#linuxcnc | Logs for 2015-10-16

Back
[00:00:16] <MacGalempsy> when I click the spindle + button, the abs.spindle.is-negative = true does this mean my encoder count is inverted?
[00:58:56] <Jymmm> http://www.cbc.ca/news/technology/hoverboard-duru-1.3270569
[01:15:58] <MacGalempsy> thats pretty cool
[01:16:30] <MacGalempsy> actually pretty exciting
[01:32:13] <Praesmeodymium> my sister told me... and now i am watching theres 3d printers on project runway
[01:32:18] <MacGalempsy> anyone have an easy way to calculate an encoder count on an unknown encoder?
[01:33:18] <MacGalempsy> i went to encoder.5.counts, but when I tweek the number, the actual counts change per 1 rotation
[02:28:50] <Deejay> moin
[02:29:24] <MacGalempsy> hi dj
[03:31:05] <fenn> spin it 100 times by hand and divide counts by 100?
[03:32:33] <fenn> you'll have to manually tweak the scale parameter anyway in an actual machine, so don't worry too much about the actual number
[03:55:25] <MacGalempsy> YAY!
[03:55:30] <MacGalempsy> finally working!!!
[03:55:50] <MacGalempsy> homed and spinning. OMG!!!
[03:56:06] <Deejay> sounds good :)
[03:57:47] <MrSunshine> hmm i should order a breakoutboard with 0-10V out
[03:57:56] <MrSunshine> so spindle turns on and off automaticly ..
[03:58:39] <MacGalempsy> Well, im so happy, I could give EVERYONE A HUG
[03:59:19] <MacGalempsy> well, i better go to bed before something breaks and I cry
[03:59:25] <MacGalempsy> have a good day everyone
[03:59:41] <fenn> quite the emotional rollercoaster
[03:59:59] <fenn> last time we were cut out of his will
[04:01:46] <Deejay> hihi
[04:48:58] <XXCoder> wheee
[04:53:15] <jthornton> morning
[04:53:26] <XXCoder> whats up
[04:53:29] <jthornton> nice it's only 55F here
[04:54:01] <jthornton> heading to Wears Valley Tn on the GoldWing
[04:56:56] <XXCoder> cool
[04:57:09] <XXCoder> chilling now, tired from work lol
[04:58:16] <jthornton> it's 4:30am here lol
[04:58:37] <XXCoder> later than me then lol 2:34 am
[05:03:43] <jthornton> it's warmer than I thought this morning, no long johns needed for the trip
[05:04:34] <XXCoder> long johns at 55f? quite a overkill lol
[05:05:27] <jthornton> yesterday they said low 40's
[05:06:10] <jthornton> so I had them out
[05:06:22] <XXCoder> lol ok
[05:06:30] <XXCoder> I dont wear em even at subzero
[05:06:49] <XXCoder> that is, if it ever happens here. it is quite rare
[05:07:24] <XXCoder> I have thick "extra cold" coat I hyave worn a total of around 10 times in 10 years
[05:07:37] <jthornton> I don't wear chaps on the bike so low 40's I wear long johns
[05:07:46] <jthornton> only had to do that a couple of times
[05:07:55] <XXCoder> ahh bicycling
[05:08:00] <XXCoder> makes more sense
[05:08:03] <jthornton> motorcycle
[05:08:08] <XXCoder> or that lol
[05:09:34] <XXCoder> working on anything cnc-wise?
