Back
[00:09:18] <Jymmm> NIST Chemistry Webbook
http://webbook.nist.gov/chemistry/
[00:10:10] <Jymmm> Thermophysical Properties of Fluid Systems
http://webbook.nist.gov/chemistry/fluid/
[02:06:35] <Loetmichel> morinin'
[02:06:49] <Loetmichel> -i
[02:26:30] <Deejay> moin
[05:28:25] <jthornton> morning
[05:28:40] <archivist> moaning
[05:31:41] * Loetmichel hands archivist a triple espresso
[05:36:03] <archivist> I just had mid morning coffee to warm my fingers up
[05:40:11] <jthornton> we have an unseasonable high of 52F today
[08:38:07] <R2E4> What is involved in creating a linuxcnc sim? Can I install it from the live cd and then upgrade/change?
[08:46:54] <FinboySlick> R2E4: Technically, you don't have to install it, you can run it straight from the livecd.
[08:47:18] <FinboySlick> (or at least you could last time I played with the livecd)
[08:48:28] <R2E4> The simulator?
[08:49:27] <FinboySlick> I'm not familiar with the simulator bit.
[08:49:52] <R2E4> I would rather install it. I found the docs for building it. I have the real linuxCNC in the shop now connected to my control panel so I wont be able to work on it dureing the day, so at work I would like to test and program in a simulator enviroment.
[08:53:08] <archivist> I have played with it straight from the livecd too
[08:57:23] <R2E4> I have the computer and 7i77 installed on control panel and wiring ins and outs. THats a long process to make it nice and neat
[08:59:10] <R2E4> Can I install linuxcnc from live cd then compile the Linuxcnc SIM on top of it?
[09:28:31] <cradek> sim can mean two things: a nonrealtime build of linuxcnc that doesn't require a special kernel, or a config that does not control any hardware
[09:28:57] <cradek> you can certainly install or just boot the livecd, and then run any of the sim configs
[09:29:27] <cradek> or, if you already have a normal nonrealtime linux machine, you can install our sim-built debs or build a sim mode linuxcnc for yourself
[09:38:37] <R2E4> So the linuxcnc install from the Livecd will work if using sim configs?
[09:41:28] <cradek> yep it'll work fine
[09:42:52] <R2E4> Whats the diff if you build it with --enable-simulator or from live CD?
[09:43:16] <cradek> --enable-simulator causes a build that doesn't need a realtime kernel
[09:43:27] <R2E4> ok
[09:43:35] <cradek> or you can use the -sim debs; you don't have to build it
[09:44:01] <R2E4> I'm installing now from livecd so I'll just locate the sim configs
[09:47:42] <R2E4> cradek: Is there a document that outlines the basic logic process for linuxcnc with 3 axis machine or any of that matter/ Re: powerup, e-stop set before power up, Drive(axis) ready before power machine on, etc...?
[09:49:01] <R2E4> I'm getting close to interfacing with the drives, and want the logic setup before.
[09:49:09] <archivist> have you looked at the integrators manual
[09:49:13] <cradek> I'm not sure. Some of those things are going to be different based on your machine's needs
[09:50:41] <R2E4> I'll gander through the integrators manual and see if theres a generic basic process.
[09:51:22] <R2E4> Normally do you have an input "Drive Ready" before you enable the drives?
[09:51:29] <cradek> it's hard to write overview-type docs, we'd rather answer specific questions
[09:51:32] <R2E4> and if not dont enable?
[09:52:04] <cradek> what would drive ready represent? you mean a signal from the drives themselves?
[09:52:23] <cradek> I have not seen a drive that has a "I'm ready to be enabled" signal
[09:52:29] <R2E4> yes..... There is a drive ready signal output of the drive, going to the old control.
[09:53:25] <R2E4> http://irmtl.com/LinuxCNC/VM40/Docs/sanyodenkibl827.pdf pins 5,6,7 on CN1
[09:53:30] <cradek> that might mean there is no error - like "I haven't faulted"?
[09:53:38] <emcPT> Good afternoon. Yestarday I asked for help here, and I would just like to inform that all is working as we wanted. Thank you to all.
[09:53:48] <cradek> yay!
[09:53:58] <archivist> another happy bunny
[09:54:11] <R2E4> alot of bunnies here....
[09:54:15] <cradek> R2E4: that pdf is broken
[09:54:34] <R2E4> What do you mean broken?
[09:54:45] <cradek> Error (18737): Unexpected EOF in JBIG2 stream
[09:54:57] <cradek> it's only 19 kbytes and doesn't display anything for me
[09:55:05] <R2E4> IT comes up for me....lol
[09:55:17] <archivist> I get the image ok and no error
[09:55:22] <cradek> huh
[09:55:33] <cradek> f7e168459513d627d84eec8871af7ad5 sanyodenkibl827.pdf
[09:55:38] <R2E4> http://irmtl.com/LinuxCNC/VM40/Docs/
[09:55:42] <archivist> you need to be on an older OS :)
[09:55:49] <cradek> do you have that same sum?
[09:56:19] <cradek> ugh, none of those open
[09:56:44] <archivist> f7e168459513d627d84eec8871af7ad5 sanyodenkibl827.pdf
[09:57:08] <cradek> ha, xpdf opens them fine
[09:57:13] <cradek> you're right about needing older
[09:57:22] <archivist> I am using evince
[09:57:36] <archivist> I was joking :)
[09:57:45] <archivist> but am on 8.04 here
[09:57:46] <R2E4> They are really old schematics, so you need an older pdf reader....lol
[09:58:19] <cradek> R2E4: not much clue about what those READY lines are except it looks like one is NO and one is NC
[09:58:27] <cradek> I would guess that's the fault output
[09:59:11] <R2E4> fault should be clear before you enable them I would think or something wouldnt work and you would be wondering why......
[09:59:20] <cradek> and what is OH1/OH2
[09:59:34] <R2E4> could give some kind of lite on panel saying drive not ready?
[09:59:51] <R2E4> OH! OH2 is overheat alarm
[10:00:16] <cradek> motion has axis.N.amplifier-fault-in pins, it'll cause a stop and disable, just like a following error
[10:01:16] <pcw_home> depending on the drives you may not get drive ready unless you enable the drives
[10:01:29] <R2E4> IS there some kind of indication in axis, or should I do that with a lite on panel?
[10:01:45] <cradek> R2E4: yes AXIS will give a message if the amplifier faults
[10:02:40] <R2E4> pcw_home: I had that thought also,..... but didnt know. i can check cause it powers up without RGSO which is ENA, I'll check status of those.
[10:03:31] <pcw_home> (Fanuc is like that)
[10:03:52] <pcw_home> (disabled is a fault)
[10:03:57] <cradek> sometimes getting boostrap can be tricky. especially if each device's fault output can disable all the other devices (like if you want amp fault to disable the spindle vfd, and also vfd fault to disable the amps)
[10:04:27] <R2E4> pcw_home: Mounted 7i77 on panel last nioght and started wiring..... coming along nicely untill I powered up an no lites on 7i77. So I quickly shut off power and realized I wired power in backwards cause I installed the 7i77 upside down.
[10:05:04] <cradek> yikes
[10:05:09] <R2E4> yeah, thats why I was looking for logic scenario..... hehe
[10:05:28] <R2E4> I corrected the polarity and all is well.
[10:05:35] <cradek> that's some stout hardware
[10:06:19] <R2E4> I thought for sure I blew it up. Its solid.....
[10:06:53] <pcw_home> It has protection diodes (but there are always ways)
[10:07:12] <R2E4> Thank god for those babies!!!
[10:07:19] <cradek> nah, thank pcw
[10:07:37] <R2E4> smae thing....lol
[10:07:46] <pcw_home> we had a customer that blew up 4 7I77s Still cant quite figure out how
[10:07:54] <R2E4> ouch!
[10:07:57] <archivist> determination
[10:08:16] <cradek> more like hubris, I bet
[10:08:48] <Jymmm> pcw_home 220VAC on the 5VDC line?
