Back
[01:25:36] <alex_joni> Jymmm: pong
[01:25:46] <Jymmm> LAG
[01:26:27] <alex_joni> connection issues..
[03:15:59] <chaudhary> hi
[06:30:38] <Jymmm> SWPadnos: Save 20%
http://sfbay.craigslist.org/sby/sad/2824426031.html
[06:33:33] <TekniQue> LED and black hat in the same context
[06:33:34] <TekniQue> lolling
[06:33:50] <Jymmm> Who's LED ?
[06:37:14] <TekniQue> Jymmm: light emitting diodes
[06:38:14] <Jymmm> I said WHO, not what. I guess Black Hat is not a company then.
[06:38:23] <Jymmm> in this context
[07:48:32] <Jymmm> GoSebGo: <----- Is that a hint, or a reminder?
[07:52:00] <Jymmm> VINTAGE Shabby Chic my ass, try FUGLY instead...
http://sfbay.craigslist.org/nby/for/2789250403.html
[08:09:42] <alex_joni> fugly indeed
[10:21:09] <skunkworks> http://www.cnczone.com/forums/linuxcnc_formerly_emc2/
[11:06:41] <mrsun> hmm, enmills with relief above the cutting edges, where the heck do you find that? :)
[11:06:52] <mrsun> or maybe one should just buy extra long endmills :P
[11:07:49] <syyl_> you mean with a thinner shank above the flutes?
[11:08:10] <mrsun> yes
[11:08:21] <mrsun> i know some sizes have that, like 7mm has 8mm shank or something
[11:08:51] <mrsun> no 7 has 6 :P
[11:21:47] <syyl_> those exist
[11:21:50] <syyl_> but are not cheap
[11:22:05] <syyl_> mostly that are hpc/mtc carbide cutters
[11:22:08] <syyl_> with a long shank
[11:22:49] <syyl_> http://www.hoffmann-tools.com/Zerspanung/20-VHM-Fraeser/VHM-Fraeser-HPC-ZOX-10-mm-GARANT-oxid.html
[11:22:58] <syyl_> you have something like that in mind, right?
[11:22:59] <Loetmichel> re @ home
[11:26:32] <mikegg> mrsun: chuck the tool up and spin it slowly, then use a grinding wheel to carefully remove some material.
[11:27:01] <syyl_> thats the cheaper way :D
[11:27:14] <Loetmichel> *baaaaahhh* this PVC foam is the grandfahter of Dirt. its electrostatic, very light and flies EVERYWHERE :-( ->
http://www.cyrom.org/palbum/main.php?g2_itemId=12748
[11:27:15] <syyl_> with a cordless drill on the benchgrinder you get a pretty good reult
[11:27:32] * Loetmichel is making more cable spools
[11:27:59] <skunkworks> vacuum!
[11:28:20] <syyl_> shopair ;)
[11:31:33] <pcw_home> The leaf blower theory of cleanup --> blow the problem somewhere else
[11:31:40] <Loetmichel> skunkworks: will do... AFTER the part is done
[11:32:08] <syyl_> thats whot we do at work, pcw_home ;)
[11:32:19] <Loetmichel> syyl_: compressed air? so i have the green/red swarf EVERYWHERE in the machine and the shop?
[11:32:31] <syyl_> in YOUR shop :D
[11:32:32] <syyl_> not mine
[11:32:34] <syyl_> :P
[11:32:45] <Loetmichel> syyl_: BAD syyl_!
[11:34:30] <archivist> machine should be faster so it throws the chips further away
[11:34:52] <syyl_> good technique
[11:35:17] <Loetmichel> hrhr, the G64P0.05 generates some strage noises. switching very fast between accelerating and breaking, like a tremolo ;-)
[11:35:37] <Loetmichel> archivist: already going 1200mm/min
[13:01:50] <mrsun> hmm, in what configuration is steppers the strongest?
[13:01:56] <mrsun> full step ? half step etc?
[13:02:07] <jdhNC> 0 step
[13:02:18] <jdhNC> it's all downhill from there.
[13:02:44] <mrsun> haha =)
[13:06:54] <mrsun> oh damn, goest fast down when you try and step fast =)
[13:07:04] <jdhNC> yep
[13:07:44] <jdhNC> goes down fast from full->half
[13:08:10] <mrsun> but say what would be best for a cnc machine then, half step ? 1/4 ?
[13:08:53] <mrsun> or doesnt microstepping make a difference in the power of the motor ?
