#emc | Logs for 2009-04-12

[00:00:04] <tomp> never heard of 'rolling average period' but many electronics guides on net.
[00:00:20] <tomp> the 'rc' constant ?
[00:08:01] <maddash> yeah
[00:08:28] <maddash> it's easy to see how the rc controls the slope of the exponential charging rate
[00:08:55] <maddash> but you only get exponentials when you conduct a one-shot 0 to 1
[00:09:07] <maddash> what's the effect of rc on a freakin' sine curve?
[00:17:49] <tomp> hmm, internet is down, i can ping google, stream video, ftp, but firefox wont get a page, progress bar is empty. rebooted, same!
[00:59:44] <eric_unterhause1> dns server down
[01:00:15] <toastydeath> just look at it as being up in reverse
[01:00:37] <eric_unterhause1> I like the fact that you can talk on IRC without dns
[01:01:48] <toastydeath> hahah
[01:02:33] <eric_unterhause1> never really made sense to me that browsers are continually doing dns lookups
[01:03:25] <fenn> only stupid browsers
[01:04:22] <fenn> it used to be you had to press ctrl-shift-r to get netscape to clear the dns cache for the page you were reloading
[01:05:22] <eric_unterhause1> firefox must do a dns lookup for each page
[01:09:17] <fenn> that would be really slow
[01:09:48] <eric_unterhause1> you would think so, but then why does the dns server going down kill the any browsing
[01:10:07] <eric_unterhause1> you should be able to reload a page, for example
[01:10:29] <fenn> reload might cause a dns lookup
[01:10:44] <eric_unterhause1> what doesn't then?
[01:10:53] <fenn> clicking on a link
[01:11:10] <fenn> (not that i actually know, just guessing)
[01:11:16] <eric_unterhause1> link is a needs a lookup anyway
[01:11:41] <SkinnYPuP> there are a few local dns caching progs if your isp's dns server is lagging
[01:11:43] <eric_unterhause1> I'm inserting random articles because I just got back from a bike ride
[01:12:10] <eric_unterhause1> I've thought about local dns caching, it seems like comcast likes to play games with dns a lot
[01:14:21] <fenn> eric_unterhause1: did you finish that mill yet?
[01:14:34] <eric_unterhause1> I'm resolved to finish it any day now
[01:14:40] <fenn> har har
[01:14:41] <eric_unterhause1> :)
[01:14:57] <eric_unterhause1> smashing my finger really put a damper on it
[01:15:24] <SkinnYPuP> that'll do it
[01:15:55] <eric_unterhause1> there were other factors as well, now that I think about it, work related
[01:19:02] <tomp> i can ping www.google.com so i guess dns is running
[01:19:12] <eric_unterhause1> actually, I got my lathe turning under it's own power the other night, that's this weekend's project
[01:19:26] <tomp> but, i gotta goto work anyways, bye bye
[01:19:35] <fenn> a perpetual motion machine!
[01:19:48] <fenn> i've always wanted to build one of those
[01:19:58] <eric_unterhause1> perpetual motion machines almost always involve batteries
[01:20:47] <fenn> yes, well, the cycle of violence will just continue unless you stop it, that's why it makes a great energy source
[01:21:15] <eric_unterhause1> usually, you have to read through to the end of the article to find out there are batteries in there for "power smoothing" or some such
[01:24:59] <fenn> anyone remember what micges' DXF importer was called?
[01:27:17] <fenn> ah i found it http://sourceforge.net/projects/vec2ngc/
[01:27:27] <eric_unterhause1> there's an intuitive name
[01:40:55] <fenn> dammit. i'm stuck on this stupid code thing and i refuse to do it the ugly way
[03:20:03] <dareposte> hi all
[08:33:07] <pjm__> morning
[09:51:05] <alex_joni> http://gawker.com/5206458/watch-a-google-street-view-car-hit-a-bridge <- LOL
[11:34:06] <BigJohnT> Dallur: hows the boat coming along?
[11:35:40] <Dallur> BigJohnT: Well
[11:36:18] <Dallur> BigJohnT: I had to replate half of the deck to get the sheer in a better shape, I expect all plating will be done in one week
[11:36:34] <Dallur> BigJohnT: Then it's 2 weeks of welding and grinding
[11:36:34] <BigJohnT> ouch
[11:36:51] <Dallur> BigJohnT: It's not a big thing, about 1 day of work to redo, it was only tackd
[11:37:02] <BigJohnT> tacked is good
[11:37:22] <BigJohnT> is your camera on?
[11:37:26] <Dallur> http://dallur.com/index.php?id=129
[11:37:26] <Dallur> yup
[11:38:24] <BigJohnT> nice big :) on the side
[11:38:31] <Dallur> yeah :)
[11:38:32] <BigJohnT> looking good
[11:38:41] <Dallur> thanks
[11:38:55] <Dallur> It's moving fast now, I quit my job so i'm building full time
[11:40:06] <BigJohnT> how will it be rigged?
[11:40:25] <Dallur> cutter, 13m mast
[11:40:43] <Dallur> in boom furling and furling on the forestay
[11:40:50] <Dallur> no furling on the staysail
[11:40:53] <BigJohnT> nice!
[11:41:56] <Dallur> I should probably model the rig on the boat and do a rendering or two :)
[11:42:34] <BigJohnT> that would be cool
[11:43:04] <Dallur> how are things on your front ? Spring there yet ?
[11:43:46] <BigJohnT> it is getting close
[11:44:17] <BigJohnT> I have some tomato plants starting from seeds about ready to plant in the garden.
[11:44:44] <BigJohnT> We have a few nice days that the wife and I could take a ride on the bike
[11:45:35] <BigJohnT> I have the V-F board from Peter and am working on the THC with it :)
[11:45:44] <Dallur> BigJohnT: nice
[11:46:43] <BigJohnT> doing some soldering on a test board this morning to hook up and start work on the THC component for EMC
[11:47:00] <Dallur> BigJohnT: you writing your own C component to interface ?