[10:08:50] <pcw_home> (field power switching regulators blown, rated at 50V in customer says only 24V)
[10:09:07] <archivist> I used to repair kits that our customers had built, I have seen some customer caused problems too
[10:09:15] <pcw_home> maybe no output capacitor on PS
[10:09:44] <CaptHindsight> spikes to >100V for several uS might be enough to do it
[10:09:52] <pcw_home> yep
[10:10:10] <archivist> I had all the gnd track blown off one board
[10:10:39] <CaptHindsight> people think automotive is only 12V but on older cars it would easily spike to >50V
[10:11:04] <pcw_home> we had someone blow up a series of 7I29s (no PS output capacitor)
[10:11:06] <archivist> only 50 :)
[10:11:09] <Jymmm> pcw_home: 120VAC --> [24V Sec / 120 Pri ] --> 600VAC =)
[10:12:35] <pcw_home> just a normal 24V control transformer into a bridge will generate really large spike if turned off at max current
[10:13:57] <CaptHindsight> If there is a way to hook something up wrong people will find a way
[10:14:14] <Jymmm> pcw_home: Heh, I remember talking to Marriss once about his "Vampire" (you can't kill them) drives.... Guarantee to kill them by unplugging motors while energized =)
[10:15:12] <R2E4> My 24v power supplie when shutoff shuts off, the relay clicks and red lite follows the click. Should I attempt to fix that?
[10:15:48] <Jymmm> R2E4: rephrase that please
[10:16:24] <pcw_home> well arcs are really nasty (charge up wiring capacitance with stored inductive
[10:16:26] <pcw_home> energy and then you get a picosecond rise time HV spike )
[10:16:54] <R2E4> 24vdc power supplies, when I remove AC power, dc stays in the caps, the ac fail and batt fail relay cycles causing dc spike
[10:17:21] <Loetmichel> *gna* i should get new tires ASAP... every time there is as much as 2 drops of rain on the street i see that darn yellow triangle with "ASR" in it blinking at the dash as soo as i accelerate :-(
[10:17:23] <R2E4> which the 7i77 sees it
[10:17:38] <CaptHindsight> Jymmm: should questions begin with "Hey Scro"? :)
[10:17:54] <Jymmm> Loetmichel: removeing the bulb would be cheaper
[10:18:29] <Jymmm> R2E4: is there a diode across the relay coil?
[10:18:49] <R2E4> ITs onboard, I can check, I have a boat load of them here.....
[10:19:12] <Jymmm> R2E4: Do you have a bleed resisotor on the PS ?
[10:20:09] <Loetmichel> Jymmm: will not prevent the car from drifting around every corner. thats the front motor and rwd "problem" ;-)
[10:20:15] <R2E4> no bleed resistor, guess I could install one or put relay on the AC power and pass dc voltage though it.
[10:20:30] * JT-Shop just discovered in python you can store objects in a dictionary :)
[10:20:42] <Jymmm> Loetmichel: Hey, ppl pay good money for drifting, you're getting it for free!!!
[10:21:05] <R2E4> We put bleed resistors on tube amps I use to build, so when power shuts of it discharges caps.
[10:21:25] <kbarry> The other day i tried machining some poplar (a 3d relief)
[10:21:36] * JT-Shop discharges caps with his finger usually
[10:21:44] <kbarry> Poplar seemed too coft to make clean cuts, and the whole thing was very very fuzzy,
[10:22:05] <kbarry> i tried the same thing in alder i had onhand and it came out perfect,
[10:22:11] <R2E4> I've been bit by that, 400+vdc from the tube VCC PS
[10:22:14] <CaptHindsight> can't people figure out when their wheels are starting to slip on their own? Do they really need a light on the dash?
[10:22:20] <R2E4> scared the be-jesus out of me.
[10:22:40] <archivist> kbarry, blunt cutters/spindle not fast enough ?
[10:22:55] <kbarry> Looking for recommendations for 3d carving wood, paint gradable.
[10:22:56] <Jymmm> kbarry: Use a scotchbright (kitchen green) pad to remove the fuzzys
[10:23:06] <kbarry> Jymmm: well, i thought that might be the case, except it cut fine in alder,
[10:23:18] <Loetmichel> CaptHindsight: the light is a circuit that will choke the engine.
[10:23:23] <R2E4> kbarry, babmoo is amazing
[10:23:33] <archivist> also was the wood green or dried
[10:23:35] <kbarry> Well, its for a client, not myself,
[10:23:41] <R2E4> sands easy and carves really nice.
[10:23:44] <kbarry> It was dried
[10:23:50] <Loetmichel> its just to inform jou that the wheels would have made a burnout if the electronics would not have intervened
[10:23:56] <kbarry> I was thinkign Beech or Birch
[10:24:12] <kbarry> just wanted to know what other were using when 3D carving for a paint-grade application
[10:24:25] <kbarry> I'm ni California (IF that helps in making a recommendation)
[10:24:42] <R2E4> Birch carves nice.
[10:24:44] <Jymmm> sign foam
[10:25:31] <kbarry> Its for a freize, would Signfoam be appropriate? I'd be affraid it might be broken too easily
[10:25:47] <kbarry> But, what is sign foam in comparison to wood pricewise?
[10:25:52] <Jymmm> a what?
[10:26:04] <CaptHindsight> Loetmichel: what if you want to heat up your rear tires or wear them down to the belts? It stops you?
http://image.modified.com/f/images/17364769/0309_sccp_03_z%2Bdrag_racing_car%2Bburnout.jpg
[10:26:21] <Loetmichel> it can be switched off ;-)
[10:26:24] <CaptHindsight> ah
[10:26:30] <kbarry> http://mediacdn.shopatron.com/media/mfg/11672/product_image/c3fe44ffe2dedeb43c653db28fc15230.jpg%3F1383449262
[10:26:35] <kbarry> carved molding
[10:26:50] <R2E4> sign foam is hi density urethane
[10:27:05] <Loetmichel> can you imagine this "old fart car" on a racetrack? i cant ;-) ->
http://www.cyrom.org/palbum/main.php?g2_itemId=11969
[10:27:39] <jdh> MDF
[10:27:42] <kbarry> R2E4, yeah, not a lot of epxerience with it,
[10:28:09] <kbarry> jdh with MDF how would you recommend i clean it up?
[10:28:11] <Jymmm> jdh: The issue is that MDF can't be painted where it has exposed edges
[10:28:26] <kbarry> I always get fuzzies, and its difficult to clear them without sanding down details.
[10:28:41] <jdh> Jymm: why not?
[10:28:55] <jdh> good sealer first
[10:29:42] <Jymmm> jdh: Sealers dont work either. IT's like a sponge. The surfaces are compressed to 120PSI, but no thte "meat"
[10:29:43] <R2E4> MDF is nice to work with. I use sealer, and a clear coat, same stuff they use on cards.
[10:30:02] <Jymmm> Yeah, bondo is the only way
[10:30:25] <Jymmm> but with the intricate detailed of a 3D carving, I doubt it would fill properly
[10:31:16] <Jymmm> Well, ok, you can paint it with black (eventually), but that's about it for some odd reason.
[10:32:14] <Loetmichel> Jymmm: MDF can be painted perfectly
[10:32:26] <Jymmm> Loetmichel: not the edges
[10:32:30] <Loetmichel> you just have to use so much paint that it gets soaked
[10:32:31] <Loetmichel> ;-)
[10:32:43] <Jymmm> surface sure, it will even take a thread and hold it well
[10:33:38] <Jymmm> I spoe to one ot the guys in the lab at a MDF Mill is the only reason I know about bondo
[10:33:40] <Loetmichel> and use a 2component paint if you want it to dry eventually ;-)
[10:33:54] <Jymmm> epoxy?