[05:10:43] <jthornton> the dxf to G code converter
[05:10:53] <XXCoder> ahh yea
[05:10:59] <XXCoder> wanted to ask if its 2d only or 3d
[05:11:17] <jthornton> it just 2d with dxf files
[05:11:33] <XXCoder> still quite useful
[05:12:12] <jthornton> yea, I'm doing it for two things, brain exercise and plasma
[05:14:24] <jthornton> http://gnipsel.com/images/GL1800/GL1800%2001.jpg
[05:14:55] <jthornton> that's what we are riding today
[05:15:05] <XXCoder> cool
[05:16:00] <jthornton> we haven't been able to do things like this for a long while
[05:21:26] <jthornton> I just quoted a CNC nailing machine to build special pallet tops lol
[05:21:54] <XXCoder> cnc nailing machine? interesting
[05:22:47] <jthornton> the nailing table moves in the Y axis and the nail guns move in X and Z
[05:24:07] <XXCoder> interesting
[05:24:08] <XXCoder> http://www.amazon.com/gp/product/B002YVD3KM?psc=1&redirect=true&ref_=oh_aui_detailpage_o01_s00
[05:24:20] <XXCoder> I plan to try see if it works as pendant lol
[05:24:27] <XXCoder> arrows for x, y
[05:24:39] <XXCoder> A, B for Z I guess
[05:25:19] <jthornton> cheap enough to try
[05:25:30] <XXCoder> maybe bit different, select for axis, A, B for one step in pos or neg
[05:26:03] <XXCoder> really its for nes emu fun
[05:26:10] <XXCoder> but same time curious lol
[05:27:26] <jthornton> I use this on my plasma http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Simple_Remote_Pendant
[05:27:57] <jthornton> left stick is XY and right is Z 1,2,3 are speeds
[05:28:03] <XXCoder> nice nice
[05:28:19] <XXCoder> honestly I would buy cheap one from aliexpress
[05:28:24] <XXCoder> one with clicky wheel
[05:29:30] <C_P-Away> Sent letter to walmart co on new-egg order! ohhh boy you guys get ready!
[05:30:31] <XXCoder> http://www.aliexpress.com/item/Free-Shipping-60mm-Manul-Pulse-Generator-for-CNC-machine/584411423.html?spm=2114.01020208.6.7.YndqOL&s=p
[05:30:34] <XXCoder> interesting
[05:31:04] <jthornton> I use that on my CHNC and BP
[05:31:19] <jthornton> with two selector switches for speed and axis
[05:31:45] <jthornton> that's a good price for MPG
[05:31:52] <XXCoder> ...
[05:32:05] <XXCoder> pulse generator manual generates lkots sexy underwear produces in results
[05:32:10] <XXCoder> gonna add -sexy
[05:33:01] <XXCoder> cheapest one of all http://www.aliexpress.com/item/5V-CNC-numerical-control-manual-pulse-generator-RGT600-001-100B-5L-hand-CNC-machine-tools-encoder/32452353366.html
[05:36:28] <XXCoder> got idea
[05:36:35] <jthornton> well time to get dressed and roll the wing out of the garage and get it packed up
[05:36:45] <XXCoder> nes controll select can be switch between constant and one step a time
[05:37:01] <XXCoder> and start for .01, .001, .0001 mode
[05:37:22] <jthornton> that should work
[05:45:10] <XXCoder> yeah
[06:45:25] <jthornton> see you guys later
[07:21:06] <ganzuul> w00t!
[07:21:22] <ganzuul> I got accepted to the metalworking program!
[07:21:44] <ganzuul> I start in February. :D
[08:55:25] <ktchk> Hello any user of debian 3.2.4 rt linuxcnc??
[09:05:32] <cradek> ktchk: what's your question?
[09:06:44] <ktchk> I am using a computer using ubuntu 10.04 the latency is ok, change to debian 3.2.4 rt the latency is 10 times. not too good.
[09:07:03] <ktchk> Is there any way to lower the latency??
[09:07:38] <cradek> do you mean the debian rt-preempt kernel?
[09:07:52] <cradek> you could try rtai instead
[09:08:07] <ktchk> yes 3.2.4 kernel
[09:08:08] <cradek> some kernels work better on some hardware and it's pretty unpredictable
[09:08:34] <ktchk> the new 3.4.9 is more bad
[09:09:15] <cradek> you could boot the linuxcnc debian+rtai live cd and check latency on that
[09:09:47] <ktchk> yes I did that on the same hardware the best is 10.04
[09:10:00] <cradek> which is best doesn't matter; you only need good enough
[09:10:08] <cradek> are you using software stepgen?
[09:10:34] <ktchk> yes parellel port
[09:14:07] <cradek> then you probably want < 30000 or so
[09:14:49] <ktchk> way over 30000
[09:15:14] <cradek> did you change any hardware?
[09:15:28] <ktchk> no thesame
[09:16:23] <ktchk> I have 2 hard disk and installed all 3
[09:20:26] <cradek> what's the result of the linuxcnc rtai kernel on debian?
[09:21:45] <ktchk> I have to boot again and write it down and will come back af=gain....
[09:21:50] <ktchk> see you
[09:29:12] <MacGalempsy> morning
[09:45:31] <ktchk> smp prempt RT debian 3.2.68-1+deb7u5 max jitter 34019 micro sec but system will stop respond after some click of muse
[09:46:47] <ktchk> smp preempt debian 3.4.55=linuxcnc 12370883 micro sec other programs working fine
[10:12:43] <Erant> ktchk: Did you try any boot-args?