[11:47:12] <BigJohnT> I'm using comp
[11:47:31] <Dallur> BigJohnT: ah OK
[11:48:02] <BigJohnT> it will be similar to your work but all rolled up in a component
[11:48:31] <Dallur> BigJohnT: Probably makes more sense as a component :)
[11:50:39] <BigJohnT> the V-F board generates a frequency based on input voltage that is fed to an encoder input, the velocity output from the encoder will be my source for the torch voltage
[11:50:51] <BigJohnT> in EMC
[11:51:50] <Dallur> BigJohnT: What really seems to cause most people a whole lot of problems is the HF/RF stuff though
[11:52:18] <Dallur> BigJohnT: I have helped a couple of guys with their THC and it's always the same, noise causing problems
[11:53:01] <BigJohnT> yea, HF start is not easy to deal with... that's why I got a Hypertherm 1250
[11:53:21] <Dallur> BigJohnT: yeah, the newer non HF machines are easier to deal with
[11:53:32] <Dallur> BigJohnT: I have wondered if there is a simpler way to get around the problem as a whole though
[11:53:55] <Dallur> BigJohnT: like placing the encoding mech on a table and connecting it to the computer with a fiber
[11:54:37] <Dallur> using standard networking fiber optics should keep the costs down
[11:56:42] <BigJohnT> does the HF mess with the electronics of the THC's?
[11:57:57] <Dallur> BigJohnT: everything pretty much
[11:58:15] <BigJohnT> Peter said that the board would have enough capability for a HF start torch...
[11:58:18] <Dallur> BigJohnT: seen blown components on the pcb, computer lockups, motors going nuts, drivers going crazy
[11:58:41] <BigJohnT> wow, makes you not want a HF start
[11:59:23] <Dallur> BigJohnT: I don't know the board but I'm very sceptical when people say HF should work, if they are 100% willing to garantee it will work I would beleave them but "should work"
[11:59:45] <Dallur> to me that means that with extensive modifications and debugging it can be made to work :)
[12:00:35] <BigJohnT> the board is brand new and I'm testing it and working out the EMC part (I hope) I have no plans on testing it with a HF torch...
[12:00:46] <Dallur> BigJohnT: don't :)
[12:01:08] <BigJohnT> the way Peter put it was it was built way overkill :)
[12:01:09] <Dallur> BigJohnT: my current advice to people with HF is to buy a new machine without HF, it's cheaper than getting a THC to work with the HF torch
[12:01:39] <BigJohnT> and causes less road rage :)
[12:02:00] <Dallur> BigJohnT: I don't remember the exact number or math but I recall that a typical HF start can arc across about 1/2 inch of air
[12:02:30] <Dallur> So every connector needs to have it's poles spaced by that, every cuircuit needs to be spaced by that, unless everything is MIL spec coated
[12:02:46] <BigJohnT> mean stuff
[12:03:20] <Dallur> BigJohnT: yeah, it's nasty to deal with and the only people that seem to really know anything about it are radio amatures
[12:03:43] <Dallur> BigJohnT: but then I start telling them how many amps I work with and they get scared
[12:03:57] <BigJohnT> Hypertherm has HF THC but they are the industrial (expensive) ones.
[12:04:39] <Dallur> BigJohnT: yeah, it seems it's hard to get "touch start" working above 100A
[12:05:10] <Dallur> BigJohnT: but in every other way touch start is better, it's easier on the consumables for one thing
[12:09:35] <BigJohnT> my current thought for removing the offset is to gradually remove it after the torch turns off and as the Z moves up...
[12:13:17] <Dallur> BigJohnT: are you talking about the cut thickness compensation offset ?
[12:14:39] <Dallur> BigJohnT: The other way to deal with offsets is to compensate for them in the CAM work and keep EMC unaware
[12:14:45] <BigJohnT> the wavy material compensation
[12:15:33] <Dallur> BigJohnT: The Z axis up/down ?
[12:15:46] <BigJohnT> yes as you cut
[12:16:24] <Dallur> BigJohnT: Ahh so your thinking people can set their own height and you will just offset from that ?
[12:16:35] <Dallur> BigJohnT: and the height they set is the "travel height" ?
[12:18:50] <BigJohnT> once the cut has started and your velocity is up the THC will move the Z up and down as needed to maintain proper voltage
[12:19:18] <BigJohnT> at the end of a cut you might have a negitive offset in there if your sheet is warped down
[12:19:21] <Dallur> and once the cut stops it will move back to the original position ?
[12:19:45] <Dallur> BigJohnT: the only really safe way is to put on the feed hold and move the Z up
[12:19:52] <BigJohnT> as you command the torch to move up to clearance height any Z offset would be removed
[12:20:15] <Dallur> BigJohnT: the problem is you might have edges or other obstacles that the torch would hit as it moves
[12:20:19] <BigJohnT> no, I don't want it moving at all once the cut stops
[12:20:39] <Dallur> BigJohnT: so basically feed hold, move Z up and then allow the f0 ?
[12:20:59] <BigJohnT> yes
[12:21:20] <Dallur> BigJohnT: I think that's the only safe way to do it
[12:21:25] <BigJohnT> offset freeze, move Z up, bleed off any offset
[12:22:34] <BigJohnT> also offset freeze if your below the velocity tolerance you set (corner height lock)
[12:22:44] <BigJohnT> or any other problem
[12:23:29] <BigJohnT> and somekind of offset freeze if the torch voltage moves too fast I think
[12:23:30] <Dallur> yup
[12:23:49] <Dallur> BigJohnT: you need some "ballast" on the movement for sure
[12:24:17] <Dallur> BigJohnT: but it's hard to have it ballasted, because if something pops up due to warpage you want to react immediately
[12:24:23] <BigJohnT> some kind of response rate or gain
[12:25:04] <BigJohnT> and the thinner it is the worse it warps and the faster you go :)
[12:26:00] <Dallur> it's tough, but I would say that the best way is to have the speed of the z movement adjusted in proportion to the voltage offset
[12:26:20] <Dallur> so If the voltage is way under, you move fast, if it's just a bit under, you move slowly
[12:26:29] <BigJohnT> ok, I had not thought about that relation
[12:27:11] <BigJohnT> a PID loop :)
[12:27:17] <Dallur> :)
[12:27:58] <Dallur> I had some problems with oscilatinos initially, but I was to lazy to implement this so I just enabled the user to control the adjustment speed
[12:28:14] <Dallur> err oscillations
[12:36:01] <BigJohnT> I have tested it with a BOB and the parallel port but the speed of the BOB or the parallel port was too slow to read the pulse freq from the V-F board
[12:36:20] <BigJohnT> as the input voltage got past 3 or 4 volts
[12:39:16] <BigJohnT> so now I'm moving on to the 5i20
[12:42:33] <BigJohnT> that's a good start... I plugged my adapter board in and no magic smoke came out and the blinky leds came on :)
[12:45:00] <Dallur> :) bbl, need to to some boat building
[12:45:22] <BigJohnT> ok thanks
[12:50:17] <tomp> is there some expectations about how fast the THC can adapt to the surface? like, can it travel over a wasbboard & track well, or maybe smooth curves of +/-12" radius? (at velocities suited to the process)
[12:51:46] <tomp> uh, what happens if it goes over a hole! oops
[12:52:06] <BigJohnT> you slap the programmer!