[13:09:00] <jdhNC> you have to trade off torque for speed, both of which are driven by size and cost
[13:09:01] <mrsun> read that like in half step you lose about 30%
[13:09:50] <jdhNC> search for 'microstepping myths'
[13:11:09] <jdhNC> you could over-spec your motor 10x, use 16/steps and still have enough power, but you lose speed
[13:11:41] <mrsun> 9,8% holding torque at 1/16 .. (thats what im running)
[13:12:24] <mrsun> tho holding torque does that have anything with how much it can handle to move ?
[13:15:04] <jdhNC> yes, but you can still trade speed for power
[13:17:15] <mrsun> by just adjusting down the maximum speed you mean ?
[13:18:37] <jdhNC> if you can live with slow, then you can have more power
[13:19:09] <mrsun> so lowering the microstepping + the speed would kick ass then ? :)
[13:20:11] <jdhNC> full-step, 0 RPM is max torque
[13:20:28] <jdhNC> increase the speed, or the stepping and torque drops
[13:37:57] <JT-Shop> mrsun: holding torque is how hard it tries to not move
[13:38:27] <mrsun> JT-Shop, ye but all steppers are specified in that, dont know how that is realy relevant information imo =)
[13:38:34] <mrsun> as i want to know how hard it tries to move =)
[13:38:58] <JT-Shop> depends on the voltage as well as a zillion other things
[13:39:28] <JT-Shop> good steppers have a torque voltage speed chart
[13:40:06] <PCW> Beyond the 1 to 1/2 step change higher microstepping ratios do not result in much torque loss
[13:40:07] <PCW> but because the flatness of the holding torque vs position which is determined by the fixed 50 pole
[13:40:09] <PCW> mechanics of the step motor, you dont gain much accuracy at higher microstepping ratios except
[13:40:10] <PCW> at very light loads
[13:41:04] <JT-Shop> http://www.automationdirect.com/static/specs/surestepmotors.pdf
[13:41:30] <JT-Shop> mrsun: look at page 2
[14:34:35] <mrsun> http://www.lathes.co.uk/stepperhead/img17.jpg damn i like that "screw" =)
[14:34:41] <mrsun> leadthingie for the table =)
[14:35:08] <jdhNC> like a worm gear thing in there?
[14:36:10] <mrsun> yeah, the wormwheel insted of a leadscrew
[14:36:36] <mrsun> http://www.lathes.co.uk/stepperhead/page2.html its a quite amazing machine overall =)
[14:39:06] <jdhNC> I like the toolpost
[14:40:57] <mrsun> theres a page3 and page1 also
[14:43:37] <JT-Shop> pcw is there a layout of the mounting holes for the 7i47 somewhere?
[14:44:01] <gene__> I'm back, with another headache
[14:44:56] <PCW> should be in the 7I47 manual (last page)
[14:46:43] <gene__> Its catching the probe break as a program stopping error on the next move away after the G38.2 has completed??????? IMO it should ignore the probe when not executing a G38.2
[14:46:56] <gene__> Ideas
[14:47:00] <gene__> Ideas?
[14:48:32] <JT-Shop> I must need to download the manual again it is missing from my copy
[14:49:38] <JT-Shop> yep, there it is in the manual :)
[14:56:09] <JT-5i25> PCW on page 8 there is a tiny typo "If you wish to use a single power supply for the 7I76s field outputa and field logic power"
[14:57:26] <gene__> Humm, just found another problem, its all running on cpu0 & its pegged out
[15:00:57] <gene__> isolcpus=1 is in the grub.cfg for this kernel, this is a showstopper I didn't have with emc-2.6.0-pre
[15:01:46] <mrsun> is it possible to use a lower resolution encoder compared to the stepping ?
[15:01:57] <mrsun> 200p encoder for 800 step/revolution stepper =)0
[15:11:11] <PCW> thanks JT-5i25, fixed
[15:16:05] <gene__> Humm, just found another problem, its all running on cpu0 & its pegged out
[15:16:13] <gene__> isolcpus=1 is in the grub.cfg for this kernel, this is a showstopper I didn't have with emc-2.6.0-pre
[15:16:17] <gene__> Ideas?
[15:18:53] <mrsun> hmm, was thinking of making a manual control panel for the mill, using 3 rotary encoders for the moving then buttons to select to move like 1mm/step of the encoder, 0.1mm etc etc ... but how the heck would one get all that info into emc ...
[15:19:18] <mrsun> *linuxcnc
[15:19:19] <mrsun> :P
[15:20:30] <mrsun> only got 5 inputs from the breakout board =)
[15:22:52] <mrsun> hmm, aparently can use usb interface for stuff like that ?
[15:22:55] <mrsun> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Simple_Remote_Pendant
[15:38:40] <gene__> In fact, nothing I do is using cpu1. WTHeck?
[15:39:53] <Valen> isocpus is good as a rule
[15:39:58] <Valen> improves your latency
[15:40:09] <Valen> how many cores do you have?