[10:13:00] <ktchk> no
[10:13:19] <Erant> I had little trouble getting my shitty Dell system down to sub 30us.
[10:13:35] <ktchk> is it help?
[10:13:46] <Erant> So wait, you're having 12370883ns jitter?
[10:14:00] <ktchk> yes
[10:14:19] <Erant> That sounds like SMM is ruining the party, but I don't get why it wouldn't under Ubuntu.
[10:14:44] <Erant> Is this a single core, dual core, quad core?
[10:15:07] <ktchk> my 10.04 is 7452 ns
[10:15:22] <Erant> That's really good. Why don't you use that one?
[10:15:43] <ktchk> single core centrino
[10:16:32] <ktchk> ubuntu is not supporting 10.04 any more, have to move some time.
[10:17:09] <Erant> Have you tried the 2.7 release of LinuxCNC?
[10:17:26] <CaptHindsight> ktchk: use whatever works, there is no one size fits all real time kernel
[10:17:32] <ktchk> yes I am running 2.7 linuxcnc
[10:17:57] <Erant> Yeah, I'm not sure why you want to move off 10.04.
[10:18:14] <Erant> Is there a package you desperately need that isn't available on 10.04?
[10:18:28] <ktchk> can not have to run also heekscnc 0.18 at the same time
[10:20:57] <Erant> I'm not entirely sure, but have you tried updating to the latest Ubuntu and nabbing the RTAI kernel?
[10:21:27] <Erant> smp preempt debian 3.4.55=linuxcnc 12370883 micro sec other programs working fine
[10:21:32] <Erant> That's not an RT kernel
[10:21:39] <Erant> No wonder the jitter's so high.
[10:22:13] <ktchk> the kernal is from linuxcnc.org
[10:23:05] <cradek> I suspect he means this one: Linux emc 3.4-9-rtai-686-pae #1 SMP PREEMPT Debian 3.4.55-4linuxcnc i686 GNU/Linux
[10:23:16] <cradek> if so, yes it is an rtai kernel
[10:24:05] <ktchk> the kernel name is from uname -v
[10:27:04] <ktchk> the micro sec is ns
[10:29:26] <ktchk> which boot-args will help??
[10:29:55] <Erant> ktchk: So, there's a couple you can try
[10:30:08] <ktchk> Please
[10:30:19] <Erant> ktchk: Seeing as you have a single core system, isolcpus=0
[10:30:34] <Erant> Then nohlt and idle=poll
[10:30:54] <Erant> idle=poll will make your system run pretty hot, but it's a decent test to see if it's the c-states causing problems.
[10:31:23] <ktchk> put them all in grub default?
[10:31:50] <Erant> Eh, just try them manually first.
[10:32:12] <ktchk> ok will come back later
[10:32:19] <ktchk> bye
[10:32:20] <Erant> Boot into grub, press e on the LinuxCNC line, go to the linux line
[10:32:27] <Erant> Press Ctrl+E and add them.
[10:52:07] <enleth> Is there a significant rigidity difference between a hollow and a solid spindle?
[11:00:03] <archivist> not really, mostly it is the diameter that defines it
[11:00:42] <archivist> frame is usually the significant problem
[11:03:39] <archivist> a dti placed in the right place can show how flexible a machine is, also a machinists level
[11:03:49] <ssi> morn
[11:04:03] <archivist> moan
[11:11:21] <ktchk> isolcpus=0 nohlt idle=poll no help to debian 3.4.55 and 3.2.68
[11:11:34] <ktchk> Any more args??
[11:17:44] <CaptHindsight> ktchk: isolcpus was broken in that version of RTAI anyway, it might help or it might not and it's a crapshoot what setting might work, plus you have a single core cpu so it shouldn't even matter if it is used
[11:24:02] <ktchk> audio people have some args to lower latency
[11:26:00] <ssi> Wolf_: alive?
[11:26:29] <archivist> every application is fighting for their bit of your latency, not giving a f*** how it affects other apps
[11:26:31] <Wolf_> yo
[11:27:47] <ssi> https://octopart.com/bom-lookup/RMRKGPe0
[11:28:48] <Wolf_> nice
[11:29:49] <Wolf_> that RJ45 from TE?
