Back
[02:15:18] <DJ9DJ> moin
[02:29:51] <t12> morning
[02:29:55] <t12> scraping isnt so hard it seems
[03:04:03] <mrsun> i didnt find it to hard, how good the quality of my work is is another thing tho :P
[03:09:04] <t12> hah yes
[03:09:04] <t12> same
[03:10:15] <mrsun> scraped almost every surface on my X1 ... 1 month of work for a beginner :P
[03:10:23] <mrsun> cost more then the mill does in worktime ;P
[03:10:31] <mrsun> but damn its smooth running after it =)
[03:12:40] <t12> the g0704 conversion is tempting
[03:12:46] <t12> but it def seems like when its done
[03:12:51] <t12> all that was purchased is the castings
[03:13:12] <t12> not that i see any alternative besides casting them myself
[03:14:58] <mrsun> i wish i could cast big parts myself =)
[03:15:14] <mrsun> i would soo make me a killer milling machine ;P
[03:16:29] <t12> i figure welding a frame together adds too much stress
[03:16:31] <t12> but maybe not?
[03:16:40] <t12> or maybe just hack it up and go find someone with a furnace to anneal it
[03:18:33] <mrsun> stress releief it? :)
[03:18:42] <mrsun> yes
[03:18:45] <mrsun> you said that already :P
[03:18:52] <mrsun> but then if its steel you get it in its "soft" state
[03:19:07] <mrsun> and steel isnt very good compared to cast iron in machines as i understand it :P
[03:20:42] <mrsun> if i had bigger burners i could cast it myself ... so far ive been able to melt 700g of cast iron ... far from a mill table or base but :P
[03:22:10] <t12> could always go the opp direction
[03:22:13] <t12> make the _tiniest_ mill
[03:22:36] <mrsun> pfft :P
[03:23:27] <mrsun> need to make a bigger and better built furnace
[03:23:46] <mrsun> the one i have is an old alu casting furnace that ive cut down and converted for a home heating oil burner
[03:23:51] <mrsun> very small volume
[03:23:58] <mrsun> might be why i manage to melt cast iron in it :P
[03:25:48] <mrsun> but need to get other projects done :P making new gib for Y axis of the mill .. scraping them flat
[03:25:55] <mrsun> or lapping ...
[05:52:31] <jthornton> good luck trying to lap anything flat
[06:56:13] <Ranewen> g day, is anyone into servo motors?
[06:58:41] <archivist> a few, ask the real question
[06:58:41] <jthornton> just ask the question
[07:01:17] <Ranewen> o
[07:01:18] <Ranewen> k
[07:02:04] <Ranewen> i got 12v PSU, and is it ok to connect my 2 servos (6v) in series... because I'm affraid of destroying them.
[07:02:42] <Ranewen> ind i am using an avr (5V) to signal wires
[07:04:45] <cncbasher> Ranewen: i presume these are RC Servos ? in that case no do not connect to greater than 6v
[07:05:14] <Ranewen> yes they are rc servos
[07:05:22] <Ranewen> so i cant connect them in series?
[07:05:33] <Loetmichel> unless the servos are desiogned to meet an external battery
[07:05:34] <cncbasher> no
[07:05:46] <Ranewen> thx
[07:05:52] <Ranewen> im glad i didnt connect them
[07:06:31] <Loetmichel> Ranewen: get yourself an 78S06 regulator and 2 capacitors
[07:06:38] <Loetmichel> and make the 12V to 6V
[07:06:56] <Loetmichel> careful: the regulator gets HOT
[07:07:00] <Ranewen> i know
[07:07:06] <Ranewen> i got
[07:07:19] <Ranewen> like dc-dc switching from ebay few $$
[07:07:32] <Loetmichel> ... or use a R/C-mode "battery elemination ciruit"
[07:07:34] <cncbasher> use a 5v pc power supply
[07:07:42] <Ranewen> dont have pc
[07:07:51] <Loetmichel> thats exactly that 78Sxx in ready to go
[07:07:57] <Ranewen> im using scalvaged parts
[07:08:01] <Ranewen> like lifepo4
[07:09:03] <Ranewen> batteries, and cheepo phone chargers... ima poor man
[07:09:36] <Loetmichel> hmmm, is there by any chance an Brushless ESC where you salvaged the servos from?
[07:09:46] <Loetmichel> usually they have internal BEC
[07:13:10] <Loetmichel> so you can use the 12V PSU at the battery input of the esc and the servo input will spit oit your needed 5/6V for the servos
[07:13:13] <Loetmichel> out
[07:16:45] <Ranewen> Loetmichel, i ordered the servos alone from ebay
[07:16:59] <Ranewen> black ones.... metal gearboxes
[07:17:28] <Loetmichel> then you are out of luck ;-)
[07:19:17] <Ranewen> Loetmichel, guess so... i got some 7805
[07:19:36] <Ranewen> but since the motors are +5W, 7805 will burn...
[07:20:24] <Loetmichel> not neccesarily
[07:20:47] <Loetmichel> the servos need the 2-3A only when blocked
[07:20:57] <Ranewen> x5 volts
[07:21:00] <Ranewen> =10W
[07:21:06] <Loetmichel> free runnung or with light load 1A should be sufficient
[07:21:06] <Ranewen> you know the cooler for 10W
[07:21:22] <Ranewen> hmm
[07:37:02] <mrsun> hmm how to heck squareness of something without having a super duper square for thousands of dollars :P
[07:41:34] <Ranewen> mrsun, indeed
[07:41:34] <mrsun> s/heck/check/g
[07:41:45] <jthornton> a test indicator and a surface plate
[07:42:04] <Ranewen> anyone got cheap ultrasonics from ebay
[07:42:11] <mrsun> got both, but how do i check the actual squareness of something using it ?
[07:42:22] <mrsun> i need to surfaces square .. like in an 90 degree angle plate
[07:43:07] <jthornton> looking for the link
[07:44:22] <mrsun> http://www.youtube.com/watch?v=baqCWk4OmwU
[07:44:23] <mrsun> hmm
[07:48:31] <jthornton> http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CEMQFjAB&url=http%3A%2F%2Fwww.neme-s.org%2FModel_Engineer_Files%2F3065-Basic%2520Squareness.pdf&ei=YGbxUPXhJ8nligKn4IDYCw&usg=AFQjCNFVnPSX9aVZnf_JgCM1UOJzY_M1og&bvm=bv.1357700187,d.cGE&cad=rja
[07:48:40] <bedah> moin
[07:49:36] <jthornton> mrsun, found it
http://www.thetoolanddieguy.com/archives/115
[07:53:16] <archivist> mrsun, you can make a square on your lathe if you have a micrometer
[07:55:50] <jthornton> how do you do that?