[12:52:10] <tomp> abandon hover mode
[12:53:24] <tomp> how fast are youmoving in these cuts, and how wavy is wavy?
[12:53:36] <BigJohnT> at this time my expectations are to be able to handle normal plasma cutting
[12:54:05] <tomp> uh, no idea what normal plasma cuttin is like, tho at imts, they honk
[12:54:10] <BigJohnT> the speed is determined by the material thickness... in my case between 10 IPM and 400 IPM
[12:54:44] <BigJohnT> wavy would be + .5 or so
[12:55:05] <tomp> ok, and whats a maximalish surface variation per inch?
[12:55:10] <tomp> ok
[12:55:25] <tomp> opps, no dZ/x
[12:55:34] <BigJohnT> depends on how rough the fork lift guy is on your material
[12:55:55] <tomp> oh, so its bendy not crinkly
[12:56:12] <BigJohnT> normally it is pretty flat to within a 1/4" or so
[12:56:34] <BigJohnT> if it is crinkled then you don't take it off the trunk :)
[12:56:41] <tomp> doe sit move while its cutting? i bet it can 'fall' as stock is removed then
[12:57:20] <tomp> is the feedback voltage drop?
[12:57:36] <BigJohnT> it might move up for thin material but the slats contain larger parts
[12:57:47] <BigJohnT> voltage variation
[12:58:23] <tomp> ok, like edm, the desired voltage is dialed in, and the servo maintains that, if the steel moves away the tool chases it
[12:58:32] <BigJohnT> yep
[12:59:04] <BigJohnT> you dial in voltage and velocity and velocity tolerance
[12:59:37] <tomp> and set some base velocity that gets fiddled ?
[12:59:43] <tomp> for Z
[13:00:03] <tomp> ohm, you said vel, sorry
[13:00:42] <tomp> we also set up a gain... kind of an accell, how FAST you want to make the change if theres a change to be done
[13:00:59] <tomp> it makes the head go from sluggo to hyper nervouis
[13:01:25] <BigJohnT> yea, some kind of gain on the Z correction would be in there
[13:02:02] <tomp> its like driving in traffic, and keeping YOUR space ( like, maintaining the path with emc's 'tolerance' )
[13:04:08] <tomp> thanks for the info, very interesting & good luck. there's a whole world of 'gap controls' as opposed to 'feed/speed/position' controls. its another 'discipline'
[13:04:32] <BigJohnT> yes, it will be interesting for sure
[13:04:53] <tomp> oh, another thing might be of tinterst... the 'curve'
[13:05:23] <tomp> when you get real close to correct, you change the velocity and accels and gain, when you are way out of line, you move fast
[13:05:56] <tomp> the curve has a poin t in the center where it doesnt move, becasue it's where it shopuld be ( the pole )
[13:06:16] <tomp> and left and right of that is infeed and outfeed
[13:06:25] <tomp> and above and beow that is fast and slower
[13:06:48] <tomp> but, the line that is tracked can be bent, and you make a sweet zone near the perfect point
[13:07:19] <tomp> and you make these steepish ramps outside of the sweet zone (run away and get back to work ;)
[13:07:37] <BigJohnT> the I part of the PID loop I think
[13:07:52] <tomp> the voltage is a lookup index into this curve, you solve fgor the new velocity
[13:08:15] <tomp> so, the velocity result is not linear
[13:08:19] <BigJohnT> nope I was wrong it is the D part of the PID
[13:08:48] <tomp> ah, derivatives in process control, kill process, only P!
[13:09:54] <tomp> ok, avoiding rants :)
[13:10:50] <tomp> bestoluck, what the H are you doing up ... must be ... oh 8 am usa
[13:11:25] <BigJohnT> been up since 5am
[13:11:49] <tomp> you can next get a strain gauge and do force feedback polishing
[13:12:12] <tomp> use the V from it same way
[13:12:21] <BigJohnT> now if I can just get my pulse signal into the 5i20 this morning I'll be off and running
[13:12:50] <tomp> from the F/V? from some pwm,.... so from emc?
[13:14:23] <tomp> do you have a dummy signal souce just for tests? maybe a 555, resistor and cap ... heck you can use hal for a waveform generator ( cool stuff that hal)
[13:14:26] <BigJohnT> from the V-F board
[13:14:43] <BigJohnT> heh I know what's wrong :/
[13:15:11] <tomp> oh the F is not a number, its a wave, good
[13:15:27] <BigJohnT> I started counting pins from the wrong end... slap on forhead
[13:15:54] <BigJohnT> * BigJohnT fires up the soldering iron to move the wire
[13:16:59] <tomp> argh, had same thing, some connectors were on bottom of pcb, i had hairy board with so many untwisted busses hand soldered! take care usually well into the 'toodamnsmall' size
[13:19:12] <tomp> connectors on botoms of pcbs have to be mirrored when placed onto the pcb
[13:19:30] <tomp> else pin 1 is wrong
[13:19:48] <BigJohnT> that made a hellofadifference
[13:19:55] <tomp> got pulses?
[13:20:15] <BigJohnT> I have the velocity in EMC :)
[13:20:21] <BigJohnT> and it is steady
[13:20:21] <tomp> goog o
[13:20:25] <tomp> good o
[13:22:13] <BigJohnT> nice I can see a good span on velocity with a 0.01 voltage change
[13:23:19] <BigJohnT> velocity changes about 5 for every 0.01 volt and I just noticed I'm still on divide by 128!
[13:23:40] <tomp> 5ipm?
[13:23:53] <tomp> or squats
[13:24:43] <BigJohnT> I think it is user units per second
[13:28:42] <tomp> 10mV has noticeable result... and the range is +/- 10V? thats a lot of rez
[13:28:43] <tomp> scrolling back...., at 0-3or4V its 3 or 400 usable velocities.