[11:30:00] <ssi> ya
[11:30:22] <Wolf_> for the first ones, suggest TE sample request :D
[11:30:29] <ssi> hah that's a good idea
[11:30:37] <ssi> I wonder if I can sample the igbt module and the arms too
[11:30:40] <ssi> that'd cut a PILE of cost out
[11:30:50] <Wolf_> maybe…
[11:31:36] <Wolf_> I have a big ass box of stuff from TE, sample order that has just about all the VW connector pins in it
[11:37:17] <CaptHindsight> archivist: what made things worse is the video driver, X and kernel devs no longer caring about adhering to any rules to insure real time
[11:37:52] <archivist> CaptHindsight, yup terrible design
[11:38:05] <CaptHindsight> they started writing directly to hardware and as long as their app works F*** everyone else
[11:38:15] <archivist> wasting your tome with fading bling
[11:38:18] <archivist> time
[11:39:21] <CaptHindsight> memleak has kept his RTAI fork going up until 3.18 or 3.16 just for Linuxcnc
[11:40:04] <CaptHindsight> but 4.x kernel have changed so much it's going to take a complete rewrite if it's even possible
[11:40:43] <archivist> at least the kernel devs have set someone on for the realtime problem recently
[11:41:25] <archivist> but dunno how long before anything gets fixed
[11:41:55] <CaptHindsight> you can unscrew a light bulb, not sure about the kernel
[11:42:23] <archivist> could we moan at them to resurrect 10 and backport security etc
[11:43:01] <archivist> I really hate the cut everyone off attitude to older versions
[11:43:27] <CaptHindsight> <-- puts back on tinfoil hat and returns Gentoo with custom kernels
[11:44:26] <archivist> moooooore madness
[11:44:56] <CaptHindsight> maybe once the madness ends with android and useless mobile desktops there will be a return to sanity
[11:45:25] * Jymmm hits CaptHindsight on the head with a sledge hammer.... BONGGGG Did you forget about the steel plate in your head? No need for tin foil hat
[11:45:37] <CaptHindsight> or maybe they will add a registry to Linux
[11:46:28] <Jymmm> Are HDMI cable not all the same? HDMI to mini HDMI, some with ethernet (built in?) ?
[11:47:34] <Jymmm> Example: http://www.walmart.com/ip/Tripp-Lite-High-Speed-with-Ethernet-HDMI-to-Mini-HDMI-Cable-6-Universal/23316695
[11:47:52] <CaptHindsight> same connectors just some new pins are used
[11:48:22] <Jymmm> where does the "ehternet" come into play here?
[11:48:59] <CaptHindsight> https://en.wikipedia.org/wiki/HDMI#Version_1.4
[11:49:11] <CaptHindsight> adds an HDMI Ethernet Channel (HEC),
[11:49:35] <CaptHindsight> + audio return channel (ARC)
[11:50:34] <CaptHindsight> IIRC it's a one pin difference
[11:50:51] <CaptHindsight> the new spec uses a previously unused pin
[11:51:17] <Jymmm> Hmmm, still no clue where the data comes into play here. Are they just explaioting ehternet between the two connected devices, or does it actually get an ip address too?
[11:55:45] <CaptHindsight> http://www.hdmi.org/manufacturer/hdmi_1_4/hec.aspx
[11:56:01] <Jymmm> thanks
[11:57:01] <Jymmm> Just ordered one of these http://www.walmart.com/ip/44656207
[11:59:19] <Jymmm> benefit of a tablet, and more functional that a roku
[11:59:33] <Jymmm> xmas gift
[12:31:31] <CaptHindsight> Walmart also has the $150 ARM Chromebooks that are are easily reloaded with Linux
[12:32:23] <Loetmichel> link?
[12:32:45] <CaptHindsight> think they use coreboot or some form of uboot
[12:33:42] <CaptHindsight> Loetmichel: http://www.walmart.com/ip/Hisense-Chromebook-11.6-Quad-Core-Processor/44389793
[12:34:21] <CaptHindsight> the google chromebook wiki has all the rooting/unlocking info
[12:34:38] <CaptHindsight> also source to the bootloader and EC firmware
[12:35:25] <Loetmichel> nice one
[12:35:51] <CaptHindsight> https://chromium.googlesource.com/chromiumos/platform/ec/
[12:35:59] <CaptHindsight> EC firmware source ^^^
[12:39:16] <CaptHindsight> https://www.chromium.org/chromium-os/developer-information-for-chrome-os-devices
[12:44:19] <CaptHindsight> heh someone used Archlinux http://archlinuxarm.org/platforms/armv7/rockchip/hisense-chromebook-c11 I wonder how many minutes it was a stable release
[12:51:20] <archivist> only in the USA https://www.google.com/patents/US5551495 they think they can patent making stairs using PC-APT for the CAM
[12:55:45] <Jymmm> archivist: If they can patent a person's DNA, they can patent anything.