[10:34:07] <Loetmichel> car paint works well
[10:34:20] <Jymmm> the epoxy paint would cost you more than the raw material
[10:34:32] <Loetmichel> it does ;-)
[10:34:34] <R2E4> http://irmtl.com/cnct/pics/molspn.jpg mdf painted rusic kind of
[10:34:35] <Jymmm> at that rate, sign foam would be cheaper
[10:34:40] <Loetmichel> but you cant sai it cant be painted
[10:34:45] <Loetmichel> its just expensive ;-)
[10:34:46] <R2E4> Oak is amazing with 3d carvings
[10:34:54] <R2E4> http://irmtl.com/cnct/pics/deer_scene.jpg
[10:35:13] <Jymmm> Loetmichel: Fine, go paint some and post a pic of the results
[10:35:28] <Loetmichel> let me search a bit, maybe i have one
[10:35:46] <R2E4> The edges are an issue with mdf though
[10:36:19] <Jymmm> except for black
[10:37:10] <CaptHindsight> UV cured urethane, epoxy or polyester will fill MDF
[10:37:30] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=7338&g2_imageViewsIndex=1 <.this one is "painted" with pu resin and then milled over. a better example isnt in stock right now
[10:37:30] <Jymmm> CaptHindsight: at what cost
[10:37:50] <CaptHindsight> I can make it for a few $$/ Kg
[10:38:02] <CaptHindsight> 1 L = ~1Kg
[10:38:37] <CaptHindsight> maybe I should start selling it for handywork
[10:39:02] <Jymmm> Loetmichel: PU Resin, so you are painting plastic
[10:39:08] <R2E4> https://www.facebook.com/media/set/?set=a.165527346907089.34646.165499736909850&type=3 The signs near the bottom are MDF.
[10:40:04] <Loetmichel> Jymmm: was a little "mishap" when milling the tiptank mold
[10:40:17] <Jymmm> This is MDF too...
http://i52.tinypic.com/r1lwra.jpg
[10:40:20] <Loetmichel> hence the yellow stripes in the mdf ;-)
[10:40:29] <Jymmm> No resins, epoxies, etc
[10:40:57] <Loetmichel> was the only example i could find on the instant
[10:41:55] <Jymmm> I tried everything,, you just can't paint mdf edges as you would paint "wood"
[10:42:20] <Loetmichel> tha may be so
[10:42:34] <Loetmichel> because it soaks the paint like a sponge
[10:42:55] <Loetmichel> thats why i said: use a 2k paint/epoxy as "primer"
[10:42:59] <archivist> use a filler first
[10:43:24] <Jymmm> archivist: try it =)
[10:43:50] <CaptHindsight> 2K paint can be 100% solids, they don't have to have any solvents that just evaporate out after applied
[10:43:54] <Jymmm> Loetmichel: but you would never use epoxy on wood normally
[10:44:29] <Loetmichel> mdf is no wood
[10:45:02] <Loetmichel> mdf is dirt from the woodshops flor pressed into something resembling a plank ;-)
[10:45:22] <Jymmm> Loetmichel: "Yes Dear"
[10:45:49] <CaptHindsight> I have oligomers that are very high viscosity, like peanut butter. Smooth on, won't soak in. Cure in seconds and then paint.
[10:46:04] <Loetmichel> but its great for making molds BECAUSE it has no "fibre orientation"
[10:46:32] <Loetmichel> or for making a sacrificial plate on the mill
[10:46:53] <Loetmichel> because you can screw in spax an nearly infinite amounts
[10:49:48] <Jymmm> CaptHindsight: bondo is cheaper
[10:49:56] <Jymmm> and more readily available
[10:50:16] <Jymmm> and less toxic =)
[10:51:00] <kbarry> Hmmm. HDU isn't terribly priced.
[10:51:16] <kbarry> Makes a lot more sense than gluding up a boatload of wood.
[10:51:28] <kbarry> twice the cosft of MDF,
[10:51:46] <kbarry> Question on MDF cleanup from experience.
[10:51:52] <Jymmm> will not rot or mildew, accepts paint readily
[10:51:59] <kbarry> Are you sealing it after machining, then sanfing off fuzzies?
[10:53:03] <Jymmm> I've never had fuzzies on MDF
[10:53:42] <kbarry> But you have 3d machined it?
[10:53:56] <Jymmm> the harley sign is one example
[10:54:06] <Jymmm> it's V-carved
[10:54:35] <Jymmm> I just use a scotchbright pad
[10:54:48] <Jymmm> green
[10:56:22] <kbarry> You ever used Extira? ITs like...... Exterior-grade MDF
[10:58:35] <kbarry> Water PROOF (As in, i cut a piece off, and left it in a texas rainstorm for 3 weeks, didnt change it a bit)
[10:58:58] <kbarry> well, not in the rain for 3 week, its rained several time, but it stayed outside for 3 weeks.
[11:00:40] <Jymmm> Heard of it, never used it
[11:01:09] <Loetmichel> kbarry: sounds interesting.
[11:01:14] <Loetmichel> where to get that in germany?
[11:01:46] <Jymmm> Loetmichel:
http://ExteriorMDFStore.de
[11:02:43] <Loetmichel> doesnt exist ;-)
[11:02:55] <Jymmm> Loetmichel: try harder
[11:03:06] <kbarry> http://www.medite-europe.com/de/products/medite-exterior/
[11:04:32] <Loetmichel> kbarry: nice
[11:04:36] <Loetmichel> will look into it
[11:04:46] <kbarry> I don't knwow what its like, i have only used Extira brand,
[11:04:53] <kbarry> But if its similar, its pretty nice.
[11:04:58] <Loetmichel> may be curing some "coolant issues" on the mill at the company ;-)
[11:05:10] <Jymmm> Loetmichel: Ah, here we go...
http://bit.ly/1dnFxwk
[11:05:11] <kbarry> Your shop won't look like chernobyl
[11:05:39] <Loetmichel> ?
[11:06:00] <kbarry> When I use MFD, my shop ends up with a thin dust layer thicked closest to my amchine
[11:06:11] <kbarry> but still a thin layer even 20 feet away :)
[11:06:40] <Loetmichel> hrhr
[11:07:37] <Loetmichel> you mean like this:
http://www.cyrom.org/palbum/main.php?g2_itemId=4821
[11:08:52] <Loetmichel> (after removing the to thin layer of MDF left
[11:08:55] <Loetmichel> )
[11:09:21] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=4935 <- after glueing on a new layer of 40mm mdf, just milling the 50mm grid back into it ;-)
[11:10:59] <kbarry> http://i.dailymail.co.uk/i/pix/2010/10/27/article-0-0BC95754000005DC-362_634x422.jpg
[11:11:23] <Loetmichel> what happened THERE?
[11:11:31] <kbarry> volcanic ash :)
[11:11:34] <kbarry> not my picture :)
[11:11:34] * archivist admits nothing
[11:12:12] <R2E4> There was a cnc in there carving mdf without a dust collector.
[11:12:35] <archivist> angle grinder with a knotted brush and rusted items and anywhere will look like that
[11:14:17] <Loetmichel> hrhr
[11:14:24] <Loetmichel> i second that, archivist
[11:15:20] <archivist> plus a dose of ethnic snot
[11:17:10] <archivist> although I actually used overalls and mask etc when I did a lorry trailer chassis last year
[11:31:18] <Loetmichel> archivist: tell me about it
[11:31:49] <Loetmichel> i worked nearly four years at that cnc router milling carbon fibre plates...
[11:32:14] <Loetmichel> even though it had a dust collector you sneezed black days after ;-)
[11:43:06] <Jymmm> sounds like you needed a hepa filter and a respirator
[11:45:53] <IchGuckLive> Hi all B)
[11:48:35] <IchGuckLive> Loetmichel: you gone die early if you dont look after your health carfuly
[11:49:53] <Loetmichel> IchGuckLive: much to late i think
[11:50:10] <Loetmichel> Jymmm: the dust vaccum HAD a hepa filter
[11:50:29] <Loetmichel> just the dust ectractor foot didnt get all of the carbon dust
[12:10:23] <tehcereal> hello
[12:10:31] <tehcereal> im having issues with registration
[12:11:10] <tehcereal> i dont get the activation email
[12:11:53] * archivist points at JT-Shop for that
[12:12:19] <archivist> the forum has a lot of spam protection
[12:12:32] <tehcereal> i noticed that :D
[12:13:00] <IchGuckLive> maybe we can help what is your gues
[12:13:10] <archivist> it got upto a few hundred spam messages a day
[12:13:42] <archivist> but we answer live in here
[12:14:13] <tehcereal> i messaged jt-shop
[12:14:14] <IchGuckLive> tehcereal: where are you from in the world USA Europ
[12:14:19] <tehcereal> europe
[12:14:25] <IchGuckLive> im in germany
[12:17:57] <tehcereal> while im here i may ask
[12:18:11] <tehcereal> if someone can help with my hal config
[12:18:17] <tehcereal> and stepgen functions
[12:19:29] <tehcereal> i need to run 3 axis with 12 lpt outputs
[12:19:54] <tehcereal> and 3 endstops
[12:27:38] <IchGuckLive> np
[12:28:05] <IchGuckLive> ther is a stepconf wizard
[12:28:10] <IchGuckLive> did you use it
[12:28:16] <IchGuckLive> tehcereal:
[12:28:26] <tehcereal> i used the stepconf
[12:28:36] <IchGuckLive> so where is the hint
[12:28:38] <tehcereal> but ist only for step/dir?
