#linuxcnc | Logs for 2012-03-12

Back
[00:16:26] <ssi> too quiet in here tonight!
[00:21:08] <frysteev> yes
[00:25:34] <ssi> I have everything working on the HNC so far (that I can test) except the spindle brake release
[00:25:47] <ssi> I'm pretty sure that's a dead SSR
[00:26:04] <ssi> the SSR for the brake is 5CR
[00:26:07] <ssi> but the one in that spot is marked 9CR
[00:26:20] <ssi> where 9CR should be, there's one marked "BAD (WAS 5CR)"
[00:26:33] <ssi> where 6CR should be, there's another one marked "BAD (WAS 5CR)"
[00:26:45] <ssi> 6 and 9 are the cutoff slide and part chute relays, neither of which this machine has
[00:27:00] <ssi> I tested the one that was in the 5CR slot, and it seems bad
[00:27:05] <ssi> so for some reason, this tihng is eating SSRs
[00:27:28] <ssi> just ordered two more of them, but I'd like to figure out why they're dying
[00:40:26] <flyback> back-emf?
[00:40:40] <ssi> dunno
[00:40:43] <ssi> it's a mechanical brake I think
[00:41:58] <flyback> well what does the ssr drive then
[00:42:14] <ssi> the brake is energize-to-release
[00:42:15] <ssi> 110v
[00:42:24] <ssi> SSR takes a control signal and energizes the brake
[00:45:20] <flyback> mabye it's getting inductive kickback
[00:45:33] <flyback> and i's exceeding what the ssr can take
[00:46:04] <ssi> maybe it's getting blasted by rayguns from aliens
[00:47:20] <flyback> do you even know what I am talking about
[00:47:30] <ssi> yes I do
[00:47:38] <ssi> but like I said, I'm pretty sure it's a mechanical brake
[00:47:48] <flyback> well the ssr still drives a coil
[00:47:48] <ssi> and it's a commercial, well designed system
[00:48:28] <flyback> is the coil dc
[00:48:48] <ssi> I'm pretty sure the ssr is switching 110v ac
[01:15:44] <pingufan> Good morning. I want to update my linux-cnc version. Currently I use v2.3.5. The new system will be installed to a new harddisk. Which files can/shall I backup from the old system? I had issues with the display resolution, also the hardware setup of motor control was a bit tricky as far as I remember back.
[01:26:07] <Thetawaves> should i get this http://www.ebay.com/itm/CNC-Spindle-300W-Motor-ER11-kit-with-360w-power-Mounting-bracket-/200716868642?pt=BI_Tool_Work_Holding&hash=item2ebba85c22
[01:26:21] <Thetawaves> or this http://www.ebay.com/itm/CNC-Spindle-300W-Motor-ER11-kit-with-PWM-Speed-governor-Mounting-bracket-/200720206451?pt=BI_Tool_Work_Holding&hash=item2ebbdb4a73
[01:28:29] <psha[work]> pingufan: emc configs
[01:29:28] <pingufan> Hi! Both look nice. But their shafts abre both a bit thin, you cannot really expect to do harder work with them.
[01:31:06] <Thetawaves> i am looking to upgrade from my 200w spindles
[01:31:27] <Thetawaves> -s
[01:31:33] <pingufan> What are you milling?
[01:31:38] <Thetawaves> aluminum
[01:32:10] <Thetawaves> i am worried that the first one is lies
[01:32:16] <Thetawaves> 48v * 4a = 192w
[01:32:23] <Thetawaves> not 300 as advertised
[01:32:24] <pingufan> I would prefer then a spindle with a really thick shaft. This prevents vibrations
[01:33:34] <Thetawaves> i think the bigger second one too
[01:33:39] <Thetawaves> too bad it's refurbished
[01:34:43] <pingufan> I bought PC power supplies with 650 W some weeks ago. They usually should be sufficient for a gaming PC when I sum up all power consumption. Though, they overheated. So I asked the manufacturer, and - suprprise - this is a peak value only for 1-2 minutes.
[01:35:07] <pingufan> So don't trust.
[01:35:33] <pingufan> I would not buy a spindle with a shaft diameter below 8mm.
[01:35:46] <SadMan> who was the manufacturer?
[01:36:42] <pingufan> I guess, it is TAMYA rc cars. ;)
[01:38:03] <Thetawaves> i have an ER11 collet, can i get a larger bore that will fit in the same nut/spindle bore?
[01:40:11] <pingufan> Which is the current version of linux-cnc ?
[01:40:34] <pingufan> Is it 2.4.7 ?
[01:41:20] <Thetawaves> http://littlemachineshop.com/products/product_view.php?ProductID=4257
[01:44:46] <pingufan> psha[work]: New emc understands directly the old configs?
[01:46:47] <psha[work]> pingufan: you should read page on wiki about upgrading
[01:46:59] <psha[work]> there are several common problems that you'll encounter
[01:47:18] <psha[work]> for example with emc.nml file
[01:47:34] <pingufan> I will do a fresh install after backing up relevant files from the old system.
[01:47:34] <psha[work]> but mostly old configs will work
[01:49:05] <pingufan> Does ubuntu 10.x support meanwhile power-off? My old Ubuntu finishes runlevel 0 and then remains powered on. I definitely know that ACPI works.
[01:54:40] <psha[work]> acpi/apm works and current board is fully supported is not same :)
[01:54:53] <psha[work]> if it's not powered off then it does not work )
[01:55:42] <pingufan> It definitely works. But not with Ubuntu (in this "special emc version)
[01:56:00] <pingufan> I use an old 2.3.5 until now.
[01:57:51] <pingufan> This is one of the few reasons to switch to a newer system, in the hope that this will help.
[01:58:31] <Jymmm> cradek: http://oi44.tinypic.com/15yenaq.jpg
[02:03:07] <Thetawaves> what cam software do you guys use
[02:20:02] <Thetawaves> <secret>
[02:52:29] <DJ9DJ> moin
[03:13:03] <pingufan> Moin
[03:26:33] <pingufan> Does this new Ubuntu 10.04 run on a VIA C7 CPU? It seems to me that there are problems.
[03:28:40] <pingufan> During installation, it does not get to step 3 of 7
[03:29:06] <Jymmm> Via? as in a Pico board?
[03:45:28] <Loetmichel> moin
[03:45:38] <Loetmichel> s/moin/nornin'
[03:45:43] <Loetmichel> mornin'
[03:45:57] <Loetmichel> <- untangles his fingers ;-)
[03:59:39] <pingufan> Jy
[04:00:41] <pingufan> Jymmm: Mini-ITX Board. Meanwhile I found out the reason. Ubuntu does lots of nasty things with USM DVD-ROM drives on tis machine. I connected now a IDE drive and it works.
[04:05:23] <pingufan> I meant USB drives
[04:05:47] <Jymmm> Verify BIOS settings in regards to Legacy USB support.
[04:06:37] <Jymmm> Usually you MUST have a USB optical drive connected PRIOR to power on, not plug in afterwards.
[04:22:10] <pingufan> I know. It starts to work, but somewhere within the whole installation process, it seems that the communication to the drive gets lost. Anyway, Through good old IDE cable it wrks.
[04:22:50] <pingufan> And after installation I will not need the drive any more at all.
[04:23:09] <pingufan> Meanwhile 61% are installed.
[04:24:03] <pingufan> Can I switch Ubuntu then from UTF-8 to ISO-8859-1 and also change my UID? This makes NFS working better.
[04:24:45] <pingufan> iso-8859-1 is what I use on all other machines in my network.
[06:08:04] <pingufan> was offline for some time (ISP trouble)
[06:10:57] <pingufan> This new Ubuntu appears to be noticable slower. How much RAM does it need? I currently have 1GB in this machine.
[06:59:50] <pingufan> Please, can somebody help me? I downloaded the latest ISO for EMC2, installed it, and now I see that I cannot run any CNC related software. When I start latency-test, it tells me that ther eis no rtapi kernel module installed.
[07:01:02] <pingufan> But kernel _is_ 2.6.32-122-rtai
[07:03:04] <jthornton> did you somehow not do a clean install?
[07:03:58] <pingufan> It is not possible to install "cleaner". This is a new harddisk.
[07:04:03] <pingufan> :)
[07:04:26] <awallin> and you are booting the rt-kernel from grub?
[07:04:49] <pingufan> It boots automatically whatever it installed.
[07:05:00] <pingufan> Kernerl is a -rtapi one
[07:05:12] <pingufan> -rtai I mean
[07:06:04] <pingufan> It wanted to do an online update. I didn't allow it. Was this ok?
[07:07:00] <pingufan> I must say that I am no real Ubuntu man. I only use SuSE. So i have no experience with Ubuntu.
[07:07:14] <jthornton> you can update everything but Ubuntu iirc
[07:07:25] <pingufan> Shall I install again from scratch?
[07:07:35] <jthornton> when you boot from the live cd can you run the latency test?
[07:08:03] <pingufan> I don't know. I pressed ESC on the boot CD and chose then "install"
[07:08:39] <jthornton> did you check the md5sum of the downloaded livecd?
[07:09:10] <pingufan> no. But it could read without any troubles. I never had issues with wrong md5sums in the past.
