#emc | Logs for 2011-08-25

[00:02:13] <Guest91935> scale.0.in is showing response and scale.0.out is showing as well with my offset
[00:03:34] -!- theorb [theorb!~theorb@] has joined #emc
[00:03:48] <andypugh> stepgen.4.enable and stepgen.4.velocity-cmd?
[00:04:23] -!- theorbtwo has quit [Ping timeout: 264 seconds]
[00:04:25] <andypugh> I am starting to suspect hardware/wiring of the C6
[00:04:30] theorb is now known as theorbtwo
[00:05:19] <Guest91935> they are showing good as well
[00:05:45] <Guest91935> if those were off i would just see no activivity, right?
[00:06:05] <andypugh> Last thing would be the parallel port pin outputs, those probably suit halscope better.
[00:07:14] <Guest91935> should i be able to see activity on halscope with scale.0.in?
[00:08:01] <andypugh> You won't _necessarily_ see step pulses on the parport pin in Halscope, due to sampling time coordination. So, if you see step pulses, they have got that far. I you don't, then it is not conclusive "absence of evidence is not evidence of absence)
[00:08:43] <Guest91935> well, i ask because i havent seen anything on halscope at all... the line shoud show a step or wave or something?
[00:09:34] <andypugh> Yes, but only when triggered, and zoomed correctly, and scaled rightetc etc
[00:09:41] <Guest91935> just wondering about my settings. i havent run any oscilliscope for years
[00:11:16] <jdhNC> can you see step/dir for any of your real axes?
[00:11:29] <andypugh> depending on where in the realtime thread your Halscope sample is, you are likely to see nothing.
[00:11:43] <andypugh> You _should_ be able to see the direction pin flip though.
[00:12:44] <Jymmm> how big of a hole I need for a 1/4-20 bolt to go through? .260" ?
[00:12:54] <Guest91935> ok, i changed to roll run mode and can see pin 14 jumping and spindle-cmd go hi
[00:13:33] <ds3> 1/4
[00:13:35] <jdhNC> .250"
[00:14:24] <andypugh> Guest91935: I think we can conclude that the problem is in the hardware layer then
[00:15:07] <andypugh> Have you wired all the terminals on the C6 , or have a good reason why any one is not connected?
[00:16:23] <Guest91935> everything is wired, correctly to the best of my knowledge. if you think things are good in my hal file then i'll follow up on the hardware end
[00:16:58] <Guest91935> my axes are working fine, so it has to be in my connection with the c6
[00:17:03] <andypugh> it might be that the "enable" line needs inverting, and/or the "step" line.
[00:17:34] <Guest91935> oh..... is that easy info to find?
[00:17:55] <andypugh> Just try it.
[00:18:57] <Guest91935> ok, i found info on my own for once! i think, its the addition of -invert to the parport pin?
[00:19:26] <andypugh> Indeed. Invert 1 to invert, invert 0 to not invert
[00:19:55] -!- Valen has quit [Quit: Leaving.]
[00:21:42] <Guest91935> oops, i think i mis understood, i need a line "setp parport.0.pin-14-out-invert 1" added?
[00:22:58] <andypugh> Yes
[00:23:20] <andypugh> (well, it's worth a try)
[00:23:30] <Guest91935> :)
[00:24:09] -!- bootnecklad has quit [Read error: Connection reset by peer]
[00:24:52] <Guest91935> is "spindle-cw" the right entry, or should it be "spindle-on"?
[00:26:10] <andypugh> Am am beginning to wonder what my almost-famous lathe/mill is worth on eBay. In comparison to http://www.ebay.co.uk/itm/3-axis-CNC-vertical-milling-machine-/260836382740 is it similar, but with YouTube evidence of greater capability. If I can get that sort of money then a conversion of a proper lathe might be easy to fund.
[00:26:29] <andypugh> Guest91935: Probably spindle-on
[00:26:49] <andypugh> spindle-cw would drive a direction relay.
[00:39:46] -!- YK [YK!18f65bc3@gateway/web/freenode/ip.] has joined #emc
[00:40:50] <YK> andypugh: In response to your post, question: what elements of the control system are expected to be slow?
[00:41:33] <andypugh> 0 velocity to full velocity, mainly.
[00:42:09] <andypugh> And I don't actually know what a good machine will need.
[00:42:45] <YK> andypugh: if the flow control valve could go from 0 to full velocity within 0.006s, would that alleviate this concern?
[00:43:51] <YK> andypugh: because then it seems to be just getting a fast stepper with on-load rpm of 1000 or so.
[00:44:54] <andypugh> Yeah, that would seem more than adequate, actually. You said that your notional time-constant (accuracy/speed) was 4mS?
[00:45:43] <YK> andypugh: Yes, 4ms was the desired value.
[00:46:36] <andypugh> I have a not-very-well-thought-out feeling that a machine with 4ms accuracy/speed will be OK with a 40mS zero-full speed But this is entirely guesswork. I am not an expert on control systems.
[00:48:27] Computer_ is now known as Computer
[00:48:33] -!- Computer has quit [Changing host]
[00:48:33] -!- Computer [Computer!~Computer@unaffiliated/computer] has joined #emc
[00:48:57] <YK> andypugh: I'm thinking that deceleration will aid stopping scenarios significantly. Given that, my concerns are mainly focused on the software-side of the control system (and now the hardware -side too, because something called the PCI was brought up)
[00:50:52] <YK> andypugh: To get an idea of how the PCI would replace the microcontroller's functions, can you tell me a bit about this hardware component?
[00:50:53] <andypugh> As I said, it makes a lot of sense to build one axis and run it from a parport as a feasibility test
[00:51:01] -!- danielfalck [danielfalck!~danfalck@static-50-53-1-104.bvtn.or.frontiernet.net] has joined #emc
[00:51:25] <YK> andypugh: What would the parport connect to?
[00:54:10] <andypugh> In reverse order: The parport would connect direct to a stepper driver and the axis encoder. You will be limited in step rate and encoder count rate, but your hardware investment is $50 plus the hydraulics. (and you can re-use all of it if it works)
[00:54:28] <andypugh> The answer to the first question is probably here: http://www.linuxcnc.org/docview/devel/html/drivers/hostmot2.html
[00:55:06] <andypugh> Assuming that by PCI you mean the PCI card, and that isn't a typo for PID?
[00:55:52] <YK> andypugh: alright, thanks- I'll be sure to take a look at those plus the HAL guide. I'll be back with more info at a later time, not today though.
[00:56:06] <YK> gnight
[00:56:55] -!- YK has quit [Quit: Page closed]
[01:00:50] <Guest91935> andypugh: one thing i havent understood yet is setting the signal timing, the board apparently needs around 25 khz....max?
[01:12:19] <andypugh> That's set by the stepgen scale and the stepgen velocity.
[01:13:01] <andypugh> (nominal pulses per inch, and nominal inches per minute == Hz)
[01:21:03] -!- WalterN [WalterN!~walter@host-174-45-106-117.bzm-mt.client.bresnan.net] has joined #emc
[01:23:11] -!- andypugh has quit [Quit: andypugh]
[01:30:12] -!- funkster [funkster!~IceChat77@216-189-172-036-dhcp.mia.fl.atlanticbb.net] has joined #emc
[01:34:39] <funkster> looking to purchase a cnc machine hardware - anyone point me in the right direction
[01:35:51] <ds3> anyone know if all current CNC'ed EDM setups are wire edm?
[01:36:33] <Tom_itx> funkster, 'hardware' is pretty vague
[01:37:10] <funkster> sorry - looking for a cnc machine to start cutting up plexiglass projects
[01:39:20] <funkster> a desktop cnc router would be ideal.
[01:39:35] <funkster> only need to work with 12x12
[01:39:49] <Tom_itx> http://www.probotix.com/
[01:40:16] <Tom_itx> ask a dozen ppl and you'll likely get a dozen answers
[01:40:46] <Tom_itx> http://www.kelinginc.net/CNCmachines.html
[01:41:08] -!- theorb [theorb!~theorb@] has joined #emc
[01:41:21] -!- theorbtwo has quit [Ping timeout: 258 seconds]
[01:41:31] theorb is now known as theorbtwo
[01:41:35] <ds3> is there such a thing as a single point, CNC EDM machine?
[01:43:23] <funkster> was looking at http://lumenlab.com/d/ - okay cool let me check those links out
[01:55:38] <funkster> damn all these is confusing, lol
[01:58:33] <mikegg> andy, what does 3pwmgen.scale do?
[01:59:17] <mikegg> hows the enduro doing, btw?
[02:02:16] <funkster> Tom_itx: im looking at the v90 now ;) looks nice
[02:03:16] <jdhNC> funk: I have an extra cnc router frame, just needs steppers/electronics
[02:04:15] <funkster> im completely new to this - looking to just purchase the whole complete set
[02:05:19] <funkster> software as well.
[02:05:29] <jdhNC> I'll include software :)
[02:06:02] <funkster> all I have is my cad/solidworks file that was made for me - trying to cut down on costs with my own cnc
[02:06:11] <funkster> so what do you suggest then.
[02:06:56] <jdhNC> for what?
[02:08:27] -!- ve7it has quit [Remote host closed the connection]
[02:09:02] <funkster> should I buy piece by piece, or a the whole shabang
[02:09:21] <funkster> as inbeing that im completely clueless, lol
[02:09:37] <jdhNC> got more time or money?
[02:09:49] <jdhNC> what do you want to cut?
[02:10:30] <funkster> plexiglass and posible sheet metal - its cases for electronics about 10x8
[02:10:49] <jdhNC> sheetmetal is a different class
[02:11:12] <funkster> aluminum.. any metal but ill stick with plexiglass for a while
[02:11:35] <funkster> i have about 1500-2000 budget
[02:11:49] <Tom_itx> if you plan to cut metal later, you need something heavier i think
[02:11:54] <jdhNC> yep
[02:12:12] <funkster> yeah i will stick with acrylic then
[02:12:13] <Tom_itx> and you can buy it once or you can buy it twice
[02:12:16] <jdhNC> mine could aluminum, but really slow
[02:12:50] <funkster> you guys suggest a specific setup for me?