[12:56:19] <Jymmm> Gives new meaning to "Your ass is mine biotch!"
[12:59:56] * Loetmichel s company had fun with siemens lately
[13:00:39] <Loetmichel> we do a certain shielded USB flashdrive drawer since 2006 for a french customer in the 100s
[13:01:28] <Loetmichel> 2008 the german BSI said that they had a similar looking device in their measurement chamber but with a "siemens" print on it...
[13:02:20] <Loetmichel> last year we got a mail from siemens lawyers... "cease and desist to produce our 2008 patented drawer and pay royalities!"
[13:02:25] <ssi> resold? or copied
[13:03:56] <Loetmichel> boss lhao. our lawyer also qoute:"and when they dont shut up immediatly they'l get a nice royalities laqwsuit ont top of the patent negation case!";)
[13:04:04] <t12> put an adc on it and call it an antenna
[13:04:15] <Loetmichel> blatantly copied to the last bolt ;)
[13:04:20] <t12> lol
[13:05:23] <Loetmichel> it seems thats what you get if your law dept dont knows where the design dept steals ;)
[13:06:34] <_methods> anyone in here ever get quotes on the cost of eltool cat40 right angle coolant driven spindle heads before?
[13:07:26] <ssi> "coolant driven"?
[13:07:45] <_methods> yeah it uses through spindle coolant to drive the tool
[13:07:53] <ssi> weird
[13:07:59] <_methods> http://www.eltool.com/products/angleheads.aspx
[13:08:10] <ssi> ohhh I see
[13:08:18] <ssi> it's like a pneumatic angle drill but coolant pressure driven
[13:08:22] <_methods> yeah
[13:08:28] <ssi> that's pretty awesome
[13:08:44] <_methods> yeah just found one in a tool cart and i was gonna put it on ebay
[13:08:45] <ssi> are they for drilling only, or can they mill?
[13:08:49] <ssi> doesnt' look like it'd mill well
[13:08:52] <Loetmichel> if your coolant pump is big enough ;)
[13:08:55] <_methods> i guess i'll just put it up there for like $5k
[13:08:57] <_methods> see what happens
[13:10:38] <_methods> can't find a price for them anywhere
[13:10:48] <_methods> and all mechanical right angles on ebay
[13:11:09] <_methods> i think they run about $12k new
[13:11:25] <ssi> dang
[13:11:53] <_methods> the old invoice was with it and it was $24k for 2 of them
[13:11:57] <_methods> but i only found this one
[13:12:31] <_methods> went to quote some stuff and make some drawings at this place and they were like hey you want some cat 40 tooling
[13:12:47] <_methods> got 2 carts full of cat40 holder for $2500
[13:12:51] <_methods> and the carts
[13:13:04] <_methods> that eltool was in there lol
[13:13:11] <ssi> that's awesome
[13:13:15] <_methods> yeah
[13:13:31] <_methods> they're gettin rid of all their old cat40 and cat50 stuff
[13:13:36] <_methods> going to all heatshrink tooling
[13:13:47] <_methods> some of it's brand new too lol
[13:17:52] <_methods> https://www.dropbox.com/s/hb1g8purp0zhlte/2015-10-16%2013.50.44.jpg?dl=0
[13:18:05] <ssi> did I tell you I got a 2 post car lift?
[13:18:06] <_methods> https://www.dropbox.com/s/cpm2vgrzijj2vm6/2015-10-16%2013.50.34.jpg?dl=0
[13:18:11] <_methods> ah nice
[13:18:12] <ssi> I have to figure out how to get it installed :P
[13:18:20] <_methods> screw it to the floor lol
[13:18:22] * Wolf_ suggests bolts
[13:18:40] <ssi> _methods: dang that's a lucky lucky find
[13:18:41] <_methods> https://www.dropbox.com/s/jt68bj3j77aqyzc/2015-10-16%2013.50.54.jpg?dl=0
[13:18:45] <_methods> yeah man
[13:19:01] <_methods> they left all the tools in and everything lol
[13:19:14] <ssi> _methods: hey do yall have a rotary on one of your lasers?