[15:40:29] <gene__> 2, and isolcpus=1 is in the grub.cfg
[15:40:47] <Valen> the isocpus things puts all the realtime tasks onto one cpu and leaves the other cpu to do silly things like put stuff on your screen
[15:40:57] <gene__> it is an atom board, a D525
[15:41:12] <Valen> the behavior is by design
[15:41:22] <Valen> turning isocpus off will give you worse latency
[15:41:44] <gene__> Taint working, all of linuxcnc is running on cpu0, and its plumb tapped out
[15:42:02] <Valen> thats normal
[15:42:14] <Valen> well everything running on one core is normal
[15:42:26] <Valen> you don't see anything running on the other cpu
[15:42:53] <PCW> shouldn't userland processes run on the other?
[15:42:58] <Valen> are you actually having a problem?
[15:43:20] <Valen> i'm pretty sure thats what he's seeing, one cpu is maxed out running linuxcnc stuff
[15:43:51] <gene__> Is .5 seconds response to a keypress, or release, normal? I set the base period out to 65 u-secs without seeming to restore real time control
[15:44:10] <Valen> you dont have hyperthreading or anything on?
[15:45:13] <gene__> Yes, one cpu0 is running 75% when linuxcnc is running, cpu0 is 0% full time, hyperthreading is off in the bios or gkrellm would be showing me 4 cores
[15:45:29] <Valen> no half a second of lag in the UI is abnormal
[15:46:09] <Valen> you can turn the isocpus thing off as a one off by editing the line in grub when it boots but i don't think it will solve your problem
[15:46:29] <Valen> you shouldn't be using 75% cpu doing nothing
[15:46:56] <gene__> I did not have this with emc2-6.0.pre2442, I do with linuxcnc-2.6.0-pre2442
[15:47:06] <Valen> sounds like a bug
[15:47:19] <gene__> How can I revert
[15:47:38] <Valen> how are you installing it?
[15:47:48] <gene__> installed last night
[15:47:49] <Valen> if stability is important you shouldn't be running pre-release lol
[15:47:56] <gene__> :-)
[15:48:06] <Valen> where did you get the pre-release emc from?
[15:48:14] <Valen> did you install it from source
[15:48:20] <Valen> nightly build bot?
[15:48:24] <gene__> so far its been nrock solid. From the repo
[15:48:39] <Valen> so you built it from source then?
[15:48:49] <Valen> running make and make install etc?
[15:48:54] <gene__> No, synaptic installed it
[15:50:09] <gene__> from buildbot lucid master-rt
[15:51:12] <Valen> probably easiest to uninstall then go to the 2.5 branch. thats the current release isnt it?
[15:51:57] <gene__> bad dog, no biscuit? Not sure, will I lose my data dirs?
[15:52:13] <Valen> generally shouldn't
[15:52:17] <Valen> might be an idea to take a copy
[15:52:36] <Valen> running nightly builds isn't for rookies lol
[15:52:53] <Valen> alternatly mention your problem on the mailing list
[15:53:09] <Valen> and see if anybody notices and makes a fix ;->
[15:53:56] <gene__> what do I edit that last line to? Who you callin a rookie, anyway I'm 77 and been using emc since 2.0 days
[15:54:31] <Valen> then quit yer wingin and learn how to edit a sources file ;-P
[15:54:32] <gene__> I forgot the smiley, sorry :-)
[15:54:41] <Valen> (note mine had a wink)
[15:55:14] <Valen> your in lucid i presume?
[15:55:16] <gene__> duly noted, gotta pull any low hanging chains ;)
[15:55:18] <Valen> 10.04 i think
[15:55:50] <gene__> I think I'm now using something that new in 2.6
[15:56:01] <Valen> something like "deb
http://buildbot.linuxcnc.org/ hardy v2.5_branch-rt" i believe
[15:56:05] <gene__> But I'll try, bbl
[15:56:09] <gene__> Thanks
[15:56:24] <gene__> yes, 10.04
[15:56:40] <Valen> so you had
http://buildbot.linuxcnc.org/ hardy master-rt before?
[15:57:19] <gene__> no, regular 2.4.7 something
[15:58:16] <Valen> hold up
[15:58:38] <Valen> oh you went from regular releases to the master branch?
[15:58:44] <Valen> in the sources file
[16:04:30] <gene__> yes, about a month ago
[16:05:31] <gene__> this seems to have happened with the name change
[16:06:55] <gene__> I had to rename some dirs and fix paths in the ini file, did I miss adjusting something else?