[07:56:31] <archivist> circular square :)
[07:56:42] <jthornton> ah
[08:03:19] <archivist> there were some 1/2 " thick foot square black granite tiles in the hardware shop today, looked just right for baby surface plates
[08:04:27] <archivist> no wavyness of the roof reflection so reasonably flat
[08:04:59] <jthornton> how would you check the flatness?
[08:06:02] <archivist> could use a mirror and collimator
[08:06:36] * jthornton looks up what a collimator might be
[08:07:40] <jthornton> is that something you can make in the shop?
[08:08:53] <archivist> yes but not that easy
[08:11:56] <jthornton> have you seen this?
http://bbs.homeshopmachinist.net/threads/18917-Qualitative-flatness-testing-cheap-%28pics%29?highlight=flatness+salt
[08:12:28] <jthornton> well minus the auto highlight
http://bbs.homeshopmachinist.net/threads/18917-Qualitative-flatness-testing-cheap-%28pics%29
[08:14:48] <archivist> yes I did the other week
[08:16:29] <jthornton> I'm still not sure how he generated the line instead of a spot
[08:17:08] <archivist> there are line lasers, one goes through a circular lens
[08:17:17] <jthornton> I'll have to go look to see if my laser makes a line too
[08:18:43] <archivist> jthornton, as found on the net the collimator and mirror method
http://www.archivist.info/cnc/pdf/
[08:21:40] <jthornton> thanks
[08:22:35] <archivist> I have a poor mans collimator, called an angle dekkor
[08:25:52] <archivist> I have a soft spot for fleabay 170972503003 when measuring relative angles
[08:26:54] <jthornton> looks like it has been well cared for
[08:27:39] <archivist> cost me £80 to replace the bubble in mine
[08:31:01] <archivist> cannot afford a mirror fleabay 281046619929
[08:32:08] <awallin> looks like a silver mirror? you can probably build the mount for less than 700
[08:32:42] <archivist> jthornton, I should call them autocollimator not just collimator see 380531770126
[08:33:33] <awallin> what are those used for?? :)
[08:33:53] <archivist> awallin, I have a precision polygon which has 8 mirror, I suppose I could just make a mount
[08:34:22] <archivist> awallin, machine testing see
http://www.archivist.info/cnc/pdf/
[08:36:09] <archivist> add a 90 deg prism and you can check squareness and z travels etc
[08:36:41] <Ranewen> using rc servos under heavy load, but they just cant stop getting warmer, how can i chill them ?
[08:37:41] <awallin> archivist: someone must have DIYed a modern version with a diode-laser and a webcam??
[08:38:11] <archivist> not seen yet :)
[08:39:43] <awallin> should build an interferometer for checking screw-comp too :) that might be harder..
[08:41:58] <archivist> ones list of "should do" exponentially grows with age
[08:45:10] <mrsun> jthornton, haha that tool and die guy realy thinks big of himself :P
[08:45:59] <mrsun> archivist, yeah read about them in a book but that requires my lathe to turn perfectly parallel and wont bother trying to make it do that on the bench its on right now :P
[08:46:49] <archivist> mrsun, adjust with wet and dry to size :)
[08:47:04] <mrsun> archivist, huh ? :)
[08:47:13] <pfred1> kitty litter?
[08:47:38] <archivist> fine sandpaper
[08:48:12] <pfred1> I use sawdust as speedy dry myself
[08:48:27] <pfred1> I usually have a bunch in my garage
[08:51:40] <mrsun> archivist, so if its the same dia at top and bottom and then i face of the bottom it will be a near perfect square then
[08:52:23] <archivist> should be better than near
[08:52:55] <archivist> depends how well you can measure the diameter
[08:53:08] <mrsun> micrometer is the closest i have .. 0.01
[08:53:21] <mrsun> mm
[08:57:07] <mrsun> i wonder how good my granite stone is =)
[08:57:13] <mrsun> in flatness
[08:57:27] <mrsun> according to the paper its within grade 00 tolerances .. but if that is correct i dont know :P
[08:57:30] <pfred1> surface plates are usually pretty accurate
[08:57:52] <mrsun> its grade 0 on the place i bought it from, then it came with a "test" paper and everything was bellow 00 maximum spec
[09:01:35] <mrsun> tho it was a cheap surface plate :P
[09:01:56] <mrsun> so never know how those things are realy real :P
[09:02:32] <pfred1> I never make anything that needs to be super accurate by design
[09:03:45] <mrsun> heh, found granite angles for like $130 or so ... i wonder if that is anything to have =)
[09:03:53] <mrsun> so i can spot stuff to scrape them to squareness
[09:03:59] <mrsun> hard to spot with a cylindrical square
[09:04:26] <pfred1> angle plates are usually made out of steel
[09:04:28] <mrsun> oh but they were very thin .. :/
[09:04:52] <mrsun> pfred1, yes but this is for work on the surface plate, to check work/spot them for scraping
[09:05:06] <mrsun> but at 20mm thick its kinda useless anyhow :P
[09:05:17] <pfred1> steel angle plates work on surface plates
[09:05:30] <pfred1> it is how we used to test our wheel dressers
[09:06:35] <pfred1> I rebuilt one of those once and it didn't move a 0.0001 indicator going through its range of motion
[09:07:11] <pfred1> I showed it to another machinist and he suggested I ask for a raise :)
[09:07:30] <mrsun> 0.015mm/100mm for a square i guess thats kinda accurate also
[09:07:43] * pfred1 doesn't do metric
[09:07:48] <mrsun> hehe
[09:08:05] <pfred1> well OK I do Imperial metric I suppose
[09:08:15] <mrsun> 0.00254mm is 0.0001 inch
[09:08:48] <pfred1> yeah to me 0.0001 is a nice round number but 0.00254mm is a bit clumsey to work with
[09:08:51] <mrsun> 0.00059inch/4 inches then
[09:10:16] <pfred1> I have a micrometer set I never break the thing out and use it though
[09:12:08] <pfred1> mrsun whenever we wanted anything closer than 0.003 of an inch we ground it we never tried for that kind of accuracy on a lathe or a mill
[10:12:47] <rizo> I have a problem with parport pci card. The lpci-v shows different I/O ports, first at dc00 (this is the one i used for hal settings). I tested the simple pwm at 1hz and halmeter shows the parport.1.pin is changing its value, but the led on that pin is not plinking.