[07:09:26] <jthornton> just tossing out ideas...
[07:09:39] <pingufan> Sure. I agree.
[07:10:08] <pingufan> Do you install like I did, or from the booted live-ISO
[07:10:23] <pingufan> (Then I try it in the same way)
[07:10:52] <jthornton> I've installed it many ways
[07:11:32] <pingufan> You perhaps can tell me a few things about config:
[07:12:05] <pingufan> How do I get ssh-server running, and how do I change my UID from 1000 to 501 ?
[07:12:29] <pingufan> Then I will do another fresh install.
[07:12:32] <jthornton> that is above my pay scale :)
[07:23:10] <pingufan> I boot now the live system from CD and try it as you suggested. Hope that 512MB RAM is sufficient.
[07:25:18] <jthornton> might be a little on the light side, check the Ubuntu web site for the minimum RAM suggested
[07:29:56] <pingufan> Ok, now it starts to become weired... Even the original ISO with live-system booted from it, gives me the same error.
[07:30:07] <jthornton> fun fact... 96% of the spammers prefer to use gmail
[07:32:39] <pingufan> Please, can someone help? I would like to use the latest version to be on safe side for neht time, but this exceeds my knowledge.
[07:33:41] <archivist> pingufan, ask better questions
[07:33:44] <jthornton> is this a new computer or one that is older?
[07:33:45] <skunkworks__> pingufan: is it older hardware?
[07:33:49] <skunkworks__> heh
[07:33:54] <jthornton> heh
[07:34:22] <pingufan> Via EPIA-M Mainboard with VIA CPU and 512MB RAM
[07:34:41] <jthornton> how old is it?
[07:34:56] <skunkworks__> could you be getting http://wiki.linuxcnc.org/cgi-bin/wiki.pl?TroubleShooting#emc2_doesn_t_run_missing_lapic
[07:34:58] <skunkworks__> ?
[07:35:03] <pingufan> From technology, it is surely 2-3 years.
[07:35:24] <jthornton> that should not be the problem then
[07:36:15] <pingufan> skunkworks__: I get "Point 3" when try to run latency-test from a terminal
[07:37:16] <pingufan> First, I will now re-install this 10.04 system as it tried to do updates. To get a defined starting point.
[07:56:37] <pingufan> This is a VIA C3 CPU (Nehemiah) with 1GHz.
[07:57:35] <pingufan> My old EMC2 2.3.5 worked fine on this hardware. Should then the new one also work?
[08:03:05] <JT-Shop> old install was running what version of Ubuntu?
[08:03:23] <cradek> I agree that is the important question
[08:03:41] <cradek> but it was too early for me to type it myself
[08:04:08] <JT-Shop> yea, an hour too early unless your in Arizona
[08:04:25] <Jymmm> heh
[08:04:41] <pingufan> The old one - I guess(!) it was 8.x, but I am not sure.
[08:05:09] <pingufan> Hello, cradek! How are you?
[08:05:41] <cradek> it's my birthday
[08:05:52] <Jymmm> pingufan: It's too early to ask cradek how he is, wait till the caffeine has kicked in.
[08:06:06] <pingufan> Really? Happy Birthday then! :D
[08:06:17] <Jymmm> cradek: H A P P Y B I R T H D A Y !!!
[08:06:18] <Jymmm> cradek: H A P P Y B I R T H D A Y !!!
[08:06:20] <Jymmm> cradek: H A P P Y B I R T H D A Y !!!
[08:07:08] <pingufan> Does Ubuntu 10.04 enforce lapic, or is it only better to have it?
[08:07:32] <cradek> the rtai kernel in the linuxcnc version of 10.04 requires lapic
[08:07:49] <pingufan> What this also the case in Ubuntu 8 ?
[08:08:05] <cradek> no
[08:08:25] <pingufan> Can I install EMC 2.4.7 with Hardy?
[08:08:26] <cradek> unfortunately the tradeoff is to support SMP you must require lapic
[08:08:38] <cradek> we changed our minds about which was more important for the ubuntu 10 version
[08:08:56] <cradek> yes you can update hardy to emc 2.4.x, see instructions on the first page of the wiki
[08:09:13] <pingufan> There is no ISO I can download and install?
[08:09:34] <cradek> no, see the instructions please
[08:11:58] <pingufan> So i give it a final try with 10.04 and see if I get it working. If not, the bloody thing will start ... You don't know if CLE266 chipset supports lapic?
[08:13:31] <cradek> no idea, but you should be aware that for some motherboards there's a certain problem you can fix: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?TroubleShooting#emc2_doesn_t_run_missing_lapic
[08:14:15] <Jymmm> alex_joni: http://i41.tinypic.com/2hx3ry8.jpg
[08:15:31] <pingufan> I did read that. Hope it supports lapic at all. If not, I tend more and more to insert the previous harddisk and stay at 2.3.5. Too much action.
[08:25:05] <pingufan> Yes, it is exactly as described... No local apic found. Now I edit grub,
[08:25:32] <Jymmm> make a backup first
[08:27:30] <pingufan> :) Of what?
[08:27:42] <pingufan> This is a totally new installation.
[08:27:46] <Jymmm> of whatever you edit
[08:28:37] <Jymmm> cradek: you awake yet?
[08:33:48] <pingufan> This will become an odyssee. Al this realtime-stuff is not here.
[08:34:29] <pingufan> Even after entering the lapic line .
[08:36:42] <pingufan> Has someone a ISO of Hardy with emc, or can show me a link to one, please?
[08:40:11] <archivist> did you look on http://linuxcnc.org/index.php/download/21/
[08:40:17] <Jymmm> pingufan: http://dsplabs.upt.ro/~juve/emc/get.php?file=ubuntu-8.04-desktop-emc2-aj13-i386.iso
[08:40:37] <Jymmm> pingufan: MD5SUM:21f4ecdcf9f5ab09ed64b5e76fb389e6
[08:40:54] <Jymmm> err
[08:41:06] <pingufan> This is a plain 8.04, or the rtai enabled one?
[08:41:27] <Jymmm> pingufan: http://dsplabs.upt.ro/~juve/emc/get.php?file=ubuntu-8.04-desktop-emc2-aj13-i386.iso MD5:1bab052ec879f941628927c988863f14
[08:41:36] <Jymmm> It's LTS LiveCD
[08:41:42] <Jymmm> 2009-04
[08:42:17] <pingufan> So I have to switch then to a rtai kernel after initial setup?
[08:42:36] <Jymmm> You asked for Hardy, there it is.
[08:42:54] <pingufan> Of course one for EMC.
[08:43:02] <Jymmm> Anything else, you might chec the wiki for details
[08:43:06] <Jymmm> check
[08:45:14] <pingufan> I install it. Hope you can help me with upgrading afterwards emc. :)
[08:46:18] <pingufan> I find it no good idea that there are not two ISOs available.
[08:46:55] <pingufan> (For 10.04 and 8.x, if there are so heavy restrictions)
[08:55:08] <cradek> go to linuxcnc.org, click "how to get it", click "basic installation"
[08:55:57] <cradek> er, I was scrolled back, I thought you had not found the iso yet.
[08:57:31] <Jymmm> cradek: you awake now?
[08:57:37] <cradek> obviously not
[08:57:57] <Jymmm> cradek: Heh, ok. ping me when you are
[09:04:16] <Jymmm> JT-Shop: Found a solution to any trespassing issues you may have... http://i42.tinypic.com/5cnjn.jpg
[09:07:12] <pingufan> cradek: What is your local time?
[09:08:08] <Jymmm> Hammer Time!
[09:08:26] <cradek> almost 0900
[09:08:49] <pingufan> Huh, I habe 3 pm
[09:09:43] <pingufan> Can you tell me how to upgrade then (when i have installed Hardy) to the latest EMC that runs on it?
[09:10:06] <pingufan> Is it only running the script "hardy-install", or does it differ?
[09:10:21] <cradek> see the front page of the wiki to find those instructions
[09:14:58] <Jymmm> cradek: I think there may already be a program that sorta kinda does this, but I'm not sure. Take a stencil, and project a light through it, you have a shadow on the wall. Now, if the light source isn't centered (let's say closer to the bottom) then the projection will be bigger on top. Isn't there some way to be able to have the light at various angles to be able to have the projection as intended?
[09:16:17] <cradek> do you mean you want to distort an image in this way? I'm not sure what you're wanting the program to do
[09:16:50] <Jymmm> cradek: Or a program that would alter the stencil so it's narrower at the top so the projection comes out perfect
[09:17:38] <Jymmm> basically have the program compensate for the lights angles as needed
[09:19:30] <cradek> gimp has a perspective transform that I think would be the equivalent
[09:21:43] <archivist> I use the gimp transform fiddle so I can take flash pictures off axis then make them look like I was on axis
[09:22:42] <Jymmm> cradek: Well, I can "distort" the artwork manually to guestimate the light's angles till I get it close enough, but I thought there was some program that did home-made 3D object laser scanning that did something similar. I want to be able to enter in the distance and angles and have it create the artwork to reproduce the stencils.
[09:23:48] <cradek> perhaps knowing you are looking for a "perspective transform" will help you find the software you want
[09:24:03] <cradek> alternatively, you might check to see if you can script the gimp transform plugin or similar
[09:24:44] <Jymmm> ok, cool. Didn't know what it was really called.