[16:10:52] <gene__> I renamed ~gene/emc2 to ~/gene/linuxcnc, and then fix the same thing in the ini files PROGRAM_PREFIX string. Was there something else I missed? Perhaps someoption in a loadrt statement in the .hal file?
[16:16:08] <Valen> I don't know, I'm running 2.5 because i needed some bits, i havent upgraded it in months
[16:22:58] <gene__> I just reinstalled emc-2.6.0-pre-2264 from the same buildbot and it all on cpu0 again, so that wasn't it.
[16:24:21] <gene__> I just reinstalled emc-2.6.0-pre-2264-gfecf238 from the same buildbot and it's still all on cpu0, showing about 25% of it, so that wasn't it.
[16:24:55] <gene__> Dinner time while I cogitate on this revoltin development :-)
[16:25:43] <gene__> When I get back, I'll reboot just for S&G.
[16:29:40] <Valen> its all *supposed* to be on one cpu
[16:29:54] <Valen> the other cpu is used for realtime tasks only
[16:34:51] <Valen> and unless whatever it is is multithreaded it wont use multiple CPU's anyway
[18:45:57] <JT-Shop> at least it is not sleeting, snowing and raining when the pipe broke at the pit-less adapter...
[19:04:25] <gene76> Lo everyone, back from dinner, rebooted, still nothing using cpu1 on this D525 Atom board, 10.04, hyperthreading turned off, isolcpus=1 in grub.cfg
[19:05:09] <gene76> emc-2.60-pre using 20% + of cpu0
[19:08:37] <gene76> Interesting tidbit from dmesg:
[19:09:12] <gene76> [ 105.426087] RTAI[hal]: mounted (IPIPE-NOTHREADS, IMMEDIATE (INTERNAL IRQs DISPATCHED), ISOL_CPUS_MASK: 0).
[19:09:13] <gene76> [ 105.426092] PIPELINE layers:
[19:09:15] <gene76> [ 105.426097] f9ec4e20 9ac15d93 RTAI 200
[19:09:16] <gene76> [ 105.426102] c085cb20 0 Linux 100
[19:10:31] <gene76> where did that ISOL_CPUS_MASK:0 come from?
[19:18:34] <Valen> isocpus is MEANT to stop things from using the other cpu
[19:19:12] <Valen> it tells the scheduler to ignore it basically
[19:19:24] <Valen> if you want to not use it
[19:19:25] <Valen> reboot
[19:19:30] <Valen> go into the grub menu
[19:19:41] <Valen> edit the kernel command and delete it
[19:19:43] <Valen> then boot
[19:19:52] <Valen> it'll take it out for one boot
[19:20:22] <Valen> pastebin the output of top and see whats happening
[22:03:48] <scanf> has anyone use emc2 on a usb flash drive?
[22:09:37] <skunkworks> a while ago - I made a persistant usb flashdrive and it worked ok.
[22:09:43] <skunkworks> with ubuntu
[22:09:56] <skunkworks> don't ask me how... ;)
[22:10:00] <jdhNC> I tried to make one, but coudl never get it to boot
[22:11:28] <cstop> how did you set the bios boot order?
[22:15:03] <scanf> skunkworks: how? :)
[22:16:48] <seb_kuzminsky> click on System -> Administration -> Startup Disk Creator ?
[22:18:38] <skunkworks> does that make it persistant? (you can save things and the next boot they are there?)
[22:20:10] <scanf> i should just be able to boot off the CD and install to the usb flash right?
[22:23:14] <seb_kuzminsky> https://help.ubuntu.com/community/LiveCD/Persistence
[22:23:33] <seb_kuzminsky> i've never tried it...
[23:13:23] <Jymmm> unetbutin
[23:22:19] <mikegg> jdNC: try tuxboot
[23:23:02] <Jymmm> http://unetbootin.sourceforge.net/
[23:23:13] <Jymmm> You dropped a bomb on me.. baby!
[23:24:24] <mikegg> yeah, I thought tuxboot would do ubuntu...
[23:24:36] <mikegg> i guess it's just clonezilla and some other things
[23:44:26] <Jymmm> Fiiiiiiiiiiiire
[23:44:52] <Jymmm> http://newpowerinc.com/flash_mp3_player/mp3/Ohio Players - Fire.mp3
[23:44:56] <Jymmm> bah
[23:45:18] <Jymmm> http://newpowerinc.com/flash_mp3_player/mp3/Ohio Players%+-+Fire.mp3
[23:45:32] <Jymmm> http://newpowerinc.com/flash_mp3_player/mp3/Ohio Players%20-%20Fire.mp3
[23:45:38] <Jymmm> http://newpowerinc.com/flash_mp3_player/mp3/Ohio%20Players%20-%20Fire.mp3
[23:45:42] <Jymmm> there we go