[13:19:23] <_methods> nope
[13:19:25] <_methods> i wish
[13:19:30] <ssi> yeah I wish too :(
[13:19:34] <_methods> could do some cool stuff with that
[13:19:47] <ssi> this work that I need done, we need to profile a bunch of 2" 6061-T4 0.035 wall tube
[13:20:02] <_methods> round or sq?
[13:20:04] <ssi> round
[13:20:13] <_methods> you coping?
[13:20:16] <ssi> yea
[13:20:24] <_methods> cant do it on a regular mill?
[13:20:26] <ssi> hang on, gonna screenshot you a part
[13:20:29] <ssi> yeah probably can
[13:20:30] <_methods> i mean we have a 4th axis
[13:20:36] <_methods> i could mill it for you
[13:21:31] <ssi> https://pbs.twimg.com/media/CRdP4uOXAAAvyxL.jpg:large
[13:21:42] <ssi> the flat flange is an example of one of the things I was gonna have you quote
[13:21:51] <ssi> they get counterbored on the backside for the weld
[13:21:56] <_methods> cool
[13:21:57] <ssi> some of them have o-ring grooves trepanned
[13:22:01] <_methods> yeah we can do that all day
[13:22:16] <ssi> the fittings on the other two tubes are adel/wiggins band-clamp fluid couplings
[13:22:17] <_methods> those are just regular copes we can do for you too
[13:22:20] <ssi> off the shelf parts
[13:22:38] <ssi> ok some of the parts have flat patterns that get rolled into cones for diameter transitions
[13:22:48] <_methods> are you rolling?
[13:22:51] <ssi> I dunno yet! :)
[13:22:54] <_methods> or you need us to roll
[13:22:55] <_methods> ok
[13:22:57] <ssi> I know all the welding is gonna be done by a guy here
[13:23:03] <ssi> it's all gotta be certified welding and xrayed
[13:23:10] <_methods> https://www.dropbox.com/s/f79voxyfa9z711u/2015-10-16%2013.51.37.jpg?dl=0
[13:23:17] <_methods> oh our welders suck ballz
[13:23:22] <_methods> so you might want to do it yourself
[13:23:26] <ssi> yea we have a guy
[13:23:40] <_methods> i wouldn't trust our welders with a plastic bag full of shit
[13:23:41] <ssi> the other thing that sucks about this job is it's like 35 different parts, with a part being like that complete manifold
[13:23:51] <ssi> each "part" is actually an assembly of 3-6 parts
[13:23:59] <ssi> and the total qty of sets of assemblies is like 9
[13:24:04] <ssi> so lots of one-off :(
[13:24:19] <ssi> AND they want to pay to have ONE made, then do a run of 8
[13:24:22] <_methods> well that's waht we do so it's cool
[13:24:26] <_methods> we're a job shop
[13:24:32] <ssi> and they don't seem to understand that it's way cheaper to cut all the flanges in one run
[13:24:46] <ssi> cool... as soon as I get pdfs of the engineered drawings I'll get them to you
[13:24:53] <ssi> I have some printouts but they haven't sent me the digitals yet
[13:25:17] <_methods> what kind of tolerances?
[13:25:24] <ssi> tolerances are mostly reasonable
[13:25:27] <ssi> they're called out everywhere
[13:25:48] <ssi> these are proper engineered drawings, they've been certified by a faa designated engineering rep
[13:25:54] <_methods> ahh good
[13:25:58] <_methods> thank god no napkins lol
[13:25:59] <ssi> :)
[13:26:14] <ssi> the duplication quality isn't great on all of them though
[13:26:16] <ssi> tehy're circa 1988
[13:26:31] <_methods> ah cool so it's an older part
[13:26:41] <_methods> i like that
[13:27:17] <ssi> https://pbs.twimg.com/media/CRdRO3_WEAAGoKT.jpg:large
[13:28:16] <_methods> yeah lookos good
[13:29:19] <_methods> easy peasy
[13:29:24] <ssi> cooool
[13:29:30] <ssi> as soon as I have the pdfs I'll start getting them to you
[13:29:38] <ssi> I'm trying to get them to give me a priority on which parts they want to start with also
[13:29:46] <_methods> sounds good we're kinda slow right now so i'll take it
[14:13:52] <Erant> Has anyone used the National Instrument PCI cards with Comedi for step generation?
