSPI interface and 7I65 component)
[18:19:18] <pcw_home> s/tha created/that creates/
[18:19:21] <Connor> ssi so, 60W max would work.
[18:20:20] <Connor> pcw_home: 5i25 ability to drive a standard hobby servo via PWM?
[18:20:30] <diepchess> i just figured out my Xeon servers do not have a parallel port at all :)
[18:20:42] -!- bedah [bedah!~bedah@g224074058.adsl.alicedsl.de] has joined #linuxcnc
[18:20:57] <diepchess> let's dig in the manual of it :)
[18:21:36] <Connor> diepchess: again, that's a Interface card.. not a stepper driver.. not designed to drive steppers direct.. at any voltage
[18:22:24] -!- asah has quit [Quit: asah]
[18:23:02] <diepchess> what stepper drivers you'd advice when bought stand alone?
[18:23:29] <pcw_home> Connor: its not optimum since it has full 0-100% range and 12 bit resolution so you only get about 200 steps from 1-2 ms
[18:23:31] <diepchess> i have entire cluster here of those Xeons crunching at prime numbers
[18:23:42] <diepchess> so using a fwe of them to also steer CNC machines not a problem
[18:23:47] -!- syyl_ws has quit [Remote host closed the connection]
[18:23:53] <diepchess> let me figure out whether there is pci-e cards with a parallel port :)
[18:25:10] <skunkworks> if you had feedback it could dither between steps...
[18:25:32] <CaptHindsight> ~$10 on ebay for 1 LPT PCIe card
[18:25:41] <Connor> was just thinking about ability to use it to drive a PWM to Brushless motor driver..
[18:26:17] <pcw_home> for a spindle 200 steps is probably OK
[18:26:24] <skunkworks> sure
[18:26:43] <pcw_home> for a servo, 50 Hz update rate is useless
[18:27:27] -!- Flipp-_ [Flipp-_!~Flipp@c-50-181-211-86.hsd1.wa.comcast.net] has joined #linuxcnc
[18:27:45] -!- patrickarlt has quit [Remote host closed the connection]
[18:28:12] <automata> i found andys UART driver at https://github.com/sittner/linuxcnc/blob/master/src/hal/drivers/mesa_uart.comp
[18:30:08] <ssi> Connor: yea... I was able to cut 5mm plywood quite well with a cheap 40W tube
[18:30:11] <ssi> Connor: and it's TINY
[18:30:44] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[18:31:02] <CaptHindsight> diepchess: http://www.ebay.com/itm/170988778068 $9.46 Parallel Port DB25 LPT Printer to PCI-E PCI Express Card
[18:31:08] <automata> I also found the hm2_uart_send and hm2_uart_read functions at http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=blob;f=src/hal/drivers/mesa-hostmot2/uart.c;h=e3c7fc33d9daf0852ba1a7a02367fceb7a3af498;hb=ac8d297d9e1c2a84d906a43f498d9064698d0e1c
[18:32:21] <automata> what firmware to load on the FPGA. I have 7i43, 7i90, 7i80, 5i25 and 5i24 available with me and can use any of them to test
[18:32:48] <automata> And how to define the baud rates of the uarts?
[18:33:01] -!- anth0ny has quit [Quit: anth0ny]
[18:33:05] -!- patrickarlt has quit [Ping timeout: 260 seconds]
[18:33:30] <pcw_home> I think theres a hm2 UART man page somewhere
[18:34:17] -!- rythmnbls [rythmnbls!~rythmnbls@] has joined #linuxcnc
[18:34:48] -!- rob_h [rob_h!~robh@] has joined #linuxcnc
[18:35:10] <automata> http://linuxcnc.org/docs/html/man/man9/hostmot2.9.html#UART
[18:35:37] <automata> that says to look at the mesa_uart subdriver for more details
[18:36:02] <automata> subdriver is loacted at: https://github.com/sittner/linuxcnc/blob/master/src/hal/drivers/mesa_uart.comp
[18:36:35] <Connor> ssi: If I do one, it'll be a 60w tube I think.. Just wondering how thick of MDF it could do.
[18:36:58] <ssi> Connor: I'll keep you informed throughout my experiments :)
[18:37:04] <ssi> Connor: there's a million variables
[18:37:11] <ssi> optics is a big one
[18:37:16] <ssi> and I'm still waiting on a 3" lens
[18:37:31] <ssi> the way my machine is built, a 4" lens is difficult to use
[18:37:41] <jdh> time for rev2
[18:37:55] <ssi> yeah I've thought about it
[18:38:00] <Connor> ssi Is that because you don't have Z movement ?
[18:38:06] <ssi> yeah
[18:38:12] <ssi> and there's not much space below the gantry carriage
[18:38:35] <Connor> Mine will have Z
[18:38:44] <automata> what is the fifo depth for the uart? Is the fifo depth configurable?
[18:40:24] -!- sheppard has quit [Quit: Ex-Chat]
[18:41:34] <automata> does the fifo depth have to be configured in firmware?
[18:41:43] <automata> in the FPGA?
[18:42:32] -!- asheppard_ has quit [Quit: Ex-Chat]
[18:45:47] <pcw_home> RX depth is 16 doublewords TX depth is 64 doublewords
[18:48:13] <diepchess> you go build a CNC milling machine, don't factor in the price of the electronics too much, then you go worry about it, and you figure out you're gonna get robbed...
[18:48:50] <pcw_home> sorry RX depth is on ly 16 bytes
[18:49:40] <pcw_home> for greater depth theres the PKTUART (2K deep max RX/TX) but different interface
[18:51:29] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[18:52:36] <ssi> diepchess: feel free to get cheap crap
[18:52:44] <ssi> we're just telling you from experience what works and what doesn't
[18:54:38] <diepchess> i raelly appreciate all the feedback
[18:56:25] -!- grummund has quit [Ping timeout: 260 seconds]
[18:56:43] <CaptHindsight> diepchess: you also need a PC with low enough latency to use software stepping, run the latency test on the LiveCD
[18:56:59] <diepchess> yeah everything i have is low latency
[18:57:10] <diepchess> yet we speak about mellanox infiniband cards then
[18:57:18] <diepchess> not parallel port equipped machines :)
[18:57:40] <jdh> low latency for RT?
[18:57:43] <ssi> a complete mesa interface kit costs about as much as a mach3 license
[18:57:49] <diepchess> RT is too slow for what i used to do
[18:58:01] <MrSunshine> https://www.youtube.com/watch?v=qMAv85za8MY anyone can figure out the operational principle of that machine?! =)
[18:58:03] <ssi> RT latency isn't the same thing as bandwidth
[18:58:08] <diepchess> RT kernel is locking with spinlock in kernel every transaction
[18:58:24] <diepchess> RT is not same thing as bandwidth
[18:58:35] <diepchess> RT schedules dead slow
[18:58:54] <ssi> MrSunshine: I dunno but I want it!