[13:29:00] <BigJohnT> 0-10 or +-5
[13:33:40] <tomp> ok, now run it over the bastard file and see what happens :)
[13:35:39] <BigJohnT> the voltage to velocity ratio is flat as a pancake
[13:37:14] <BigJohnT> the chart http://imagebin.ca/view/Jv6CxyD7.html
[13:37:53] <BigJohnT> bbl breakfast calls
[13:38:22] <tomp> have a good un, thx
[13:40:27] <tomp> (wow bigjohn has breakfast that talks, mine cant even hum)
[13:58:33] <Garak> anyone live near a lexus dealer, there has been recall on the electric power stearing, the assembly has a nice but short ballscrew in it http://designerthinking.com/projects/Lexus%20steering/Lexussteering.html
[14:06:49] <tomp> nice, looks like 0.2" pitch, 1"dia, rolled ballscrew with a rack on one end
[14:07:10] <tomp> 10" of thread (less len of ball nut=travel)
[14:09:11] <Garak> not alot of travel, but for %
[14:09:18] <Garak> $5, its not bad
[14:09:34] <tomp> wah! you got a whole sh&*load of 'em, nice score ( nice rack )
[14:09:49] <Garak> thats not my page, thats kats'
[14:10:04] <tomp> thanks for the tip anywho
[14:10:05] <Garak> I don't live anywhere near a lexus dealer
[14:10:39] <tomp> and i cant put em in my baggage :( the taiching lexuz dealer is only a couple kilo away
[14:10:41] <Garak> if anyone finds some and are willing to ship some of them to me
[14:11:27] <Garak> ballscrews are bloody expensive
[14:11:43] <tomp> might be nice for a hexapod, with the motor, they're more like 'thrusters'
[14:12:22] <Garak> yea, I bet they have some force behind them, I think thats what kats' is going to use them for
[14:13:00] <Garak> They may just be 1tpi, its something like 4 parallel threads
[14:13:27] <Garak> for a cnc application one would have to gear it way down
[14:15:14] <tomp> oh multiple lead
[14:16:01] <tomp> am i getting this? the motor spins and shaft(screw/rack) does not, right?
[14:16:25] <tomp> else the rack would be pretty useless (doh)
[14:17:00] <Garak> yea, the ball nut is the spinning part of the motor
[14:17:23] <Garak> its a neat design in someways
[14:17:27] <tomp> i've only seen that on tiny steppers, cool
[14:17:35] <TaigGecko5404axi> Hello, I'm looking for information on how to configure a Gecko540 driver on a taig 4-axis cnc mill
[14:18:28] <TaigGecko5404axi> I have a configuration file for Mach3, I wonder if information can be dug from it
[14:19:35] <tomp> sorry, dunno geckos
[14:19:35] <tomp> configure meaning what? a cfg file for emc to use the gecko?
[14:20:02] <TaigGecko5404axi> well, if its sufficient
[14:20:12] <TaigGecko5404axi> I'm rather looking for a guide
[14:20:16] <tomp> i know people here have used geckos
[14:20:47] <jlmjvm> i use gecko drives
[14:21:06] <TaigGecko5404axi> ok, any link for me?
[14:21:26] <BigJohnT> I use gecko 203v's
[14:21:49] <BigJohnT> did you use the stepconf wizard to generate a configuration?
[14:22:02] <TaigGecko5404axi> what is in the 540? there are 4 drivers
[14:22:25] <TaigGecko5404axi> no, right now I can't because I'm on 8.10
[14:22:43] <jlmjvm> arent they like a small 201 drive
[14:22:46] <Garak> anyone know of an affordable source for ballscrews, cheapest I've found is over $100 for a short little rig
[14:23:04] <TaigGecko5404axi> I'm gathering information before I install a 8.04
[14:23:19] <BigJohnT> what info do you need?
[14:23:50] <SWPadnos> TaigGecko5404axi, I can send you a G540 config, if you like
[14:24:12] <TaigGecko5404axi> a guide to setup a 3 or 4 axis mini mill with gecko540 and 280oz steppers
[14:24:14] <SWPadnos> you'll still need to set the step resolution and accel/vel limits
[14:24:40] <TaigGecko5404axi> do you think I can dig it from a mach3 config file
[14:24:40] <tomp> Garak new ball screws aint cheap , like what you found is not bad price. maybe you can find used from a machine, but skip piles of ball screws in boneyards, theyre in piles for a good reason
[14:25:16] <TaigGecko5404axi> the step resolution I have it
[14:25:28] <SWPadnos> TaigGecko5404axi, I have a stepconf file for the G540, and no, you won't be able to convert a mach file
[14:25:51] <TaigGecko5404axi> no
[14:25:58] <TaigGecko5404axi> I have the mac file
[14:26:01] <TaigGecko5404axi> mach
[14:26:10] <TaigGecko5404axi> http://www.deepgroove1.com/tutorials.htm
[14:26:25] <SWPadnos> the mach file is useless for EMC2
[14:26:40] <SWPadnos> however, I'm telling you that I have the EMC2 equivalent, which I will send you
[14:26:49] <TaigGecko5404axi> okay, I though some info could have been extracted from it
[14:26:59] <TaigGecko5404axi> great!
[14:27:04] <TaigGecko5404axi> need my email?
[14:27:16] <SWPadnos> no, we'll try dcc here - one sec
[14:27:17] <tomp> no just concentrate
[14:27:29] <tomp> :)
[14:27:32] <SWPadnos> I'm booting the machine it's on :)
[14:27:45] <TaigGecko5404axi> I'm on the java client, though
[14:28:06] <SWPadnos> well, if it doesn't work I'll get your email address :)
[14:29:21] <SWPadnos> are you on a Windows machine right now?
[14:29:44] <TaigGecko5404axi> nope
[14:29:52] <TaigGecko5404axi> what is the server name
[14:29:56] <SWPadnos> ok, good. line endings get screwed up
[14:30:01] <TaigGecko5404axi> I have konversation
[14:30:08] <SWPadnos> ok, great
[14:30:29] <TaigGecko5404axi> brb
[14:32:18] <SWPLinux> TaigGecko5404axi: let's see if DCC works with the web client first
[14:33:09] <SWPadnos> I guess so
[14:34:29] <TaigGecko540-4ax> it worked, thank you
[14:34:33] <SWPadnos> TaigGecko540-4ax, let me know how it works for you. I should send it to Mariss some day and I'd like a few people to use it
[14:35:13] <TaigGecko540-4ax> what kind of machine do you have?