[02:12:57] <Tom_itx> not really
[02:13:09] <Tom_itx> you know your needs better than anyone
[02:13:21] <Jymmm> a brake
[02:13:24] <Tom_itx> there are also desktop mills
[02:13:30] <Tom_itx> regular mills
[02:13:44] <funkster> max 12x12 , 1/2in plexiglass thats all i will do
[02:13:53] <jdhNC> http://www.ebay.com/itm/150558232923
[02:14:07] <jdhNC> I have a spare one of those (95% the same anyway)
[02:14:39] <funkster> niiice
[02:14:58] <Tom_itx> doesn't look very rigid
[02:15:23] <jdhNC> it's plastic and wood... that's about as rigid as it is.
[02:15:42] <jdhNC> most everything is boxed though, so it is really not bad
[02:15:44] <funkster> yeah thats fine , all i need.
[02:17:05] <Tom_itx> that needs motors and control stuff
[02:17:15] <Tom_itx> and $100 shipping from canada
[02:17:38] <jdhNC> heh, but mine is half price and shipping from .nc.us :)
[02:17:43] <funkster> yeah - id like to buy it all in one shot, instead or piecing together.
[02:17:55] <jdhNC> or I'll be in Gainesville next week :)
[02:18:00] <Jymmm> not for $2000 you wont
[02:18:15] <jdhNC> http://www.ebay.com/itm/160565315893
[02:19:01] <Jymmm> so, what do you do, screw the work piece into the plywood?
[02:19:15] <jdhNC> mine came with mdf T-slot top
[02:19:37] <jdhNC> but, I just bolt down some more MDF and screw or clamp ot that
[02:19:47] <Jymmm> And I doubt a dremel is going to do 1/2" acrylic.
[02:20:08] <jdhNC> sure itwill... not in one pass though
[02:20:23] <funkster> shit! lol
[02:20:24] <Jymmm> with what tooling?
[02:20:36] <Jymmm> not anythign 1/8" shank that I know of.
[02:20:38] <jdhNC> onrud end mill
[02:21:10] <Jymmm> 1/8" shank
[02:21:17] <jdhNC> yep
[02:21:29] <jdhNC> you can mount a bigger spindle too
[02:22:52] <jdhNC> http://www.artichoke.org/arrows.jpg
[02:23:02] <jdhNC> those are only .20 acrylic
[02:23:49] <funkster> clean cuts
[02:26:30] <jdhNC> I did the cam badly, I was holding them in place by hand at the end, trying to keep my fingers clear
[02:27:13] <jdhNC> you in miami?
[02:27:44] <funkster> yeah
[02:34:44] -!- ries has quit [Quit: ries]
[02:35:57] <frysteev> ola
[02:36:45] <frysteev> servo / closed loop question, what is the signal typ that emc typicalcally receivers from ther encoders,
[02:37:12] <jdhNC> quadrature
[02:37:31] <frysteev> is there a + pin and a - pin for counts in each direction, or a count in, and a direction signal
[02:38:02] <frysteev> butthe quadrature encode tyically goes to some some of board before the computer no?
[02:38:50] <mikegg> yeah, the mesa 7i33 is a common encoder counter
[02:38:56] <funkster> fireball v90 looks like overall it will be under 2k
[02:39:37] <frysteev> i guess im wondering, how does the mesa and emc talk?
[02:39:57] <mikegg> hostmot2
[02:40:17] -!- i_tarzan has quit [Ping timeout: 252 seconds]
[02:40:18] <frysteev> i guess i need to figure out what that is then :P
[02:40:21] <mikegg> the mesa hardware interfaces through parallel port or PCI
[02:40:36] <mikegg> hostmot2 is the firmware that the FPGA runs...I think
[02:40:42] <jdhNC> http://wiki.linuxcnc.org/emcinfo.pl?HostMot2
[02:40:43] <frysteev> ahh,
[02:42:04] <frysteev> ok, so for shits and gigles, does emc take a count pulse in, and another signal telling it which direction? like the reverse of the direction and step puse it sends out
[02:42:53] <mikegg> with a quadrature encoder, the direction is built in
[02:44:23] <frysteev> we are doing a homebrew interface, all my resolvers are going into an fpga, its counting the direction and steps, im just trying to figure out how to have it communicate with emc
[02:44:48] <mikegg> whew
[02:45:07] <frysteev> ?
[02:45:23] <mikegg> I'm waaayyy to lazy to try that :)
[02:45:33] <mikegg> I just buy the mesa gear and use hostmot2
[02:45:54] <frysteev> im way to cheap to be able to afford that,
[02:46:33] <frysteev> we got a handle on talking to the servos, and all the other i/o. its just how to comunicate the feedback back into emc that is missing
[02:46:49] <jdhNC> get a 7i43, they are like $70
[02:47:07] <mikegg> someone will surely shoot this idea down, but maybe write a hal component to take the count and direction as input and output a count that goes up in one direction and down in another?
[02:47:31] <frysteev> well i need the 7149, and theanything i/o board,
[02:47:43] <frysteev> and the resolver baord isnt supported in emc yet ive ben told
[02:49:20] <jdhNC> someone posted an inexpensive resolver->quad interface the other day
[02:50:22] <frysteev> yes, there is an arduino one,
[02:51:00] <jdhNC> no, that's too slow. This was a commercial device
[02:51:03] <frysteev> but i dont wanna do mulitiple conversions,
[02:51:24] <jdhNC> then ditch the resolver and get an encoder
[02:51:34] <frysteev> i have the hardware and brains on my side to take in the real data, its just how to deliver it to emc,
[02:53:24] <jdhNC> write a driver and hal component for a 7i49?
[02:54:14] <frysteev> but why do i need any mesa gear?
[02:55:48] <jdhNC> do you have an existing resolver interface? how do you talk to it?
[02:57:19] <frysteev> right now we arent, we are replacing the olde cpu with emc,
[02:58:00] <frysteev> it has an a-d converter for the resolvers, and excites them and does all that, then we have a chip counting and processing direction etc.
[02:58:44] <jdhNC> have it spit out quad?
[02:59:12] <jdhNC> if you have spare outputs... and spare inputs on teh emc side
[02:59:48] <frysteev> i can, but emc shouldnt be taking the quad input, it can, but it sucks cpu ive been told
[03:00:08] <jdhNC> not if you feed it to a mesa
[03:01:14] <frysteev> i guess im wondering, why everyone just uses mesa's
[03:01:25] <jdhNC> they cheap easy reliable?
[03:01:53] <jdhNC> if you need a way to get a lot of pins on a PC, that seems to be the most cost effective way
[03:02:23] <jdhNC> are you going to do step/dir to the serveros?
[03:02:44] <frysteev> not sure yet, we can
[03:02:57] <frysteev> im new to servos and closed loopp
[03:04:07] <jdhNC> a 7i43 would do that, or pwm out to the servo and handle quad
[03:05:38] <frysteev> im just trying to not have like 5 steps of conversion, hmm
[03:09:41] <mikegg> what command signal do your drives use? +/- 10 VDC
[03:10:14] <frysteev> 15vdc
[03:10:20] <mikegg> 0-15?
[03:11:32] <mikegg> no, that wouldn't make sense...nevermind. Was thinking of a spindle
[03:12:01] <frysteev> +/- 15vdc
[03:44:04] -!- mikegg has quit [Quit: Leaving]
[03:50:41] -!- sumpfralle has quit [Quit: Leaving.]
[04:27:49] -!- danielfalck [danielfalck!~danfalck@static-50-53-1-104.bvtn.or.frontiernet.net] has parted #emc
[04:47:03] -!- mhaberler [mhaberler!~mhaberler@] has joined #emc
[05:04:32] <alex4nder> hey
[05:09:59] -!- WalterN has quit [Read error: Connection reset by peer]
[05:19:01] -!- Eartaker has quit [Quit: Leaving]
[05:24:20] -!- nicko [nicko!~nicko@222-155-223-27.jetstream.xtra.co.nz] has joined #emc
[05:33:30] -!- psha[work] [psha[work]!~psha@] has joined #emc
[05:36:42] <Connor> I just finished making my first Stepper motor mount for my G0704 CNC conversion.. First time making anything worth while on the mill.
[06:34:22] -!- vladimirek [vladimirek!~vladimire@adsl-dyn18.91-127-76.t-com.sk] has joined #emc
[06:44:53] -!- e-ndy [e-ndy!~jkastner@nat/redhat/x-ccqfkqdvhhqnbytk] has joined #emc
[06:50:37] -!- sumpfralle [sumpfralle!~lars@] has joined #emc
[06:50:38] -!- sumpfralle has quit [Client Quit]
[07:02:35] -!- Tom_itx has quit [Ping timeout: 264 seconds]
[07:10:38] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[07:13:58] -!- RyanS [RyanS!~ryanstruk@ppp157-163.static.internode.on.net] has joined #emc
[07:14:45] -!- Tom_itx has quit [Ping timeout: 246 seconds]
[07:18:49] <RyanS> This is kind of a machining question, where would I look for off-the-shelf RHS clamps (these appear to be for proprietary sizes): http://www.rose-krieger.com/Prod_Pgs/Clamps/Indusrial_Clamps/Flange_Clamps_RFC.htm , sure I can get some machined, but not worth it
[07:20:15] -!- psha[work] has quit [Quit: Lost terminal]
[07:31:39] -!- odiug [odiug!~guido@pD9F71247.dip.t-dialin.net] has joined #emc
[07:37:10] -!- bootnecklad [bootnecklad!~bootneckl@host-2-97-45-142.as13285.net] has joined #emc
[07:40:39] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc
[07:50:15] -!- robh__ [robh__!~robert@5ace7016.bb.sky.com] has joined #emc
[07:57:04] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[08:00:40] -!- mhaberler has quit [Quit: mhaberler]
[08:04:49] -!- mhaberler [mhaberler!~mhaberler@] has joined #emc
[08:08:29] -!- nicko has quit [Quit: Visitor from www.linuxcnc.org]
[08:10:55] -!- Danimal_garage has quit [Read error: Connection reset by peer]
[08:14:47] -!- cevad has quit [Quit: Leaving]
[08:19:53] -!- Calyp [Calyp!~Caly@unaffiliated/calyptratus-tzm] has joined #emc
[08:21:02] -!- bootnecklad has quit [Read error: Connection reset by peer]
[08:22:40] -!- bootnecklad [bootnecklad!~bootneckl@host-2-97-45-142.as13285.net] has joined #emc
[08:26:31] -!- odiug has quit [Ping timeout: 246 seconds]
[08:48:10] -!- i_tarzan [i_tarzan!~i_tarzan@] has joined #emc
[08:49:04] -!- El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]
[08:57:18] -!- cevad [cevad!~davec@host-174-45-229-40.bln-mt.client.bresnan.net] has joined #emc
[10:06:42] <Loetmichel> mornin'
[10:08:25] <Jymmm> ug
[10:10:39] <Loetmichel> *OUCH!* now i am awake... SMPS primary capacitor... 325Vdc... i should learn to wait for discharge before dismantling notebook PSUs ;-)
[10:11:13] <archivist> har har
[10:11:16] <Jymmm> lick it!