[18:59:02] <diepchess> those infiniband cards for example when connected to stock exchanges, you want transaction to be shipped within a microsecond or so (not yet ack back)
[18:59:02] <MrSunshine> me to :/
[18:59:19] <MrSunshine> then i can ask to get wood from the sides of the roads, just drive it home and throw it in the machine! =)
[18:59:27] <diepchess> RT kernel needs like order some dozens of microseconds
[18:59:35] <diepchess> which is dead slow
[18:59:50] <diepchess> that's why all those manufacturers make their own drivers that integrate into the kernel :)
[19:00:22] <diepchess> also the RT kernel can't avoid the runqueue latency by the way
[19:00:26] <diepchess> which is easily 10-20 milliseconds
[19:00:41] <diepchess> so if a thread is not scheduled yet, it has to wait
[19:01:23] <diepchess> newer server machines of course, the linux kernel, will do same thing like what IRIX and other OSes alraedy did do decades ago
[19:01:32] <mozmck> are you talking about preempt-RT?
[19:01:54] <diepchess> that's if you try to get a spinlock and do not manage to take it within say 600 ns, then the thread in question will be forced to idle
[19:02:05] <diepchess> which means you catch the runqueue latency to wake up again which is a minimum of 10 ms
[19:02:19] <automata> how do I use pktuart? and can you point out relevant firmware
[19:02:50] -!- spatialbrew has quit [Ping timeout: 272 seconds]
[19:02:50] <diepchess> of course using kernel 2.4, linuxcnc doesn't have to worry about that yet
[19:02:54] <automata> bit file for FPGA
[19:02:57] <jdh> imagine a 10mS latency for a stepgen
[19:03:31] <mozmck> diepchess: huh? linuxcnc is using 2.6 to 3.x kernels
[19:03:38] <rythmnbls> 2.4 is that a typo ?
[19:03:51] -!- f1oat [f1oat!~f1oat@AMontsouris-553-1-22-30.w92-151.abo.wanadoo.fr] has joined #linuxcnc
[19:04:00] <diepchess> oh that's what i read on homepage
[19:04:04] <diepchess> that it's kernel 2.4
[19:04:15] <jdh> and there is certainly no chance of a 10-20mS latency.
[19:04:17] <mozmck> whatever you read is a little out of date.
[19:04:45] <diepchess> jdh in linux you get the 10-20 ms latency if you have a thread that's forced back to the runqueue
[19:05:01] <diepchess> as it fires at a total outdated 100Hz
[19:05:08] <mozmck> people have run torture tests even on preempt-RT for days with nothing close to that latency
[19:05:18] <rythmnbls> or on RTAI
[19:05:44] <CaptHindsight> the current kernel on the wheezy iso is 3.4-5 pae RTAI
[19:05:45] -!- Jymmm [Jymmm!~jymmm@unaffiliated/jymmm] has joined #linuxcnc
[19:06:02] <diepchess> they don't run software that communicates to other hardware mozmck
[19:06:12] <mozmck> RTAI of course is better - max latency of 5000 ns is not uncommon.
[19:06:23] <CaptHindsight> we have RTAI working up to 3.14 but you'll have to build from source
[19:06:24] <diepchess> for example every packet shipped even RAW gets spinlocked by the kernel
[19:06:47] <mozmck> diepchess: who doesn't? I'm talking about tests using linuxcnc talking to mesa cards over ethernet
[19:07:08] <diepchess> can linuxcnc connect to ethernet?
[19:07:32] <mozmck> look at the mesa 7I80.
[19:07:57] <Connor> mozmck: That fully supported yet?
[19:08:12] <CaptHindsight> yeah hm2_eth is working
[19:08:19] <Jymmm> I thought drivers were stil srota beta
[19:08:23] -!- PetefromTn_ [PetefromTn_!~IceChat9@66-168-88-232.dhcp.kgpt.tn.charter.com] has joined #linuxcnc
[19:08:45] <Jymmm> sorta*
[19:08:45] <CaptHindsight> pretty sure they are in master
[19:09:33] <mozmck> the point is, no such latencies are necessary with either preempt-RT or for sure RTAI or Xenomai when properly set up.
[19:09:47] <diepchess> means you need dedicated box
[19:09:55] <diepchess> for linuxcnc
[19:09:59] <diepchess> which is ok
[19:10:27] <Connor> We always encourage dedicated box..
[19:10:37] <ssi> it's practical to have a dedicated box
[19:10:46] <ssi> and they're cheap
[19:10:52] <diepchess> i am running right now bunch of Xeon machines that are crunching at all 8 cores prime numbers
[19:10:59] <diepchess> i can easily add 3d printers
[19:11:08] <diepchess> as shipping some gcode to an arduino is no sweat of course
[19:11:16] <diepchess> meanwhile they are 100% loaded those machines with the crunching software
[19:11:23] <diepchess> i had hoped to do the same for the cnc mill :)
[19:11:53] <Connor> Yes, because your off-loading the toolpath processor to the arduino..
[19:12:06] <ssi> meanwhile you can't use your arduino to crunch primes
[19:12:08] <mozmck> When doing latency testing people generally run firefox, play games, watch movies etc.
[19:12:14] <ssi> so you have a dedicated computer for your 3d printer
[19:12:26] <ssi> it's the same thing
[19:12:37] <diepchess> no i don't have a dedicated machine for the 3d printer
[19:12:39] <mozmck> Not that it is recommended while cutting something...
[19:12:45] <ssi> yes you do
[19:12:47] <diepchess> it's running nonstop crunching software
[19:12:47] <ssi> the arduino
[19:13:16] <diepchess> i run the GUI on the Xeon box
[19:13:21] <SpeedEvil> The arduino to be fair has a problem crunching evens.
[19:13:22] <diepchess> so i can graphical see what happens
[19:13:30] <diepchess> and click with mouse how hot it is bla bla
[19:13:44] <ssi> so build your cnc mill with an arduino
[19:13:58] <diepchess> yes the arduino would be able to stand alone steer the cnc
[19:14:04] <ssi> sounds good
[19:14:06] <diepchess> however no good solution
[19:14:10] <diepchess> easier is use a mouse
[19:14:13] <ssi> head on over to #reprap and bug them about it
[19:14:39] <diepchess> i had hoped same is possible with cnc mill
[19:14:48] <ssi> it is possible
[19:15:10] <Connor> diepchess: It is possible.. but.. the gcode processors are limited.. and they're a PITA.