[10:13:31] <pfred1> you can't see an LED blink at 1 Hz
[10:13:38] <rizo> no
[10:13:57] <archivist> sure you can see 1hz
[10:14:19] <pfred1> wait a minute you're driving an LED off a parallel port pin?
[10:14:55] <rizo> I have 5V - led - resitor - parportpin
[10:15:15] <rizo> is it possible the parport addres is wrong
[10:16:04] <pfred1> so you just want the parallel port to sink current?
[10:16:21] <JT-Shop> http://linuxcnc.org/index.php/english/forum/18-computer/4743-parallel-port-tester?limitstart=0
[10:16:36] <rizo> Yes, but yust to test if the parallel port is working good.
[10:16:41] <pfred1> I'd have to see a schematic of exactly what you're doing but so far it doesn't sound good
[10:17:33] <pfred1> rizo this is the BOB I made:
http://www.instructables.com/id/Parallel-Port-Break-Out-Board-BOB/
[10:17:44] <pfred1> it is a bit more sophisticated than an LED on the port
[10:17:46] <rizo> The schematics is simple: +5V - led - 330Ohm - resistor. The same circuit works fine on motherboard parport connector
[10:18:48] <rizo> The problem is in hal. Maybe the parport address.
[10:18:58] <pfred1> when I first started investigating parallel ports i just used a multimeter set on the voltage scale myself
[10:19:11] <pfred1> it is a lot less of a load
[10:19:35] * pfred1 has *never* burnt a parallel port out yet either
[10:20:19] <pfred1> my BOB loads the port something like a tenthousandth of an amp
[10:20:48] <pfred1> I forget the exact vaule today it is low though
[10:21:02] <Ranewen> small oftopic: blinking LED's are must have for sucessfull consumer products
[10:21:20] <pfred1> Ranewen I never make anything without at least one LED on it
[10:21:48] <Ranewen> i don't know the link, but blinky LED's sell more
[10:21:50] <pfred1> probably because I have a drawer of the suckers i need to use up
[10:22:44] <rizo> Why are there the IO ports at pci with different length, five size=8 and one size=16
[10:23:00] <rizo> which one is the one i should use?
[10:23:21] <pfred1> the one that works
[10:24:10] <pfred1> rizo just so you know I've done all of this so it works (TM)
https://www.youtube.com/watch?v=cgbeyNNBZ68
[10:25:10] <pfred1> I am much more cautious running current into parallel ports than you seem to be though
[10:25:18] <rizo> i also have the working cnc machine
[10:25:24] <rizo> made by myself
[10:25:41] <rizo> but now i have the problem with parallel port that i can not solve
[10:25:57] <pfred1> buffer the port
[10:26:52] <rizo> I tested this circuit more than 10 times on other ports. This is not a problem for parallel port to sink.
[10:27:38] <rizo> I tried to connect the Voltage meter instead of led.... same result.
[10:29:18] <Ranewen> hey, i will dismantle my pc, i need to know can i use Copper (Carlube) Multi-purpose grease, Helps prevent disc brake squeal, anti-seize assembly compount................................ between my proc and heatsink ?
[10:29:44] <pfred1> rizo some parallel ports can sink less than 3ma
[10:29:49] <pfred1> they vary
[10:30:11] <Tom_itx> Ranewen, i doubt it
[10:30:12] <pfred1> there is no spec today for what a parallel port can sink believe me i looked for it
[10:30:44] <Ranewen> because copper has the "highest" thermal conductivity
[10:30:53] <pfred1> silver does
[10:31:19] <pfred1> rat shack sells thermal compound
[10:31:57] <Ranewen> http://en.wikipedia.org/wiki/List_of_thermal_conductivities
[10:32:00] <rizo> OK, i disconnected the led... i connected the voltage meter... same result
[10:32:19] <Ranewen> this paste is "greasy" copper.....
[10:32:27] <JT-Shop> does the parallel port tester show your port working?
[10:33:05] <Ranewen> Thermal grease, silver-based 8.89+ Copper, pure 385
[10:33:24] <rizo> still working on tester....
[10:33:38] <JT-Shop> the one I linked to you
[10:34:09] <rizo> will try...
[10:34:53] <pfred1> also some thermal compounds can damage electronics over time too
[10:35:09] <pfred1> CPUs aren't exactly as hardy as disc brake pads
[10:42:03] <rizo> JT-Shop, i tried the tester...it works fine on both parallel ports
[10:43:26] <rizo> i don't understand this...in my hal code, i can see with halmeter the value of pin is changeing with no effect on output
[10:46:56] <JT-Shop> what is the output connected to?
[10:48:42] * JT-Shop is working with a short sleeved shirt outside in January and sweating
[10:50:54] <rizo> output of what...parport...to led
[10:52:15] <rizo> i have pwmgen.0.pwm connected to parport.1.pin-17-out
[10:53:22] <JT-Shop> does the paralle port tester light up the led?
[10:53:52] <rizo> yes
[10:56:53] <rizo> sorry
[10:56:58] <rizo> found an error
[10:57:19] <rizo> i forgot to add the parport.0.read/write
[10:57:23] <rizo> :))
[10:57:36] <rizo> sorry for my stupidity
[10:58:39] <rizo> i forgot to add the parport.1.read/write
[10:59:52] <rizo> strange... halmeter did not detect this...