[10:11:52] -!- mhaberler has quit [Ping timeout: 240 seconds]
[10:13:06] <RyanS> whats the difference btw grip & non grip u bolts ? (this is the closest I could find to an engineering channel)
[10:14:37] <archivist> that would depend how they grip
[10:15:38] <Loetmichel> RyanS: no idea, but i would guess the gripped ones are knurled in the u for better grip to the mounted tube/part
[10:19:32] <jthornton> RyanS: what are you trying to do?
[10:22:59] <RyanS> just a simple bracket on length of square steel RHS, this would have been perfect http://uk.rs-online.com/web/p/products/3737569/ but alas that's not a readily available tube size
[10:23:48] <RyanS> So I thought perhaps steel plate with two square u bolts
[10:24:20] <jthornton> should work or just two steel plates with normal bolts
[10:24:22] <RyanS> I guess u bolts would be plenty strong enough..
[10:26:19] <RyanS> I was going to telescope over the tube with a short section, however I changed my design, the tube is now fixed at both ends so can't do that
[10:26:49] <jthornton> what is RHS?
[10:27:39] <RyanS> Don't know what it means but it's square steel tube
[10:27:57] <RyanS> No idea why they call it RHS
[10:28:16] <jthornton> rolled welded steel tube?
[10:28:21] <Jymmm> Right Hand Studmuffin
[10:28:27] <jthornton> you must be from the other side of the pond from me
[10:29:07] <RyanS> this stuff http://allthingsstainless.com.au/home/tube/square.html
[10:31:51] <RyanS> It is for a customised stainless steel computer desk, just trying to come up with ways to mount a removable keyboard/mouse tray... so u bolts would be pretty sturdy? not like that's carrying any weight
[10:32:13] <Jymmm> a grip u bolt has a condom
[10:33:40] <RyanS> umm, a rubber coating?? I assume
[10:33:49] <Jymmm> not necesarily
[10:34:00] <Jymmm> that's why I said condom
[10:34:30] <RyanS> Are you pulling my leg ? :)
[10:34:44] <Jymmm> one is an epoxy coated powder
[10:35:34] <Jymmm> I thought it was knurling too, but doesn't appear to be at all.
[10:35:56] <Jymmm> and it's priamrily a UK thing it appears
[10:36:18] <RyanS> not these ? http://www.ebay.com.au/itm/Wire-rope-grip-U-bolt-clamp-8mm-5-16-wire-pack-10-/110703427259
[10:38:33] <Jymmm> Gripped (Anchor) or Non-Gripped (Guide)
[10:40:35] <RyanS> ah ha! "The Pipe Grip range firmly holds the pipe to prevent movement, whilst damping vibration. All Insulated U-bolts the following benefits: Reduction of transmitted noise levels, vibration and shock loadings. Prevention of abrasive fretting between pipes and clips." they're for pipefitting!
[10:40:51] -!- Poincare has quit [Quit: changing servers]
[10:41:06] -!- Poincare [Poincare!~jefffnode@2001:470:cb24:4::2] has joined #emc
[10:42:23] <Jymmm> But that still doesn't say/show what is the difference
[10:42:54] <Jymmm> BS 3974 Part 1
[10:43:01] <Jymmm> is the UK spec for it
[10:46:57] <Jymmm> HERE WE GO... Grip U Bolts (condom) http://www.nortesco.com/industrial/anti/anti_pipe.html
[10:47:29] -!- mhaberler [mhaberler!~mhaberler@] has joined #emc
[10:48:36] <Jymmm> Lovely.... PTFE sleeves, lets home those aren't for gas pipes
[11:00:19] -!- automata [automata!~automata@] has joined #emc
[11:00:52] -!- psha[work] [psha[work]!~psha@] has joined #emc
[11:02:47] -!- mhaberler has quit [Quit: mhaberler]
[11:11:34] * Jymmm kicks JT-Shop in the paint buckets!
[11:12:07] * Jymmm kicks jthornton in the paint buckets too!
[11:13:13] <Jymmm> jthornton: I *finally* found a way to fasten the HC and have my guides aligned!
[11:13:53] -!- mhaberler [mhaberler!~mhaberler@] has joined #emc
[11:14:45] <jthornton> you don't know how glad I am
[11:14:58] <Jymmm> jthornton: bite me!
[11:15:34] <jthornton> don't kick me today, I spent all day yesterday and half the night sitting with a good friend as his wife of 40 years passed away yesterday morning
[11:15:45] <Jymmm> jthornton: You had me doing like 4 different profiles
[11:15:52] <Jymmm> jthornton: my condolences.
[11:16:21] <jthornton> yea it was not a good day for him
[11:16:57] <automata> Hi, Has anyone tried EMC2 on an intel core i3 / i5 / i7 processor?
[11:18:16] <automata> any motherboards known to work?
[11:19:29] <automata> I am aware that EMC does not need this kind of hardware and intel Atom is preferred to run EMC. However, I need to run a Win**** XP as a guest OS
[11:19:49] <automata> on a virtual Box on the same machine. (Workflow constraints).
[11:20:22] <Jymmm> jthornton: Since I'm using 3/4" MDF for the walls, and it doesn't like to be fastened on edge (only face), If I take a 1" L profile and lay it over the top edge of the MDF, I can screw it into the face of the MDF. Then the HC on top of that, and another 1" L on top of that, with the 1st L tapped. It also lets me deal with the wall thickness of 1/8"
[11:21:16] <jthornton> that sounds good, are you done yet?
[11:21:47] <Jymmm> jthornton: Nope, still need to deal with the height of the bottom ruler offset
[11:22:08] <jthornton> at least you have a plan now
[11:22:47] <Jymmm> jthornton: or at least half a plan. It was a bitch trying to overcome the wall thickness issue.
[11:23:03] <Jymmm> with available materials that is
[11:23:55] <Jymmm> It's not perfect, but I can deal with 1/8" overshoot.
[11:24:21] <Jymmm> The HC cells are 1/4" so that should be good enough
[11:24:54] <jthornton> I can't wait to see the finished part
[11:25:02] <Jymmm> me too
[11:25:23] <jthornton> how long have you been trying to sort this out?
[11:25:44] <Jymmm> Like ACTUALLY do drawings and such? 3 days
[11:25:45] <Tom_itx> this will be the most over engineered table i have ever seen
[11:26:06] <jthornton> no, like trying to figure it out
[11:26:07] <Tom_itx> i can't wait
[11:26:45] <Jymmm> jthornton: a while, but I never had photos nor seen the REAL thing I'm trying to replicate.
[11:27:26] <Jymmm> jthornton: I contacted one of the dealers that's been around for a while and he got me some photos
[11:28:23] <Jymmm> Glad I did, as he pointed out a design flaw with the actual product regarding airflow
[11:29:09] <jthornton> that helps with a little inside info
[11:29:17] <Tom_itx> i went to get my niece a tire a couple days ago and passed by one tire store stopping at another one. last night the first one burned to the ground
[11:30:08] <jthornton> insurance fire?
[11:30:14] <Tom_itx> no idea
[11:30:27] <Tom_itx> we have had those before though
[11:30:28] <Jymmm> jthornton: Yeah, so I'll try to incorporate an adjustable flap to divert the airflow as needed.
[11:30:43] <Jymmm> Tom_itx: Ew, that must stink up the neighborhood
[11:31:44] <Tom_itx> there was a salvage that had a couple acres of tires they were supposed to clean up several years back and magically they would just burst into flames from time to time
[11:32:22] <jthornton> at night too I'd bet
[11:33:42] -!- ries [ries!~ries@] has joined #emc
[11:43:14] -!- awallin_ has quit [Remote host closed the connection]
[11:54:06] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc
[12:03:17] -!- mhaberler has quit [Quit: mhaberler]
[12:09:16] -!- rooks has quit [Quit: So long, and thanks for all the fish.]
[12:17:38] -!- awallin_ has quit [Remote host closed the connection]
[12:19:10] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc
[12:22:58] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[12:33:07] <automata> Has anyone tried to use EMC with a core i3/i5/i7
[13:03:07] -!- syyl [syyl!~syyl@p4FD132D6.dip.t-dialin.net] has joined #emc
[13:10:00] <A2Sheds> do any PCIe PP cards work well enough to use the Mesa 7i43 for use with servos?
[13:11:10] -!- FinboySlick [FinboySlick!~shark@] has joined #emc
[13:13:40] <awallin_> no i3/5/7 on the real-time latency-test page on the wiki
[13:18:25] <automata> Thats what I saw...
[13:19:34] <awallin_> is it possible to run latency-test just with a liveCD, no installation required? I have at least two i7's I could try on..
[13:19:54] <Tom_itx> should be able to
[13:20:28] -!- theos has quit [Disconnected by services]
[13:20:41] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[13:22:31] <awallin_> is there a difference btw 32-bit and 64-bit?
[13:22:36] <awallin_> is this one 32-bit? http://www.linuxcnc.org/lucid/ubuntu-10.04-linuxcnc1-i386.iso
[13:24:35] <awallin_> eh, 47 minutes to download the cd
[13:25:14] <awallin_> hm, slightly better from alex jonis mirror, 2 minutes!