[19:15:21] <diepchess> oh i'm not bound to gcode
[19:15:32] <diepchess> is same solution possible with linuxcnc
[19:15:45] <ssi> in a manner of speaking
[19:15:46] <diepchess> that some embedded thing is running the linuxcnc
[19:15:54] <diepchess> and then a GUI runs on the Xeon box?
[19:16:10] <diepchess> where i can click with mouse?
[19:16:37] -!- tris has quit [Ping timeout: 272 seconds]
[19:16:57] <diepchess> from what i understand the embedded box needs a parallel port
[19:17:08] -!- tronwzrd [tronwzrd!~tronwizar@] has joined #linuxcnc
[19:17:12] <MattyMatt> there's a russian port to an STM32
[19:17:16] <Connor> linuxcnc = motion controller.. (Use to be called Enhanced Motion Controller).. in 3D printers, Arduino = motion controller.
[19:17:48] <MattyMatt> but STM32 can also run grbl, which is what smoothie is based on
[19:18:07] -!- tronwizard has quit [Ping timeout: 245 seconds]
[19:18:20] <jdh> all of this is to avoid having to use a dumpster PC?
[19:18:40] <MattyMatt> mine literally is :)
[19:18:53] <MattyMatt> A1400 clean inside just needed new psu
[19:19:01] <jdh> heh
[19:19:33] <ssi> jdh: exactly
[19:19:45] <ssi> tigerdirect sells off lease dells that make excellent dedicated machines for ~$200
[19:20:04] <jdh> sounds like linuxcnc is pretty much exactly what you don't want. for whatever reason.
[19:20:07] <MattyMatt> or a J1900 mobo for $50
[19:20:14] -!- tris [tris!tristan@2001:1868:a00a::4] has joined #linuxcnc
[19:20:29] <ssi> jdh: I've had this exact same argument with countless manufacturers of "cnc hardware"
[19:20:36] <jdh> yeah, and here.
[19:20:46] <jdh> usually involves an arduino though.
[19:20:52] <SpeedEvil> Or an edison.
[19:20:54] <ssi> for some reason the people that come up from the reprap trenches are just in love with the idea of offloading everything to an arduino
[19:21:21] <MattyMatt> arduino is still the convenient way to get a/d for linuxemc
[19:21:38] <diepchess> it's not about the arduino, that's a junk thing (16Mhz huh)
[19:21:46] <diepchess> it's about how you connect to it.
[19:21:56] <ssi> sure
[19:22:02] <ssi> and there's good ways to do that
[19:22:22] <MattyMatt> so if you're running something with thermistors on linuxemc, you still need an arduino, unless someone's figured out how to read the audio line in
[19:22:47] <jdh> you mean a cheap way?
[19:22:54] <PCW> just use a modulator
[19:23:00] <MattyMatt> yeah mesa is overkill
[19:23:20] <MattyMatt> modulator like a 55 timer driven PCjunior joystick port?
[19:24:03] <PCW> like a modulator (multiply by 1 or -1)
[19:24:31] <MattyMatt> mesa is overkill for a reprap of course, not in all cicumstances :)
[19:24:41] <PCW> so you get a AC signal with the amplitude of your original DC signal)
[19:25:01] <diepchess> anything you have to buy yourself is overkill for a riprap anywy
[19:25:12] <ssi> but we're not talking about repraps
[19:25:21] <MattyMatt> yep repraps should be made entirely of junk
[19:25:51] <MattyMatt> it's kinda the point
[19:27:06] <MattyMatt> PCW you'd need a comparitor or just let a logic pin do that?
[19:27:07] <diepchess> i want to build a simple CNC, yet as it seems, i need some ancient parallel port once again. Even the printer here has its own ethernet address for 12+ years now...
[19:27:33] <ssi> you don't need a parallel port
[19:27:33] <PCW> just logic/analog
[19:27:37] <ssi> I don't know how many ways to say it
[19:28:26] <MattyMatt> I got a dual parallel card for $12
[19:28:29] <SpeedEvil> diepchess: modern motherboards often have them even
[19:28:30] <PCW> A parallel port just gives you a low latency way to do I/O directly from the PC
[19:28:39] <SpeedEvil> http://www.ebuyer.com/630828-gigabyte-ga-j1900n-d3v-intel-celeron-j1900-2ghz-vga-dvi-hd-audio-mini-itx-ga-j1900n-d3v
[19:28:45] -!- micges-dev has quit [Ping timeout: 264 seconds]
[19:29:47] <diepchess> let's not get into a motherboard discussion here.
[19:30:26] <PCW> Almost all the J1800/J1900 MBs have parallel ports (sometimes just headers)
[19:30:31] <MattyMatt> http://www.ebay.co.uk/itm/NEW-Siig-Dual-Profile-PCI-Express-Parallel-Port-Adapter-JJ-E01011-S1-/281359685477
[19:30:43] <ssi> what exactly is your goal?
[19:30:43] <MattyMatt> and 2 more if you want em :)
[19:30:48] <ssi> to spend as little money as possible?
[19:30:50] <PetefromTn_> jeez we still talking about parallel ports >
[19:30:56] -!- Tecan [Tecan!~fasdf@unaffiliated/unit41] has joined #linuxcnc
[19:30:59] <jdh> Pete: not still... again.
[19:31:09] <ssi> or are you actually philosophically tied to using or not using specific kindso f hardware
[19:31:33] <PetefromTn_> just say die.... the manufacturers did LOL
[19:33:02] <MattyMatt> it's convenient and cheap cabling
[19:33:17] <diepchess> i assume no one here has tried to run linuxcnc under Xen?
[19:33:28] <MattyMatt> compared to 50 pin headers needing ribbon cables
[19:33:29] <diepchess> well there is always a first time :)
[19:34:01] <rythmnbls> I've run the iso under xen, just to grab kernel configs, not to do latency checks
[19:36:26] <jdh> I've run the sim in a vm
[19:37:39] <diepchess> here: it has a board with lpt port. Can i connect this to linuxcnc?
[19:37:51] <diepchess> the link http://www.ebay.com/itm/Germany-Ship-Promote-4Axis-Nema-23-Stepper-Motor-290oz-in-Driver-CNC-/271436341842?pt=LH_DefaultDomain_0&hash=item3f32dde252
[19:39:34] <jdh> it's a sad board, but yes.
[19:39:38] <ssi> lel
[19:39:50] <diepchess> what sort of board is it?
[19:39:54] <diepchess> doesn't deliver 3.5A?
[19:40:30] <MattyMatt> TB6560 is a sick chip, and that board doesn't help
[19:40:42] -!- norbert__ has quit [Quit: Konversation terminated!]