[09:33:54] <JT-Shop> so Jymmm I don't think they won the spelling bee...
[09:37:06] <Jymmm> JT-Shop: That's only because be has a purty mouth.
[09:44:15] <pingufan> How do I make a perfect backup of an UBUNTU 8 machine when I configured everything? Doing a "dd" is no good way as it also writes empty data blocks. Is there a tool which maches a copact recovery DVD ?
[09:44:25] <pingufan> ...makes ...
[09:45:29] <pingufan> As new releases appear to be incompatible with my hardware, this is nore important as I initially thought.
[09:48:42] <pingufan> Hello Loetmichel! Wie geht's?
[09:51:50] <Loetmichel> pingufan: english channel
[09:52:12] <Loetmichel> but : good, even have time to look in the IRC @ work ;-)
[09:52:55] <pingufan> I try to get a newer AXIS running.
[09:55:47] <JT-Shop> what do you mean by "perfect backup"?
[09:58:54] <pingufan> compact enough to fit on a DVD, but containing everything (including bootloader) and possibly bootable, to restore a backed up system.
[09:59:33] <JT-Shop> google "how to create a live cd with ubuntu"
[10:01:18] <pingufan> Ok. I'll do so.
[10:32:02] <pingufan> How do i enable ssh-server on ubuntu 8, please?
[10:51:31] <ssi> any of you hardinge folks have any idea why the spindle brake might eat SSRs?
[10:52:14] <Loetmichel> ssi: forgotten to fit a RC filter?
[10:52:29] <ssi> Loetmichel: I haven't forgotten to do anything, it's as received :)
[10:52:37] <archivist> I dont have a hardinge but is it a solenoid
[10:52:46] <ssi> I'm not sure what it is to tell you the truth
[10:53:00] <ssi> all I know is there's a dc controlled 110VAC SSR that runs it
[10:53:03] <ssi> and the SSR is dead
[10:53:13] <ssi> and there are two more in the cabinet marked "DEAD, USED TO BE 5CR" (which is the brake SSR)
[10:53:29] <archivist> 110v I remember you mentioning 220 the other day
[10:53:46] <cradek> pingufan: http://www.google.com/search?q=enable+ssh-server+on+ubuntu+8
[10:53:54] <cncbasher> ssi> the ssr's are a pain especialy the brake
[10:54:04] <ssi> cncbasher: can it be improved?
[10:54:14] <cncbasher> i disconnected mine
[10:54:27] <cradek> you could replace it with a contactor
[10:54:34] <archivist> or a better ssr
[10:54:37] <cncbasher> the back emf blows them
[10:55:14] <Loetmichel> thats why i suggested a RC filter parallel to the Brake
[10:55:14] <archivist> that means you are using one of too low a voltage
[10:55:18] <pingufan> cradek: It claims that it cannot find it.
[10:55:40] <cradek> what claims that what cannot find what?
[10:55:43] <pingufan> It also cannot install joe, when I enter "apt-get install joe"
[10:55:50] <pingufan> ... as root.
[10:55:55] <ssi> hrm
[10:56:12] <ssi> I ordered two replacements, one that's the same as what I have, another that's made by opto22 but spec'd the same
[10:56:26] <pcw_> is there a flyback diode across the brake?
[10:56:35] <cradek> pcw_: it's ac
[10:56:55] <ssi> maybe it needs a varistor across it
[10:57:13] <ssi> I need to check the drawing and see if there's supposed to be there
[10:57:23] <pcw_> probably need a bigger SSR (one rated for inductive loads)
[10:58:19] <ssi> cradek: is your machine an hnc or a chnc?
[10:59:06] <cradek> ssi: mine's HNC
[11:01:20] <pingufan> Why does this ubuntu 8 continuously try to update to 10.04 ?
[11:01:49] <cradek> it gives you the option to do that
[11:02:01] <ssi> so I removed the contactors for the coolant and spindle motors
[11:02:11] <ssi> there's varistors and RC filters across all the phases for both motors
[11:02:19] <ssi> should I remove them when I put the VFDs in, or leave them?
[11:06:18] <pcw_> i would remove the RC snubbers as they will probably do a slow burn if powered by a VFD
[11:06:24] <ssi> ok
[11:06:31] <ssi> leave the varistors?
[11:06:48] <pcw_> RC snubbers were likely for the contactors
[11:06:54] <ssi> yeah probably so
[11:07:25] <JT-Shop> pingufan: because you have not turned that feature off
[11:07:32] <pcw_> varistors should be OK (though not useful)
[11:07:57] <ssi> if not useful, they're easy enough to remove
[11:08:42] <pingufan> Nice. I will screw in the old drive. I have enough troubles with this crap for next weeks.
[11:08:50] <ssi> so what I ended up doing was, rather than try to provide the whole cabinet 3ph, I looked at the diagram and it just takes a single phase off to power everything except the coolant/spindle contactors
[11:09:05] <ssi> so I removed the contactors and am feeding L1/L3 with single phase
[11:09:13] <ssi> now everything comes up and works... everything but the spindle brake
[11:09:39] <cradek> yeah, mine's all single phase now too
[11:09:54] <cradek> I put a run cap on the coolant pump - starts a little slow but it works
[11:10:02] <ssi> I bought a VFD for the coolant pump
[11:10:13] <ssi> GE 1/4hp VFD is $33 at dealerselectric.com :D
[11:11:05] <ssi> I had to futz with the signal wiring in the power cabinet again
[11:11:27] <ssi> there's one relay on that side that seems to want to be energized before anything will run, and it had been connected to the contactors
[11:12:57] <ssi> someone also put in a relay that runs a 110v outlet on the back side
[11:13:11] <ssi> that's no longer hooked up, but I'll shoehorn it in at some point
[11:16:29] <flyback> <ssi> GE 1/4hp VFD is $33 at dealerselectric.com :D
[11:16:31] <flyback> whoa
[11:16:51] <flyback> anyone know of a cheap one with rs232 interface?
[11:17:02] <awallin> 1/4hp is not a whole lot...
[11:17:11] <flyback> it's enough for a blower at work
[11:17:12] <pingufan> JT-Shop: Could you, please, tell me how to get the normal repositories in and how to get the upgrade to AXIS 2.4.x installed?
[11:17:12] <flyback> :P
[11:17:44] <flyback> my boss asked about this a while ago and I couldn't find s good solution
[11:18:59] <JT-Shop> pingufan: like cradek said it is on the wiki... here is the link http://wiki.linuxcnc.org/cgi-bin/wiki.pl?UpdatingTo2.4
[11:19:22] <ssi> hrm so from the drawings it looks like the AC from the brake SSR goes through a bridge rectifier and then has a varistor across it
[11:19:25] <ssi> odd
[11:19:49] <JT-Shop> dc brake maybe
[11:20:33] <JT-Shop> my Samson lathe puts DC on the spindle motor windings for a brake
[11:21:02] <cradek> ssi: can you just put a nice big braking resistor on your vfd and forget that mess?
[11:21:50] <ssi> yeah probably!
[11:21:58] <ssi> where's your sense of adventure? :D
[11:22:19] <ssi> cradek: I'm guessing you removed your brake?
[11:22:45] <cradek> it didn't have one that I know of, unless the old control would engage both gear-change clutches together to brake
[11:23:03] <cradek> I kind of doubt it did that
[11:23:08] <ssi> it's supposed to have a brake on the same shaft as the gear clutches
[11:23:25] <cradek> oh, maybe mine was gone already then!
[11:23:27] <ssi> thing is, it's energize-to-release
[11:23:32] <ssi> so I can't turn the spindle until I get it free
[11:23:44] <cradek> yeah mine sure doesn't have that
[11:23:50] <cradek> and I didn't remove it, someone else must have
[11:23:56] <ssi> gotcha
[11:24:05] <ssi> what about the variable speed mess
[11:24:30] <cradek> it's electrically unhooked and a 2x4 holds it in the right spot
[11:24:37] <ssi> hahahah
[11:25:07] <cradek> with two gears and a decent vfd, I don't miss it
[11:25:16] <ssi> yeah I'm sure
[11:25:24] <ssi> I'm wondering if it's worth screwing with
[11:25:48] <cradek> nah, vfd control is 100x better
[11:26:06] <ssi> what sort of sorcery do you need to make the gear changes happen in hal?
[11:26:25] <cradek> I used two custom M codes for selecting high and low
[11:26:40] <ssi> I see
[11:26:47] <cradek> I did interlock so they could only change once the spindle stops
[11:26:51] <ssi> so it's not as clever as figuring out which gear to use based on requested speed
[11:26:59] <cradek> no, I don't believe in that
[11:27:05] <ssi> why's that?
[11:27:34] <cradek> because on a lathe speeds change throughout a program
[11:27:41] <cradek> imaging a facing cut
[11:27:48] <cradek> you don't want to stop and change gears in the middle!
[11:27:53] <ssi> oh right, that makes sense
[11:27:54] <cradek> so pick the one you want and program it
[11:28:16] <ssi> here's the scheme of the brake, btw
[11:28:16] <ssi> https://p.twimg.com/AnzJGT8CAAAGK2R.jpg:large
[11:28:52] <ssi> not sure what the pots in there are for
[11:34:51] <cradek> bizarre
[11:35:18] <cradek> to keep the cabinet warm?