[13:27:05] -!- automata has quit [Remote host closed the connection]
[13:28:34] <A2Sheds> I'm going to finish the coreboot port for an AMD Fusion board http://www.asrock.com/mb/overview.asp?Model=E350M1&cat=Specifications
[13:29:12] <A2Sheds> then we won't have to bother with Intel boards, EFI, SMI and latency issues
[13:31:09] <A2Sheds> I'll try to get the jitter well under 1ms
[13:39:26] <alex_joni> awallin_: 32bit
[13:39:39] <alex_joni> there was a 64bit version for 8.04, but it doesn't really help for emc2
[13:40:01] <alex_joni> A2Sheds: 1ms?
[13:40:23] <awallin_> so will 32/64 will have any difference for latency-test?
[13:42:10] <funkster> just talked to probotix, guys are awesome
[13:42:31] <funkster> going to order a v9- with bosh colt - cutting up som e 1/2" acrylic
[13:43:03] <A2Sheds> alex_joni: there were some Atom 945 boards that had jitter in the .3-.5ms range
[13:45:09] <awallin_> ?? 10us seems to be a typical jitter value
[13:47:03] <alex_joni> awallin_: nope, no difference
[13:47:18] <alex_joni> and since you're getting the lucid one, there's no 64bit version anyway
[13:56:10] <awallin_> burning the CD went smoothly first, but now it's taking forever at "Creating Image checksum" just at the end...
[13:59:21] -!- Loetmichel has quit [Ping timeout: 252 seconds]
[14:04:35] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has joined #emc
[14:04:58] -!- Loetmichel [Loetmichel!Cylly@p54B151E8.dip.t-dialin.net] has joined #emc
[14:26:17] -!- pcw_home [pcw_home!~chatzilla@ip-66-80-167-54.sjc.megapath.net] has joined #emc
[14:28:54] <pcw_home> "<automata> Has anyone tried to use EMC with a core i3/i5/i7 "
[14:28:56] <pcw_home> I'm pretty sure Tormach's EMC/Mach control computer is an I5
[14:29:54] <pcw_home> but it might be just a newer core duo
[14:30:11] -!- funkster has quit [Ping timeout: 264 seconds]
[14:30:22] <psha[work]> btw with amd phenoms it's working quite well
[14:32:28] <psha[work]> i've used asus M4A88T-M matx board
[14:33:29] <psha[work]> pcw_home: btw he left channel hour ago :)
[14:34:14] <pcw_home> Yeah maybe he will read the log
[14:35:15] <pcw_home> I just found our that may test computer for EMC has better latency if it has a good harddrive
[14:35:24] <pcw_home> my
[14:35:56] <pcw_home> What did I type, more coffee...
[14:37:31] <tom3p> psha[work], if you could tell us... what kind of latency figures for what length of testing?
[14:37:42] -!- FinboySlick has quit [Ping timeout: 240 seconds]
[14:40:03] <psha[work]> length of testing not less then ~24h
[14:40:20] <psha[work]> numbers... hm, that's harder
[14:40:47] <psha[work]> there _was_ problem with acpi (need to disable as much as you can in bios) that raised after some time of testing
[14:41:11] -!- Valen [Valen!~Valen@2001:44b8:3199:9100:21e:8cff:fe63:55df] has joined #emc
[14:41:44] <psha[work]> but finally i've solved it and everything went smooth
[14:42:23] -!- FinboySlick [FinboySlick!~shark@] has joined #emc
[14:43:15] <tom3p> thx
[14:43:52] <pcw_home> My test computer now only has bad latency when it beeps, (like backspacing at beginning of line)
[14:43:54] <pcw_home> I remember fixing this before the hard drive died but dont remember how
[14:43:56] <pcw_home> wiki instructions dont seem to work for me (pcspkr not found)
[14:44:29] <psha[work]> pcw_home: i've finally banned 'snd' module and thus whole alsa thingie is not loaded
[14:45:06] <psha[work]> however pcspkr is loaded independently
[14:46:18] <tom3p> haha """Controlling a laser with Linux is crazy, but everyone in this room is crazy in his own way. So if you want to use Linux to control an industrial welding laser, I have no problem with your using PREEMPT_RT." -- Linus Torvalds """
[14:47:17] <tom3p> psha interesting, the audio latency folk are having problems.. trashing audio is fer sure a solution, now how to get rid of video and hd? :)
[14:47:39] <pcw_home> Ill try disabling snd thanks psha
[14:48:15] <tom3p> Real-Time Linux Wiki https://rt.wiki.kernel.org/index.php/Main_Page
[14:50:06] <psha[work]> tom3p: heh, most of installations have disabled hardware 3d :)
[14:50:10] <psha[work]> so it's already trashed
[14:50:19] <pcw_home> I thought that video on my system was a problem so I tweaked a bunch of things
[14:50:21] <pcw_home> but latency made it biggest improvement when I replaced the hard drive (and went back to stock settings)
[14:50:23] <psha[work]> either via 'vesa' driver or using libgl1-mesa-swx11
[14:52:31] <tom3p> yeh i had to use swx11 to get rid of weird artifacts on a TFT ( tears in AXIS / anyOpneGL)
[14:55:19] -!- Valen has quit [Quit: Leaving.]
[14:57:53] <A2Sheds> awallin: sorry i was off by 3 orders of magnitude, meant nS vs uS earlier
[14:59:16] <psha[work]> A2Sheds: nS?!
[14:59:39] <psha[work]> i bet now you are 3 orders on the other side :)
[15:00:08] <A2Sheds> yeah, SWP found some Atom boards with jitters below 1 microsecond
[15:00:27] <A2Sheds> I have a link somehwere
[15:01:07] <psha[work]> heh, below and 0.001us are different things :)
[15:01:08] -!- e-ndy has quit [Quit: Ex-Chat]
[15:01:24] * archivist looks for the fempto seconds error
[15:03:30] <A2Sheds> they were IIRC in the .3-.5us range
[15:03:58] <A2Sheds> so 300 - 500 nS
[15:05:14] -!- sumpfralle [sumpfralle!~lars@] has joined #emc
[15:06:02] <tom3p> the reported >number< from latency test was 300 to 500 ? please forget the units just for clarity, the tool reported the number 300 to 500?
[15:07:13] <A2Sheds> I didn't run the test and my memory hasn't had the rest it's needed to properly recall :)
[15:08:22] <A2Sheds> tom3p: are the latency results in nanoseconds?
[15:10:14] <tom3p> A2Sheds, dunno, i just wanted to get the raw data, then look at the units of measure
[15:11:18] <tom3p> it reports in nS
[15:14:00] -!- adb [adb!~Moldovean@] has joined #emc
[15:14:50] <tom3p> my experience ( limited ) suggests 10uS is pretty good, and 20 to 10uS is common. less than 10uS is not off the shelf, and i'd say 5uS is a damn lie.
[15:17:36] <A2Sheds> SWPadnos: might recall the vendor he tested , they were Atom + i945 industrial PC's
[15:19:48] <A2Sheds> I can't find the link right now
[15:20:49] <tom3p> SWPadnos, no offense, i doubt you said 3-500 nS , and if you did, i got the CC ready to buy you AND me one as apology :-!
[15:21:19] <A2Sheds> what I'm up to is: coreboot replaces the vendor BIOS, so no EFI, no SMI and we can tweak the chipset register settings for what we want
[15:21:53] <tom3p> cool, anything to get tighter response! thx
[15:22:10] <A2Sheds> i really tired today :(
[15:22:34] <tom3p> we all would like to be able to output a 1uS sqr wave under software loop
[15:25:09] <A2Sheds> I have't tested RTAI on AMD Fusion yet, not sure how well the IRQ arbiter handles activity
[15:26:03] -!- Paragon-ws [Paragon-ws!~chatzilla@cpc13-bexl7-2-0-cust145.2-3.cable.virginmedia.com] has joined #emc
[15:26:46] <A2Sheds> my guess is that the irq's for a Mesa PCIe card on the Fusion's PCIe port will be much tighter than if the Mesa card is connected through a PCIe HUB
[15:27:38] <A2Sheds> I don't know if any off the shelf Fusion boards have a PCIe slot available that way
[15:31:59] -!- odiug [odiug!~guido@pD9F71247.dip.t-dialin.net] has joined #emc
[15:34:56] -!- psha[work] has quit [Quit: Lost terminal]
[16:06:58] -!- kljsdfhklj [kljsdfhklj!~ln@] has joined #emc
[16:14:22] -!- awallin_ has quit [Remote host closed the connection]
[16:20:26] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc
[16:20:53] -!- odiug has quit [Ping timeout: 252 seconds]
[16:22:24] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has parted #emc
[16:24:48] -!- isssy [isssy!~isssy@78-83-51-185.spectrumnet.bg] has joined #emc
[16:26:43] <Connor> I just finished making my first Stepper motor mount for my G0704 CNC conversion.. First time making anything worth while on the mill.
[16:27:09] -!- automata [automata!~automata@triband-mum-] has joined #emc
[16:27:22] <jdhNC> cool, jpegs?
[16:27:37] <jdhNC> what motors are you going to use?
[16:27:44] <automata> anyone has use a core i3/i5/i7 for running EMC2?
[16:28:02] <Connor> not yet. I need to de-burr it.
[16:28:18] <Connor> Probably The 540 Oz-Inch Nema 23 steppers.
[16:28:37] <jdhNC> for Z also?
[16:28:49] <Connor> Yea. More than enough for the Z
[16:29:16] <Connor> Direct drive, with ballscrews or original screws.
[16:29:43] <jdhNC> make two of everything and send to me !
[16:29:53] <Connor> ROFL.
[16:30:14] <Connor> I'm doing this manually.. took me 3 hours to make the first mount plate.
[16:30:21] <jdhNC> I have 3 425's on my router. Wonder how well they would work.
[16:30:23] <Connor> learning while I'm going.
[16:30:31] <Connor> Probably be fine.
[16:30:39] <jdhNC> that's just the first one... by the time you get to mine, it should be only 2 hours each
[16:31:32] <jdhNC> $310 shipped for PS/drive/motors from xylotex
[16:31:53] <Connor> I used my boring head even.. :)
[16:32:07] <jdhNC> did you drill first?