[19:40:52] <diepchess> oh dear all those online offers use this tb6560
[19:40:58] <diepchess> only the leadshines you like?
[19:41:07] <ssi> leadshines are decent
[19:41:12] <MattyMatt> it's solid, but falls over if you go too fast, and explodes if your psu is even a sniff too high volts
[19:41:13] <ssi> tb6560 are junk
[19:41:33] <rythmnbls> nicely packaged smoke
[19:41:42] <jdh> on the bright side, you can get the boards dirt cheap
[19:41:43] <diepchess> this box however http://www.automationtechnologiesinc.com/products-page/digital-stepper-motor-driver/3-axis-dsp-based-digital-stepper-drive-max-60-vdc-6-0a
[19:41:55] <MattyMatt> TB6600 is fine, on paper. the faults are gone from the datasheet. the DRV-268 uses those
[19:41:56] <diepchess> it's 260 dollar. and you only have 1 piece of the puzzle
[19:42:22] <diepchess> if i use the A4982 to steer somehow the steppers is that an ok driver?
[19:42:25] <PCW> The Leadshine MX3660 looks nice
[19:42:27] <diepchess> texas instruments
[19:42:29] <jdh> it is 3 decent drivers and a breakout board.
[19:43:05] <Connor> MX3660 is good. combine with pport or 5i25/6i25 and you have a winner.
[19:43:10] <jdh> what is your goal?
[19:43:32] <diepchess> simple alu milling cnc machine
[19:44:34] <diepchess> microstepping not done ok by the tb6560 driver?
[19:44:59] asheppard is now known as sheppard
[19:45:02] <MattyMatt> yeah although you can't afford to use it much, with the horrid stepping rates you can achieve
[19:45:29] <jdh> it would be good to match the level of drivers/etc to the level of machine they will be driving.
[19:45:31] <MattyMatt> I can get 200 rpm max out of my motors
[19:45:54] <jdh> a tb6560 is probably fine for a huge class of machines. Especially first machines.
[19:45:56] <MattyMatt> with M8 leadscrews, it's painful
[19:46:01] <diepchess> jdh in that case i would use a file and hacksaw to cut out a driver board from junkwood drifting here on the rhine river
[19:46:24] tjb11 is now known as tjb1
[19:46:32] <jdh> tool selection is also important. I wouldn't use a hacksaw on wood.
[19:46:55] <MattyMatt> unless there's nails in it
[19:47:44] <jdh> so, sounds like a tb6560 + arduino being drip-fed over infiniband from a xeon is the way to go.
[19:48:10] <ssi> jdh: I bought a tb6560 4axis kit for my first machine, and it let the smoke out pretty quickly
[19:48:12] <MattyMatt> jdh that's almost exactly how my printer was :)
[19:48:14] <ssi> I soon replaced it with a g540
[19:48:32] <jdh> ssi: I went to a xylotex. Not much better, but it still works fine.
[19:48:38] <MattyMatt> http://imgur.com/a/id9ye#0
[19:48:52] <diepchess> what was the problem you observed in the produced objects ssi?
[19:49:02] -!- tmcw has quit []
[19:49:33] <ssi> the problem I observed was a burned pc board and magic smoke in the air, and zero motion
[19:50:17] <jdh> was it preceded by an incredibly annoying high pitched whine?
[19:50:35] <ssi> I don't recall
[19:50:39] <ssi> it's been like six years now
[19:50:39] <MattyMatt> I messed one up bypassing the optos, but they've been solid on 12V, stolid rather than solid I should say
[19:51:10] <ssi> these days I pretty much go straight to 48V, leadshine drives, and 7i76 kit
[19:51:19] <jdh> but, you are rich
[19:51:22] <ssi> it's not the cheapest way to go, but I know it's going to work well
[19:51:25] <ssi> yeah, so rich
[19:51:30] -!- motioncontrol [motioncontrol!~io@host54-73-dynamic.246-95-r.retail.telecomitalia.it] has joined #linuxcnc
[19:51:32] <ssi> heh
[19:51:37] <ssi> if I was rich I'd buy a tormach
[19:51:46] <jdh> with super steppers!
[19:51:54] -!- skunkworks has quit [Read error: Connection reset by peer]
[19:51:55] <jdh> I would too. steppers and all.
[19:51:55] <diepchess> if i was rich i'd be laying in the sun on a beach
[19:52:02] <ssi> not me
[19:52:08] <jdh> teh beach is half a mile away
[19:52:17] <diepchess> Scotland?
[19:52:40] <jdh> does scotland have beaches? sounds cold and rocky
[19:52:52] <ssi> lol
[19:53:03] <jdh> beach water temp is still 28c here
[19:53:10] <ssi> I just revised my estimate downward for my friends
[19:53:24] <diepchess> yeah 200 dollar is too much says my bank account
[19:53:27] <ssi> I told them to get that mx3660 and a 5i25
[19:53:30] <diepchess> i have to look for other solution regrettably
[19:53:34] <diepchess> and 3 axis solution not 4
[19:53:37] <MattyMatt> has anyone tried those DRV-268 TB6600 drivers yet? I've got 2 here I got super cheap but not tried yet, and they're cheap enough anyway
[19:53:43] -!- chillly has quit [Quit: Ex-Chat]
[19:54:15] -!- dr0w has quit [Quit: Leaving]
[19:54:16] <MattyMatt> DIV268N sorry
[19:54:20] -!- patrickarlt has quit [Remote host closed the connection]
[19:54:39] <MattyMatt> http://www.ebay.com/itm/HY-DIV268N-5A-Single-Axis-5A-CNC-Stepper-Driver-for-CNC-Milling-Machine-ge-/281115000196
[19:54:57] <Connor> I tell you what is the hard part.. picking the stepper.. and understanding the torque and inductance etc etc. pick the wrong stepper.. and you end up with stepper that can't go very fast without stalling out.
[19:55:13] <diepchess> wow those are just 10 euro MattyMatt
[19:55:46] <Connor> MattyMatt: Looks like a leadshine knock off.
[19:55:46] <diepchess> here MattyMatt http://www.ebay.co.uk/itm/0-2-5A-Single-Axis-TB6600-CNC-Engraving-Machine-Stepper-Motor-Driver-Controller-/121364943800?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item1c41ea13b8
[19:55:51] <ssi> Connor: it's not hard once you know the secret :)
[19:56:01] <jdh> I never did open up my chinese 6040 drivers to see what was in there.
[19:56:04] <MattyMatt> they're getting cheaper all the time :) either they're rubbish, or toshiba's reputation hasn't recovered yet
[19:56:19] <ssi> I usually end up with steppers around 2.5mH inductance, and run 48V
[19:56:21] <diepchess> how many amps would you advice Connor?