[11:25:31] <tjtr33> http://pastebin.com/Kh4kHnPB kinda complement to weighted_sum, often we use groups of i/o in chunks of 4/8/16, so this is a "bytewrite"
[11:42:41] <rizo> JT-Shop, can i program the partest program pins to hold its state
[11:45:18] <JT-Shop> yea, change the button to a check box
[11:45:44] <JT-Shop> http://linuxcnc.org/docs/html/hal/pyvcp.html#_buttons
[11:48:35] <rizo> i see... thank you
[11:54:21] <tjtr33> linuxcnc is just so flexible
http://videobin.org/+6ad/720.html next... Mxxx to set values
[12:12:04] <bpuk> Well, that looks like a slightly more reasonable jitter. 9249 (instead of ~28-30k)
[12:13:21] <IchGuckLive> bpuk: yes you got it
[12:13:51] <IchGuckLive> run latency with 4 glxgears and pull a fle > 400MB around then do a USB into
[12:15:19] <bpuk> 4 glxgears, plus compiling lcnc 2.6, and plugging in a usb brings it up to 9432
[12:15:37] <IchGuckLive> then you are the SERVO king
[12:15:51] <IchGuckLive> with 200ksteps per second
[12:15:59] <bpuk> Not put any servo's in yet - this is just a motherboard upgrade :D
[12:16:57] <bpuk> or connected this PC to the lathe yet - still doing all the software installs inside where it's warm
[12:24:42] <archivist> what is this warm you speak of
[12:25:42] <bpuk> about 5 degrees - think that's about 40 in english. Warmer than where the lathe is
[12:26:40] <archivist> I got my lathe indoors, hoping the pc will warm the place up a bit
[12:27:08] <bpuk> darn, changing to a low-power, high efficiency PC was probably a bad idea then ;)
[12:27:25] <bpuk> and yeah, PC's do tend to produce a suprising amount of heat
[12:27:55] <archivist> 5 deg out doors 6.6 in so the pc's are helping
[12:28:42] <kwallace> Off topic but, what is the modern equivalent to the WRT54g for hacking low power Linux apps?
[12:37:58] <ReadError> whats a good aluminum cutting fluid
[12:38:05] <ReadError> that i could get from mcmaster or amazon
[12:41:10] <kwallace> I still think kerosene works well on aluminum. I put a little bit of oil in it.
[12:42:53] <kwallace> Diesel fuel might work well.
[12:43:13] <bpuk_> tried WD40?
[12:43:43] <ReadError> yea i been using WD40 and 3-in-1 mix
[12:43:53] <ReadError> just wasnt sure if there was any better option
[12:44:52] <bpuk_> I use WD40 on the router - and normal soluble oil on the mill - I get a better finish with WD40 - but there is still the occasional problem of BUE and chip welding
[12:45:29] <Loetmichel> bpuk_: i cant stand the smell of WD40 in high volumes at milling
[12:46:11] <Loetmichel> i just use a mix of water, dish detergent and castor oil ;-)
[12:46:31] <Loetmichel> in a Flower pump spray can
[12:47:05] <t12> i've started to like the smell of tapmagic
[12:49:23] <kwallace> Water based fluids made a smelly, rusty mess of my mill, so I avoid water based fluids, like the plague, unless the machine is made specifically for water based fluids.
[12:50:22] <Loetmichel> my mill is made from aluminium
[12:50:47] <Loetmichel> and using the pump spray means low volume, so no mess
[12:51:05] <Connor1> http://www.tormach.com/store/index.php?app=ecom&ns=prodshow&ref=31750
[12:51:20] <Loetmichel> fro time to time one has to use alcohol to get the smeary castor oil from the components, thats all
[12:51:22] <Connor1> I got some of that to use.. I've ran it through a bit... but, haven't used it to cut yet..
[12:51:32] <Connor1> no signs of rust on anything.. and I ran it through months ago.
[12:57:26] <kwallace> My machines were made before water based fluids became the standard. I would think newer higher quality machines should be okay. In one shop, we had a 500 gal. of coolant which was used to fill the machine sumps. It needed attention a few times a day every day.
[13:00:21] <Loetmichel> kwallace: normal industrial cooling fluids have a rust inhibitor so the machine wont rust even though the fluid is < 50% water
[13:01:07] <Loetmichel> and a bacteria inhibitor to conteract the "gets smelly" effect
[13:02:21] <mrsun> even if the fluid is less then 50% water? :P
[13:05:38] <kwallace> You're right, but my experience has been that if one doesn't pay proper and constant attention to the water based fluids, they can be a disaster. The marketing claims only seem to apply to ideal conditions, so the buyer needs to be aware.
[13:08:14] <Loetmichel> kwallace: right
[13:08:37] <Loetmichel> but not applicable if one uses minimum amounts by hand from a spray gun
[13:16:19] <kwallace> Even then, if the machine sits for a while, the water in the slides can separate out and rust. If the machine is going sit for a while, all of the water needs to be removed which is next to impossible. I just avoid ever having any water on my machines in the first place.
[13:19:05] <devilindisguise> hello!
[13:19:09] <gene__> Hi guys, got a prob with the or2 module. Either input=true should give a true out, not so, both have to be true
[13:19:44] <devilindisguise> i am having issues running linuxcnc within qemu
[13:20:21] <gene__> Linuxcnc only runs well with its own kernel
[13:20:38] <cradek> gene__: works for me...
[13:20:39] <cradek> 32769 bit IN TRUE or2.0.in0
[13:20:40] <cradek> 32769 bit IN FALSE or2.0.in1
[13:20:40] <cradek> 32769 bit OUT TRUE or2.0.out
[13:21:22] <gene__> I can't get a true out of it for one input true.
[13:21:54] <cradek> gene__: I don't see how that can be!
[13:22:14] <gene__> me either Chris
[13:22:16] <devilindisguise> gene: mhm, ok. well, its the rtai patched kernel delivered with linuxcnc based on ubuntu 10.04...
[13:22:41] <gene__> CXhris that ones yours
[13:22:52] <gene__> Chris that ones yours
[13:23:42] <devilindisguise> the rtapi setup seems to fail
[13:23:58] <gene__> humm, lemme check my addf statement
[13:24:10] <devilindisguise> is this known, or is it my specific qemu that doesnt work?
[13:24:53] <gene__> IIieeeeeuuuuup, missing. I'll get me coat
[13:25:23] <bpuk> who the heck decided that kettle leads were a good connector for computers. *mutter*
[13:25:57] <christian___> hallo zusammen bin linucnc neuling gibt s hier hilfe fg
[13:26:14] <tjtr33> appliance cord = kettle leads :) never heard that one
[13:26:27] <cradek> gene__: we've all done it! [and more than once]
[13:27:13] <bpuk> or IEC C13 I think. Stupid connector for a computer, should have a locking clip of some kind. Yes I just knocked the lead out ;)
[13:28:26] <devilindisguise> christian: kommt auf dein problem an. englisch ist die sprache der wahl.