[12:28:47] <tehcereal> i need a phase control
[12:28:57] <tehcereal> 4 pins per motor
[12:29:00] <tehcereal> :D
[12:29:01] <IchGuckLive> no way with parport
[12:29:10] <jthornton> tehcereal, you forum account is approved
[12:29:28] <IchGuckLive> tehcereal: what driver do you got
[12:30:47] <tehcereal> i have built my own haha
[12:31:00] <tehcereal> i mean "built"
[12:31:09] <IchGuckLive> ist schon klar !!
[12:31:17] <tehcereal> its only 3 l293ds and some resistors and switches
[12:31:34] <tehcereal> the motors are week
[12:31:38] <IchGuckLive> i see
[12:31:53] <IchGuckLive> why not the old fecion L296 /L297
[12:32:02] <IchGuckLive> fashion
[12:32:10] <tehcereal> i had the l293d around
[12:32:16] <IchGuckLive> i guess 3D printer
[12:32:24] <tehcereal> nah
[12:32:30] <tehcereal> just a small plotter
[12:32:43] <tehcereal> its a student project
[12:32:48] <tehcereal> nothing serious
[12:32:59] <tehcereal> it needs just to move
[12:33:00] <tehcereal> haha
[12:33:01] <IchGuckLive> so you need to put 1 more IC infrotz of the L293
[12:34:07] <tehcereal> to get step/dir
[12:34:08] <tehcereal> ?
[12:34:12] <IchGuckLive> yes
[12:34:16] <archivist> tehcereal, you just need to change the stepper type and configure the pins
[12:34:25] <tehcereal> i thought that
[12:34:40] <archivist> bit of hal tomfoolery
[12:34:49] <IchGuckLive> archivist then he needs all 12 outputs
[12:35:05] <archivist> I know, add another parallel port
[12:35:09] <tehcereal> type 10 would be good i guess
[12:35:49] <tehcereal> why another parallel port
[12:36:01] <archivist> for a "getting to know" it will work but not the right way for real work
[12:36:11] <IchGuckLive> tehcereal:
http://www.datasheetcatalog.com/datasheets_pdf/L/2/9/7/L297.shtml
[12:36:22] <IchGuckLive> this woudt do the job for you
[12:36:48] <IchGuckLive> then you also get enable pins for Now extra cost
[12:37:10] <tehcereal> doesn't the parallel portss have 12 output pins
[12:37:31] <tehcereal> but some are inverted
[12:37:35] <archivist> cant remember how many can be out
[12:37:39] <IchGuckLive> tehcereal: Why did you build your wn and not going to use a TB6560 3Axis all in one board for 25Euros
[12:38:10] <IchGuckLive> tehcereal: tha parport is 12/5 so it is ok
[12:38:29] <tehcereal> we cant buy anything else
[12:38:34] <tehcereal> just raw chips
[12:38:37] <IchGuckLive> tehcereal: but then you run out of pins with your system
[12:39:24] <IchGuckLive> then if raw chips go for L296/297 Combination for your layout
[12:39:25] <ktchk> TB6560 have raw chips
[12:39:36] <IchGuckLive> hi to HK
[12:39:42] <IchGuckLive> how is the weather
[12:39:44] <ktchk> Hi
[12:39:50] <ktchk> cold
[12:39:54] <Loetmichel> and tb6560 are MUCH easier to interface
[12:40:07] <CaptHindsight> Jymmm: that's what Bondo is. unsaturated Polyester + talc, clay etc
[12:40:23] <Loetmichel> but ai bit more "wimp" when it comes to loose contacts/overvoltage and stuff
[12:41:34] <tehcereal> we considered about TB6560
[12:42:12] <tehcereal> but their price is the price is higher
[12:42:25] <IchGuckLive> so go for the L297
[12:42:31] <IchGuckLive> 38cent
[12:43:07] <tehcereal> we will do the l297/296 combo if the l293 fails :D
[12:43:20] <tehcereal> we have teams of 2
[12:43:32] <tehcereal> and the cheapest hardware
[12:43:51] <IchGuckLive> L297/L298 !!!
[12:43:53] <tehcereal> wins
[12:43:57] <tehcereal> :D
[12:44:03] <tehcereal> ok ok :D
[12:44:11] <IchGuckLive> 4,5Euros combi but the TB6560 is also on 4Euros
[12:44:56] <tehcereal> but im sure lot of my friends in other teams wil go the l297 combo
[12:44:59] <IchGuckLive> up to you to waste all the pins
[12:45:28] <tehcereal> all it need it to be functional and the cheapest hardware wins
[12:46:04] <IchGuckLive> then china is your friend HK will sell it to you for 15USD
[12:46:30] <ktchk> TB6560 from china is Yuen 11 us$2.00
http://item.taobao.com/item.htm?spm=a230r.1.14.35.XJsYfV&id=14718754978
[12:46:55] <IchGuckLive> oh you bastard you told me 3USD
[12:46:58] <tehcereal> yeah taobao :D but how can i get it from there
[12:47:08] <ktchk> new price
[12:47:14] <IchGuckLive> O.O
[12:47:19] <tehcereal> you own a store ktchk?
[12:47:34] <ktchk> No but I can help
[12:48:04] <IchGuckLive> tehcereal: where are you located then i can send you one
[12:48:14] <tehcereal> croatia
[12:48:21] <IchGuckLive> oh thats far
[12:48:32] <tehcereal> far far away :D
[12:48:33] <IchGuckLive> ask HK for fast delivery
[12:48:39] <ktchk> How many needed
[12:48:53] <tehcereal> 3 i guess
[12:49:15] <ktchk> ok email me
[12:49:23] <IchGuckLive> tehcereal: there are 3 axis on one board
[12:50:08] <ktchk> some single axis board is us$10
[12:50:44] <IchGuckLive> ktchk: i think he needs to build his own board around
[12:50:57] <tehcereal> yeah
[12:51:26] <IchGuckLive> tehcereal: use 150pf for the clock and you are up and running the fastest
[12:53:36] <tehcereal> the problem is the professor looks the price on rs components/ farnel
[12:53:39] <tehcereal> not taobao :D
[12:54:14] <tehcereal> but we can get plus points for the board being able to run proper motors
[12:54:19] <IchGuckLive> tehcereal: show him the China billl
[12:54:20] <tehcereal> over 2A
[12:54:23] <tehcereal> :D
[12:54:25] <ktchk> look pcb only
http://item.taobao.com/item.htm?spm=a230r.1.14.23.yk6i96&id=13243136565
[12:55:32] <ktchk> TB6560 is 3A
[12:57:37] <tehcereal> and we need to design our own scheme and be able to explain what all the components do on the board :D
[12:58:07] <tehcereal> ktchk can you order me something else from taobao
[12:58:09] <tehcereal> haha
[13:02:06] <IchGuckLive> tehcereal:
http://www.pollin.de/shop/dt/NzgwOTk4OTk-/Bauelemente_Bauteile/Aktive_Bauelemente/IC/Schrittmotor_Treiber_L298.html
[13:02:06] <ktchk> go private
[13:02:22] <IchGuckLive> tehcereal: 2,45Euros
[13:02:41] <tehcereal> :D
[13:03:18] <tehcereal> thank you we will sure consider it
[13:03:37] <IchGuckLive> rs croatia 42KN
[13:04:17] <tehcereal> dankeschon
[13:04:25] <IchGuckLive> for better boardlayout HN is 48kn
[13:06:02] <IchGuckLive> the l2971 is 90kn
[13:06:22] <IchGuckLive> http://hr.rsdelivers.com/product/stmicroelectronics/l2971/stepper-motor-controller-l297-1-dip20/0636362.aspx
[13:06:59] <tehcereal> i found that
[13:07:07] <tehcereal> but not the item for 42kn
[13:07:18] <IchGuckLive> and the total TB6560 is 80kn
[13:07:29] <IchGuckLive> so best price on that one chip version
[13:08:21] <IchGuckLive> http://hr.rsdelivers.com/product/toshiba/tb6560ahqo-8/step-motor-driver-25a-pwm-5v-hzip25/7703680.aspx?query=tb6560
[13:08:57] <tehcereal> hmm
[13:09:14] <IchGuckLive> that woudt be your best choice and price winner