[19:56:26] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[19:56:49] <Connor> diepchess: For the Steppers, or the drivers ?
[19:56:57] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[19:56:58] <ssi> amazon carries a meanwell 48V switcher for $56 now
[19:57:00] <MattyMatt> on paper tho, I repeat. the TB6560 faults are gone
[19:57:03] <diepchess> my budget is tad limited Connor
[19:57:12] <diepchess> the steppers
[19:57:31] <Connor> 3Amp for a small machine is pretty typical.
[19:57:46] <jdh> find a junked large printer, steal parts.
[19:58:00] <ssi> Connor: if I come up there sometime soon, will you help me put my g0704 spindle back together, and get that dang pulley made at pete's?
[19:58:10] <diepchess> yeah i was aiming for 3 amps, though i'm looking at my lathe now whether it is worth producing my own linear rails somehow
[19:58:14] <MattyMatt> hell £10 that's even less than I paid, but I did get a free breakout board. it's some guy who bought a 3 axis kit and blew up one driver and quit
[19:58:51] <Connor> ssi A) I can help, but, probably need to be a pete's as he has the press. B) Need to make the pulley to match your spindle.
[19:59:04] <ssi> well let's find the time to go to pete's
[19:59:05] <Connor> We just made the 1 for mine.. and had my spindle on site.
[19:59:07] <ssi> I'll fly up with the spindle
[19:59:19] <ssi> I put the new bearings on it but I could never get it to have no play
[19:59:23] -!- patrickarlt has quit [Ping timeout: 240 seconds]
[19:59:41] <SpeedEvil> ssi: hot-glue
[20:00:01] <ssi> seems legit
[20:00:36] <ssi> I sold my clausing
[20:00:41] <ssi> I have some room for another machine now
[20:00:44] <ssi> I really need some kind of mill
[20:00:59] <diepchess> 5 axis ssi
[20:01:03] <diepchess> easy to build
[20:01:10] <ssi> tell you what
[20:01:18] <ssi> build me a 5 axis mill, and I'll provide your electronics
[20:01:29] <diepchess> problem is getting them accurately working
[20:01:40] <ssi> I don't want to deal with 5 axis cam :P
[20:01:41] -!- anth0ny has quit [Quit: anth0ny]
[20:01:56] <PetefromTn_> ssi Anytime you guys want to come over and work on those pullies and whatnot you are more than welcome.
[20:02:08] <diepchess> i was aiming for 4 axis ssi
[20:02:14] <diepchess> yet the 4th being a C axis on the bed
[20:02:24] * MattyMatt dumps rocks and barrel of epoxy at ssi's. some assembly required
[20:03:04] <MattyMatt> mm barrel of epoxy, I wish
[20:03:12] <MattyMatt> real epoxy is spendy
[20:03:45] <MattyMatt> £45 l was the best price I've found for assuredly pure stuff
[20:03:51] <MattyMatt> per litre
[20:04:18] <MattyMatt> I've seen plenty of good machines made with £10 a litre stuff tho
[20:04:38] <diepchess> i investigated 5 axis ssi - it's not easy to produce it accurately. it's easy to build one though. it will sit idle however.
[20:04:48] <MattyMatt> like the famous german machine. €12/l for that
[20:04:58] <SpeedEvil> I wonder about PU
[20:05:34] <ssi> the epoxy I use for composite work is around $115/gallon
[20:05:35] -!- anth0ny has quit [Client Quit]
[20:05:42] <ssi> west systems
[20:05:44] <MattyMatt> does the german machine have a proper name?
[20:05:45] <ssi> it's good stuff
[20:05:58] <MattyMatt> ssi yep that's my 2nd choice
[20:06:04] <MattyMatt> available in UK
[20:06:28] <ssi> it's easy to work with because of the volumetric ratio pumps they sell for it
[20:06:47] <SpeedEvil> Err - polyester
[20:07:04] <ssi> polyester resin is cheaper... not sure it gets as hard though
[20:07:10] <SpeedEvil> http://www.cfsnet.co.uk/acatalog/General_Purpose_Resin.html
[20:07:16] <SpeedEvil> 220kg/450 quid
[20:07:21] <Loetmichel> soooo, the tinkerPC has a new psu... thjat was a cheap buy.. got a new returned one for 13 eur instead of 45 eur new. got it today: no fuction... opened it: as i thought, just the primary filter choke ripped loose. resoldered it: perfect 530W psu for 13eur (+4.30 pp) ;-) -> http://www.cyrom.org/palbum/main.php?g2_itemId=15271
[20:07:31] <ssi> vinylester might be a better choice, but it requires MEKP which scares the hell out of me
[20:07:37] <MattyMatt> polyester requires thinners, and tyhinners are bad that spoil the zero cure shrinkage
[20:07:40] <SpeedEvil> http://www.cfsnet.co.uk/acatalog/West_Resin.html
[20:07:46] <SpeedEvil> MattyMatt: true, I guess
[20:08:36] <diepchess> would it be interesting to buy TB6600 drivers for 3 motors and a break out board ?
[20:08:37] <MattyMatt> if the rocks are tight packed, and you allow for a bit of shrinkage, it'd perform OK I guess
[20:08:39] <Loetmichel> ... aaaand: nor more shutoffs/chirping when gaming Elite ;-)
[20:08:47] <CaptHindsight> epoxy is only 10-15% of the mix
[20:09:06] <SpeedEvil> CaptHindsight: I've been wondering about something related
[20:09:12] <MattyMatt> diepchess yeah that's what I got. you can have the breakout board if you like
[20:09:21] <SpeedEvil> CaptHindsight: concrete mixing under vacuum - for really, really low cement content
[20:09:26] <SpeedEvil> CaptHindsight: well - steam
[20:10:13] <diepchess> wow cool MattyMatt let me message you my mail address
[20:10:25] <CaptHindsight> acrylated concrete
[20:10:40] Guest19842 is now known as phantoxeD
[20:11:05] <CaptHindsight> there are published studies of their qualities for machine bases and frames
[20:11:38] <CaptHindsight> acrylated concretes, resin aggregate mixes etc
[20:12:46] <CaptHindsight> I doubt that any DIYer is going to see much difference between using marine polyester resin and epoxy
[20:13:40] <ssi> what's the benefit of using epoxy in aggregate rather than just concrete in aggregate?