[16:32:25] <Connor> yup. I used my 3/4" End mill to start the whole.
[16:32:39] <Connor> then opened it up using the boring head to 1"
[16:33:10] <jdhNC> going to direct-connect to the screws?
[16:33:45] <Connor> yea. for now.. When I go ballscrews, the Y will be belt connected rear mounted.
[16:34:00] <Connor> http://www.kelinginc.net/NEMA23Motor.html
[16:34:18] <Connor> 64.95 each for 570 Oz-in nema 23 steppers
[16:34:48] <jdhNC> big shaft
[16:34:56] <Connor> Big Motor.
[16:35:26] <Connor> http://www.g0704.com/G0704_electronics.html
[16:35:46] <Connor> That's his break down of various electronic packages.
[16:35:57] <jdhNC> yeah.... but he's a Mach user so you have to question his judgement
[16:36:13] <Connor> with what, the motors, or the drivers ?
[16:36:36] <jdhNC> that was a joke
[16:36:39] <jdhNC> nevermind.
[16:36:42] <Connor> oh. ha. :)
[16:37:24] -!- e-ndy [e-ndy!~jkastner@static-84-242-102-36.net.upcbroadband.cz] has joined #emc
[16:37:36] <count> Dear EMC2 guys: thanks! (http://www.youtube.com/watch?v=WBDTU-sSixo)
[16:37:45] <count> our American robots Merlin is now operational again
[16:37:52] <count> via parallel port
[16:38:19] <jdhNC> count: cool, you can send donations to my paypal account.
[16:38:37] <count> I just donated thanks :)
[16:40:31] <A2Sheds> any suggestions for a PCIe to EPP card to use up my stock of 7i43 cards?
[16:43:06] <pcw_home> We've got experience with PCIE Parallel ports
[16:43:18] <pcw_home> got no experience
[16:43:30] <A2Sheds> ahhah
[16:44:50] <A2Sheds> I got rid of all the VIA epias with EPP and not ready with a Fusion PCIe + 3x20 yet
[16:44:59] <pcw_home> I know MOSCHIP (who bungled the EPP interface on their PCI chips) has a single chip part, but i have not wanted to spend the 30 or so bucks to see if it works
[16:45:31] <A2Sheds> when is that lower cost PCIe IO card due? I think i asked you last week :)
[16:45:36] -!- automata has quit [Ping timeout: 260 seconds]
[16:45:44] -!- automata [automata!~automata@triband-mum-] has joined #emc
[16:46:18] <Connor> I used a little cheapo PCI 2 port parallel card just fine.. I suspect any of them will work just fine.
[16:46:30] <pcw_home> Pretty soon I will use a small SP6 + TIs PCIE-PCI bridge (basically a 5I25 with the bridge added)
[16:46:47] <A2Sheds> Connor: was it PCI or PCIe?
[16:46:58] <pcw_home> should be like $99
[16:47:10] <Connor> I think it was PCI.. but, that was a year or two ago..
[16:47:25] <pcw_home> since the TI bridge is only $2.50 in quantity
[16:47:33] <A2Sheds> Lattice has their PCIe IP on sale, but I know you only use xilinx
[16:47:58] <pcw_home> yeah like to keep all the IP GPL-able
[16:48:46] <A2Sheds> pcw_home: which TI part#?
[16:50:23] <A2Sheds> http://focus.ti.com/docs/prod/folders/print/xio2001.html ?
[16:50:35] <pcw_home> I forgot the name its a PCIE-PCI bridge (6 masters 32 bit up to 66 MHz)
[16:50:37] <pcw_home> Yeah thats it
[16:50:51] -!- mhaberler [mhaberler!~mhaberler@62-47-129-200.adsl.highway.telekom.at] has joined #emc
[16:50:53] <A2Sheds> great, 1 less board i have to make :)
[16:51:33] <A2Sheds> can you have it by mid october? :)
[16:51:37] <pcw_home> 'bout 1/3 the price of the competition
[16:51:55] -!- odiug [odiug!~guido@pD9F71247.dip.t-dialin.net] has joined #emc
[16:52:03] -!- psha [psha!~psha@] has joined #emc
[16:52:04] <pcw_home> sorry vacation (dictated by SO)
[16:52:51] -!- Danimal_garage [Danimal_garage!~kvirc@ip68-111-139-153.sd.sd.cox.net] has joined #emc
[16:53:03] <Danimal_garage> hi
[16:53:34] <A2Sheds> I'm only controlling a 2 axis servo table, so PCIe --> EPP with 7i43 should be good enough for now
[16:53:57] <A2Sheds> and it's slow
[16:54:04] <pcw_home> PCIE version has advantage over 5I25 (PCI core in FPGA) that if you break the config you have the bridges GPIO to reprogram the SPI flash
[16:54:34] <pcw_home> so you dont need JTAG if you botch a firmware upgrade
[16:54:45] <A2Sheds> great!
[16:54:51] <Connor> http://www.ivdc.com/cnc/mill/cnc_conversion/STEPPER-Y-MOUNT1.JPG
[16:54:53] <Connor> http://www.ivdc.com/cnc/mill/cnc_conversion/STEPPER-Y-MOUNT2.JPG
[16:56:33] <Connor> re-doing the top-down photo.. it sucked.
[16:57:42] <Connor> okay, http://www.ivdc.com/cnc/mill/cnc_conversion/STEPPER-Y-MOUNT1.JPG
[16:57:45] <Connor> new picture.
[16:58:21] <Connor> So, I used my calipers, and made little X's were the holes went, and then used a center punch..
[16:59:28] <Danimal_garage> hack!
[16:59:35] <Danimal_garage> j/k, i do that all the time
[16:59:48] -!- IchGuckLive [IchGuckLive!~chatzilla@95-89-104-26-dynip.superkabel.de] has joined #emc
[16:59:53] <IchGuckLive> Hi all
[16:59:53] <Danimal_garage> the old Polocks i used to work with would castrate me for that though
[17:00:02] <Danimal_garage> hi
[17:00:09] <jdhNC> I put crosshairs in my circles and punch through the drawing
[17:00:10] <IchGuckLive> jdhNC: wind around ?
[17:00:13] <Connor> then.. I used a center drill on them.. but, I started off with the center hole as the reference... then used the dials on the mill to measure out the correct location.. using the detents to make sure I was in the ball park.
[17:00:43] <jdhNC> Ich: not that I know of... won't get anything real until tomorrow
[17:00:44] <IchGuckLive> jdhNC: expect a Cat3 full hit
[17:00:46] <Connor> I would have too.. but, the drawing didn't have them.. and alum scratches pretty easy.
[17:01:11] <jdhNC> those are hoss dwgs? or pdfs? or both?
[17:01:22] <Connor> Yea. Hoss's
[17:01:51] <jdhNC> ich: I'm going with "light brushing up the coast vs. direct hit"
[17:02:34] <Connor> but, my point is, I used the dials on the mill and even compensated for the backlash.. first time doing accurate milling..
[17:02:38] <Connor> manually.
[17:03:02] <IchGuckLive> accurat milling use CRC
[17:03:27] <Connor> I have a new found respect for those guys who doing manual milling.. with and without a DRO.
[17:03:29] <jdhNC> I think I'd take scribed marks over chinese dials with backlash for accuracy
[17:03:46] <Connor> jdhNC: They were dead on.
[17:03:51] <Connor> every time.
[17:04:07] <IchGuckLive> http://linuxcnc.org/docs/html/gcode_main.html#sec:G41,-G42
[17:04:09] <jdhNC> I don't even have my X handwheels mounted
[17:04:37] <Connor> I need to get myself a belt sander..
[17:04:52] <Connor> and Need to polish the mount
[17:05:10] <jdhNC> I have a 4xsomething blet sander, works great for deburring Al
[17:05:40] <Connor> My standoffs didn't turn out too bad for making them on a mill vs a lathe.. Though, they could stand to be turned down just a bit and polished again.
[17:07:44] <jdhNC> did you use tubing or solid?
[17:07:52] <Connor> solid.
[17:07:59] <Connor> 1/2" Alumn stock.
[17:08:16] <Connor> Center drilled, and tapped for #10-32
[17:09:12] <IchGuckLive> tapped G83 ?
[17:10:19] <IchGuckLive> i need to go sorry Bye
[17:10:22] -!- IchGuckLive [IchGuckLive!~chatzilla@95-89-104-26-dynip.superkabel.de] has parted #emc
[17:11:11] <Connor> I put my pointed center finder in the mill all the way up so it didn't wiggle.. and used it and the quill with the tap handle to keep the tap from wobbling.. Worked pretty good.
[17:11:35] <Connor> though I have to say, Alumn isn't very fun to tap.. Very gummy.
[17:11:54] <Connor> had to run the tap in and out 4-5 times.. cleaning the tap each time..
[17:15:34] -!- automata has quit [Ping timeout: 245 seconds]
[17:16:59] -!- ssloy [ssloy!~ssloy@ANancy-555-1-175-131.w90-33.abo.wanadoo.fr] has joined #emc
[17:18:12] <ssloy> hello, would anyone please be so kind to help me a bit with my configuration?
[17:18:30] <ssloy> i have a hobbycnc card coupled to emc2
[17:18:51] <ssloy> steppers work just fine, however i cannot get limit/home switches working
[17:19:30] <ssloy> no matters what tension (+5/gnd) i put onto pins 10-12, hal reads them as log1....
[17:19:38] -!- automata [automata!~automata@triband-mum-] has joined #emc
[17:20:56] <ssloy> i'm really out of ideas
[17:21:10] <ssloy> tried all configurations in bios (ecp/epp/...)
[17:21:34] <jdhNC> are they configured as input pins?
[17:21:54] <ssloy> good question, how do i check?
[17:22:43] <ssloy> i have parport.0.pin-XX-in in halshow
[17:23:04] <ssloy> so i guess it has work
[17:23:15] <ssloy> but is there any specific place to configure?
[17:23:39] <jdhNC> I just took the defaults from teh stepconf wizard
[17:23:47] <ssloy> me too
[17:24:03] <jdhNC> and none of them ever show up as off?