[14:14:05] <Erant> There's some passing references to it (old) on the Wiki, but nothing really new.
[14:17:16] <archivist> there was a user who used linuxcnc with comedi but dont think he used NI cards for the motor driving
[14:18:10] <archivist> or I dont remember that, was using comedi for the rest of his data capture I think
[14:19:09] <Tom_itx> Jymmm,
[14:25:42] <ktchk> I reload the debian 3.4.55 and linuxcnc 2.6.10, the latency goes down to <4000 ns
[14:26:06] <cradek> yay
[14:51:52] <ganzuul> https://www.mazakeu.com/quick-turn-nexus-250-ii-msy/
[14:52:13] <ganzuul> my school has one of these
[14:55:59] <_methods> those are nice lathes except the mazak part lol
[14:58:24] <ganzuul> gotta earn it to get to work on it... I'm gonna do it.
[15:08:22] <aventtini6> hi guys
[15:10:10] <ganzuul> harro
[15:13:04] <aventtini6> guys still having problems on arcs
[15:13:29] <aventtini6> the machine is running real slow on cycles
[15:13:56] <aventtini6> end some another shapes
[15:14:08] <aventtini6> why is this happening
[15:14:32] <Wolf_> pic?
[15:14:46] <Wolf_> more info needed
[15:15:17] <aventtini6> for example when you make a rectagle with radius corners
[15:15:22] <aventtini6> mill is 6mm
[15:15:35] <aventtini6> radius is 10 mm
[15:15:45] <aventtini6> and it gos in steps
[15:15:58] <Wolf_> is the straight line distance correct?
[15:16:00] <cradek> share example gcode of a rectangle?
[15:16:03] <aventtini6> yes
[15:16:18] <aventtini6> i cant give gcode im home
[15:16:37] <Wolf_> can it do circles?
[15:16:38] <aventtini6> it slows on g0 g1 g2 and 3
[15:16:43] <aventtini6> yes
[15:17:06] <Wolf_> I would blame the cam soft then
[15:17:10] <cradek> this is really not enough information
[15:17:11] <aventtini6> but if i go fast its not a perfect circles
[15:17:40] <cradek> it's not clear what problem you're reporting OR what you're doing to see it
[15:17:42] <aventtini6> if i slow down the feed it make corect
[15:17:53] <aventtini6> o sec let me find a video
[15:18:03] <cradek> if I had to guess, you have incorrectly low acceleration settings
[15:18:17] <cradek> yes video would be great
[15:18:26] <Wolf_> could be loss of steps, stiction, gibs, acceleration, and about 10 other things
[15:19:47] <aventtini6> im uploding
[15:19:49] <aventtini6> 10 sec
[15:20:39] <aventtini6> https://www.youtube.com/watch?v=P8_hKA4cGGc
[15:20:57] <aventtini6> in straight line it gos 2000mm/m
[15:21:08] <aventtini6> this is a 16 mm endmill
[15:22:22] <aventtini6> max speed is 80
[15:22:39] <aventtini6> and acceleration is 60 and tested to 150
[15:23:06] <aventtini6> linuxcnc is 6.10
[15:23:23] <Tom_itx> that version hasn't been written yet
[15:23:24] <cradek> acceleration setting is only 60mm/sec^2?
[15:23:38] <aventtini6> yes
[15:23:57] <Tom_itx> 2.6.10?
[15:24:01] <aventtini6> yes
[15:24:04] <cradek> that will really limit any non-straight path
[15:24:24] <cradek> make sure your acceleration is as high as possible, and update to 2.7
[15:24:32] <aventtini6> the servo manual it was 5000mm/m
[15:24:34] <aventtini6> max
[15:24:54] <aventtini6> so in max velocity i have 5000mm
[15:25:06] <cradek> I don't understand mm/m
[15:25:15] <aventtini6> mm per minute
[15:25:24] <Tom_itx> he's merican
[15:25:34] <Tom_itx> thinks ipm
[15:25:38] <Tom_itx> not mpm
[15:25:42] <aventtini6> sorry did no know
[15:25:43] <cradek> Tom_itx: no, I think m = meter
[15:25:47] <Tom_itx> oh
[15:25:52] <Tom_itx> m=mile
[15:25:56] <cradek> oh stop
[15:26:00] <ssi> mm/m is dimensionless
[15:26:00] <Tom_itx> :)
[15:26:00] <aventtini6> :)))
[15:26:09] <aventtini6> machine is metric
[15:26:16] <aventtini6> to be clear
[15:26:27] <Wolf_> millimeters divided by meters?