[14:35:19] <TaigGecko540-4ax> is it similar?
[14:35:25] <SWPadnos> no :)
[14:35:37] <SWPadnos> I have a G540 with a stepper motor clamped to a bench :)
[14:35:45] <TaigGecko540-4ax> heheh
[14:35:56] <SWPadnos> with a laser on it so I can see (on the far wall) if the motor is in the right spot
[14:36:22] <TaigGecko540-4ax> right spot for what?
[14:36:56] <SWPadnos> I would line up the spot somewhere, then do a bunch of G0 moves, and make sure the motor didn't lose any steps
[14:37:14] <TaigGecko540-4ax> I see, some sort of encoder
[14:37:17] <SWPadnos> with the settings I had, I needed one turn per 0.2"
[14:37:26] <SWPadnos> well, a "lost step indicator"
[14:37:52] <jlmjvm> is it positioning good
[14:38:04] <tomp> a 'pointed stick'
[14:38:14] <SWPadnos> it didn't at first, but I eventually got it (needed to invert the step pulses)
[14:38:22] <TaigGecko540-4ax> anyway, I wont be able to give you feedback before I receive the machine
[14:38:41] <SWPadnos> no problem. it's been a few months already, a little while longer won't hurt much
[14:39:11] <TaigGecko540-4ax> I think such info if present in a mach3 file
[14:39:22] <SWPadnos> ?
[14:39:44] <TaigGecko540-4ax> step direction..
[14:40:06] <tomp> accel? top velocity, step format?
[14:40:25] <SWPadnos> sure. but I'm the guy that makes the config files, so there's nobody to tell me how to do it ;)
[14:40:26] <TaigGecko540-4ax> http://www.geckodrive.com/upload/540B.xml
[14:40:38] <SWPadnos> (Mariss sent me the G540 so I could make an EMC2 config)
[14:40:56] <TaigGecko540-4ax> what stepper do you have?
[14:41:11] <SWPadnos> I bought 3 of the Keling ones that are "made for the G540"
[14:41:19] <SWPadnos> I don't remember the model number
[14:41:37] <TaigGecko540-4ax> That's the ones I'll get, I think
[14:41:54] <TaigGecko540-4ax> but they are not branded Kelig (rebranded)
[14:41:57] <SWPadnos> they seem to be well matched to the G540, and aren't too expensive
[14:42:08] <TaigGecko540-4ax> yeah, they say
[14:42:17] <SWPadnos> Keling is a reseller, I think he imports direct from China
[14:42:25] <SWPadnos> so there are no "keling brand" motors AFAIK
[14:43:09] <TaigGecko540-4ax> well, anyway, I have these guys
[14:43:10] <TaigGecko540-4ax> http://www.deepgroove1.com/280oznema.htm
[14:43:42] <SWPadnos> oh, here it is: KL23H284-35-4B
[14:44:20] <SWPadnos> these are 387 oz-in
[14:44:59] <SWPadnos> http://www.kelinginc.net/KL23H284-35-4B.pdf
[14:45:15] <SWPadnos> $49 instead of $39 though
[14:45:31] <TaigGecko540-4ax> yeah, they are bigger
[14:46:02] <TaigGecko540-4ax> but I've been told that they are a bit slower
[14:46:06] <SWPadnos> hmmm. actually, one thing about that file - you may have to fiddle with the A axis settings a bit
[14:46:11] <TaigGecko540-4ax> for G0 moves
[14:46:39] <TaigGecko540-4ax> yeah well, I'll begin with 3 axes
[14:46:39] <SWPadnos> could be. deepgroove doesn't seem to list inductance, which is the limiting factor for speed (combined with supply voltage)
[14:47:32] <TaigGecko540-4ax> anyway thank you, are you here often?
[14:47:53] <SWPadnos> I think I had this motor (unloaded) running at 50 kHz step rate
[14:48:00] <SWPadnos> yes, I'm almost always logged in, and I
[14:48:13] <SWPadnos> I'm in front of the computer ~12 hrs/day
[14:48:18] <SWPadnos> maybe more ;)
[14:48:29] <SWPadnos> unless I'm getting coffee, like now
[14:48:52] <TaigGecko540-4ax> hehe, fine, I'll be back when I receive the beast
[14:50:16] <BigJohnT> tomp: at divide by 64 I get a 0.02 volt resolution http://imagebin.ca/view/lmbHH38e.html
[14:53:36] <tomp> yes dbld the velocity at every value
[14:54:54] <BigJohnT> everything I've read says if you can get 0.1 volts of resolution it is good enough
[14:55:07] <tomp> you got 14000userunits/sec / 60 = 233ipm?
[14:55:18] <BigJohnT> It is a nice straight line so that make the math easier
[14:55:48] <tomp> yes it sounds like you got tons of rez ( before you get excited, the smallest discernable unit shoudl be above the noise level ;)
[14:56:31] <BigJohnT> now that you mention that it is time to fire up halscope I think
[14:56:49] <BigJohnT> and I have divide by 32 and divide by 1 left to test out
[14:57:51] <tomp> my edms can discern 10mV ( well heck they're analog ) but, they're also 0-5V and have 100mV of noise that i filter ( i chuck everyuthing above 500hz)
[14:58:54] <tomp> you cant really move that fast, so ignore the fast stuff
[15:00:12] <tomp> theres frogs in the rice fields here, its a warm spring nite, windows open, the sound of scootters climbing the hill
[15:02:13] <BigJohnT> at least you don't have to listen to coqui's
[15:02:26] <BigJohnT> that drove me nuts in puerto rico
[15:03:25] <tomp> just looked em up, no these sound just like the ones in illinois
[15:03:40] <BigJohnT> that ain't bad then
[15:03:45] <tomp> nice
[15:06:32] <BigJohnT> every now and again a whippoorwill will take up residence outside my window and they can wake the dead
[15:10:36] <BigJohnT> * BigJohnT goes back to working on the THC component
[15:13:00] <alex_joni> hi BigJohnT
[15:15:24] <BigJohnT> hi alex_joni
[15:22:44] <alex_joni> BigJohnT: seems you've got competition
[15:22:45] <alex_joni> http://www.youtube.com/watch?v=Wo77Neoec0Y
[15:24:32] <jlmjvm> how can i get my counters to reset after homing,is it possible without an index pulse?