[13:09:22] <tehcereal> i mean i dont have to buy from rs
[13:10:18] <IchGuckLive> then buy from HK and get one PCB to feed your chematic
[13:11:59] <IchGuckLive> tehcereal: you also dont need external diodes
[13:12:25] <IchGuckLive> if you want use UDF
[13:12:43] <IchGuckLive> UFD Ultra fast not the F304
[13:14:10] <tehcereal> all the used items must be available to buy from croatia
[13:14:21] <tehcereal> so every webshop that shipps item to croatia
[13:14:25] <tehcereal> or a local store
[13:15:54] <IchGuckLive> http://hr.rsdelivers.com/product/vishay-semiconductor/uf5404-e3-54/rectifier-3a-400v-ultrafast-do-210ad/6289984.aspx?query=uf%205404
[13:16:47] <tehcereal> we will for sure consider all the boards
[13:17:07] <IchGuckLive> as teams work
[13:17:27] <tehcereal> the l297/l298 board the tb6560 and the cheapest l293d :D
[13:17:52] <tehcereal> and will choose the best price/performance
[13:17:56] <tehcereal> board and motors
[13:18:30] <IchGuckLive> motors is best to use the Epson laserprinter ASTRO 44W
[13:19:37] <tehcereal> yeah we already went to the used item makret
[13:19:39] <IchGuckLive> 23LM-C355-44W
[13:19:42] <tehcereal> to find some broken printers
[13:20:01] <IchGuckLive> this works for me on 100 mashines Tb6560
[13:23:08] <tehcereal> thank you for the help
[13:23:12] <tehcereal> :D
[13:23:29] <tehcereal> but first i want to try to make the l293d work
[13:23:37] <tehcereal> i already built the pcb :S
[13:23:44] <tehcereal> so i need to test it
[13:23:48] <tehcereal> :D
[13:24:13] <tehcereal> the next is the l297 combo
[13:24:50] <tehcereal> thats i think the best solution indeed
[13:25:01] <tehcereal> more pins left
[13:25:44] <IchGuckLive> http://foengarage.de/cnc4.jpg
[13:26:35] <tehcereal> yours?
[13:26:42] <tehcereal> looks professional
[13:26:51] <IchGuckLive> 100 pice of them in education use
[13:27:07] <IchGuckLive> 400Euros Etch
[13:27:29] <tehcereal> can it mill aluminium
[13:27:32] <IchGuckLive> incl PC !
[13:27:40] <IchGuckLive> no pcb chees
[13:27:52] <IchGuckLive> at 1200mm/min
[13:28:23] <tehcereal> the driver if i see good is the 3 axis tb6560?
[13:28:31] <IchGuckLive> you will not mill alu with a L297/298 that is quit a chalange
[13:28:40] <tehcereal> i know that :D
[13:28:45] <IchGuckLive> yes HK ones
[13:28:57] <ktchk> Right
[13:28:59] <IchGuckLive> not in direct mode
[13:29:01] <tehcereal> and the motors are from the printer or?
[13:29:06] <IchGuckLive> yes
[13:29:15] <IchGuckLive> 100 steppers for 99Euros
[13:29:25] <tehcereal> they look quite big
[13:29:34] <tehcereal> where did you buy them
[13:29:35] <IchGuckLive> 160x160mm
[13:29:59] <IchGuckLive> the mill lathe and the Pc fit a Education table
[13:30:20] <ktchk> IchGuckLive: I have nema17 0.9 deg for US 20 each
[13:30:55] <IchGuckLive> i see but im out of the business now
[13:31:13] <IchGuckLive> will only buld plasmas 3m now
[13:31:20] <ktchk> good for 3d printer
[13:31:35] <IchGuckLive> ok im off BYE for today
[13:31:40] <IchGuckLive> O.O ;-)
[13:32:15] <ktchk> Bye
[14:45:16] <kengu> what is it with those 23LM-C355-44W stepper motors
[14:47:14] <Tom_itx> does anyone have any sample code using 'logic.9' and/or 'lut5.9'?
[14:47:30] <Tom_itx> i've read the docs but it doesn't make a whole lot of cents
[14:49:14] <skunkworks> like this?
http://electronicsam.com/images/emco/linuxcnc_configs/Latest/Half-fullstepMorph/emco-lathe.hal
[14:50:13] <skunkworks> :) I don't know if you should ask me to explain that...
[14:50:45] <skunkworks> it switches between 2 differentstep gens... one half stepping - one full stepping
[14:51:14] <Tom_itx> seems the 2 functions are quite similar
[14:51:41] <Tom_itx> the syntax was mainly what i was after
[14:52:47] <Tom_itx> what function does 202 equate to?
[14:54:33] <Tom_itx> and is 202 a hex or decimal value there?
[14:55:22] <CaptHindsight> carbon fiber brushless gimbal shows up today, looks good but not docs included, you have to search for all the drivers, applications and firmware
[14:56:15] <skunkworks> jepler calculated it for me.. It has 3 inputs where the 3rd switches between 1 and 2 inputs. (if that makes sense)
[14:56:50] <Tom_itx> i wish there was a little better explanation of both those functions
[14:57:22] <pcw_home> the lut5 makes sense to me. its just a lookup table
[14:57:37] <Tom_itx> so what would the 202 equate to?
[14:58:28] <pcw_home> 00000000000000000000001000000010
[14:59:04] <Tom_itx> i mean, how do you determine from the number, how many inputs there are and what function the output performs?
[14:59:33] <skunkworks> I think there are always 5 inputs..
[14:59:52] <Tom_itx> says 'up to 5'
[15:00:04] <pcw_home> 0 => 0
[15:00:06] <pcw_home> 1 => 0
[15:00:07] <pcw_home> 2 => 1
[15:00:09] <pcw_home> 3 => 0
[15:00:10] <pcw_home> etc
[15:00:19] <pcw_home> (upto 31)
[15:00:28] <andypugh> lut5?
[15:00:31] <pcw_home> the input is just the bit address
[15:00:50] <Tom_itx> so is the 202 representd there hex?
[15:00:53] <pcw_home> the output just returns the bit at the bit address
[15:00:59] <pcw_home> yes
[15:01:11] <pcw_home> (if entered in hex)
[15:01:16] <Tom_itx> so on the hal line you don't need to write 0x202
[15:01:24] <pcw_home> you do
[15:01:32] <Tom_itx> ok well he didn't so it must be decima
[15:01:32] <Tom_itx> l
[15:01:49] <pcw_home> that would be odd
[15:02:17] <andypugh> No, it's even in any base
[15:02:22] <Tom_itx> setp lut5.0.function 202
[15:02:22] <Tom_itx> setp lut5.1.function 202
[15:02:34] <Tom_itx> from skunkworks' example
[15:02:40] <andypugh> That will be interpreted in decimal.
[15:02:52] <pcw_home> I assumed all hal parsing of constants was the same (decimal if no 0x)
[15:03:11] <andypugh> 11001010
[15:03:29] <Tom_itx> so what would the value for a 3 input AND gate be?
[15:04:07] <pcw_home> connected to which inputs?
[15:04:13] <Tom_itx> i don't care
[15:04:18] <Tom_itx> probably the first 3
[15:04:45] <pcw_home> bit 7 = 1
[15:04:49] <Tom_itx> in0..2
[15:04:56] <pcw_home> all others 0
[15:05:19] <pcw_home> (assuming the unused pins are 0)
[15:05:46] <pcw_home> (you can make the unused pins dont-care by duplicating the pattern)
[15:05:47] <Tom_itx> 64 dec
[15:06:10] <pcw_home> decimal just makes this hard
[15:06:43] <Tom_itx> i still don't get the association of the logic 'and or xor etc' with the table
[15:07:48] <pcw_home> so "or" is 0000 0000 0000 0000 0000 0000 0000 1110
[15:08:58] <pcw_home> you can generate at the ~4 billion logic functions of 5 inputs with the lut5 (this is how the logic in FPGAs works)
[15:10:27] <Tom_itx> i'm not getting it :(
[15:10:34] <skunkworks> don't feel bad...