[17:24:09] <ssloy> nope
[17:24:23] <ssloy> tried on two computers...
[17:24:39] <ssloy> i check the actual tension with a multimeter
[17:25:05] <ssloy> but halshow just tells me it is on regardless what i actually have
[17:25:08] <jdhNC> ground one of the pins and check it with halscope?
[17:26:47] <jdhNC> I know enough to make mine worth through extensive trial and error, but not enough to tell you what is wrong with yours :)
[17:27:11] <jdhNC> got a funky breakout board that decides some pins are output only?
[17:28:32] <jdhNC> On board connections for home and limit switches with 10K pullup resistor provided to each. No need for a separate .break out. board
[17:28:49] <jdhNC> looks like they will all read 1 until you ground them.
[17:32:24] <automata> On the forums or wiki, it is not clear that a core i processor is good choice for running RTAI
[17:34:08] <jdhNC> is there soemthing that would make it bad, or just overpriced?
[17:34:19] <automata> There seems to be an ASMP-Linux which is designed for hard real-time performance on multi core processors
[17:34:51] <automata> I am not sure about that, but is it possible to ensure that only one core is dedicated to the realtime process?
[17:35:01] <jdhNC> there is a way to lock it ot one
[17:35:18] <automata> using the isolcpu flag in grub.
[17:35:53] <ssloy> mm, nothing useful (to me) from halscope
[17:36:33] <automata> That ensures that the Linux kernel does not use the isolated CPU. Howwever, I have not yet seen any mention that the isolated CPU is used for realtime
[17:36:55] <automata> Does something tie the realtime process to the isolated CPU?
[17:37:26] <cpresser> automata: there is a kernel-command-option (isolcpus or similar)
[17:37:35] <cpresser> but dont ask me about the details :)
[17:38:02] <cpresser> ah damm, i am just repating your statement. didnt read all chat lines
[17:41:28] <automata> I have been working with RTAI from way back on the Linux 2.2 Kernel... but have been out of touch since 2005.
[17:41:55] <automata> http://downloads.hindawi.com/journals/specialissues/0022008003.pdf has more info on the usage of isolcpus flag..
[17:42:07] <Danimal_garage> yay, my linear rails got here
[17:42:14] <Danimal_garage> they should work nicely
[17:42:23] <automata> BUt this multi-core development oin the last 6 years has really passed me by...
[17:42:37] <automata> Time to roll up sleves and start digging in again..
[17:43:18] <automata> BTW isolcups flag has been around since 2004 since linux 2.6.9
[17:45:23] <ssloy> halscope shows log1
[17:45:50] <ssloy> it seems like the port is one-way-through
[17:46:05] <automata> As I thought with the use of isolcpus, we need to us the sched_setaffinity function in the realtime process
[17:46:24] <automata> to bind the realtime process to the isolated CPU...
[17:46:50] <automata> Now where would we do that?? Which module should we set affinity for??
[17:48:16] <automata> motion module?
[17:48:55] <automata> how do we set the servo thread in HAL to have affinity to a single processor?
[17:50:13] -!- psha has quit [Quit: Lost terminal]
[18:00:40] -!- sumpfralle1 [sumpfralle1!~lars@g230217250.adsl.alicedsl.de] has joined #emc
[18:03:08] -!- psha [psha!~psha@] has joined #emc
[18:03:38] <Loetmichel> re @ home
[18:04:13] -!- sumpfralle has quit [Ping timeout: 240 seconds]
[18:28:12] -!- kljsdfhklj has quit [Quit: kljsdfhklj]
[18:28:28] -!- kljsdfhklj [kljsdfhklj!~ln@189-30-67-56.fnsce704.dsl.brasiltelecom.net.br] has joined #emc
[18:36:29] -!- Danimal_garage has quit [Read error: Connection reset by peer]
[18:37:08] -!- Danimal_garage [Danimal_garage!~kvirc@ip68-111-139-153.sd.sd.cox.net] has joined #emc
[18:41:58] -!- mhaberler has quit [Quit: mhaberler]
[18:42:24] <ssloy> [censored]
[18:42:29] <ssloy> jdhNC: found the problem
[18:42:45] <ssloy> jdhNC: it was a brand new db25 cable
[18:42:52] <ssloy> broken one, apparently
[18:43:19] <ssloy> i plugged my card into lpt port directly and it works perfectly
[18:44:01] <ssloy> thanks for the hint about halscope though, the difference between log levels pushed me to check the cable
[18:45:14] <Jymmm> ssloy: Not all DB25 cables have all 26 connections
[18:45:39] <archivist> they cannot have 26!
[18:45:45] <Jymmm> archivist: shield
[18:46:41] <Jymmm> At least my DB25 cbales have shield connected the the connector shells.
[18:46:43] <archivist> the shield is connected to one of the 25, do not count the shield
[18:47:49] <Jymmm> Mine aren't. It's 26 isolated connections.
[18:48:08] <Danimal_garage> well you're just weird
[18:48:36] <Jymmm> woohoo!
[18:48:59] <archivist> the shield may also not be connected both ends to avoid a ground loop
[18:49:53] <Jymmm> Yeah, I have it isolated at the far end, that's actually how I found out there are 26 connections in the first place.
[18:50:12] <PCW> Yep IEEE 1284 have shield independent of all signals (and all signals with twisted pair gnd connected to shield)
[18:52:06] <PCW> Shield intended for ground to case at both ends for RFI
[18:54:59] -!- sumpfralle1 has quit [Ping timeout: 258 seconds]
[18:54:59] <Danimal_garage> whoops, drilled a hole in the wrong place...
[18:55:10] <PCW> Kinda surprised me too. It was a good surprise as I had though the might common some of the GND wires in IEEE1284 cables,
[18:55:11] <PCW> since they dont, IEEE-1284 cables work with our scheme of using 1/2 the GND wires for 5V supply to daughtercards/breakout boards
[18:55:31] <PCW> thought they
[18:57:27] -!- sumpfralle [sumpfralle!~lars@g230217250.adsl.alicedsl.de] has joined #emc
[18:59:03] <Danimal_garage> PCW: i'd like that 7i48 back since you won't warranty it.
[18:59:13] <Danimal_garage> how much would it cost to fix?
[19:04:13] -!- isssy has quit [Quit: Visitor from www.linuxcnc.org]
[19:06:27] <PCW> Well the 5V PIC chip had a hole in it so it looked like power was applied backwards or overvoltage was applied and its was modified (mounting holes drilled out)
[19:06:29] <PCW> so I guess they considered it out of warranty for those reasons. We can check if its repairable
[19:07:13] <Danimal_garage> they said it was because i drilled the holes out for some standoffs
[19:07:31] <Danimal_garage> which apparently means all my cards are out of warranty
[19:08:21] <Danimal_garage> cant find standoffs with screws small enough to fit through the holes anywhere
[19:09:11] <PCW> You cant find 4-40 standoffs?
[19:09:33] <Danimal_garage> nope
[19:09:48] <Danimal_garage> not locally anyways
[19:09:50] <Connor> Were are you located?
[19:09:56] <Danimal_garage> San Diego
[19:10:17] <Connor> McMaster has them, as does Grainger.
[19:10:35] <Connor> Heck, I think even Jameco and DigiKey have them too.
[19:10:43] <Danimal_garage> i said locally
[19:10:47] <Jymmm> Danimal_garage: http://www.murphyjunk.bizland.com/
[19:10:51] <Connor> I missed the local part.
[19:11:30] <Jymmm> http://www.sdelectronicsupply.com/
[19:11:38] <Danimal_garage> thats not local
[19:11:48] <Danimal_garage> san diego is friggin huge
[19:12:10] <Jymmm> Danimal_garage: whaaaa whaaaa whaaaa
[19:12:45] <Connor> go to hardware store, buy alum round stock.. and 4-40 taps. and make your own.
[19:13:16] <Danimal_garage> and i've been there... it's basically a garage full of surplus stuff he got from somewhere
[19:14:31] <Jymmm> you want threaded? stud? height?
[19:14:54] <Danimal_garage> Jymmm: i don't want anything, my machines have been built for years
[19:15:06] -!- sumpfralle1 [sumpfralle1!~lars@] has joined #emc
[19:15:24] <Danimal_garage> but at the time, there was nothing but 6-32 standoffs locally when i needed them
[19:16:09] <Jymmm> Grainger
[19:16:10] <Danimal_garage> so i opened up the holes in the boards .02" or so to make room for them
[19:16:44] <Jymmm> http://www.grainger.com/Grainger/MICRO-PLASTICS-Round-Standoff-3ZMX2?cm_mmc=GoogleBase-_-Fasteners-_-Spacers%20and%20Standoffs-_-3ZMX2&ci_src=14110944&ci_sku=3ZMX2
[19:17:01] <Connor> Sorry, reading back through the thread..
[19:17:11] <Jymmm> Connor: dont bother
[19:17:20] <Connor> shrug
[19:17:26] <Connor> moving on.. :)
[19:17:27] -!- Eartaker [Eartaker!~Eartaker@c-67-185-76-100.hsd1.wa.comcast.net] has joined #emc
[19:17:31] -!- sumpfralle has quit [Ping timeout: 250 seconds]
[19:18:01] <Danimal_garage> the odds og grainger having anything in stock that i actually need locally is slim to none
[19:18:05] <Danimal_garage> of*
[19:18:06] <Jymmm> Connor: Danimal_garage couldn't find 4-40 standoffs years ago and is still whining about it to this day =)
[19:18:17] <Connor> Yea.
[19:18:44] <Connor> Okay, can someone suggest ways of finishing aluminum parts?
[19:18:53] <Danimal_garage> i'm not whining about anything, just saying i had to drill the boards out at the time so i could use the standoffs that i could get when i needed them
[19:18:54] <Jymmm> wire brush
[19:19:04] <Connor> I almost wish I had a tumbler.. I would thrown them into that..
[19:19:13] <PCW> Lye, that will finish them...
[19:19:13] <Danimal_garage> anodizing
[19:19:24] <Danimal_garage> drain cleaner
[19:19:33] <Connor> Danimal_garage: Something I can do at the house..