[11:35:59] <ssi> mebbe!
[11:36:25] <cradek> R1 certainly sets the "on" current
[11:36:32] <cradek> R2 seems to set the "off" current
[11:36:41] <ssi> ah
[11:36:52] <cradek> ("off" current also interacts with R1)
[11:36:53] <ssi> I guess I can test the bridge and replace the varistor
[11:37:05] <ssi> I'll have six spare varistors comin off the contactors :)
[11:37:10] <cradek> I bet you'll find R2 is open
[11:37:20] <ssi> hrm maybe so
[11:37:37] <cradek> (looks like an awful setup to me)
[11:38:17] <jdhnc> I often feel that way looking at other peoples code, until I try to re-implement something.
[11:38:29] <pingufan> The update ran now, but now EMC (2.4.6) aborts with an error. Message size of 10748 exceeds maximum of 10180.
[11:39:53] <cradek> go back to http://wiki.linuxcnc.org/cgi-bin/wiki.pl?UpdatingTo2.4 and read the instruction that's in boldface
[11:41:06] <pingufan> You mean the first line? The following instructions only apply to Ubuntu 8.04 "Hardy Heron". emc 2.4 is not available for older releases of Ubuntu.
[11:41:27] <cradek> PLEASE READ THE WHOLE PAGE
[11:41:33] <cradek> there are exactly two lines in boldface
[11:41:51] <cradek> you want the one that's obviously applicable to your problem
[11:42:16] <JT-Shop> Once you have done the upgrade, update any custom configurations by following these instructions: UPDATING
[11:42:27] <JT-Shop> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?UPDATING
[11:42:48] * JT-Shop is worn out now and retires for a nap...
[11:43:12] <cradek> maybe I should do that too.
[11:43:38] <JT-Shop> I always have a better outlook when I take a nap
[11:44:00] <djdelorie> success! http://www.youtube.com/watch?v=4IhEbEsjm-Q
[11:44:42] <ssi> JT-Shop: I ended up buying some soluble coolant
[11:44:47] <ssi> JT-Shop: I think it's mobilcool 102
[11:45:07] <ssi> mobilcut 102 it is, rather
[11:45:19] <ssi> I bought a gallon of that and a gallon of vactra
[11:50:15] <pcw_> looks like the were trying to tweek around a design problem
[11:50:28] <pcw_> (brake resistors)
[11:50:30] <cradek> yeah
[11:56:40] <pingufan> Thank you, AXIS runs now again. Now I also notice that the area right of the coordinates-display is appearing (at least) black and not showing yellow. Should I see there the workpiece, too, or is this part of the screen entirely reserved from left to right?
[11:57:11] <cradek> not sure - can you take a screenshot of that?
[11:57:45] <pingufan> sure. Where do I paste pictures?
[11:58:18] <cradek> try imagebin.org
[11:59:33] <djdelorie> question: once I get this frimware installed, what motor characteristics will linuxcnc want? I mean, as far as fastest step speed, acceleration, top speed, etc. I figure, I need to have the controller test the motors and report the numbers to type in...
[11:59:54] <ssi> hrm I probably didn't need to buy new SSRs... I think there are SSRs that use low voltage dc to switch 110VAC, and the 110VAC switches the spindle/coolant contactors
[12:00:03] <pingufan> cradek: Here it is: http://imagebin.org/203056
[12:00:16] <ssi> I probably need to run the straight DC control signals to the VFDs
[12:00:33] <cradek> pingufan: that looks correct to me
[12:01:05] <cradek> pingufan: I don't understand what yellow you expect to see?
[12:01:16] <pingufan> I am a bit surprised that I do not see the red dashed lines of moving space in the screenshot.
[12:01:47] <awallin> adjust your monitor ... :)
[12:01:51] <pingufan> Right of the coordinates, the whole area was is a strange yellow background with random pixels inside.
[12:02:19] <cradek> I see the red dashed lines but no yellow weirdness in your screenshot
[12:02:27] <cradek> I wonder if you have a video card problem
[12:02:43] <cradek> or specifically opengl problem
[12:03:02] <DJ9DJ> pingufan, do you see any pink elephants around? ;-)
[12:03:07] <pingufan> But you still see here the black stripe through the "M" of EMC ?
[12:03:26] <cradek> yes there is an incorrect black stripe I didn't see at first
[12:03:38] <pingufan> It is always there, but only in this window.
[12:03:40] <cradek> pingufan: consider http://article.gmane.org/gmane.linux.distributions.emc.devel/2399
[12:04:07] <mrsun> hmm dave at eevblog doesnt seem to know that stepper motors can lose steps if stepper doesnt have the torque it takes
[12:04:17] <mrsun> "belt must be slipping" ... (its a cogged belt)
[12:05:28] <pingufan> Well, I can live with that. It is still much better than before. So the area right of the coordinates is always black and not showing the workpiece?
[12:09:40] <awallin> if you don't see a preview of the g-code then you have a video/opengl problem
[12:10:29] <pingufan> That's only right of the coordinates.
[12:10:40] <pingufan> below the coordinates I see them.
[12:11:51] <awallin> oh, maybe AXIS reserves a big area of the screen for the DRO-numbers. I'm not sure if the numbers are supposed to be "transparent"
[12:12:15] <pingufan> I assumed that, too.
[12:12:52] <awallin> on 2.6.0-pre which I have here the black box is a lot smaller. but not transparent
[12:13:25] <awallin> hm, no. wait.
[12:14:00] <pingufan> The only remaining thing (then I have the mill running again (and with 2.4.6), is: With ubuntu 10.04, the machine turned off when I chose shutdown. With Hardy (8.0x) it only enters "System halted" but does not power off. Can this be fixed somehow?
[12:14:21] <awallin> the background of the numbers is black (non-transparent). just to the right of the numbers there's a region which is slightly transparent (a bit hard to see the toolpath through this). This is on a laptop with integrated intel-video and 2.6.0-pre in simuilator mode
[12:15:45] <awallin> pingufan: shutdown stuff might be a bios-setting (acpi?)
[12:15:57] <pingufan> I can live with this lost area. Power-off would be more helpful.
[12:16:32] <awallin> do you have a very small screen? 800x600 or something? or unusually large/small numbers specified in .axisrc?
[12:16:37] <pingufan> Sure. But why does it work with every Linux i try, except Hardy ?
[12:17:01] <pingufan> Screen is 1280x1024
[12:17:35] <pingufan> Look at the screenshot I posted. above. http://imagebin.org/203056
[12:18:08] <awallin> yeah, something makes AXIS calculate size of the DRO-box to be a bit large..
[12:18:21] <pingufan> :)
[12:18:46] <pingufan> Funny, that the yellow toolpath appears there.
[12:19:03] <pingufan> (When I move around with the cursor keys.
[12:19:33] <awallin> that sounds like the order of opengl calls matters..
[12:19:45] <awallin> if the yellow is drawn last or somethn
[12:20:11] <pingufan> Pherhaps. I am no developer. But that sounds reasonable.
[12:21:21] <pingufan> Can I enforce support of ACPI in the kernel? I remember that one said, it was left away because of some reason.
[12:46:04] <mikegg> durr. if you're not careful removing the encoder end cap from a servo, it costs four weeks and $150.
[12:47:05] <JT-Shop> ssi: I hope your coolant tank is easier to get out and clean than mine was... it had 100lbs of fines in it and some really alien growth in it...
[12:47:44] <pingufan> :D
[12:48:46] <pingufan> What does this "RTAPI: ERROR: Unexpected ... " message mean? I loose no steps, machine works smoothly, nut this message sometimes appears.
[12:49:09] <pingufan> ... but ...
[12:50:15] <pingufan> Can I disable this message in a config?
[12:50:55] <micges> unfortunately no
[12:51:00] <JT-Shop> that means you have a problem and should do an extended latency test to see how bad the problems is
[12:52:11] <pingufan> I did an latency test and then added 1/3 of the value shown in addition. This message still appears.
[12:53:08] <JT-Shop> how long did you let it run and did you also do other things on the computer while it ran like it recommends?
[12:53:43] <pingufan> At least 10 minutes. I played music (streaming audio), surfed the WEb, ...
[12:53:54] <JT-Shop> run it for a few hours
[12:54:22] <pingufan> I currently use the PC almost always through tightvnc, so i can be away from the noisy machine.
[12:54:23] <JT-Shop> some problems only show up at 30 minutes or other time intervals
[12:54:56] <JT-Shop> I have no clue what tightvnc is
[12:55:30] <pingufan> See the whole desktop through the network from another PC.
[12:56:12] <pingufan> "Remote Desktop" in -> System -> Settings
[12:56:23] <JT-Shop> ah, ok
[12:56:45] <pingufan> So this makes in addition a lot of stress.
[12:59:04] <pingufan> You surely can answer this wuestion to me: With AXIS 2.3.5, "G38.2 z0 f20" worked in user cordinate system. When I entered z0, it only moved down to workpiece's z0.
[12:59:25] <pingufan> Is this unchanged, or are now G53 coordinates used?