[13:29:05] <christian___> ja da hammas des problem
[13:32:11] <devilindisguise> has anyone run linuxcnc 2.5.1 within qemu?
[13:32:44] <cradek> devilindisguise: you oughta run the simulator version, not the realtime
[13:33:22] <gene__> Ok, now, is motion.probe-input level sensitive or edge?
[13:33:41] <devilindisguise> cradek: ok. is tehre a deb with with linuxcnc compiled for sim?
[13:33:51] <cradek> devilindisguise: yes
[13:34:23] <devilindisguise> cradek: ok, thanks. another thing. is it rtai or do you know why it fails to start the rtapi part?
[13:34:25] <cradek> http://linuxcnc.org/dists/lucid/linuxcnc2.5-sim/binary-i386/
[13:35:02] <cradek> no idea, especially since you haven't said what the error is. but it's silly to worry about running rtai in a virtual machine anyway
[13:36:09] <devilindisguise> cradek: well, for testing i really like virtualization.
[13:36:20] <IchGuckLive> cristian wars off befor i coudt jump in im from germany !
[13:36:24] <devilindisguise> cradek: but ill go into that and post some deatils
[13:36:27] <IchGuckLive> as may here 2
[13:37:50] <IchGuckLive> if he cristian comes in ones more plesa direct him to y german tutorials
http://www.youtube.com/watch?v=SjXwTWWHULc
[13:38:05] <IchGuckLive> and i am alwas on at 19-20 Berlin time
[13:38:13] <IchGuckLive> BY for today
[13:43:40] <devilindisguise> ok __rtai_lxrt_init is making trouble, liblinuxcnchal.so functions cause a segfault because of that. says the kernellog.
[13:49:30] <mrsun> to check tailstock alignment of the lathe i can just what ain indicator in the chuck and sweep it around a dead center in the tailstock right? :)
[13:49:54] <AR__> yes
[13:51:13] <archivist> mrsun, get the bed level (both ends, remove twist caused by bench) first
[13:51:24] <mrsun> ye i guess :P
[13:51:39] <mrsun> need to get the new lathe bench finnished :/
[13:52:00] <gene__> cradek? could I get enough bounce contact between the tool and the face of a piece of pcb to effectively make motion.probe-input edge sensitive?
[13:52:02] <mrsun> sucks that the concrete floor in the cellar is just thrown in there, so someplaces its like 2cm thick, others 5, others 10 etc :/
[13:52:36] <mrsun> http://www.practicalmachinist.com/vb/attachments/f45/65754d1356634551-what-unusual-uses-there-cylindrical-square-cynsq.jpg now THERES a way to check a angle plate for level!
[13:52:40] <mrsun> :P
[13:52:42] <archivist> I went to a grinding place to ask about straightening and the first thing the bloke asked was about my bench :)
[13:52:53] <cradek> gene__: if your bounces last for more than a servo cycle, certainly that can cause you trouble
[13:53:00] <mrsun> level->square
[13:53:27] <gene__> 2khz servo-thread
[13:54:03] <mrsun> thats freakin brilliant .. if cylindrical squares actualy is that square that you say that must be the ultimate way to check for square =)
[13:54:17] <cradek> you could debounce with a 1-3ms or so time constant in hardware or software
[13:55:46] <gene__> Looks like I'll have to, darnit
[14:02:30] <devilindisguise> i put the debug output to:
http://nopaste.info/f345655868.html
[14:03:13] <devilindisguise> i tried without apic, hpet... no difference, also qemu or qemu with kvm doesnt make a difference
[14:03:25] <MercuryRising> is there a good reason to have the heated bed below the glass transition temperature of abs? i have been printing at ~120C, and it sticks great
[14:05:53] <devilindisguise> something completly different. has anyone built or used a rl interface to axis? like rotary encoder knobs, jog/shuffle dials, knobs to control the UI?
[14:06:11] <MercuryRising> whoops, wrong one!
[14:06:27] <devilindisguise> i have some of those&some avr microcontrollers laying around and wonder if they could be put to use
[14:13:21] <gene__> This crazy Chris. Put a clip lead held against the copper of the pcb, solid contact. put clip lead holdown bolt of qc tool, noisier than hell on halscope
[14:15:34] <cradek> gene__: my very good commercial probe has a 3-4 ms time constant. you sure have to probe slowly, at least on the final probe move.
[14:20:10] <gene__> I can hold the probe solidly on the tool tip, and the led on the C1G is not out, but dim. So I need to do some filtering. My big scope says its the nominally 40 khz motor switching as they sit there idele.
[14:21:08] <gene__> That is essentially random noise at a 2khz servo-thread rate
[14:22:57] <gene__> could be frozen in either state as thread run time.. what would be a good 'gain' for a lowpass in series with that?
[14:23:06] <gene__> s/as/at
[14:27:05] <tjtr33> how to write the mu symbol into a vcp? ( eg: for micro inch or micro second )
[14:28:52] <devilindisguise> ok, jeff eppler did it, but is blog is down. well, a good starting pint.
[14:28:56] <devilindisguise> cu folks
[14:33:31] <gene__> Humm, need a one logic line debounce, man page "loadrt debounce cfg=" is not well defined, please elaborate
[14:35:26] <ReadError> i learned the handy feature of "Run from here" last night
[14:35:31] <ReadError> after i broke and endmill
[14:37:11] <gene__> I told it cfg=1, then get debounce.0 does not exist?
[14:37:53] <pcw_home> gene what about a .01 capacitor...
[14:39:01] <gene__> I have done that too, unhandy but helped, didn't solve
[14:40:50] <gene__> what I need is a cfg line for debounce's loadrt that sets up one line of debounce.
[14:41:29] <gene__> or, how do I state it in the net statement since debounce.0.in is an error
[14:45:22] <gene__> ahh, found it, comment troubling line out, then inspect with halconf
[14:46:14] <Tom_itx> net in-home-z debounce.0.3.in <= hm2_[HOSTMOT2](BOARD).0.gpio.043.innet out-home-z debounce.0.3.out => axis.2.home-sw-in axis.2.pos-lim-sw-in
[14:46:25] <gene__> debounce.0.0.in and debounce.0.0.out. Consistency, would be nice if there was some... :)
[14:46:35] <Tom_itx> ^^
[14:47:05] <Tom_itx> except the lines ran together
[14:47:11] <Tom_itx> innet should be split
[14:47:17] <Tom_itx> in net
[14:51:05] <tjtr33> putting \1 thru \7 in vcp labels is interesting, but more digits dont seem to work, even \8 \9 are jusy literal
[14:52:07] <JT-Shop> dang internet is lumpy when it rains
[14:58:30] * Tom_itx spreads it smooth with a butter knife
[15:11:41] <gene__> and just had another crash quitting linuxcnc
[15:16:18] <gene__> It gets all cleaned up, and crashes about 3 seconds later. Flakey damned machine!