[20:13:47] <ssi> er, cement in aggregate rather
[20:13:51] <CaptHindsight> polyesters and epoxies come in all kinds of hardness, tensile and compressive strengths
[20:13:52] <SpeedEvil> I'm in the mddle of working out how to cast high performance reinforced panels
[20:14:11] <CaptHindsight> the epoxy adds some dampening
[20:14:19] <SpeedEvil> Fibreglass 'mesh' - stressed at ~1 ton/m, with 2cm panel poured round it
[20:14:19] <CaptHindsight> since it can flex a little
[20:14:30] <SpeedEvil> with integral decoration
[20:14:44] <ssi> I'd be interested in trying to build a mill using something like that
[20:19:15] <CaptHindsight> testing out acrylated concrete is pretty low cost
[20:20:16] <CaptHindsight> Home Depot and similar sell high compressive strength concrete mixes and also gallon jugs of acrylate mix
[20:21:14] -!- MC500 [MC500!43352076@gateway/web/freenode/ip.] has joined #linuxcnc
[20:21:29] <CaptHindsight> http://www.sakrete.com/products/detail.cfm/prod_alias/5000-Plus-Concrete
[20:21:35] -!- Roguish [Roguish!~chatzilla@c-67-188-44-176.hsd1.ca.comcast.net] has joined #linuxcnc
[20:21:53] -!- skunkworks_ [skunkworks_!~chatzilla@str-broadband-ccmts-ws-26.dsl.airstreamcomm.net] has joined #linuxcnc
[20:21:55] <CaptHindsight> http://www.sakrete.com/products/detail.cfm/prod_alias/Maximizer
[20:22:17] -!- SpeedEvil has quit [Ping timeout: 245 seconds]
[20:22:49] <CaptHindsight> http://www.sakrete.com/products/detail.cfm/prod_alias/Concrete-Bonder-Fortifier
[20:23:45] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[20:23:48] <CaptHindsight> http://www.quikrete.com/productlines/ConcreteAcrylicFortifier.asp
[20:23:50] -!- patrickarlt has quit [Remote host closed the connection]
[20:24:11] <MC500> I had a problem with my DAC, I noticed I had to set P in my PID loop to a negative number for it to work. I fxed it by changing my OUTPUT_SCALE from 1 to -1 wich is the same resolution but negative now, is that the right way to fix it?
[20:25:36] <PCW> probably, the PID comp doesn't like negative I terms so you cant just invert all the PID terms without trouble
[20:25:49] -!- rythmnbls has quit [Quit: Leaving]
[20:25:57] -!- Nick001-shop has quit [Read error: Connection reset by peer]
[20:26:25] -!- Nick001-shop [Nick001-shop!~chatzilla@50-32-101-83.adr01.dlls.pa.frontiernet.net] has joined #linuxcnc
[20:26:45] <MC500> well the axis is rotary and it now spins backwards to what it was.. + is right hand - is left... dosent matter but is that right???
[20:27:13] <MC500> I kinda think its right.. but dono
[20:28:02] -!- kfoltman [kfoltman!~kfoltman@] has joined #linuxcnc
[20:29:17] -!- patrickarlt has quit [Ping timeout: 260 seconds]
[20:29:57] -!- jasen has quit [Quit: Page closed]
[20:30:24] <CaptHindsight> http://www.cnczone.com/forums/epoxy-granite/ one discussion has 4,844 replies
[20:31:35] <roycroft> i remember reading part of that thread
[20:31:47] <roycroft> i think well before it hit 4k+ posts
[20:31:57] -!- automata has quit [Ping timeout: 245 seconds]
[20:32:00] * roycroft doesn't have a spare week to read one thread
[20:32:41] -!- automata_ has quit [Client Quit]
[20:32:43] -!- anth0ny has quit [Client Quit]
[20:33:23] <CaptHindsight> "when you exceed 8% (epoxy), cost for the epoxy keeps rising without adding much strength."
[20:34:38] <CaptHindsight> and degas using a vacuum bag, were the two main lessons learned
[20:37:42] -!- automata_ [automata_!~Amit@triband-mum-] has joined #linuxcnc
[20:37:50] <ssi> CaptHindsight: any thoughts on designing machines using that process?
[20:39:06] -!- anarchos2 has quit [Ping timeout: 272 seconds]
[20:41:41] <CaptHindsight> http://www.youtube.com/watch?v=dHzkeK_AXH0
[20:42:02] -!- tris has quit [Ping timeout: 272 seconds]
[20:42:22] -!- tris [tris!tristan@2001:1868:a00a::4] has joined #linuxcnc
[20:43:32] <CaptHindsight> http://www.icecoldcomputing.com/misc/portaalfrees_Zkolom_resonantie.png "Resonance is measured using a 5um linear scale (using linear scales in the feedback loop), Doing the same with EC shows a flat line..."
[20:46:35] <Connor> ssi, chk pvt msg
[20:47:31] -!- zeitue [zeitue!~z@] has joined #linuxcnc
[20:54:44] <CaptHindsight> Repeatability and precision of movement to 0.0005" and better - objectives for MadVac CNC http://oneoceankayaks.com/madvac/madvac_index.htm
[20:54:58] <CaptHindsight> PetefromTn_: ^^ similar to the machine you liked earlier
[20:55:29] <PetefromTn_> nice man...
[20:55:33] <PetefromTn_> I need two of those LOL
[20:55:47] <CaptHindsight> but they used Mach2 :(
[20:57:08] <PetefromTn_> BASTARDS!!
[20:57:09] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[20:57:32] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[20:57:55] -!- zeitue has quit [Quit: Leaving]
[20:58:23] -!- sirdancealot has quit [Ping timeout: 240 seconds]
[20:59:35] -!- FinboySlick has quit [Quit: Leaving.]
[21:03:04] * JT-Shop needs to kick the shipping managers ass... that place is a mess
[21:03:10] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[21:05:01] <Jymmm> I'd fire the bastard, but slaves need to be sold, not fired.
[21:05:14] -!- anarchos2 [anarchos2!~mike@S010600259ce59399.vn.shawcable.net] has joined #linuxcnc
[21:07:28] <JT-Shop> I can't even dock his pay... he works for free
[21:08:45] <Jymmm> Maybe take away his beer for a month?
[21:09:48] <JT-Shop> that would only be a 30 pack of Corona
[21:09:56] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[21:10:19] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[21:10:21] anth0ny_ is now known as anth0ny
[21:10:43] -!- anth0ny has quit [Client Quit]
[21:15:20] -!- a_morale has quit [Ping timeout: 250 seconds]
[21:15:23] -!- f1oat has quit [Ping timeout: 240 seconds]
[21:16:41] -!- a_morale [a_morale!~quassel@2-227-115-13.ip186.fastwebnet.it] has joined #linuxcnc
[21:18:34] postaL is now known as postaL_offline
[21:21:31] * likevinyl is away: aburrís.
[21:24:41] <Deejay> gn8
[21:25:00] <PetefromTn_> GN8 DEEJAY!!