[15:25:13] <alex_joni> jlmjvm: what counters?
[15:25:19] <BigJohnT> which counters?
[15:25:30] <BigJohnT> heh
[15:25:36] <jlmjvm> my dro for x y and z
[15:25:55] <alex_joni> once you home you should be in 0
[15:26:06] <alex_joni> if other numbers are displayed, then you probably have an offset applied
[15:26:15] <alex_joni> either G92 or G54 offset
[15:26:18] <BigJohnT> unless you left an offset in there
[15:26:56] <alex_joni> jlmjvm: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?CoordinateSystems#So_if_you_re_lost_what_should_you_do
[15:27:11] <jlmjvm> this is different,if you turn power to motor off and on they will move slightly
[15:27:17] <JymmmEMC> Paint a face on a soccer ball
[15:27:20] <BigJohnT> try a G92.1 in the MDI window
[15:27:54] <jlmjvm> and that diff wont go away after homing,unless you turn emc off and rehome again
[15:28:25] <tomp> metalgeek used two rack on his gantry, no crabbing there!
[15:30:27] <alex_joni> jlmjvm: if the motor moves (on a stepper) there is no way of emc2 to know about it
[15:30:35] <alex_joni> but the numbers in the display shouldn't change
[15:31:16] <tomp> metalgeek doesnt say anything about a floating head height (THC)
[15:31:23] <jlmjvm> emc reads the encoder when it moves
[15:32:07] <jlmjvm> if i unhook encoders it always homes to zero
[15:32:37] <jlmjvm> but they have no index
[15:33:04] <JymmmEMC> steppers or servo?
[15:33:15] <jlmjvm> doesnt the index enable reset the couners after the latch during homing
[15:33:21] <jlmjvm> stepper
[15:35:13] <tomp> so you have steppers and encoders, right?
[15:35:46] <jlmjvm> yep
[15:36:00] <alex_joni> are you homing using the encoders?
[15:36:11] <alex_joni> using a HOME_SEARCH_VEL & such?
[15:36:19] <alex_joni> or just take current location, and use as home?
[15:36:42] <jlmjvm> no,use HOME_SEARCH_VEL & such
[15:36:55] <jlmjvm> want me to paste config
[15:37:47] <alex_joni> you can use pastebin
[15:38:20] <JymmmEMC> http://codepad.org/ <-- is AWESOME!
[15:38:22] <alex_joni> this is cool: http://www.youtube.com/watch?v=QYBAH6E8IZs
[15:40:46] <jlmjvm> http://codepad.org/hrw0ltob
[15:43:26] <alex_joni> http://www.youtube.com/watch?v=Nn1bJ3YAQdI&feature=related <- does anyone spot something at 4:51 ?
[15:44:08] <jlmjvm> also my encoders are always powered,motor power off i can still rotate screw and see dro move on emc
[15:44:19] <skunkworks> where it moves the controller? ;)
[15:44:58] <jlmjvm> would be a dro if i had handles
[15:45:21] <skunkworks> LawrenceG: how is the circuit working..
[15:45:59] <jlmjvm> wired straight into a parport
[15:46:21] <alex_joni> skunkworks: yeah
[15:47:14] <LawrenceG> skunkworks, milled pcb yesterday... it looks ok... have to get a digikey order done today as I dont have some of the parts
[15:49:10] <LawrenceG> http://imagebin.ca/view/0-6vZMtf.html
[15:50:39] <LawrenceG> http://imagebin.ca/view/hS96Ah1.html
[15:51:40] <tomp> hahah the head moves the operator cabinet
[15:57:42] <tomp> gnite
[16:12:26] <BigJohnT> LawrenceG: you make the mill that cut that board?
[16:14:58] <skunkworks> LawrenceG: very nice
[16:21:44] <BigJohnT> * BigJohnT ponders about feeding a PID component from the THC component to do the actual offset...
[16:29:57] <BigJohnT> I think I'll take a nap and think about it
[16:35:13] <alex_joni> heh
[16:35:16] <alex_joni> sounds good :)
[16:51:15] <jlmjvm> i changed my wiring,encoder power comes from a pin linked to machine on,homes to zero every time now
[16:52:04] <alex_joni> jlmjvm: cool
[17:00:59] <jlmjvm> gotta run,thanks
[17:09:24] <LawrenceG> BigJohnT, no, It is a shoptask lathe mill.... the backlash even with ballscrews is pretty bad (the circle on the bacl of the pcb is closer to square!
[17:13:42] <pjm__> afternoon all
[17:37:16] <JymmmEMC> LawrenceG: I picked up a 20 ft push-up yesterday, Think I might get one of these http://www.arrowantennas.com/j-pole.html
[18:02:43] <pjm__> JymmmEMC ah u must be a ham too perhaps?
[18:36:20] <BigJohnT> * BigJohnT decides to move the 5i20 to the third computer on the left as it has a better monitor
[18:39:53] <alex_joni> isn't it easier to move the monitor?
[18:42:09] <skunkworks__> logger_emc: bookmark
[18:42:09] <skunkworks__> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2009-04-12.txt
[19:05:20] <BigJohnT_> BigJohnT_ is now known as BigJohnT
[19:48:47] <JymmmEMC> If you had a choice of materials for a 8/32" screw 1" long, what has the best shear strength ?
[19:49:31] <BigJohnT> steel
[19:49:43] <JymmmEMC> SS?
[19:50:20] <BigJohnT> depends on the flavor
[19:50:34] <JymmmEMC> well, HSS -vs- SS ?
[19:51:20] <BigJohnT> I don't think you will find a screw made from HHS
[19:51:56] <toastydeath> that'd be an expensive screw.
[19:52:18] <JymmmEMC> Ok, take two 3/8" diameter aluminum rods, use an 8/32 set screw to attech them end-to-end. What material would you use?
[19:52:31] <JymmmEMC> solid rods
[19:52:51] <BigJohnT> with a rod coupling?