[15:18:39] <tjtr33> /xxx escape format... like /085 should print the mu symbol. maybe my font aint got one :p
[15:20:52] <tjtr33> er \xxx
[15:23:25] <gene__> this time it was a little faster, 200 millisecs after clickung quit
[15:59:32] <gene__> And I think I have the probe whipped, it trips off when moving left, and ignores the noise when moving right. Except now its disabled for x motion. Grrr
[16:06:11] <gene__> Sigh, gotta think up a way to bypass this stuff when x is moving, either way. And it HAS to be bidirectional for x as it could be measureing a tool facing either way.
[16:15:10] <MattyMatt> yay I got me a dinky knee mill base ->
http://www.ebay.co.uk/itm/milling-machine-base-/261151455705
[16:16:02] <pjm> that is a nice purchase!
[16:16:44] <MattyMatt> gotta get it up stairs tho. I hope the table winds up and off
[16:16:49] <Tom_itx> not the same one you were looking at
[16:17:06] <MattyMatt> nope that went for more
[16:17:11] <MattyMatt> and was too far away
[16:17:31] <MattyMatt> but was a far more sensible size for upstairs :)
[16:17:52] <bpuk> looks like a nice winning bid too - should strip down into a few fairly manageable parts without too much trouble
[16:20:33] <MattyMatt> 2 I hope
[16:21:12] <MattyMatt> I'll take planks and ropes. the base looks about 100lbs of iron
[16:21:54] <MattyMatt> and another 60 for the works, which I can lift by hand hopefully
[16:22:25] <MattyMatt> hernia fun
[16:23:00] <MattyMatt> my usual buddy has just had a transplant op, so he's out. all my other friends are weaklings
[16:23:04] <DJ9DJ> gn8
[16:25:42] <MattyMatt> knee Z won't be easy to motorise. I cant see an easy way to counterweight it
[16:25:56] <andypugh> MattyMatt: Add a chinese spindle and you could have a decent machine
[16:26:15] <MattyMatt> yeps that's exactly what I was thinking
[16:26:37] <MattyMatt> I'll improvice with one of my router spindles for now tho
[16:26:52] <bpuk> could skip the Z counterweight and stick a larger motor on there - depends what sort of power you're willing to use
[16:27:19] <MattyMatt> 240V 13A is my limit, so 3kW
[16:27:39] <bpuk> or add a separate Z slide for the spindle and leave the 'knee' as a manual adjustment
[16:28:05] <bpuk> the latter probably being a much better option in that case
[16:28:05] <andypugh> I got a new motor today. It looks ideal for a lathe spindle:
https://picasaweb.google.com/lh/photo/YWXJhBzhPrr4BMY-hFvHwNMTjNZETYmyPJy0liipFm0?feat=directlink
[16:28:10] <MattyMatt> yeah that's more sensible, but that'd leave me wishing I'd just got an X-Y table instead
[16:28:42] <andypugh> My mill has a motorised knee and fixed head.
[16:28:54] <bpuk> andypugh: that's... different. BLDC?
[16:29:15] <andypugh> Deckel CNC machines move the table for Z. (but they slide the head for Y, so are just strange)
[16:29:16] <bpuk> manual knee and motorized quill here
[16:29:35] <andypugh> bpuk: Yes, it's a bldc from a direct-drive washing machine.
[16:29:37] <bpuk> that sounds more like a grinder design
[16:29:58] <andypugh> 27 quid for the parts, £20 postage. I was drunk and it was on eBay
[16:30:04] <MattyMatt> I thought those washing machines were only found in US
[16:30:30] <MattyMatt> it's 12 pole isn't it?
[16:30:32] <bpuk> heh. I came home one day to find a black london taxi in the drive. Turned out my dad had got drunk and gone on ebay. Always a mistake
[16:31:06] <andypugh> I don't think I will be able to resist seeing how it works on my current lathe, even though it is intended for a decent lathe.
[16:32:04] <bpuk> must have been a pretty beefy washing machine
[16:32:31] <andypugh> I wonder what voltage it uses? Almost certainly rectified-mains I expect.
[16:33:41] <MattyMatt> http://hackedgadgets.com/2008/12/16/cnc-machine-built-using-washing-machine-motors/
[16:36:11] <andypugh> It sounds horribly like he is using them as stepper motors.
[16:38:19] <MattyMatt> is there a diff between VFD and a microstepping stepper driver?
[16:38:43] <bpuk> looks that way. Funny how some people see PM motors as low pole count steppers, and others see steppers as really high pole count PMSM
[16:38:46] <MattyMatt> ah his is not microstepping :)
[16:39:41] <MattyMatt> 42 coils in 6 poles. I wonder if that will work on a 3 pole VFD
[16:41:37] <MattyMatt> presumably 7 or 14 times slower than a straight 3 phase motor
[16:42:20] <andypugh> You could run it on a VFD, but that wouldn't be ideal.
[16:43:01] <andypugh> The phase angle will tend to be very small with a VFD, and the current too high.
[16:44:54] <pcw_home> yes a vfd would mainly pull inwards, instead or tangentialy
[16:47:46] <MattyMatt> it looks like it' be nice in a bicycle wheel
[16:49:27] <MattyMatt> so, custom 6 pole stepper driver is the only option?
[16:49:45] <pcw_home> no, BLDC servo
[16:51:47] <tjtr33> \x7f is end of the line for escape codes in vcp, cant set that hi bit
[16:58:55] <bpuk> hmm. I should have both a base_period and a servo_period in my hal files? (5i20 based system, feeding steppers, encoder on spindle and a handwheel)
[16:59:59] <pcw_home> No base period needed
[17:01:06] <bpuk> oh good. New computer, copied the old configs over - jitter has gone from ~30k to ~10k. Checking through the hal and ini files to sanity check them before reconnecting it
[17:01:36] <bpuk> servo perios is 1000000 - so 1ms there. Which sounds about right.