[15:10:48] <Tom_itx> are you reading the table top to bottom with your string of numbers?
[15:11:15] <pcw_home> No I am looking at bit(n)
[15:12:02] <pcw_home> so an input of 0 returns an output that the bit value of bit 0
[15:12:14] <pcw_home> that is
[15:13:33] <pcw_home> so on your 3 input and gate, only address 111 (bit 7) returns 1
[15:13:43] <Tom_itx> it might be easier for me to start with 'logic9'
[15:13:50] <heathmanc> got a bitfile/firmware issue with a 4i69 card
[15:13:55] <heathmanc> anybody interested?
[15:14:08] <Tom_itx> heathmanc elaborate
[15:14:30] <andypugh> pcw_home: You take the 4i69 and I will try to explain to Tom :-)
[15:14:48] <Tom_itx> as JT-Shop would say... it's nap time
[15:15:06] <heathmanc> error when loading linuxcnc is board has fpga 6slx25fg, but the firmware in the bit is for fpga 6slxftg256
[15:15:27] <heathmanc> just received the 4i69 today
[15:15:30] <Tom_itx> heathmanc did you compile it?
[15:15:55] <heathmanc> lol, you lost me there, my abilities lie elsewhere. this bit file was provided by pcw last week
[15:15:58] <Tom_itx> andypugh i'll bug ya later about the logic
[15:16:05] <Tom_itx> i gotta get caught up here a bit
[15:16:07] <heathmanc> was supposed to be for a 4i69 and a 7i48
[15:16:18] <andypugh> Tom_itx: Let me try quickly. If you see input 0 as the "1s" bit of a binary number, and input 1 as the "2s" and 3 as the "4s", then the inputs are combined to make a binary number that can have values from 0 to 31.
[15:16:43] <heathmanc> it is the 4i69-25 card
[15:17:22] <pcw_home> how big is the bitfile
[15:17:41] <andypugh> And those values of 0 to 31 are used to pick one bit of a 32-bit number that you setp into the function in HAL.
[15:18:07] <heathmanc> 801563
[15:18:11] <Tom_itx> andypugh, so looking at the chart on the doc page, 0x80 would be a 3 input and?
[15:18:39] <andypugh> So, in the AND3 case, bits 0, 1 and 2 are all set, = 1 + 2 + 4 = 7. So the output value of the function is whatever the value of bit 7 of the parameter is set to.
[15:19:16] <andypugh> Tom_itx: Yes. Because bit 7 is set.
[15:19:24] <pcw_home> so I probably forgot to change the card type when I built the bitfile (thats a X25 bitfile but the IDROM card type tag is wrong)
[15:19:57] <heathmanc> if you say so, this portion is foreign to me
[15:20:00] <Tom_itx> ok can you have a multi input or just the same?
[15:20:06] <Tom_itx> say a 3 input or
[15:20:30] <andypugh> Tom_itx: The calculators in Linux and MacOS let you toggle individual bits in a number and just read out the result ;-)
[15:20:51] <pcw_home> 3 input or is 1111 1110
[15:21:15] <pcw_home> 0xFE
[15:21:43] <andypugh> Though that is also (0 or 1 or 2) xor (3 or 4) (I think :-)
[15:22:12] <pcw_home> 3 input nor is the inverse 0000 0001
[15:23:23] <pcw_home> Whats nice about luts5 is that _any_ function of 5 inputs is possible
[15:23:41] <Tom_itx> if you know how to calculate them :)
[15:23:48] <heathmanc> is there a way around this issue, or do i need a new bitfile
[15:24:12] <pcw_home> new bitfile
[15:24:36] <pcw_home> (or edit the driver source and have it complain but muddle-on)
[15:25:15] <heathmanc> lol, if i could stick a wire in it and make it work, that would be better, your computer jargon might as well be klingon
[15:25:34] <heathmanc> should have bought the 4i65
[15:25:38] <heathmanc> wasn't using my noggin
[15:28:02] <Tom_itx> pcw_home, you can mix logic with it as well? say 3 input AND with an OR on 4 & 5?
[15:28:20] <pcw_home> Yes
[15:28:26] <andypugh> Tom_itx: Yes, except there is only one output bit.
[15:28:34] <Tom_itx> ok, i'll try to wrap my head around it this evening
[15:28:42] <Tom_itx> andypugh yeah i got that part
[15:28:50] <pcw_home> yes, except for that small detali
[15:29:19] <andypugh> It's great for things like door/pause/start interlocks, that sort of thing. You just need to write out what the output should be for every situation, and calculate the parameter.
[15:29:28] <Tom_itx> so how are logic and lut different? they seem quite similar
[15:29:37] <Tom_itx> other than the 5 input restricton on lut
[15:30:09] <Tom_itx> andypugh that's kinda what i'm looking at
[15:30:10] <andypugh> logic only does and / or / xor. lut5 does any possible input to output mapping
[15:31:01] <Tom_itx> the kick to it is that i have a button that will toggle 2 outputs
[15:31:05] <andypugh> And if you find that yout lut5 logic is wrong, you can just alter one number, rather than try to figure out how your chain of 37 and2 and or2 blocks fit together.
[15:31:18] <Tom_itx> but needs certain criteria on one but different on the other
[15:32:20] <pcw_home> LUT5 can do any function, logica is hardwired to and or xor etc
[15:32:52] <Tom_itx> i'll draw a flow chart first to make sure i have the logic part figured out
[15:33:10] <Tom_itx> then see if this is gonna work with lut or logic
[15:33:49] <pcw_home> for example you can make a 3 input mux from lut5 but not from logic
[15:34:52] <andypugh> Tom_itx: To do that, I would use lut5 to calculate the logic, and use a latch to latch the output of the lut5 to the output. Use the lut5 (earlier in the the thread) to calculate the output value, and if there is only one button toggle the latch on that button.
[15:35:49] <andypugh> If you need to recalculate the output on a change in more than one input then you can OR those inputs together into the latch toggle.
[15:36:07] <Tom_itx> i did it with and & or gates in hal and it _almost_ works the way i expect
[15:36:27] <andypugh> Or it might be better to feed the latched input from the button into the lut5. It would depend on what is needed.
[15:36:29] <Tom_itx> there is one hitch i haven't quite figured out
[15:36:37] <cradek> or use ladder!
[15:36:43] <Tom_itx> oh poop
[15:36:46] <cradek> non-ladder gets silly fast
[15:36:50] <Tom_itx> i'm not ready for ladder!
[15:36:59] <Tom_itx> i did look at it a bit
[15:37:03] <pcw_home> heathmanc send an email to mesa (either my address or just tech) and
[15:37:05] <pcw_home> I will send back a fixed bitfile (remind me what daughterboards are connected)
[15:38:49] <heathmanc> Awesome, done!
[15:38:52] <heathmanc> thanks for your help
[15:52:13] <Err> so, per the previous discussion, is there a utility that can compile arbitrary LUT tables to stuff into a Xilinx FPGA - or is there a block of the FPGA programmed to evaluate LUTs, and you're loading constants or something similar?
[15:52:39] <Err> (I've never worked with Xilinx parts - only Altera - and I don't think they have any customer-facing compiler that will let you generate a programming file from LUTs)
[15:52:45] <andypugh> Err: We were discussing HAL functions
[15:52:58] <Err> oh, those LUTs are evaluated in software, then?
[15:53:02] <andypugh> Yes
[15:53:23] <PCW> (the HAL lut5 comp just happens to work like a FPGA lut)
[15:53:24] <Err> their description sounded very much like FPGA implementation - I made a poor assumption, I guess
[15:53:40] <Err> sorry for the interruption :-)
[16:16:40] <CaptHindsight> http://machinedesign.com/automotive/odd-ball-vehicles-week
[16:16:57] <CaptHindsight> most of us around here had cars like that last week
[16:28:13] <JT-Shop> when you run LinuxCNC headless can you run without a keyboard and mouse as well?