[13:00:13] <mikebronner> Does anyone have specs on CW230 and CW8060 drivers for Step Time, Step Space, Direction Hold, and Direction Setup in the configuration wizard of EMC2?
[13:00:39] <JT-Shop> have you checked the list on the wiki to see if they are listed there?
[13:00:43] <cradek> pingufan: that is unchanged
[13:02:20] <pingufan> But this is really nasty. I have the tool-sensor mounted on the machine desk. As the workpieces vary in height, it is not easy to tell the right height. It happens repeatedly that AXIS then cries "this exceeds the machine limits".
[13:03:44] <mikegg> make your machine limit bigger?
[13:03:48] <cradek> no
[13:04:15] <cradek> you could probe in an unoffset coordinate system
[13:04:21] <mikebronner> @JT-Shop: not that I have been able to find. We have cross-referenced the wiki multiple times, and scoured google, and got the machine running with settings of 10000, 10000, 20000, 20000 (respectively). But it makes me feel uneasy not knowing if I'm using intended numbers, or something just close enough to get them working.
[13:04:28] <cradek> or use g91
[13:05:48] <frysteev> anyone familair with control techniques products? mainly the fnc5
[13:07:49] <JT-Shop> mikebronner: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing
[13:08:20] <JT-Shop> who makes the CWxxx drivers?
[13:09:21] <cncbasher> JT-Shop: their standard leadshine drives , 10us for pulse
[13:10:07] <JT-Shop> ah ok
[13:10:18] <cncbasher> so 10000,20000 should be fine for them , their quite wide on tolerance settings
[13:10:27] <pingufan> cradek: What does G91 help? The tip sensor is ~20mm abowe the machine desk. Always. I touch it with the first tool, then again for compensationg the tool length later. When the workpiece has a height of 30mm, I can move down 30 mm in absolute machine coords. But I cannot use the same command (Z word) when the workpiece is only 5mm now.
[13:11:45] <JT-Shop> cncbasher: if you have the info on them can you add that to the wiki page?
[13:12:26] <pingufan> Actually I would like to store two user-functions, one for touching the first tool, one for compensating the next tool, in an external file (or in emc directly), so all I have to do is calling one M-function and I am done.
[13:12:26] * JT-Shop heads to town now
[13:12:27] <cradek> you should write the tool length probe in such a way that your workpiece coordinate system doesn't matter
[13:12:53] <cradek> go up all the way (g53 g0 z0)
[13:12:54] <djdelorie> JT-Shop: in that chart, does "x Edge" mean "both edges" or "I don't know" ?
[13:13:06] <cradek> move over the probe switch (g30 or g50 g0 x... y...)
[13:13:26] <cradek> now do a probe move that's less than your full travel (g91 g38.2 z-...)
[13:13:51] <cradek> there are a hundred ways to do it
[13:14:06] <cradek> oops I meant: move over the probe switch (g30 or g53 g0 x... y...)
[13:14:30] <pingufan> And this works?
[13:15:25] <cradek> you try it and tell us
[13:16:02] <mikebronner> @JT-Shop: Thanks for the info :)
[13:16:47] <mikebronner> @cncbasher: thanks for the info. Are there ways to optimize performance, or are these drivers either working or not working?
[13:17:41] <pingufan> G91 G38.2 .... does not revert to G54 coordinate system (which I normally use)?
[13:17:42] <cncbasher> mike they usually work or not as the inputs are optoisolated
[13:18:03] <mikebronner> @cncbasher thanks a lot :) you put my mind at ease
[13:18:20] <cncbasher> just looking for my spec sheet
[13:19:21] <IchGuckLive> cradek: Truetype Tracer 4.0 question on your side
[13:20:02] <IchGuckLive> cradek: isent there a "-o" for generating a file
[13:20:21] <cncbasher> mike: pulse > 1.5us and direction of at least 5us
[13:20:24] <IchGuckLive> cradek: or can i call this direkt out of axis
[13:20:43] <mrsun> hmm, how much off can cheap endmills be? ...
[13:20:45] <IchGuckLive> awALLINMS IS A PYTHON THAT IS CALLABLE
[13:20:58] <mrsun> got a 4mm, but the pieces i pieces i cut becomes 0.1mm oversized
[13:21:32] <IchGuckLive> mrsun: new ones 50% max
[13:21:44] <mrsun> IchGuckLive, 50% max? :P as in a 4mm becomes 2mm ? :P
[13:21:57] <syyl> yeah...
[13:22:02] <mrsun> ;P
[13:22:07] <syyl> i would change my tool supplier ;)
[13:22:10] <IchGuckLive> if the grinder does it why not
[13:22:13] <mrsun> IchGuckLive, nah seriously what do you mean? :)
[13:22:31] <archivist> mrsun, are your spindle and collet concentric
[13:22:37] <syyl> a endmill is normaly a few 1/1000 under its nominal diameter
[13:22:40] <IchGuckLive> im in the wrong thinking its on 0.1 as you got it
[13:22:46] <mrsun> archivist, ahh havent checked that either, might be something there also =)
[13:22:48] <syyl> maybe 1..2/100mm
[13:23:20] <archivist> mrsun, some of the cheap sh*t is very wobbly
[13:23:21] <IchGuckLive> resellers offen do grinding then resell them
[13:23:51] <mrsun> IchGuckLive, ahh these were some kind of a set i bought, new
[13:23:59] <syyl> oh
[13:24:06] <syyl> the goldbronce painted ones?
[13:24:09] <syyl> from china?
[13:24:10] <mrsun> nah
[13:24:20] <mrsun> its just steel, but i would guess its from china :P
[13:24:24] <syyl> :d
[13:24:26] <mrsun> no coating
[13:24:35] <IchGuckLive> i got 50pcs this week from 5-1mm for 25Euros but there is no one in messerment as it is to read on it
[13:25:06] <mrsun> when micing the endmill i get it to be 0.3mm undersized i think it was
[13:25:13] <mrsun> so i guess, its due to that =)
[13:25:29] <mrsun> wait
[13:25:30] <mrsun> 0.03
[13:25:31] <IchGuckLive> mrsun: so use CRC and your fine
[13:25:37] <mrsun> IchGuckLive, crc? :)
[13:25:42] <syyl> 0,03 is not to uncommon
[13:25:44] <syyl> can happen
[13:25:54] <IchGuckLive> Cutter Radius Compensation G41/G42
[13:26:06] <syyl> he uses a HMC
[13:26:08] <mrsun> IchGuckLive, ahh yes, but more work .. heh :P
[13:26:18] <syyl> human machine controll
[13:26:24] <mrsun> syyl, huh ? :P
[13:26:28] <mrsun> no ... ? :)
[13:26:28] <syyl> or am i wrong?
[13:26:33] <syyl> sry :D
[13:26:37] <mrsun> the the computer running the machine :P
[13:26:43] <syyl> :)
[13:26:46] <IchGuckLive> hes on 2.6.pre with G2 options
[13:27:49] <pingufan> Somebody here showed me a milled PCB. I damaged last week 2 of my V-shaped cutters (10°, 0.2mm ø). They lost their tip after ~1 meter of milling length. I run ~18000rpm and moved only with 40mm/min, so speed cannot have been to high. Is my tool pherhaps too slim? Should I use a bigger angle, i.e. 30° ?
[13:28:09] <syyl> your feed is way to slow
[13:28:20] <syyl> that tool will rub itself to death
[13:28:28] <mrsun> pingufan, what material in the tool ? :)
[13:28:32] <mrsun> what material in the pcb?
[13:28:42] <mrsun> you know that pcbs are usaly made of glass fiber right? :P
[13:28:49] <IchGuckLive> pingufan: im on mm with 30deg and i go about 8meters per bit
[13:28:52] <cradek> how's your runout?
[13:29:00] <djdelorie> FR4 is highly abrasive to cutting tools
[13:29:03] <pingufan> I know! but I started at the slow side. It is a 2.5 Euro tool, entirely "hard metal" not better specified.
[13:29:15] <cradek> yes pcb tools are carbide for that reason
[13:29:23] <cradek> oh "hard metal"? hmm
[13:29:30] <IchGuckLive> i use VHM
[13:29:34] <syyl> "hard metal" should be carbide
[13:29:41] <djdelorie> I even looked into some diamond grit blades for my scroll saw to cut them. Ended up getting a metal shear instead.
[13:29:42] <pingufan> I don't know the right english name.
[13:29:42] <cradek> try again with known tooling?
[13:29:59] <syyl> and with more feed
[13:30:05] <mrsun> sure its not "hss" ? :P
[13:30:06] <IchGuckLive> cradek: on the Truetype Tracer 4.0 ?
[13:30:11] <syyl> 40mm/min is not slow, its almost fullstop ;)
[13:30:15] <cradek> what's your measured spindle runout?
[13:30:19] <pingufan> Higher speed is better than lower one?
[13:30:22] <syyl> go for 800...1000mm/min
[13:30:28] <mrsun> pingufan, you have to cut
[13:30:29] <syyl> at leath the tool will cut, not rub
[13:30:34] <mrsun> not abrase the material away
[13:30:48] <mrsun> to slow feed you will just rub the tool against the material, not cut anything =)
[13:30:55] <pingufan> My fast advance is only 1080 mm/min!