[17:03:15] <tjtr33> the rest of these escape chars may be useful to gui designers
http://imagebin.org/242661
[17:10:27] <bpuk> I have max_acceleration for each axis set at 200 mm/sec/sec - from memory I dropped these down after having occasional stalls (which may have been PC/jitter related) - do they sound typical for a stepper system?
[17:13:29] <MattyMatt> those chinese spindles all seem to be either ER11 or ER20, but I have a full ER16 collet set already :p
[17:14:43] <Tom_itx> better trade it in
[17:15:30] <bpuk> http://www.ebay.co.uk/itm/FOUR-BEARING-1-5KW-WATER-COOLED-SPINDLE-MOTOR-ER16-VFD-DRIVE-INVERTER-AT1-/260963656443?pt=LH_DefaultDomain_0&hash=item3cc2a562fb
[17:15:53] <bpuk> gah, I'll look again with US orders removed
[17:16:43] <bpuk> or you can search ebay for 'er16 spindle' about 20 results here
[17:16:48] <tandoori> hi. i assume some of you lot have experience in cnc mills. I am trying to purcase my first. This is what I found:
http://www.aliexpress.com/item/Low-price-high-quality-3020-4-axis-mini-pcb-cnc-milling-machine/612311017.html
[17:18:13] <bpuk> MattyMatt: Also, ArcEuroTrade's 0.75kW spindle uses ER16
[17:21:59] <bpuk> tandoori: What do you want to make?
[17:22:03] <MattyMatt> tandoori that's a nice looking complete set, even the 4th axis with chuck which is nothing to do with pcb milling :)
[17:22:29] <tandoori> MattyMatt: I have other ideas to use that thing ;-)
[17:22:42] <tandoori> bpuk: mainly pcb milling but other stuff too
[17:23:31] <tandoori> what size traces can i expect to be able to make with something like that?
[17:26:29] <bpuk> On a stepper based system like that you can probably expect to have a maximum positioning accuracy of around 0.1mm - you should be able to cut finer than that, but it won't be repeatable from one board to another - which might not matter so much.
[17:27:30] <bpuk> endmills are available from 0.1mm upwards - which tells you your minimum creepage. I'd assume a minimum trace of 0.15 mm width as a fairly safe bet. Ideally slightly larger, but for a one-off you may be able to push things a bit
[17:28:09] <tandoori> so all the boards will be slightly differnet, but the eye may not be able notice?
[17:28:44] <tandoori> i dont care if they are differnet so long as they have the same operation
[17:29:36] <bpuk> All the boards may be slightly different unless the raw stock is placed in exactly the same place - but we're talking about a minute difference - unless you are attempting to mill very fine traces or do pick and place it's irrelevant
[17:30:21] <andypugh> MattyMatt IRAMS module is ideal. and a uP to convert hall sensor inputs to phase excitations
[17:31:03] <tandoori> mmm. that cnc mill will do just fine. I also have plans to try and mill stone applications ;-) someone told me i would need a water cooled addition
[17:31:34] <bpuk> depends on the stone - I cut soft stones with nothing more than a spray bottle full of water
[17:32:11] <tandoori> im talking about beryl and corundum
[17:33:50] <andypugh> Aren't they what you use for cutting stone?
[17:34:00] <bpuk> That's... pretty darn hard. Diamond burrs and lots of water are likely to be your best chance
[17:34:07] <tandoori> aye
[17:42:24] <bpuk> andypugh: emerald, ruby and sapphire - the latter two are occasionally used as an abrasive, but yeah - hard.
[17:43:56] <Valen> I have milled glass with diamond burrs if thats any help
[17:44:38] <tandoori> thats a whole lot of help
[17:44:41] <Valen> big thing to watch for is the stainless steel ones seem to work harden and snap after a fairly short while
[17:44:48] <Valen> what are you trying to do?
[17:46:22] <tandoori> i plan on ordering clourless corundum hemisphere for an....experiment
[17:47:22] <tandoori> Valen: if i am successful, I will show you the result of my toil
[17:47:34] <Valen> I wanted to make a solid saphire ring for the missus, figured it'd be too fragile
[17:47:50] <Valen> got any more details, size, how much you want to remove, a source perhaps?
[17:50:17] <tandoori> oh my project? well i hope to order flat sliced sheets of beryl and corundum and cut them like any other material
[17:50:34] <andypugh> This was todays work:
https://picasaweb.google.com/lh/photo/BE1WrWEhMirG0ADLY4Lc7NMTjNZETYmyPJy0liipFm0?feat=directlink
[17:51:08] <andypugh> Just needs the arc-shaped holes for locking the adjustment now.
[17:51:09] <Valen> The only places i could find saphire from wanted to charge me 2x the cost of the saphire for shipping
[17:51:25] <tandoori> for a small thing like that? thats crazy
[17:51:32] <tandoori> unless you are ording a brick of it
[17:51:39] <andypugh> Maybe it was a 20kg sapphire?
[17:51:51] <tandoori> haha maybe
[17:52:14] <bpuk> andypugh: Not going to try for a disc brake on there? :P
[17:52:24] <Valen> it was 25mmx8mm i think it cost like 30 euro
[17:52:42] <andypugh> bpuk: It's mainly intended for gear hobbing.
[17:52:51] <Valen> shipping was only by super express premium courier for 100 euro
[17:52:57] <Valen> (to australia)
[17:53:18] <tjtr33> andypugh, very nice. like the locking connectors. could you put something we know near for scale? i guess its like 350/400 mm high?
[17:53:43] <andypugh> I got a bit creative with the way that the motor couples to the shaft, in a quest for ultimate shortness:
https://picasaweb.google.com/lh/photo/lYuP0nddcE8oukjhT66EdtMTjNZETYmyPJy0liipFm0?feat=directlink
[17:53:47] <bpuk> andypugh: I had guessed that - looks like a nice unit - encoder on the motor only (nothing on the output shaft) I guess?
[17:53:54] <Valen> so how much and where were you getting your chunks from tandoori?
[17:54:09] <andypugh> 100mm centre height
[17:54:18] <Valen> and what sort of total lenght of cut were you thinking?