[19:19:38] <Jymmm> Connor: http://image.made-in-china.com/2f0j00NeMtZKOFQkoQ/Crimed-Wire-Cup-Brush-with-Shank.jpg
[19:19:39] <Connor> Drain Cleaner? what will that do ?
[19:19:41] <Danimal_garage> i do anodizing at my house
[19:19:47] <Danimal_garage> disolve it lol
[19:20:17] <Danimal_garage> all you need is battery acid and a power supply to anodize
[19:20:37] <Jymmm> and a damn good exhaust fan
[19:21:27] <Danimal_garage> not really
[19:21:49] <Danimal_garage> it's very diluted battery acid
[19:22:02] <Danimal_garage> so far less bad than charging a battery
[19:22:08] <Jymmm> Connor: wire cup brush works if it's a one off project
[19:22:41] <Connor> It's my parts for my standoffs.. I kinda want a dull look, and want to match the color of my machine, either the green, or tan.
[19:22:49] <Connor> and stepper mounts.
[19:23:00] <Jymmm> Danimal_garage: I don't muck around with acid without a exhaust fan in place, at least indoors.
[19:23:21] <PCW> Danimal: I thought they swapped a 7I33TA for the 7I48, did they charge you for the 7I33TA? . If so we can see if the 7I48 is repairable
[19:23:49] <PCW> No salad dressing for jymmm
[19:23:50] <Danimal_garage> PCW: yes i was charged because of the holes drilled out
[19:24:43] <Danimal_garage> according to the email i got anyways
[19:24:45] <PCW> I think also because you wanted a different card
[19:25:04] <Jymmm> pcw_home: No, I have the exhaust fan running anytime I make Basalmic vinaigrette, you should see the spinach go flying!
[19:25:46] -!- automata has quit [Ping timeout: 250 seconds]
[19:26:20] <Jymmm> Connor: Send your parts to Danimal_garage, he'll annodize them Hulk Green for ya!
[19:26:40] <frysteev> ooooooo
[19:27:09] <frysteev> i have a billet aluminum toilet seat i have been looking to get anodized
[19:27:10] <Danimal_garage> well the 7i48 wouldnt work, for my machine, which is why i wanted the 7i33, it kept pulsing while feeding
[19:27:26] <Jymmm> frysteev: I don't want to know.
[19:27:57] <Danimal_garage> i couldn't risk having my machine down for days while i tried to figure it out.
[19:28:22] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc
[19:29:09] <Jymmm> pulsing?! consistent, or like funky power?
[19:29:50] <Danimal_garage> consistent
[19:30:04] <Danimal_garage> at higher velocities it would ferror
[19:30:09] <Jymmm> Danimal_garage: You're not building a second controller? But modifying the existing one?
[19:30:54] <Danimal_garage> Jymmm: i was going to swap out the 7i33 in my mill for the 7i48 and put the 7i33 from the mill into my little cnc engraver i was building
[19:31:03] <frysteev> is the 7148 pretty plug and play with emc?
[19:31:11] <Jymmm> Danimal_garage: ah
[19:31:29] <Danimal_garage> i figured down the road it would be nice to have the extra channels for a 4th/5th axis
[19:32:05] <Jymmm> OH... Popeyes is having 8 pcs of chicken for $4.99 today only
[19:32:28] <Danimal_garage> but i need my machine running so i can pay the bills, i can't afford to have it down to figure out a glitch
[19:33:57] <Jymmm> Danimal_garage: What if you build a new controller box/pc/etc and just swap out cables between the mill and mini mill as needed? Then you won't have to have so much down time while figuring it out?
[19:34:40] <Jymmm> If you get the new one worked out, then you can just move the mill controller to the mini mill
[19:35:11] <Danimal_garage> because the terminals are not layed out the same, it's a PITA to rewire the machine every time
[19:35:12] <PCW> I suspect you had insufficient power at the 7I48 (causing the "pulsing") and something went wrong when you tried to apply 5V to the 7I48
[19:35:25] <Jymmm> Danimal_garage: Ah, ok.
[19:35:52] -!- psha has quit [Quit: leaving]
[19:35:57] <Danimal_garage> PCW: i don't see how, the only power leaving the 7i48 is for the analog outputs
[19:36:05] <Jymmm> Danimal_garage: YAY for Standards!
[19:36:25] <Danimal_garage> there's no power going to any encoders or anything
[19:36:44] <Danimal_garage> the drives supply the 5v for that
[19:36:53] <PCW> This was the 5V power on th e7I48 5V chips dont get holes in them without some provocation
[19:37:04] <Danimal_garage> and it did it with only one channel hooked up
[19:37:27] <Danimal_garage> i'm refering to the insufficient power you suggested
[19:37:28] <Jymmm> It's easy enough to check the current draw
[19:38:20] <PCW> The 7I48 draws enough power that it needs almost always needs he external supply
[19:39:00] -!- ssloy has quit [Quit: leaving]
[19:39:08] <Jymmm> Danimal_garage: What is your 5v source?
[19:39:35] <Danimal_garage> it was from the ribbon cable from the 5i20
[19:39:42] <Danimal_garage> just like the 7i33
[19:41:41] <PCW> It takes somewhere around 300 mA with no external load. whether you can get away with flat cable power also depends on how long your flat cable are
[19:42:22] <Danimal_garage> if the 7i48 can't even run one channel, even without having to supply power to any encoders, i don't see the point of having the option to run from the 5v off the ribbon cable
[19:42:41] <Danimal_garage> my ribbon cable was about a foot long
[19:42:51] <Danimal_garage> 18" max
[19:43:34] <Danimal_garage> not trying to bust your balls, but just saying, it can cause confusion
[19:43:46] <PCW> It may also have been damaged somehow we cant duplicate your problems with the 7I48
[19:44:10] <Danimal_garage> is it possible it just doesn't play well with digital drives?
[19:44:41] <PCW> No its known to do just fine with digital drives
[19:45:22] <Danimal_garage> even ones that digitally produce the encoder signal?
[19:45:42] <Danimal_garage> not sure how that would matter
[19:46:26] <Danimal_garage> my servos have resolvers and the drives create a quadrature encoder output
[19:48:06] <frysteev> you have a resolver to quard converter?
[19:48:06] -!- Danimal_garage has quit [Read error: Connection reset by peer]
[19:48:11] <frysteev> err quad
[19:48:34] -!- Danimal_garage [Danimal_garage!~kvirc@ip68-111-139-153.sd.sd.cox.net] has joined #emc
[19:48:40] <PCW> Dont see how it could.
[19:48:42] <PCW> There is one thing that might cause a tuning difference, if you are using the encoders velocity output, (as opposed to the D/DT method)
[19:48:43] <PCW> theres a bug in muxed encoders that would make the derivative feedback too small (you would need to multiply D by 6 to fix)
[19:49:01] * frysteev is waiting for emc support of the 7149 :
[19:49:03] <frysteev> :P
[19:52:43] <andypugh> It's coming, I think.
[19:53:21] <frysteev> ooo when?
[19:53:36] <andypugh> Hang on, I think it is already supported?
[19:53:41] <frysteev> cuz im on the verge of using a 7148 with quad converters,
[19:54:07] <frysteev> rather the 4149, cuyz then i oculd lose a bunch of equipment
[19:54:10] <andypugh> No, sorry, getting my numbers confused.
[19:54:31] <andypugh> 7i49 is on KimK's list, I think.
[19:55:07] <PCW> No not supported yet (should be rather simple since it just has 32 bit position and velocity registers at the driver interface)
[19:55:08] <frysteev> hmm
[19:55:30] <frysteev> what does that mean? :P
[19:55:44] <andypugh> I think the debate is how to incorporate an absolute encoder scheme into EMC2.
[19:56:14] <PCW> Yeah thats the main issue (same with SSI, BISS etc)
[19:56:55] <andypugh> It can simply export rawcounts and position, and cover most uses. But how to manage index latch and reset is more inbvolved.
[19:57:15] <frysteev> h,,
[19:57:18] <frysteev> hmmmm
[19:58:15] <andypugh> PCW: Do you mind if I change your else/elesif/elseif/else port-width code to use a "switch" structure?
[19:59:30] <PCW> Sure my C knowledge consists of cut/pasting nearby things that look like they might do the right thing...
[20:02:31] <PCW> saved that patch in the nick of time (cycled power to try new code with 5I20 and hard drive died unresurrectably)
[20:02:32] <PCW>
[20:02:51] -!- isssy [isssy!~isssy@78-83-51-185.spectrumnet.bg] has joined #emc
[20:02:52] <isssy> 5
[20:05:14] <PCW> 6
[20:06:00] <KimK> If I start first with delivery of only the (default?) incremental encoder mode for the 7i49, that should make things move along a little faster. How busy are you this weekend, Andy?
[20:06:37] <Danimal_garage> not sure of the type of output from the drives as far as the encoder output goes
[20:06:42] -!- Danimal_garage has quit [Read error: Connection reset by peer]
[20:06:48] -!- El_Matarife [El_Matarife!~El_Matari@adsl-68-88-67-186.dsl.rcsntx.swbell.net] has joined #emc
[20:07:09] -!- Danimal_garage [Danimal_garage!~kvirc@ip68-111-139-153.sd.sd.cox.net] has joined #emc
[20:09:07] <andypugh> Not very busy, and it's a 3-day weekend
[20:13:05] <Jymmm> andypugh: Oh come on, you don't like 85,000 else/elseif statements going on for infinity for so long that it takes 362 hours, 27 minutes, and 19 seconds to scroll down half way?
[20:13:20] <KimK> andypugh: Excellent, I'll come back and chat with you about this as soon as I am able. I've got to run an errand first.
[20:13:24] <KimK> Danimal_garage: It's a popular scheme to have a resolver on a servomotor/drive set and then have the drive provide a simulated encoder follower output for the computer control. It solves a lot of problems with knowing the initial orientation of the servo (ask Andy about that, lol!).