[19:53:01] <JymmmEMC> set-screw coupling
[19:53:08] <toastydeath> any steel then
[19:53:24] <BigJohnT> a cup point set screw of any stell
[19:53:29] <BigJohnT> steel
[19:53:41] <BigJohnT> the aluminum is your weak link
[19:53:43] <JymmmEMC> Now toss in 100MPH wind =)
[19:53:56] <toastydeath> the aluminum will fail way before the steel does
[19:53:58] <Skullworks> yup grade 5 set screw
[19:54:04] <toastydeath> no matter the size of the aluminum
[19:54:16] <JymmmEMC> grade 5 is ???
[19:54:50] <toastydeath> it's just sort of an average-ish screw.
[19:54:54] <toastydeath> in terms of strength
[19:55:05] <Skullworks> more ductile and better shock toreant than grade 8 which would crack thru the thread root.
[19:55:24] <JymmmEMC> ah
[19:57:33] <BigJohnT> I don't think you can find set screws with a grade they typ only spec the min hardness
[20:03:58] <fenn> jymmm just use some threaded rod
[20:04:20] <BigJohnT> JymmmEMC: what in the wide wide world of sports are you doing?
[20:04:26] <JymmmEMC> fenn: It's an antenna
[20:04:52] <JymmmEMC> The long element is made in two pieces, each piece is then chucked into a lathe, A hole drilled in the center of each piece, the end turned so it is flat and square. Then both are threaded 8-32. A 1 inch long set screw is attached to the upper section so the two pieces can be screwed together. The angle bracket is also drilled and tapped so the top element can be attached to the bracket for storage or transport.
[20:04:52] <JymmmEMC> This option will add $10. per antenna.
[20:04:57] <fenn> short bits of threaded rod
[20:05:22] <fenn> you're going to have a hard time finding 1" 8-32 setscrews
[20:05:42] <BigJohnT> mcmaster carr has them up to 2"
[20:05:53] <fenn> i wonder what the hell they're for
[20:06:21] <BigJohnT> I don't want to find out or have to tap that hole :)
[20:07:27] <JymmmEMC> BigJohnT: http://www.arrowantennas.com/j-pole.html
[20:08:05] <JymmmEMC> You should see the equip he's made
[20:08:21] <JymmmEMC> http://www.arrowantennas.com/sams/sam.html
[20:09:06] <BigJohnT> so the longer one is split in two and screwed back with the set screw in the middle?
[20:09:20] <JymmmEMC> yeah
[20:09:26] <JymmmEMC> for portability
[20:09:51] <BigJohnT> I would guess that it comes with the set screw...
[20:09:58] <JymmmEMC> it does
[20:12:30] <JymmmEMC> I believe he uses SS for rust prevention, but I was mostly curious on the shear strength
[20:16:55] <fenn> i'd be more concerned about fatigue
[20:19:27] <JymmmEMC> BigJohnT: Well, here's the machine that does the tapping... http://www.arrowantennas.com/sams/sam13.html
[20:20:30] <BigJohnT> a true inventor
[20:20:58] <JymmmEMC> No doubt... I was admiring all his machines, especially the rod buffing machine he made using skate weheels
[21:36:15] <alex_joni> good night all
[21:36:23] <BigJohnT> good night alex_joni
[23:11:01] <BigJohnT> heh wildrice
[23:11:54] <wildrice> Howdy. What's up?
[23:12:56] <BigJohnT> chilling and working on my THC component
[23:13:09] <BigJohnT> hows your project going
[23:13:53] <wildrice> Slow at the moment. My wife said I had to clean the workshop and throw out anything older than her.
[23:13:56] <Optic> where to find a 470uF cap on easter sunday
[23:14:10] <Optic> heh
[23:14:21] <wildrice> Radio Shack maybe?
[23:14:34] <BigJohnT> that's pretty harsh wildrice
[23:14:45] <wildrice> I have some if you want to drive to texas!
[23:16:06] <BigJohnT> texas ain't that far
[23:16:44] <wildrice> Long ago I decided that if I need something once I will very likely need it again. So I always buy in quantity. That is why the workshop has such a big junk drawer.
[23:17:30] <BigJohnT> you should see my sho
[23:17:36] <BigJohnT> shop
[23:17:59] <wildrice> But try Radio Shack . They will sometimes have the part. Or... Start opening up things and steal one off a PC board.
[23:19:29] <wildrice> You can often pull something like that off an old power supply. Just make sure you discharge the caps before poking around.
[23:20:30] <BigJohnT> Optic: knows how to discharge things :0
[23:21:03] <Optic> hehe
[23:21:05] <Optic> bzzzt!
[23:21:08] <Optic> with my TONGUE!
[23:22:14] <BigJohnT> you could open a tatoo parlor Optic
[23:23:01] <wildrice> We were discussing that while eating Easter Brisket. Son-in-law said you could get killed by licking a 9 volt battery. I had him look it up at snopes.
[23:23:38] <BigJohnT> nope, done it many times
[23:23:56] <wildrice> Me too. That's the best test.
[23:24:03] <BigJohnT> more danger in lighting your natural methane gas
[23:25:24] <wildrice> I told him that what you have to watch out for is opening a disposable flash camera. Those can pack a wallop.
[23:25:51] <eric_unterhause1> I'm trying to read the manual for my VFD -- rough going
[23:26:16] <BigJohnT> it is in mandern right
[23:27:24] <eric_unterhause1> 'merican
[23:27:34] <BigJohnT> lol
[23:27:45] <eric_unterhause1> JMK wrote it
[23:28:13] <BigJohnT> it is in programmer tongue then
[23:28:30] <eric_unterhause1> actually, it's just very complete
[23:28:35] <wildrice> Which dialect?
[23:28:42] <Optic> my linksys switch doesn't talk to my mac anymore, it has a bulgy cap
[23:28:44] <Optic> poor thing
[23:29:10] <BigJohnT> .45acp between the eyes is the only cure for that
[23:29:23] <eric_unterhause1> routers seem to have a weakness in their power supplies
[23:29:32] <eric_unterhause1> all of mine have anyway
[23:31:09] <eric_unterhause1> Here is my favorite part, my VFD should not be on a circuit that can deliver more than 5000 amps short circuit
[23:31:29] <Optic> they are cheap caps, STORM brand
[23:31:44] <Optic> probably replace with pansonic caps and never have a problem again
[23:32:00] <eric_unterhause1> my router is heat sensitive
[23:32:30] <eric_unterhause1> I should probably replace the caps
[23:32:51] <eric_unterhause1> I tried freeze spray, but that didn't pinpoint the problem
[23:33:03] <eric_unterhause1> it did make it start working though
[23:33:57] <BigJohnT> * BigJohnT is listening to the Andrews Sisters anyone know who they are?