[17:54:41] <bpuk> heh, hex spline, nifty
[17:54:49] <tandoori> Valen: im bugging these people for my project
[17:54:55] <tandoori> http://www.geolite.com/default.htm
[17:55:00] <tandoori> oops, forgot to paste the link
[17:55:05] <Valen> andypugh: I hate your surface finishes ;-P
[17:55:33] <Valen> they make anything I do look like crap lol
[17:55:35] <andypugh> bpuk: the hex socket was bored on my lathe. Which _is_ nifty.
[17:56:17] <bpuk> yeah, I saw a video yonks back of you doing eccentric turning. At the time you had to turn the spindle by hand I think. I assume you've got past that now
[17:56:29] <andypugh> Valen: It's mainly a matter of adjusting the spindle speed over-ride until the finish looks good. And using the right inserts.
[17:56:50] <bpuk> also, yowza. You need a plunge EDM :P (not that the lathe isn't nifty, they're just... handy)
[17:56:59] <andypugh> No, I still have to turn the spindle by hand, but that washing machine motor might fix that :-)
[17:57:54] <bpuk> did you ever update the wiki page on how to do it? From memory there was *almost*, but not quite enough info to replicate it
[17:58:12] <bpuk> also, this new computer is crazy quiet
[18:00:17] <bpuk> oh - I tried to get a quote for a photo-etched encoder disk - but they decided the line count was slightly too high at 294 LPI. Which is a shame, as that would have allowed me to use the 8500 series 3 channel encoder
[18:02:58] <skunkworks> andypugh: very cool!
[18:04:53] <andypugh> I will update the Wiki. For hex-faceting I wrote a different, new, component.,
[18:05:27] <andypugh> It now knows the working diameter, so makes a correct hex of arbitrary size, internal or external.
[18:05:51] <bpuk> that's useful
[18:14:57] <bpuk> so, apart from the washing machine motor, what's the next crazy plan?
[18:15:51] <andypugh> Maybe atually starting on the clock which is why I got the first lathe?
[18:16:01] <andypugh> Three years ago.
[18:16:43] <bpuk> heh :D
[20:24:47] <MattyMatt> tandoori -> giant sapphire sheets
http://www.opticalwindows.co.uk/catalogue//SAPPHIRE_WINDOWS_AND_PLATES.aspx
[20:28:09] <Valen> andypugh: what clock you makin?
[20:28:20] <Valen> also MattyMatt thats where i think i was going to get my saphire from
[20:28:23] <andypugh> mechanical/digital
[20:30:42] <MattyMatt> 100mm max face size if you want a sapphire watchglass :)
[20:31:34] <MattyMatt> that's overkill for a clock on the mantelpiece tho, unless you party hard
[20:37:38] <tandoori> MattyMatt: man you are awesome
[21:14:35] <andypugh> OK, time to go to the airport. Back in a week. Have fun
[21:34:59] <ReadError> sup tjb1
[21:44:45] <tjb1> xbox smart glass is awesome
[21:44:51] <tjb1> for android
[21:45:26] <ReadError> i need to get a nice shopvac
[21:45:33] <ReadError> one that oil and crap wont destroy the filter
[22:12:12] <r00t4rd3d> lol
[22:12:21] <r00t4rd3d> i make my filters out of old t shirts
[22:13:08] <r00t4rd3d> i think i got an old pillow case on there now
[22:13:33] <ReadError> oh nice
[22:13:37] <ReadError> hey so
[22:13:51] <ReadError> if i get my backlash compensation under no load
[22:14:03] <ReadError> how do i determine what it would be under a cutting load
[22:14:31] <r00t4rd3d> i never did backlash compensation
[22:15:36] <r00t4rd3d> and my machine is money accurate
[22:16:37] <r00t4rd3d> at least cutting wood
[22:17:16] <r00t4rd3d> Im redoing my zaxis now
[22:21:19] <ReadError> http://dl.dropbox.com/u/25091878/Photo%20Jan%2012%2C%2011%2005%2017%20PM.jpg
[22:21:31] <ReadError> my tripod vacuum system is working ;)
[22:21:45] <ReadError> should i save this stuff to cast later
[22:21:49] <ReadError> or is it too impure
[22:22:16] <r00t4rd3d> i would save it and then scrap it
[22:22:23] <r00t4rd3d> fill a garbage can
[22:22:27] <ReadError> lol
[22:22:37] <ReadError> well i got a bunch of other aluminum too
[22:22:43] <ReadError> i figured it would be a fun project
[22:30:46] <tjtr33> i wrote custom M110 to do this 'halcmd sets onval $1'. 'onval' happens to be an S32.
[22:30:53] <tjtr33> When I exec M110 P123, I get "<commandline>:0: value '123.000000' invalid for S32"
[22:30:54] <tjtr33> Is the some bash magic to make halcmd receive an integer value?
[22:46:17] <kwallace> tjtr33: Just a shot in the dark here, but could it help to run the bash float through CONV_S32_FLOAT before connecting it to onval?
http://www.linuxcnc.org/docs/2.5/html/man/man9/conv_s32_float.9.html
[22:47:11] <kwallace> Oops
http://www.linuxcnc.org/docs/2.5/html/man/man9/conv_float_s32.9.html
[22:49:27] <tjtr33> kwallace, i just found out my scripts with #!/bin.sh didja know that thes ubuntoids made 'dash' the shell? still shocked ! yes i can add conv comps, thanks
[22:50:07] <dgarr> tjtr33: f=123.456; echo ${f%%.*}
[22:52:14] <tjtr33> and if it was passed to the script as $1 , then f=$1; halcmd sets onval 'echo ${f%%.*}' ?
[22:52:22] <tjtr33> btw thx for the idea
[22:58:47] <tjtr33> dgarr, you absolutely right :) wheni finally figgered it out & returned here , you gave me the answer
[22:58:52] <tjtr33> halcmd sets onval ${f%%.*}
[22:59:39] <dgarr> or even simpler halcmd sets onval ${1%%.*} (if $1 is the right one)
[23:00:07] <tjtr33> thx! ( ya know the web is FULL of 'you cant do that' on this subject ) , oh! even cooler
[23:48:41] <tjtr33> dgarr,
http://videobin.org/+6ag/724.html its .ogg encoded rt clk to view fullscreen
[23:50:05] <JesusAlos> hi. I'm looking for a tutorial to install Heekscnc in ubuntu
[23:58:04] <tjtr33> JesusAlos, did you see
https://github.com/Heeks ? I think the latest versions of Heeks are for Windows . I did build it on 10.04, including the CNC extensions.