[13:31:11] <pingufan> For G0!
[13:31:23] <syyl> then G1 should also work that fast ;)
[13:31:32] <IchGuckLive> go up to 200mm/min loweset
[13:31:45] <pingufan> I am so scaried about feeding through a PCB in that speed...
[13:31:59] <syyl> what could happen?
[13:32:08] <mrsun> feeding through ? .. how would that happend?
[13:32:09] <pingufan> But I will try. With a bigger angled tool.
[13:32:19] <IchGuckLive> drillbit is fixed or does it loose on turns
[13:32:25] <djdelorie> pingufan: do the math. RPM, diameter, speed = how much you're shaving off per rotation
[13:32:28] <pingufan> Cause this nasty "click" noise.
[13:33:20] <mrsun> put a thick backing behind the board if your scared to cut your table =)
[13:33:22] <pingufan> And how much do you calculate for a 0.2mm tool? 1/2 the diameter, 1/3rd,... ?
[13:33:29] <syyl> naah
[13:33:29] <pingufan> For one cut.
[13:33:43] <syyl> there is a chipload for every tool
[13:33:58] <pingufan> But not for those I bought.
[13:34:02] <pingufan> ()chuckle)
[13:34:09] <syyl> but there are numbers that work always
[13:34:21] <syyl> like 0,03mm per tooth and revolution for a small v-cutter..
[13:34:36] <pingufan> THIS is helpful. Thank you.
[13:34:44] <syyl> so you get 1800rpm * 0,03mm * 1tooth
[13:34:52] <syyl> "18000
[13:34:58] <mrsun> a calculator i found recomended 11 SMM at 18000 rpm for a 0.2mm tool =)
[13:35:28] <pingufan> What is SMM
[13:35:38] <mrsun> surface meters per minute
[13:35:48] <mrsun> so 11meters/min i guess =)
[13:35:49] <syyl> thats the cutting speed
[13:35:55] <syyl> has nothing to do with feed
[13:36:01] <mrsun> ahh =)
[13:36:35] <syyl> feed scales up and down, if you change the rpm of the spindle
[13:36:37] <djdelorie> i.e. linear speed of the cutting edge as it whirrls around ?
[13:36:41] <syyl> yeah
[13:36:52] <pingufan> And do you also have a rule of thumb for the thickness of a chip?
[13:37:04] <syyl> depends on the tool and the material
[13:37:19] <syyl> for a v cutter in pcb i go between 0,03 and 0,08mm
[13:37:25] <pingufan> Copper on fibre-glass, a usual PCB.
[13:37:26] <djdelorie> for woodturning, we deal mostly in linear "edge speed" - the speed the cutting edge sees.
[13:37:29] <mrsun> then do the trick a friend of mine got tought, take the tool, lay it on the floor, roll it forward for 11 meters and count the revolutions? :P
[13:37:42] * syyl scratches his head
[13:37:44] <mrsun> or wait, wouldnt that make 18000 again? :P
[13:37:48] <syyl> what? :D
[13:37:50] <mrsun> im to tired for this :P
[13:38:02] <syyl> that will give you a ruined endmill ;)
[13:38:27] <mrsun> sucks that machinist calculator costs =)
[13:38:34] <mrsun> great tool =)
[13:38:38] <pingufan> :)
[13:39:00] <djdelorie> do we need a web-based form for this? ;-)
[13:39:03] <mrsun> g-wizard i ment =)
[13:39:06] <pingufan> With this data you gave me, I can calculate by my own. Useful info. Thank you.
[13:39:07] <syyl> on real tools you get charts for your chiploads
[13:39:32] <IchGuckLive> pingufan: http://www.pictureupload.de/originals/pictures/120312192912_pendant_v1_pcb.jpg this is one bit V-shape path
[13:39:42] <mrsun> most people cant afford "real tools" :/
[13:39:53] <pingufan> But you agree that I should use a wider angle, 30° i.e.?
[13:39:56] <syyl> oh
[13:40:20] <syyl> a "real" endmill from a well named tool supplier doesnt cost to much
[13:40:25] <pingufan> This is exactly what I want to do and get.
[13:40:40] <syyl> a 6mm endmill, hss, 3 flute, 4,90eur
[13:40:47] <syyl> thats ok in my mind
[13:40:57] <mrsun> yeah that wasnt to bad =)
[13:40:59] <IchGuckLive> syyl: agree good price
[13:41:17] <pingufan> This is a 3-blade cutter?
[13:41:28] <syyl> same in carbide, between 7 and 10eur
[13:41:32] <syyl> yes
[13:41:35] <syyl> 3 flute :D
[13:41:52] <IchGuckLive> pingufan: mine pcb is a broken drillbit 0.8mm sharpend bymyself
[13:41:57] <pingufan> Means 3 cutters on diameter?
[13:42:27] <pingufan> Yes, i must get experience with that.
[13:42:29] <syyl> it has 3 teeht
[13:42:47] <syyl> http://www.zps-fn.de/files/category/350/detail_preview.jpg
[13:42:49] <pingufan> And still only 0.3mm diameter?
[13:42:52] <syyl> thats a 4 flute
[13:43:07] <syyl> naa, in that example its bigger :D
[13:43:11] <IchGuckLive> pingufan: where are you located in the world
[13:43:19] <pingufan> Austria
[13:43:30] <IchGuckLive> ah man spricht deutsch
[13:43:36] <IchGuckLive> B)
[13:43:39] <pingufan> Du auch. :)
[13:44:07] <IchGuckLive> here engl im near Kaiserslautern Germany on the french border
[13:44:31] <pingufan> Oh, there! I have a friend in the "blackwood".
[13:45:09] <pingufan> There is also a company where I order my PCBs. "Becker & Müller". Great company.
[13:45:53] <IchGuckLive> pingufan: http://stores.ebay.de/VHM-Shop-gunstig-fair?_trksid=p4340.l2563
[13:46:57] <pingufan> This tools I bought at CNC-HEIZ in Vienna. They are fine. But V-Shapes are more sensitive.
[13:47:11] <IchGuckLive> pingufan: you need to wait till she got 50pcs for 25euros
[13:47:55] <pingufan> Look here: http://www.bk-faserverbundtechnik.at/shop/index.php/cPath/6_36
[13:48:37] <IchGuckLive> this is to expensiv
[13:49:16] <pingufan> But with 30° the tool is more rugged. The 10° I bought are like a sewing needle.
[13:49:25] <djdelorie> 9eu per "oh crap" :-)
[13:50:06] <syyl> the 9eur are ok
[13:50:07] <IchGuckLive> i pay 1.9euros eatch
[13:50:33] <syyl> that type of engraving cutter has a pretty agressive grind
[13:50:38] <syyl> (but i dont like em)
[13:51:06] <IchGuckLive> syyl: agree O.o
[13:51:16] <syyl> i prefer them
[13:51:16] <syyl> http://www.bk-faserverbundtechnik.at/shop/images/Stichel_Stichelsatz_Gravierstichel_CNC_Frase.jpg
[13:51:23] <syyl> as they work also in steel
[13:51:49] <Connor> I hate trying to use v-bits..
[13:52:14] <syyl> no problems here ;)
[13:52:16] <IchGuckLive> as a 0.6 endmill Diamond is at 2.2euros
[13:52:33] <Connor> My Z isn't 100% trammed to my table.. and my table is a little off.. gives me nothing but issues on trace widths
[13:52:47] <pingufan> They have one advantage: You can sharpen them by yourself.
[13:52:59] <syyl> they are also more ruged
[13:53:04] <pingufan> But this is the only advange besides of the price.
[13:53:04] <syyl> take more abuse
[13:53:08] <IchGuckLive> pingufan: the others also
[13:53:25] <mrsun> 125 euro .. damn
[13:53:37] <syyl> i do all my engraving with cutters of that type
[13:53:44] <syyl> and pcb cutting
[13:53:47] <pingufan> I know. But for first steps I use cheap tools.
[13:53:53] <syyl> and chamfering
[13:55:16] <pingufan> IchGuckLive: Your tools are 1/8" shaft?
[13:57:25] <IchGuckLive> yes
[13:57:43] <pingufan> Where do you buy them?
[13:58:01] <IchGuckLive> at ebay shops from pcb manufactures
[13:58:02] <pingufan> It possibly makes sense to try them.
[13:58:32] <IchGuckLive> but wait till she got 50 for that price then your done for a year 50pence etch
[14:06:16] <mrsun> hmm, i want a forth axis =)
[14:06:19] <mrsun> would be nice
[14:08:12] <IchGuckLive> mrsun: 11NM or more
[14:08:24] <mrsun> IchGuckLive, huh ? :)
[14:08:58] <ssi> I want a fourth axis too, but I need the first three first :(
[14:10:18] <IchGuckLive> mrsun: only positioning or do you need to mill with it
[14:10:31] <mrsun> IchGuckLive, for milling with it
[14:10:35] <mrsun> so i can reach more sides =)
[14:11:25] <ssi> that's like 1500ozin
[14:11:43] <IchGuckLive> reatch more sides does not need to mill only position and hold
[14:11:50] <IchGuckLive> mill neds more power
[14:12:02] <IchGuckLive> 30-50NM
[14:12:07] <mrsun> IchGuckLive, i want to bea ble to rotate it while i mill =)
[14:12:23] <ssi> 30NM really?