[16:29:23] <`Nerobro> yes
[16:29:38] <someone972> So how does the software handle moving to the specified position after the tp runs?
[16:30:10] <someone972> Does it just head towards that point from whereever it is until it reaches it or gets a new position?
[16:31:31] <Deejay> gn8
[16:33:48] <JT-Shop> I need to try linuxcncrsh
[16:35:22] <cpresser> someone972: that depends on the commands you input into the controller. but a basic move such as 'G0 X1' just travels to X=1 and. even if there are obstacles in the way :)
[16:35:39] <JT-Shop> can you use a push button to make the remote run a program with linuxcncrsh
[17:00:56] <Tom_itx> JT-Shop, why not?
[17:07:57] <someone972> So I'm still getting used to all the terms used, what exactly is spindle sync?
[17:09:56] <Tom_itx> in terms of what?
[17:10:18] <Tom_itx> if you were rigid tapping for instance, it would sync with the driving axis
[17:11:26] <PCW> dropped the spindle off the boat: spindle sync
[17:11:45] <someone972> Well the trajectory planner has a field synchronized, with the comment "spindle sync required for this move"
[17:12:14] <someone972> And then there's uu_per_rev which says "for sync, user units per rev"
[17:12:48] <Tom_itx> inch or metric
[17:12:51] <Tom_itx> i suppose
[17:13:05] <Tom_itx> or encoder count maybe?
[17:13:22] <someone972> It says (e.g. 0.0625 for 16tpi)
[17:13:30] <someone972> Is that threads per inch?
[17:13:33] <someone972> turns per inch?
[17:13:45] <Tom_itx> threads per inch is tpi
[17:14:01] <Tom_itx> more or less the same
[17:14:24] <someone972> Is there anything other than rigid tapping where it would need to be syncronized?
[17:14:33] <Tom_itx> lathe
[17:14:50] <Tom_itx> depends what you are doing
[17:15:12] <Tom_itx> single point threading
[17:15:36] <CaptHindsight> gear cutting
[17:15:56] <someone972> So I assume that with my diy 3-axis machine with just a dremel that I set the rpm manually, that would never be used then
[17:16:19] <CaptHindsight> yes
[17:16:45] <CaptHindsight> since your dremel can't be synced
[17:18:01] <someone972> Ok, the next thing that I see is indexrotary with the comment "which rotary axis to unlock to make this move"
[17:18:45] <someone972> Is a rotary axis just an axis that spins instead of moving linearly?
[17:18:55] <JT-Shop> Tom_itx, how would you make a program run in a remote with a push button?
[17:20:44] <Tom_itx> my pendant does it
[17:21:42] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/cnc/configs/sherline/my_jog.hal
[17:22:25] <Tom_itx> i'm working on bits of that, but the program start & pause / resume work
[17:22:59] <Tom_itx> net program-start-btn halui.mode.auto and2.5.in0 <= hm2_[HOSTMOT2](BOARD).0.gpio.034.in_not
[17:23:07] <Tom_itx> net remote-program-run halui.program.run <= and2.5.out
[17:24:09] <JT-Shop> I mean using linuxcncrsh not an input
[17:24:21] <JT-Shop> I know how to do that using halui
[17:24:26] <Tom_itx> i don't know about linuxcncrsh
[17:24:39] <Tom_itx> what is it exactly?
[17:24:43] <Tom_itx> maybe there's a solution
[17:25:03] <JT-Shop> you can run LinuxCNC via telnet
[17:25:37] <tom_R2E3> Hi All
[17:26:05] <Tom_itx> that seems a bit like asking for trouble
[17:26:18] <tom_R2E3> quick question: Is there a way to run LinuxCNC in a simulation mode
[17:26:27] <Tom_itx> sim
[17:26:33] <tom_R2E3> ?
[17:26:50] <Tom_itx> pick a sim from the menu
[17:27:04] <tom_R2E3> ah ok
[17:27:35] <tom_R2E3> I'd like to play about with my own config
[17:27:53] <tom_R2E3> without having the machine powered up
[17:28:06] <Tom_itx> JT-Shop, mode auto would be a start
[17:28:17] <Tom_itx> http://linuxcnc.org/docs/html/man/man1/linuxcncrsh.1.html
[17:30:23] <Tom_itx> run [<StartLine>]
[17:30:23] <Tom_itx> With set, runs the opened program. If no StartLine is specified, runs from the beginning. If a StartLine is specified, start line, runs from that line. A start line of -1 runs in verify mode.
[17:30:53] <Tom_itx> JT-Shop ^^
[17:31:20] <Tom_itx> you can also open a file
[17:31:48] <Tom_itx> want me to set up a terminal by your napping chair?
[17:32:59] <Tom_itx> seems most options are available with the get and set commands
[17:33:12] <tom_R2E3> that'd be lovely
[17:34:28] <tom_R2E3> Ok, i'll have a play with that tomorrow (i'm at a windows PC)
[17:39:01] <Tom_itx> JT-Shop i don't hardly see a way to tie that all to a remote button
[17:39:32] <Tom_itx> unless you had a command file and a way to execute it with a pushbutton
[17:39:41] <Tom_itx> ie Function key?
[17:39:56] <andypugh> It's possible
[17:40:04] <Tom_itx> or input from a port
[17:40:30] <Tom_itx> you'd have to be very trusting
[17:42:29] <Tom_itx> set up several command files and read a few buttons from the parallel port or such
[17:42:36] <Tom_itx> you could execute several that way
[17:46:28] <Tom_itx> http://pyserial.sourceforge.net/pyparallel.html
[17:46:38] <Tom_itx> or such to read the port
[17:50:13] <Tom_itx> http://jumptuck.com/2011/11/14/python-parallel-port-control/
[18:05:22] * JT-Shop wipes paint off my fingers so I can type
[18:06:26] <JT-Shop> back to painting
[18:35:33] <andypugh> I have been French-polishing :-)
[18:36:27] <uw> that sounds inappropriate...
[18:39:11] * JT-Shop has been having fun painting the splitter
[18:39:25] <JT-Shop> http://gnipsel.com/images/splitter/splitter-18.jpg
[18:40:43] <andypugh> I see you still have the canonical device.
[18:42:01] <Tom_itx> andypugh you gettin anywhere with your build?
[18:42:08] <JT-Shop> ah yes
[18:44:15] <andypugh> Tom_itx: Which build?
[18:47:09] <Tom_itx> the arm kernel
[18:48:14] <andypugh> Ah, no. I have been building cabinets and fixing LinuxCNC bugs tonight
[18:48:48] <Tom_itx> undocumented features
[18:48:51] <andypugh> The kernel build is stuck at "bufp.c:(.text.unlikely+0x209c): undefined reference to `dma_alloc_noncacheable'"
[18:49:25] <Tom_itx> ls -l should show the simlinks
[18:49:42] <Tom_itx> someone said keep doing that until you get to the end of them
[18:50:15] <andypugh> Which is especially interesting as there is no file named bufp.c. But there is a bufp.o
[18:51:12] <Tom_itx> not hiding somewhere?
[18:57:24] <andypugh> Nowhere to hide
[19:02:00] <Tom_itx> did you try searching for a bufp with a wildcard?
[19:02:49] <andypugh> Yes, and I found a bufp folder and a bufp.o (not in the same place)
[19:03:22] <Tom_itx> grep the make file for the same
[19:06:42] <andypugh> Actually, the file _does_ exist. But Eclipse can't see it!
[19:07:03] <Tom_itx> i never had much love for eclipse
[19:07:18] <Tom_itx> i tried it once with avr and dumped it right away
[19:08:29] <jdh> it's pretty massive
[19:09:16] <Tom_itx> well i crashed it within the first 5 minutes of using it
[19:11:14] <andypugh> Tom_itx: But it beats gedit
[19:13:00] <uw> i love gedit wtf
[19:14:34] <uw> what is that mini trailer for?
[19:16:46] <uw> and does that cannon work
[19:17:23] <eric_unterhausen> eclipse's workspace thing drives me nuts
[19:17:36] <jdh> the cannon is for Mach users
[19:17:50] <eric_unterhausen> I have a cannon
[19:17:54] <uw> lol
[19:18:03] <eric_unterhausen> afraid to use it though
[19:18:13] <eric_unterhausen> carbide
[19:18:59] <Tom_itx> you mean you have a bomb
[19:19:15] <uw> need vids of that cannon working
[19:19:23] <eric_unterhausen> wonder if the company that made it is still around
[19:19:34] <uw> cannon company?