[21:25:18] <Deejay> greetings pete :)
[21:25:23] <Deejay> have a nice day
[21:25:47] -!- Deejay has quit [Quit: bye]
[21:25:49] <PetefromTn_> you too man.
[21:27:27] -!- bedah has quit [Quit: Ex-Chat]
[21:27:43] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[21:28:10] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[21:29:58] -!- chillly has quit [Quit: Ex-Chat]
[21:32:25] <JT-Shop> PetefromTn_, now I have to re-quote a quote from March! I hate making quotes
[21:32:35] -!- FinboySlick [FinboySlick!~shark@squal.net] has joined #linuxcnc
[21:33:10] <PetefromTn_> well at least you have quotes to make. this is my first real quote and I hope I didn't screw myself with it.
[21:36:14] <JT-Shop> if you don't get rich from it you got some experience and you always have to pay for that
[21:36:54] <PetefromTn_> RICH??? Hell I would be happy to just make a few bucks here and there really...
[21:38:03] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[21:38:27] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[21:38:33] -!- sylphiae [sylphiae!~sylphiae@99-7-4-250.lightspeed.cicril.sbcglobal.net] has joined #linuxcnc
[21:39:36] <JT-Shop> yea, I have to laugh when people say "I hope you get rich making them"
[21:42:01] -!- motioncontrol has quit [Quit: Sto andando via]
[21:43:52] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[21:44:15] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[21:46:01] <ssi> I need to write a program which will cut a series of lines at varying powers, feedrates, and ppis
[21:46:07] <ssi> and focus heights
[21:46:12] <ssi> too many damn variables! :(
[21:46:16] <ssi> then I have to do that for every material
[21:46:16] <ssi> heh
[21:50:40] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[21:51:06] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[21:51:46] -!- sylphiae has quit [Read error: Connection reset by peer]
[21:54:35] <JT-Shop> heh
[21:54:54] -!- kfoltman has quit [Quit: Ex-Chat]
[22:01:59] -!- sylphiae [sylphiae!~sylphiae@99-7-4-250.lightspeed.cicril.sbcglobal.net] has joined #linuxcnc
[22:03:58] -!- automata_ has quit [Ping timeout: 272 seconds]
[22:07:15] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[22:11:30] -!- anth0ny_ has quit [Quit: anth0ny_]
[22:13:32] -!- sumpfralle has quit [Ping timeout: 245 seconds]
[22:14:38] -!- sumpfralle [sumpfralle!~lars@p578a641d.dip0.t-ipconnect.de] has joined #linuxcnc
[22:15:41] -!- mle has quit [Ping timeout: 260 seconds]
[22:22:23] -!- anth0ny_ has quit [Client Quit]
[22:23:09] -!- PetefromTn_ has quit [Ping timeout: 260 seconds]
[22:27:49] <diepchess> on youtube i see a guy load the tb6600's with no more than 2.5A
[22:29:04] -!- toastydeath [toastydeath!~toast@c-69-244-86-195.hsd1.md.comcast.net] has joined #linuxcnc
[22:36:37] -!- Cylly [Cylly!cylly@p54B11C63.dip0.t-ipconnect.de] has joined #linuxcnc
[22:37:32] -!- Loetmichel has quit [Ping timeout: 276 seconds]
[22:39:31] <diepchess> heh do i see correct that for the X Y movement i want low inductance and for the Z movement i want high inductance stepper?
[22:44:33] -!- arturia [arturia!~sylphiae@99-7-4-250.lightspeed.cicril.sbcglobal.net] has joined #linuxcnc
[22:47:15] <Tom_itx> mmm i guess he left...
[22:47:42] -!- sylphiae has quit [Ping timeout: 252 seconds]
[22:47:50] <Tom_itx> the do have software for making quotes for machine time...
[22:49:11] arturia is now known as sylphiae
[22:49:16] -!- Tecan has quit [Quit: Live Long And Phosphor!]
[22:51:18] <ssi> ok, I'm running an elaborate line test now
[22:51:23] <ssi> hopefully this'll give me some information
[22:51:53] -!- gonzo_nb has quit [Remote host closed the connection]
[22:54:19] * JT-Shop is tired...
[22:55:50] * Tom_itx is too
[22:56:17] * Tom_itx passes out the first round of beer
[22:57:04] * JT-Shop is too tired for beer and goes straight for a glass of red
[23:00:48] <ssi> holy balls this is a lot of lines
[23:00:58] <ssi> and I have to do it in steps of focus height too
[23:02:09] <ssi> https://pbs.twimg.com/media/BxNQFD7IIAA9jKt.png:large
[23:03:25] <ssi> each block is ~30 lines, from 10 to 10,000 ppi in decades
[23:03:52] <ssi> the rows of blocks are feedrates from 5 to 100ipm in 10ipm steps
[23:04:04] <ssi> and then each row is a different power in 2mA steps from 8 to 28
[23:04:28] <ssi> I am going to repeat this whole process something like eleven times for focus heights from .100 to .500 in .050 steps
[23:04:39] <ssi> and then do all of the above for different materials
[23:05:15] <JT-Shop> nice a cold front is coming in tonight
[23:05:29] -!- sumpfralle has quit [Ping timeout: 276 seconds]
[23:06:18] <JT-Shop> ssi, what is that?
[23:06:44] <ssi> JT-Shop: I'm trying to find the most efficient combination of power, feedrate, ppi, and focus height for a given material
[23:07:15] <JT-Shop> ah that makes sense
[23:07:18] <ssi> rather than just floundering around trying things, I'm doing a huge ridiculous empirical test
[23:09:36] <toastydeath> ?
[23:10:09] <toastydeath> all those tests have already been done
[23:10:23] <ssi> I'm all ears
[23:11:12] <toastydeath> provided you have enough gullet clearance, the energy required to remove a given volume is minimized around .020" per tooth
[23:11:24] <ssi> lazrbeemz has no teeths
[23:11:27] <toastydeath> oh
[23:11:29] <toastydeath> laser
[23:11:31] <toastydeath> sorry, carry on
[23:11:33] <ssi> :)
[23:11:42] -!- sumpfralle [sumpfralle!~lars@p578a641d.dip0.t-ipconnect.de] has joined #linuxcnc
[23:11:57] <ssi> there's so many freakin variables in laser, it's hard to keep a hold on
[23:12:11] -!- patrickarlt has quit [Remote host closed the connection]
[23:13:05] <ssi> but that chipload figure is a good one to store in the back of my brain
[23:14:29] * JT-Shop thinks you need to paint some shark teeth on your laser
[23:15:37] postaL_offline is now known as postaL
[23:15:42] -!- patrickarlt has quit [Read error: Connection reset by peer]
[23:16:09] <ssi> :)
[23:18:41] -!- almccon has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
[23:23:00] -!- micges-dev has quit [Quit: Wychodzi]
[23:24:20] -!- rob_h has quit [Ping timeout: 276 seconds]
[23:26:16] -!- Valen [Valen!~Valen@c211-30-128-202.blktn6.nsw.optusnet.com.au] has joined #linuxcnc
[23:27:05] -!- asdfasd has quit [Ping timeout: 260 seconds]
[23:27:55] <JT-Shop> last batch out of the powder coat oven and I'm ready to veg...