[20:16:21] <Danimal_garage> KimK: yea, it works ok, however it's not pefect. If i push the drives a little, they get warm which causes the count to go horribly wrong
[20:17:00] -!- Mjolinor [Mjolinor!~Mjolinor@cpc1-burn3-0-0-cust572.10-1.cable.virginmedia.com] has joined #emc
[20:17:03] <Danimal_garage> at least if i had real encoders going to EMC2, it wouldn't loose count because the drives got above 90f
[20:17:04] <andypugh> The resolver-style encoder could just duplicate all the pins of a quadrature encoder, but it might make sense for it to return from an index-latch immediately.
[20:17:24] <andypugh> Resolvers are meant to be tougher than encoders
[20:18:07] <Danimal_garage> the resolvers work fine, it's the drives that are flakey
[20:18:08] <PCW> Yeah the index is always latched :-)
[20:19:17] <andypugh> cradek: During lathe-threading, what would happen if the index-latch went low, and returned a count of half-a-turn? (normally I assume that the count seen on index-latch is a small fraction of a turn?)
[20:19:22] <Danimal_garage> i'm actually planning on gearing down the servos and mounting encoders to the ball screws so i can bypass the encoder output of the drives
[20:19:50] <andypugh> Linear scales might work even better?
[20:21:26] <PCW> I guess you could just simulate idex latch by checking for crossing resolver 0
[20:21:38] <PCW> index
[20:22:01] <andypugh> Yes, that is the other approach. I guess it might be safer.
[20:23:10] <andypugh> PCW: If I pass you the patch back, are you in a position to then re-test it with the 5i25?
[20:23:27] <PCW> Yep
[20:24:16] <PCW> Um i did a git commit so not sure what exactly i would need to do
[20:24:18] <Danimal_garage> don't want linear scales, too many chips and too much coolant
[20:24:38] <Danimal_garage> the coolant will eat up any rubber seals on the scales
[20:25:06] <Danimal_garage> at least encoders are easier to cover
[20:25:10] <PCW> Will a pull overwrite my changes so I can apply your patch?
[20:25:16] <Danimal_garage> do linear scales even have an index line?
[20:25:28] <PCW> Yes usually
[20:26:15] <Danimal_garage> ah cool didnt know that
[20:26:18] <andypugh> Did you create a new branch, or make changes in master (or the 2.5 branch?)
[20:28:00] <andypugh> In any case, I think git reset --hard v2.5_branch will put you back to "normal"
[20:31:15] <PCW> changes to local 2.6pre = master
[20:32:38] <andypugh> OK, I am working on it on a v2.5 base, as the hope is that it will go out with the 2.5 release.
[20:33:17] <andypugh> But you can just git reset --hard master
[20:33:32] <andypugh> And I think that the patch will work with either.
[20:35:02] <PCW> OK I'll make a local 2.5 (do you recal the full name?)
[20:35:35] <andypugh> v2,5_branch
[20:35:44] <andypugh> (sorry, v2.5_branch)
[20:36:24] <andypugh> I think that git checkout origin/v2.5_branch should create you a new tracking branch, but my git-fu is weak.
[20:36:38] <PCW> Wonder if its a vain hope that the sserial discovery stuff will get in 2.5?
[20:37:19] <PCW> Weve had about a billion bugs but will send you test cards today at long last
[20:37:29] <andypugh> It might make it, it is part of a new feature, so can't break anything that curently works.
[20:37:44] -!- isssy has quit [Quit: Visitor from www.linuxcnc.org]
[20:38:09] <andypugh> But that is a decision for Mr C
[20:38:11] -!- e-ndy has quit [Quit: Ex-Chat]
[20:38:32] <PCW> shouldn't it just means the driver needs to root around in the remotes database for name info
[20:38:51] <anonimasu> hmm, anyone got a clue who wrote my name earler this week my log dosent go far enough to show...
[20:39:55] <PCW> plus data size object size, scaling and packing order in the mailbox regs
[20:40:08] -!- Mjolinor has quit [Quit: Leaving]
[20:43:47] -!- syyl has quit [Quit: Leaving]
[20:45:59] -!- isssy [isssy!~isssy@78-83-51-185.spectrumnet.bg] has joined #emc
[20:54:54] -!- FinboySlick has quit [Quit: Leaving.]
[20:56:40] -!- PCW has quit [Remote host closed the connection]
[21:03:19] -!- isssy has quit [Quit: Visitor from www.linuxcnc.org]
[21:16:53] -!- odiug has quit [Quit: Leaving.]
[21:21:08] -!- vladimirek has quit [Remote host closed the connection]
[21:39:01] <andypugh> What's the simplest way to get the kernel sources for the RTAI/EMC2 kernel?
[21:40:04] -!- n2diy [n2diy!~darryl@] has joined #emc
[21:53:57] -!- sumpfralle [sumpfralle!~lars@g230217250.adsl.alicedsl.de] has joined #emc
[21:55:31] -!- sumpfralle1 has quit [Ping timeout: 258 seconds]
[22:20:12] <Danimal_garage> i hate buying hardware at home depot... $7 for a few scres
[22:20:23] <Danimal_garage> screws*
[22:23:13] -!- n2diy has quit [Quit: Ex-Chat]
[22:26:10] -!- n2diy [n2diy!~darryl@] has joined #emc
[22:30:54] <Danimal_garage> it's weird not seeing JT-Shop in here all day lol
[22:31:49] <andypugh> Danimal_garage: You are not paying $7 for the screws. You are paying them $7 for having the screws there for you
[22:32:08] <Danimal_garage> i know, i know...
[22:32:30] -!- bootnecklad` [bootnecklad`!~bootneckl@host-78-151-88-87.as13285.net] has joined #emc
[22:32:39] <andypugh> Right, night all
[22:32:59] -!- andypugh has quit [Quit: andypugh]
[22:34:41] -!- bootnecklad has quit [Ping timeout: 240 seconds]
[22:43:52] <Jymmm> Danimal_garage: and crappy quality ones too
[22:51:02] -!- mhaberler [mhaberler!~mhaberler@62-47-130-36.adsl.highway.telekom.at] has joined #emc
[22:59:17] <KimK> I want to change a numbered subr (o100) to a named subr (in a separate dir & file, using the new SUBROUTINE_PATH ini setting). But the numbered subr also contains a couple of while loops (o102 while, o101 while, o101 endwhile, o102 endwhile).
[22:59:20] <KimK> Is there a possibility of a name collision if somebody else uses o101 or o102 locally, or is that protected against? Is there a way to make those variables share the new name? Maybe something like <mysubr-101> while, <mysubr-101> endwhile?
[23:10:21] -!- mhaberler has quit [Quit: mhaberler]
[23:10:26] -!- kljsdfhklj has quit [Quit: kljsdfhklj]
[23:17:18] <frysteev> KimK: i believe i read ealier that you may be working on the mesa 7149?
[23:17:48] -!- kljsdfhklj [kljsdfhklj!~ln@189-30-67-56.fnsce704.dsl.brasiltelecom.net.br] has joined #emc
[23:19:28] <KimK> frysteev: Well, yes, but a lot of other stuff happened in the meantime. This weekend is looking good, though. How can I help you?
[23:20:32] <frysteev> i was looking at the 7149 for my project, but was told it wasnt emc supported yet, then today your name was mentioned,
[23:21:05] <frysteev> basically just curious on the status of that project,
[23:21:09] <KimK> What would you like to do with the 7i49?
[23:21:42] <frysteev> drive my servo boards, but more importanly it has the resolver interface,
[23:26:46] <frysteev> (right now my planned setup uses a couple of converters, so simplfying it would be aweosme
[23:27:11] -!- robh__ has quit [Ping timeout: 240 seconds]
[23:28:20] <KimK> And all you want to do is use a resolver (up to x6) for an encoder?
[23:30:35] <frysteev> correct, im working on retrofitting a robot arm to emc, (asea irb6)
[23:32:33] <KimK> Ah, a robot arm. How many joints? Only one motor/resolver per joint?
[23:33:36] <KimK> Other than kinematics, the joints/motors are otherwise unrelated/uncoupled?
[23:33:42] <frysteev> 5 axis, so 5 servos/resolvers
[23:34:03] <frysteev> one motor per axis
[23:35:03] <frysteev> http://www.flickr.com/photos/megacyclelabs/6037366638/
[23:37:12] <Danimal_garage> i think these linear slides are going to beef up my z axis a ton
[23:37:23] -!- n2diy has quit [Quit: Ex-Chat]
[23:38:11] <frysteev> mmmm beef, its whats for dinner
[23:38:12] <KimK> frysteev: OK, that sounds like a standard encoder-type application. Are the resolvers are the black cans and the motors the larger grey/brown cans? I see two black cans that look like they're directly coupled off of the motor, any "tricks" hiding in the other three joints?
[23:39:05] <frysteev> the resolv ers are in under those black plastic covers,
[23:39:22] <frysteev> all servos, and resolvers at the same model
[23:41:27] -!- n2diy [n2diy!~darryl@] has joined #emc
[23:41:41] <KimK> frysteev: OK, sounds pretty standard. Let me see what progress I can make this weekend and I'll give you an answer afterward (maybe Tuesday?). Will that do?
[23:42:23] <frysteev> totally, i was more curious if this was an active project or something that i should check back next year :P
[23:43:57] -!- El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]
[23:43:59] <frysteev> woot!
[23:44:14] <KimK> That's difficult to say at this point. But I can give you an answer on Tuesday.
[23:44:34] <frysteev> aweosme, thankyou
[23:45:33] <KimK> Sure. Sorry I don't have it done already, but other stuff has kept me from it up to now.
[23:47:35] <KimK> What do you plan to do with the robot arm?
[23:50:14] bootnecklad` is now known as bootnecklad
[23:54:22] <frysteev> plasma cutting primarily,
[23:54:48] <frysteev> and usuing it as a test bed for other robotic evilness,
[23:55:05] -!- PCW [PCW!~chatzilla@] has joined #emc
[23:56:51] <KimK> OK. But let me know if you start building lots of robots inside a hollowed-out volcano.
[23:57:14] <Danimal_garage> lol
[23:58:04] -!- WalterN [WalterN!~walter@host-174-45-106-117.bzm-mt.client.bresnan.net] has joined #emc
[23:58:20] <frysteev> KimK: maybe,,,
[23:58:43] <frysteev> i want a beer tap on the end of it too, :P