[19:19:48] <uw> i think that's a different cannon
[19:22:15] <eric_unterhausen> wow, it costs $350 new
[19:23:12] <Tom_itx> back on that lut5 a bit: expression = (i2 and i1) or (not i2 and i0) would look at inputs 1 & 2 for the AND part and 2 & 0 for the NOT part?
[19:23:17] <eric_unterhausen> no, just $160
[19:23:28] <Tom_itx> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Lut5
[19:23:30] <Tom_itx> shown there
[19:24:02] <eric_unterhausen> isn't the not modifying i2?
[19:24:29] <Tom_itx> anything i say about it would be a guess
[19:24:30] <eric_unterhausen> os if i2 is zero and i0 is one, or if i2 and i1 are one
[19:25:09] <eric_unterhausen> has to be i2=0 and i0 =1
[19:26:01] <andypugh> Tom_itx: lut5 doesn't "do" any logic at all.
[19:26:11] <andypugh> It is a look-up table.
[19:26:27] <eric_unterhausen> for logic
[19:26:34] <andypugh> You just define what output you want for every single combination of inputs.
[19:26:39] <Tom_itx> i get that but where are you looking for the output at?
[19:27:00] <Tom_itx> horizontally in the table?
[19:27:06] <andypugh> lut5.0.out?
[19:27:52] <Tom_itx> no that's not what i meant to say
[19:28:04] <Tom_itx> i realize the result will be there
[19:28:58] <Tom_itx> you add the 'weights' together for the mask for the logic you want?
[19:29:16] <andypugh> You can if you want to do it the stupidly hard way
[19:29:28] <Tom_itx> for understanding i do
[19:29:33] <Tom_itx> but not in reality
[19:29:45] <andypugh> Write out the table
[19:29:59] <andypugh> (as shown in the man page)
[19:30:04] <andypugh> Ignore the "weights"
[19:30:30] <andypugh> In the "weights" column write down a 1 or a 0 for what you want the output to be.
[19:30:45] <andypugh> then, reading up from the bottom, that binary number is your "function"
[19:31:58] <andypugh> Simple 3-input worked example
[19:31:59] <andypugh> http://www.linuxcnc.org/index.php/english/forum/10-advanced-configuration/27392-stop-coolant-during-pause#42786
[19:32:41] <andypugh> There are 10 sorts of people in the world...
[19:32:58] <uw> :|
[19:33:10] <Tom_itx> uh huh
[19:33:15] <Tom_itx> heard that one before
[19:37:35] <andypugh> So, have you written out your truth table?
[19:38:45] <Tom_itx> not yet, i'm still scratching my head over the details
[19:38:53] <Tom_itx> not the lut but the actual workings of it
[19:39:32] <andypugh> I just remembered, I made a google spreadsheet for calculating lut5:
https://docs.google.com/spreadsheet/ccc?key=0AhjJW1-T6n7CdFNhVnlhNEEwLU1lS3B5Q0FfWGtfTEE&usp=drive_web#gid=0
[19:40:28] <andypugh> Just put in the yellow column the output you want for the input combination, and read-off the function number
[19:40:39] <Tom_itx> ok
[19:40:54] <Tom_itx> i should stuff that into a spreadsheet
[19:40:58] <andypugh> In fact, if I change it, you might see it live
[19:41:46] <andypugh> Say you want a 5-input and? Then it is only the 11111 that needs to have an output of 1.
[19:41:59] <Tom_itx> yes
[19:42:23] <andypugh> But maybe all zero is good too?
[19:42:39] <Tom_itx> what do you do in the case of an OR
[19:42:42] <andypugh> In that case the "function" is 0x80000001
[19:42:46] <Tom_itx> where either one could be set?
[19:42:55] <andypugh> Any sort of or?
[19:43:08] <Tom_itx> just a 2 input or gate
[19:43:22] <Tom_itx> either one would cause it to be true
[19:43:36] <andypugh> See that?
[19:43:50] <Tom_itx> ahh ok
[19:43:58] <Tom_itx> the pc is slow
[19:44:00] <andypugh> Actually that is XOR, as the 00011 case is 0
[19:44:02] <Tom_itx> it takes a bit to refresh
[19:44:14] <andypugh> It is OR now
[19:44:26] <Tom_itx> ok
[19:44:44] <andypugh> If you wanted (0 OR 1) AND (4 AND 5) then:
[19:44:53] <Tom_itx> then if i wanted to use bits 4 & 5 for the or, i'd just move the 1's column over for each result
[19:45:19] <andypugh> 0xEE000000
[19:45:41] <andypugh> It is now showing (0 OR 1) AND (4 AND 5)
[19:46:28] <Tom_itx> i think i get it now
[19:46:41] <Tom_itx> can this be saves as a spreadsheet?
[19:47:13] <andypugh> Probably. Or you could just use the Google doc, I think it is completely public
[19:47:22] <andypugh> Can you change it?
[19:47:38] <Tom_itx> yeah
[19:47:41] <andypugh> There you go
[19:48:21] <andypugh> Do you understand what the (very simple) code in the HAL component does?
[19:48:50] <Tom_itx> you mean the inputs and outputs? yes
[19:49:04] <Tom_itx> and assigning the mask
[19:49:16] <andypugh> It takes inputs 0 to 4 as bits in a binary number (the first column of the sheet) then looks at what that numbered bit is of the 32-bit "function"
[19:50:48] <andypugh> I think it is literally out = (function & 1 << (in0 + 2* in1 + 4*in2 + 8*in3 + 16*in4) != 0)
[19:51:40] <Tom_itx> makes sense
[19:51:54] <andypugh> There is a File->download as menu item that lets you make a copy in Excel, if you want.
[19:52:29] <andypugh> It's an amazingly adaptable function. I don't know who thought of it, but it's great
[20:39:15] <eric_unterhausen> what is that int15 call anyway?
[20:39:32] <andypugh> No idea.
[20:39:53] <pcw_home> software interrupt
[20:40:14] <pcw_home> not sure that even works in protected mode
[20:40:16] <eric_unterhausen> 6F08H was really the question
[20:40:23] <eric_unterhausen> should cause a page fault
[20:41:46] <eric_unterhausen> those are apparently some neurons that died, cause I don't remember int15h at all
[20:43:18] <pcw_home> int15 function 0 is turn on casette motor
[20:44:13] <eric_unterhausen> ya, 6F08H wasn't an original bios function, looks like to me anyway
[20:46:26] <pcw_home> I vaguely remember int13 since our first product was a disk emulator
[20:46:27] <pcw_home> (that intercepted int13 and added our CMOS battery backed-up RAM ) drive
[20:46:29] <pcw_home> to the drive chain
[20:46:41] <eric_unterhausen> ya int13== msdos
[20:47:00] <eric_unterhausen> it was just an int13 handler
[20:51:35] <andypugh> 6F08H is meant to set a Windbond chip (with GPIO) to input mode
[20:52:23] <eric_unterhausen> i don't quite understand using the software interrupt though
[21:00:12] <pcw_home> do they have any example code? I would just dump the int15 and talk to the hardware
[21:01:32] <eric_unterhausen> ha, just caught a spammer
[21:02:18] <eric_unterhausen> someone once told me that posting hidden gifs had something to do with seo, but I'm not sure
[21:44:36] <Tom_itx> ok back to lut5 for a bit: can i use the same pin for an OR function as and AND function?
[21:44:50] <Tom_itx> as if they were tied together
[21:45:37] <Tom_itx> also can i feed the output of the AND to the OR or does that require a second lut5 element?
[21:45:52] <Tom_itx> probably requires 2 lut5 elements
[21:46:34] <pcw_home> you can do _any_ logical function that has 5 or fewer inputs and one output
[22:29:25] <Tom_itx> pcw_home is it illegal to feed the output to one of the input pins?
[22:29:47] <Tom_itx> meh nevermind, that wouldn't work anyway
[23:43:22] <Jentrep> hey guys :)