[23:27:59] <JT-Shop> goodnight all
[23:28:53] -!- sumpfralle has quit [Ping timeout: 276 seconds]
[23:30:13] -!- mhaberler [mhaberler!~mhaberler@194-96-112-2.adsl.highway.telekom.at] has joined #linuxcnc
[23:34:23] -!- mhaberler has quit [Ping timeout: 240 seconds]
[23:35:10] <Tom_itx> toastydeath, .020" chip load seems like quite a bit
[23:35:19] <toastydeath> yep
[23:35:40] -!- patricka_ has quit [Remote host closed the connection]
[23:35:58] <Tom_itx> most calculators for the average materials figure it around .002-010" per tooth
[23:36:15] <toastydeath> yep
[23:36:19] <Tom_itx> depending on your spindle HP
[23:36:23] -!- patrickarlt has quit [Remote host closed the connection]
[23:36:47] <Tom_itx> if i tried that on my sherline the spindle would stop :D
[23:36:49] <toastydeath> yep
[23:37:06] <toastydeath> and so your sherline can't approach maximal efficiency in terms of volume removed per minute
[23:37:24] <toastydeath> it continues downward but the gains are vanishing past .020"
[23:37:51] <ssi> my lathes can take WAY more than that as a chipload
[23:38:01] <ssi> .125" at a time isn't unheard of
[23:38:14] <Tom_itx> we're not talking lathe here... at least i'm not
[23:38:23] <toastydeath> correct, and that's the correct way to run a roughing operation
[23:38:31] <Tom_itx> i've seen deeper cuts than that on a lathe
[23:38:40] <ssi> yeah, stouter lathes than mine
[23:38:40] <toastydeath> you want to minimize the number of shear planes per inch
[23:39:00] <Tom_itx> i've also seen .250 cuts in steel with a shell mill but i'm not sure what the chip load was
[23:39:04] <toastydeath> the reason it flattens out is because the energy required to deform the chip rises
[23:39:21] <toastydeath> you have energy required to shear + energy required to deform
[23:40:54] <toastydeath> and it's the same for mills and lathes - on mills you get choked out sooner because of the flute depth limitation
[23:40:55] <Tom_itx> i've seen those high speed videos... i wonder if it would look the same for a roughing cutter
[23:40:56] <toastydeath> big shell mills usually don't have that limitation and behave more like a lathe
[23:41:11] <toastydeath> roughing cutters work exactly on that principle
[23:41:36] <toastydeath> by taking a smaller percentage of the advancing face per flute, by staggering what areas it cuts
[23:41:37] <Tom_itx> i know the chips were coming off white
[23:41:46] <toastydeath> high speed machining is different
[23:41:55] <Tom_itx> and sticking to the enclosure paint
[23:42:07] <toastydeath> high speed machining is not usually hp-efficient
[23:42:17] -!- patrickarlt has quit [Remote host closed the connection]
[23:42:26] -!- The_Ball has quit [Ping timeout: 255 seconds]
[23:42:33] <toastydeath> because you are melting material as in a grinder rather than shearing
[23:42:45] <toastydeath> for heavy roughing operations high speed machining is not used
[23:42:55] <Tom_itx> that's why generally a high speed spindle is lower hp
[23:43:04] <Tom_itx> made for aluminum not steel
[23:43:35] <Tom_itx> or lower torque
[23:44:19] <toastydeath> dunno what machines you're referring to, a lot of high speed spindles are 30-50 hp
[23:44:19] <toastydeath> and that's plenty
[23:44:48] <toastydeath> it depends on if someone is doing "high speed machining" on a budget VMC, trying to get by
[23:45:01] <toastydeath> or on an actual machine built for it
[23:45:18] <toastydeath> hsk spindle, 50+ hp at 20-30k rpm
[23:45:25] <Tom_itx> the Okumas we had didn't have 30 Hp spindles
[23:45:44] <Tom_itx> not HUGE machines but not small by any means
[23:45:56] <toastydeath> what type of tooling
[23:46:01] <CaptHindsight> we need to invent those phasers on Star Trek, move past all this cave man metal on metal
[23:46:08] <Tom_itx> 50 taper
[23:46:11] <Tom_itx> all sorts of tooling
[23:46:26] <Tom_itx> lots of insert cutters
[23:46:53] <toastydeath> unless a machine has hsk or a similar holder, it may be a fantastic machine, but is not really a high speed milling center
[23:47:06] <toastydeath> cat-50 is fantastic and is obviously super rigid, but I'd bet that machine was better with a big face cutter running slower than as a high speed machine
[23:47:21] <toastydeath> even if it had a 12k spindle or something
[23:47:36] <Tom_itx> i'd say probably half of em were 40 taper
[23:47:37] -!- patrickarlt has quit [Ping timeout: 260 seconds]
[23:47:54] <toastydeath> all nmtb-derivative tapers have trouble at speed
[23:47:58] <Tom_itx> i don't think we ever ran 12k
[23:49:26] <Tom_itx> i know at least one of them had a high / low range
[23:50:22] <toastydeath> right, generally high speed machines will be some low number like 3-5 hp until you hit some absurd number
[23:50:23] <toastydeath> and the motor drive switches over
[23:50:23] <toastydeath> geared machines (hi/lo range) are another sign of a non-high-speed machine
[23:50:40] <toastydeath> (good machines, to be sure - every mill i've ever touched that had a geared head was fantastic)
[23:51:25] <toastydeath> one day i'd like to buy a small mori seiki
[23:51:35] <toastydeath> one of their older machines without an enclosure
[23:51:35] <Tom_itx> if i only had room :D
[23:51:44] <toastydeath> same
[23:51:45] <Tom_itx> yeah they're pretty rugged
[23:56:27] <Valen> wire edm = not metal on metal right?
[23:56:44] -!- AR__ [AR__!~AR@] has joined #linuxcnc
[23:58:46] <CaptHindsight> yes, not metal on metal
[23:59:22] <Valen> there ya go all sorted then ;->
[23:59:48] <CaptHindsight> so an EDM is a phaser, kewl