[23:34:12] <eric_unterhause1> yes, they are the Andrews Sisters
[23:34:21] <eric_unterhause1> hth
[23:34:38] <wildrice> Boogie Woogie Bugle Boy
[23:34:54] <BigJohnT> LOL and don't sit under the apple tree
[23:35:08] <BigJohnT> ok, wildrice you just showed your age :)
[23:35:09] <wildrice> With anyone else but me
[23:35:17] <eric_unterhause1> I think all their songs were covers of someone else's hits
[23:35:57] <wildrice> Yep. I am older than dirt.
[23:36:05] <BigJohnT> now we are rocking to Bondie
[23:36:36] <eric_unterhause1> There was a time when I would have liked to play Bondie with Debby Harry
[23:36:36] <BigJohnT> 55 here
[23:36:54] <eric_unterhause1> yer old
[23:37:05] <wildrice> 54 next weekend here
[23:37:24] <eric_unterhause1> this place is a regular old folks home
[23:37:59] <wildrice> But we are on the bleeding edge of milling!
[23:38:06] <BigJohnT> :)
[23:38:12] <eric_unterhause1> I try to avoid the bleeding
[23:38:41] <eric_unterhause1> so far it's just the girls that call me stumpy
[23:39:02] <BigJohnT> they call me dad
[23:39:30] <eric_unterhause1> they call me grandpa, but their grandmothers used to call me "stumpy"
[23:39:59] <BigJohnT> LOL
[23:40:31] <eric_unterhause1> some young thing said "oh, it's nice that you can still ride a bicycle"
[23:43:05] <BigJohnT> I can still ride a bike http://i47.photobucket.com/albums/f163/johnplctech/HPIM1988.jpg
[23:43:32] <eric_unterhause1> do motorcyclists get right hooked?
[23:44:00] <eric_unterhause1> I know they get the left cross treatment a lot
[23:44:09] <BigJohnT> only if you can do 140 mph
[23:44:46] <eric_unterhause1> I got right hooked on my bike tonight, chased the offending driver down
[23:44:56] <eric_unterhause1> "I thought you were supposed to yield to me"
[23:45:02] <fenn_> fenn_ is now known as fenn
[23:45:40] <BigJohnT> there are so many nuts behind the wheel it is scary
[23:45:46] <eric_unterhause1> don't know how that's supposed to work, first I knew she was there was when she forced me to turn right
[23:46:07] <BigJohnT> been there before, it sucks
[23:46:17] <BigJohnT> that is why we carry in MO
[23:46:29] <eric_unterhause1> usually I beat the crap out of the car
[23:46:31] <wildrice> Yep, the first thing you need to learn on a bike is how to lay it down.
[23:46:40] <eric_unterhause1> laying it down is death
[23:46:45] <eric_unterhause1> and I'm serious
[23:46:52] <eric_unterhause1> you want to be on top
[23:47:06] <BigJohnT> the second thing you need to know is you can stop faster when you tires are on the ground
[23:47:15] <cradek> first thing you need to learn is to stop, which involves the front brake
[23:47:19] <eric_unterhause1> I'll take a note of that
[23:47:24] <cradek> so many idiot riders don't know how to stop their bike
[23:48:01] <eric_unterhause1> I need to practice my cleating
[23:48:02] <cradek> if a biker says "I had to lay it down" it means he did something dumb and lost control
[23:48:09] <BigJohnT> I was conserned when I found out that the rear brake controlled one front caliper but it still stops faster than anything else I own
[23:49:03] <eric_unterhause1> somebody runs you off the road can't exactly complain if you do $5000 damage to the side of the car, can they?
[23:49:40] <eric_unterhause1> as long as you're fast enough
[23:50:06] <BigJohnT> btw, cradek I'm getting in the mid 40's for milage on the new wing
[23:50:13] <cradek> nice
[23:50:23] <BigJohnT> much better than the 98
[23:50:27] <cradek> I only get about 35 on mine
[23:50:28] <eric_unterhause1> 48 in the Prius
[23:50:40] <cradek> is the goldwing fuel injected?
[23:50:44] <BigJohnT> yes
[23:50:53] <cradek> that explains it - totally different technology
[23:51:08] <BigJohnT> only the last few years have the good milage
[23:51:19] <eric_unterhause1> you probably have a lead ..... right hand
[23:51:23] <BigJohnT> 1832cc
[23:51:35] <cradek> that's a car engine :-)
[23:51:41] <BigJohnT> 0-140 faster than you can write the check for the ticket
[23:51:57] <BigJohnT> it's bigger than my wifes honda civic
[23:52:02] <eric_unterhause1> bigger than the engine in the prius,that's for sure
[23:52:33] <BigJohnT> but she does have a 20hp electric motor too
[23:54:59] <BigJohnT> in the integral term section can anyone explain the part after the K sub i in the formula?
[23:55:03] <BigJohnT> http://en.wikipedia.org/wiki/PID_controller
[23:55:27] <BigJohnT> the swoop de do with the t and the 0
[23:55:53] <eric_unterhause1> integrator?
[23:56:08] <BigJohnT> I don't know what you call it
[23:56:21] <eric_unterhause1> yes, that is the integral sign
[23:56:22] <cradek> the accumulated error since the beginning of time
[23:56:52] <eric_unterhause1> it's summing the error, as cradek says
[23:57:02] <wildrice> I thin that that is the sum of the value following it as it goes from 0 to the time t.
[23:57:08] <BigJohnT> ok, that makes sense to this redneck
[23:57:21] <eric_unterhause1> in a digital control, it is actually a summation
[23:57:42] <cradek> it means the longer the error has been there, the more the integral term grows, so the higher the number gets multiplied by that Ki
[23:57:45] <eric_unterhause1> unless someone gets all obsessive about it
[23:58:13] <BigJohnT> ok, thanks
[23:58:19] <eric_unterhause1> integral term gets rid of steady state errors
[23:58:29] <wildrice> You add the errors up over time and divide by the time
[23:58:32] <eric_unterhause1> and is somewhat destabilizing
[23:59:10] <cradek> yes it's because of being there for a long time that makes the Iout grow, so the steady state error gets compensated
[23:59:32] <eric_unterhause1> just a note: you don't understand the de(t)/dt term either, but there is no funny swoop