[14:12:25] <mrsun> 8.4 with a 10:1 ratio maybe? :)
[14:12:26] <ssi> oh wait
[14:12:27] <IchGuckLive> therfor i asked 11NM is Stepper 3NM at 1:38 worm
[14:12:28] <ssi> the gearing of the table
[14:12:32] <ssi> yea ok
[14:13:11] <mrsun> if i buy for a forth axis i wouldnt go for any 3Nm i would go for the biggest i can get my hands on then gear the hell out of it :P
[14:13:30] <IchGuckLive> mrsun: then you need a Nema 34 on it
[14:13:41] <mrsun> IchGuckLive, shouldnt be a problem ? :)
[14:14:26] <mrsun> but just in electronics it would cost me about 170 euro :P
[14:14:32] <mrsun> that is motor + driver
[14:14:37] <mrsun> but im far away from that still so =)
[14:14:50] <mrsun> i want to make a money machine first (gantry router or whatever its called) =)
[14:15:14] <IchGuckLive> i woudt go then for a 1:50 worm at 400Steps/rev it makes 20.000steps per 360deg
[14:15:31] <IchGuckLive> so ad another 100Euros for the worm
[14:16:48] <IchGuckLive> http://maedler.de/Product/1643/1619/705/1931/1933/794.aspx mrsun
[14:17:36] <IchGuckLive> mrsun: more precision needet well -> http://maedler.de/Category/1643/1619/705/743.aspx
[14:18:40] <IchGuckLive> times 4 the price but the quality is outstanding
[14:18:59] <IchGuckLive> http://maedler.de/Product/1643/1619/705/743/776/777.aspx move the mouse over the E
[14:19:26] <mrsun> 53% efficiency
[14:19:49] <mrsun> so a 8.4Nm would become about 4.2Nm then ? :)
[14:20:17] <mrsun> *16 tho with the lowest gearing, adds up to 64Nm in the end ;P
[14:20:18] <IchGuckLive> if you can go in the axis distance it will be less at 30mm distance http://maedler.de/Product/1643/1619/705/743/750/751.aspx
[14:21:06] <IchGuckLive> 8,4times 5 -20%
[14:22:13] <IchGuckLive> oh i see madler has a engl language also
[14:23:25] <mrsun> http://www.youtube.com/watch?v=EMaZ6NlB574 <-- thats real nice =)
[14:25:27] <IchGuckLive> mrsun: full turn is not good as a part is mounted make softlimits
[14:26:13] <IchGuckLive> http://www.youtube.com/watch?v=w0-tXDEvAqg
[14:39:44] <IchGuckLive> mrsun: he increased to 5_axis
[14:42:52] <mrsun> hmm, for this cutter radius compensation, i guess the cam program calculates the tool path (center of tool) not the actual part and turns cutter compensation on ,...
[14:42:56] <mrsun> so how would that work ? :P
[14:46:45] <morfic_> hey
[14:47:03] <JT-Shop> djdelorie: I don't have a clue what that means
[15:11:20] <morfic_> Has anyone seen Dave Cole lately? (Not sure davec is him or not :)
[15:15:28] <davec> Nope, I'm not him.
[15:21:07] <morfic_> davec: k, just checking, trying to get a hold of him
[15:31:55] <Thetawaves> mrsun, what cam do you use?
[15:32:39] <mrsun> atm heekscad and tried aspire
[15:49:14] <JT-Shop> Sweet! the plasma is all functional now... only need to de-shamblize the cables and then my only obstical is me
[15:49:59] <djdelorie> m/e once pondered a "makerbot" that worked by welding on steel pellets one at a time to build up the assembly...
[15:50:22] <Thetawaves> screw that
[15:50:28] <Thetawaves> just feed in material
[15:50:48] * djdelorie then pondered cleaning up the first car wreck with a broom and shovel
[15:50:59] <ssi> I have to de-shamblize my plasma table's cables
[15:51:04] <ssi> I bought some cable chain for it, and never installed it
[15:51:15] <ssi> plus my table leaks a little bit
[15:51:18] <djdelorie> Thetawaves: I was thinking "build car frames from scratch". Dot matrix printer with 1/2" steel dots.
[15:51:32] <Thetawaves> that is a hell of an idea
[15:51:32] <ssi> and I only have a single rail on the gantry, and it's not as rigid as I'd like... at high speed it cuts wavy lines :P
[15:51:38] <Thetawaves> but instead of cylinders, use cubes
[15:51:56] <Thetawaves> spheres*
[15:52:15] <djdelorie> the trick is getting the spot welder to melt the "dots" just enough
[15:52:29] <ssi> little device that melts metal and poops out little drops of it
[15:52:34] <ssi> :D
[15:52:52] <djdelorie> ha! a reprap with a ceramic extruder, feeding out liquid aluminum...
[15:53:27] <mrsun> problem would be that you wouldnt melt the material youve already "extruded"
[15:53:32] <Thetawaves> you'd need a heavy duty setup to do that
[15:53:37] <mrsun> hard to make it fuse :P
[15:53:48] <djdelorie> you'd need to run a current through the extruder to weld it?
[15:53:53] <Mjolinor> induction coil round hte work being produced
[15:54:03] <mrsun> keeping it molten? ;P
[15:54:07] <djdelorie> and bridging gaps would be tricky.
[15:54:12] <Thetawaves> what would keep it from flowing?
[15:54:31] <Thetawaves> you could easily just put a torch in front of the extruder to melt the top of the metal
[15:54:41] <djdelorie> you'd have to use an alloy that had a plastic state between solid and liquid, and heat it to that.
[15:54:58] <mrsun> and oxides
[15:55:01] <mrsun> big issue
[15:55:04] <Thetawaves> yeah that
[15:55:10] <ssi> shielding gas
[15:55:24] <djdelorie> it's a cnc machine. Seal it in a room and flood it with nitrogen :-)
[15:55:48] <djdelorie> or run it in a vacuum
[15:55:52] <mrsun> isnt nitrogen absorbed into alu ? :P
[15:56:06] <Thetawaves> what kind of tech would you use to heat the al extruder?
[15:56:17] <ssi> use argon, not nitrogen :D
[15:56:23] <djdelorie> probably induction heating in a ceramic nozzle
[15:56:45] <Thetawaves> aww fuck this is out of my price range
[15:57:08] <Thetawaves> you would probably need at least 10kw
[15:57:17] <ssi> yea it's gonna take a pile of energy
[15:57:34] <mrsun> resin bound steel pelets infused with brass afterwards
[15:57:42] <ssi> hell, you need 700A in an arc/tig welder to do heavy aluminum worth a crap
[15:57:47] <mrsun> like they do rapid prototyping in metal :P
[15:58:09] <djdelorie> a robot could probably use a mig welder as the extruder
[15:58:09] <mrsun> nah rapid prototyping in metal i guess would be a mill, but for very complex work
[15:58:19] <djdelorie> or laser sintering
[15:58:30] <Thetawaves> thats how they do it
[15:59:27] <djdelorie> solves the briding problem too. Not sure I want be around a computerized laser capable of building cars from dust...
[15:59:32] <djdelorie> bridging
[15:59:58] <Thetawaves> ahaha
[15:59:59] <Thetawaves> http://www.boatdesign.net/forums/boat-design/homemade-sintering-titanium-28360.html
[16:02:18] <Thetawaves> what the fuck djdelorie
[16:02:21] <Thetawaves> you read slashdot?
[16:02:27] <Thetawaves> posted a few hours ago
[16:02:27] <djdelorie> yes
[16:02:28] <Thetawaves> http://www.tuwien.ac.at/en/news/news_detail/article/7444/
[16:02:53] <djdelorie> I saw the headline go by this morning, didn't read it. Cool.
[16:05:16] <Thetawaves> also this: http://hardware.slashdot.org/story/12/03/12/1231245/mit-fiber-points-to-woven-glasses-free-3d-displays
[16:25:38] <DJ9DJ> gn8
[17:31:12] <JT-Shop> plasma deshamblization complete... now to figure out why I got 125v when I dialed in 81v :/
[17:31:28] <JT-Shop> maybe take a break and work on cannon parts
[17:31:43] <JT-Shop> yea that sounds like a good idea
[17:31:55] <jdhnc> don't point it near your girlfriend.
[17:32:37] <JT-Shop> yea, she don't like guns and hides under the stairs when we are shooting
[17:36:29] <ssi> hah
[17:36:36] <ssi> I had to break mine into guns
[17:36:39] <ssi> now she loves 'em
[17:36:41] <ssi> has three of her own
[17:36:49] <ssi> she was terrified of guns when I met her
[17:41:58] <JT-Shop> well my wife has plenty of guns and knifes and loves to shoot... but Cocoa she don't like them at all
[17:44:15] <JT-Shop> I have a powerbook g4 that someone gave me and when I press the power button a note plays but nothing happens, the battery is charged, it has been dropped more than once...
[17:44:43] <jdhnc> http://www.cbsnews.com/8301-504083_162-57394103-504083/california-man-charged-with-murder-after-homemade-cannon-kills-girlfriend/
[17:45:21] <JT-Shop> yea, you don't want to play with stuff like that
[17:45:50] <Jymmm> JT-Shop: A new computer for you http://i43.tinypic.com/vsf8z4.jpg
[17:46:12] <JT-Shop> someone took a bite out of it all ready
[17:47:33] <Jymmm> that's just quality testing
[17:49:37] <JT-Shop> hmm reaseating the memory stick didn't do anything... any secret key combos for a mac?
[17:50:14] <Jymmm> what does the LED do?
[17:50:21] <JT-Shop> jdhnc: that was not a cannon, that was a pipe bomb
[17:50:28] <JT-Shop> dont' see any led
[17:50:41] <Jymmm> Is this a desktop or laptop?
[17:50:46] <JT-Shop> the battery one turns yellow if you plug it in
[17:50:49] <JT-Shop> laptop
[17:50:56] <JT-Shop> powerbook g4
[17:52:01] <Jymmm> and there is no white LED near the latch?
[17:52:11] <Jymmm> or around the power button?
[17:52:42] <JT-Shop> no
[17:53:24] <JT-Shop> the caps lock led works
[17:58:28] <mikegg> part it out...
[18:03:31] <JT-Shop> I only know enough about macs to get a screwing and I got it free :/
[18:04:42] <mikegg> have you tried booting from a CD?
[18:06:22] <mikegg> just hold down C, I think
[18:09:27] <JT-Shop> does it have to be a mac CD?
[18:09:54] <JT-Shop> and no, I've not tried that
[18:09:56] <mikegg> oh wait, the G4 isn't intel
[18:10:16] <JT-Shop> no, it has a 1/4 eaten apple on it
[18:10:17] <mikegg> is it? I think that's a powerPC processor
[18:10:52] <mikegg> right, but the newer macs use intel processors, you can load linux pretty easily I think
[18:11:11] <JT-Shop> ha, I'll give that a shot
[18:12:48] <JT-Shop> wow something came up
[18:12:55] <mikegg> cool!
[18:13:40] <JT-Shop> I'm at some kind of dos prompt 0 >
[18:14:27] <JT-Shop> when I try dir it says it can't open the directory device
[18:14:53] <mikegg> cd .. ?
[18:15:04] <JT-Shop> I put the live cd in there
[18:15:20] <JT-Shop> changed to 1>
[18:15:20] <mikegg> oh, I meant "cd .."
[18:15:25] <mikegg> see if you can list the drives
[18:16:18] <JT-Shop> list unknown
[18:18:41] <skunkworks__> logger[psha]_:
[18:18:41] <logger[psha]_> skunkworks__: Log stored at http://psha.org.ru/irc/%23linuxcnc/2012-03-12.html
[18:18:54] <skunkworks__> cradek: Happy Birthday!
[18:30:55] <ries> KimK: Are you here?
[19:24:31] <JT-Shop> mikegg: I think it is a doa hard drive
[20:07:05] <morfic> JT-Shop: install smartctl and ask smart for some info/short test or long test
[20:07:48] <morfic> cradek: happy birthday
[20:24:24] * flyback hopes siri causes madness and cancer
[20:29:27] <ries> flyback: That's stupid of you saying that and quite offensive
[20:43:53] <flyback> yeah well, rIES I hate the world
[20:45:01] <ries> flyback: don't hate the world...
[20:45:13] <Tom_itx> flyback why are you even in this channel?
[20:45:19] <Tom_itx> you don't do cnc
[20:45:51] <flyback> I was helping someone earlier
[20:45:59] <Jymmm> flyback: You've pulled that random comment crap for years in the other channel, it's not appreciated nor tolerated in here. Please do not do it again.
[20:46:01] <Tom_itx> you're not helping now
[20:46:10] <flyback> that's fine
[20:46:20] <flyback> doens't matter anyways
[20:46:24] <flyback> I decided wha I need to do
[20:46:43] <ries> sounds like a good decision to me
[20:47:16] <Jymmm> Actually that's pretty light compared to what he typically does.
[20:47:44] <Tom_itx> there aren't too many channels left for him
[20:48:31] <Jymmm> I doubt he's band from the other channel, just most have him on ignore
[20:48:56] <Dave911> A guy on the DIY-CNC email list reports this: error 32 BROKEN PIPE come up when
[20:48:58] <Dave911> i try to test the motors in the config screen on the Step Configuration software. Has anyone heard of that before??
[20:49:34] <Dave911> It is not an error I have heard of before.
[20:50:23] <Thetawaves> do you have the latency test running?
[20:50:58] <Dave911> I have no idea.. it really is not me reporting the error.. some LinuxCNC newbie.
[20:51:24] <Thetawaves> i believe that happened to me when i tried to test the motors while the latency test was running
[20:52:17] <Dave911> OK.. cool, I'll suggest that and see what happens. I'm going to recommend that he get on this IRC also.. Thanks..
[20:56:45] <Thetawaves> is there a way i can setup linuxcnc to always run the plunge speed at 70% of the regular feed rate?
[21:03:20] <Dave911> I think you could setup your regular feed rate as a variable. Do some math to set a different variable to 70% of your feed rate and use that variable as your plunge rate..
[21:07:01] <mikegg> are you running g-code? ctrl-h works pretty well..
[21:17:32] <Thetawaves> ctrl-h?
[21:22:12] <mikegg> yeah just search and replace the g-code for the feed rate you want to change
[21:23:23] <mikegg> in the text editor that is..
[21:24:46] <Tom_itx> some gcode editors let you do math on stuff to change offsets etc
[23:23:41] <Kirk_Wallace> I'm trying to follow #include links through some source code for Modbus. I started with my demo.c, which has #include mb.h. mb.h has #include port.h, mbport.h and mbproto.h . One of my problems is, how do the .c files get included? Thanks to yesterday's IRC session, I learned about CFLAGS. Is there something else I need to learn about how .h and .c files get along?
[23:28:00] <djdelorie> .c files normally aren't "included" from elsewhere, they're where you start, usually built from a Makefile or other build system, one at a time.
[23:31:05] <Kirk_Wallace> I think the way my sample files are set up, the functions are declared in the .h's and defined in the .c's, so my guess is that they get included somehow and are not in my main demo.c .
[23:31:51] <Kirk_Wallace> Oops. I see your comment about the Makefile.
[23:37:11] <Kirk_Wallace> What I'm up against is the demo works fine with RS232 which is full-duplex. I found a RS485_ENABLE define which appears to add RTS toggling to control the RS485 transceiver. When I add the RS485_ENABLE the AVR pin gets setup but not connected to the modbus functions. I'm trying to reverse engineer the demo to see if I can fix the missing bits.
[23:39:38] <Kirk_Wallace> What's this from the Makefile? ## Other dependencies
[23:39:38] <Kirk_Wallace> -include $(shell mkdir dep 2>/dev/null) $(wildcard dep/*)
[23:41:40] <djdelorie> the $(shell) creates the directory if it doesn't exist, and the $(wildcard) lists any makefile fragments to be included in the build rules.
[23:41:56] <djdelorie> for your purposes you can probably ignore that line
[23:43:07] <Kirk_Wallace> Or how about? ## Objects that must be built in order to link
[23:43:08] <Kirk_Wallace> #OBJECTS = excoils.o
[23:43:08] <Kirk_Wallace> OBJECTS = demo.o
[23:43:08] <Kirk_Wallace> MBPORTOBJECTS = port/portserial.o \
[23:43:08] <Kirk_Wallace> port/portevent.o \
[23:43:08] <Kirk_Wallace> port/porttimer.o \
[23:43:44] <djdelorie> yeah, those are the kinds of things to look for. Elsewhere in the makefile there will be a pattern rule to say how to build *.o from *.c
[23:48:45] <Kirk_Wallace> Here is the Makefile:
[23:48:51] <Kirk_Wallace> http://www.wallacecompany.com/machine_shop/EMC2/AVR/Makefile
[23:49:54] <djdelorie> hmmm... if you're building a project that *uses* modbus, perhaps modbus is a library? Do the instructions tell you to *link* in modbus?
[23:50:59] <Kirk_Wallace> It's FreeModbus http://freemodbus.berlios.de/index.php?idx=32
[23:52:39] <Kirk_Wallace> The demo is a sample from the AVR section of the package. I've been massaging it to get the application I want.
[23:54:09] <djdelorie> well, I don't know that package, but in general, after you compile all the files to *.o (or *.a), including your demo.o, you use gcc to link all the *.o (and *.a) into the final executable.
[23:56:47] <Thetawaves> you avr users..
[23:56:55] <Kirk_Wallace> I'm just used to linking files with includes from the main .c file so this Makefile stuff is new to me. I think I can be pretty sure the .o file list will cover all of the .c files that are linked. Which means I've got to look at all of them to find my RTS problem, darn.
[23:57:04] <alex4nder> hey
[23:57:24] <Thetawaves> i wrote a 9p network fs layer to communicate with the micros as if they are files mounted
[23:57:56] <Thetawaves> could get about 440 reads/sec i believe
[23:58:42] <Thetawaves> i thought it would make it really easy to interface with emc
[23:58:56] <Thetawaves> just write your encoder targets to a file
[23:58:59] <Thetawaves> profit!