[15:26:41] <cradek> he meant max velocity 5000 mm/min
[15:26:44] <aventtini6> mm per minute
[15:26:53] <Wolf_> I know ;P
[15:27:13] <cradek> equals 83 mm/sec
[15:27:38] <cradek> then the correct acceleration should probably be 1000+ mm/sec^2
[15:27:56] <aventtini6> for example i use 800mm/m on 1 machine and the second machine same files works slow
[15:28:05] <aventtini6> i have tryed to make any pid
[15:28:14] <aventtini6> and it dont want more the 1,3
[15:28:16] <Tom_itx> you might have to retune slightly if you update to 2.7 but it's worth it
[15:28:36] <aventtini6> update is no problem
[15:28:44] <cradek> wait am I correct you have 83 mm/sec velocity and 60 mm/sec^2 acceleration?
[15:28:54] <aventtini6> im using liniar scales
[15:29:02] <aventtini6> yes
[15:29:10] <aventtini6> i dont have the hal
[15:29:16] <aventtini6> and ini sorry
[15:29:19] <aventtini6> right now
[15:29:20] <cradek> do you realize it takes 1.4 seconds to get up to speed and another 1.4 seconds to stop again?
[15:29:30] <cradek> that is grossly misconfigured
[15:29:50] <aventtini6> i have tryed 80 to 250
[15:29:57] <cradek> you need to fix those settings and probably tune the pid
[15:30:12] <cradek> I'm guessing it should be 1000+ mm/sec^2 acceleration
[15:30:17] <aventtini6> i have tryed and p is max 1,3
[15:30:32] <cradek> I don't understand
[15:30:38] <cradek> tryed what
[15:30:40] <aventtini6> PID
[15:30:46] <aventtini6> P
[15:30:50] <aventtini6> is 1,3
[15:31:05] <cradek> what happens when P is more than 1.3?
[15:31:08] <aventtini6> if i put more the that is oscilate realy bad
[15:31:26] <cradek> then scaling is probably wrong, or I,D,FF settings are wrong
[15:31:35] <aventtini6> are all 0
[15:31:42] <cradek> yes that's wrong
[15:31:46] <aventtini6> same thing hapends
[15:32:04] <cradek> it could be your velocity loop oscillating too
[15:32:29] <cradek> if it's the position loop oscillating, some D will help it not oscillate
[15:32:49] <cradek> if velocity mode, most of your pid output should come from FF1 which should be 1.0 if your scaling is correct
[15:34:29] <aventtini6> let me psot the ini
[15:34:39] <XXCoder> bow
[15:34:52] <XXCoder> whats up all
[15:37:09] <aventtini6> working
[15:37:11] <aventtini6> :))
[15:37:42] <XXCoder> cool. for a change I'm not doing overtime today
[15:37:53] <XXCoder> need to do stuff, plus bit tired of ot lol
[15:48:14] <aventtini6> http://pastebin.ca/3199909
[15:48:19] <aventtini6> done
[15:49:34] <cradek> FERROR = 10 MIN_FERROR = 5 MAX_VELOCITY = 85 MAX_ACCELERATION = 30
[15:49:41] <cradek> you need to tune and fix these
[15:50:03] <cradek> this is terribly misconfigured and untuned
[15:50:32] <cradek> it takes 2.8 seconds to get up to rapid and another 2.8 seconds to stop again
[15:50:50] <cradek> 5.6 seconds per move
[15:51:47] <aventtini6> so
[15:52:16] <aventtini6> i lower the f error lower it dont work
[15:52:23] <cradek> that's because it's untuned
[15:52:25] <aventtini6> it give flowing error
[15:52:29] <cradek> yes
[15:52:36] <cradek> you need to fix that
[15:53:32] <cradek> bbl
[15:53:36] <aventtini6> im running liniar scales with tacho
[15:54:36] <aventtini6> and if i test with the calibration i cant get any better then now
[15:54:53] <aventtini6> if there is a mecahincal problem on the ballscrew
[15:57:48] <aventtini6> for exampe FERROR must me lower then this ?
[15:57:53] <aventtini6> lower the 10 ?
[15:58:31] <cradek> FERROR should be maybe 20 encoder counts, whatever that is
[15:59:05] <cradek> MIN_FERROR maybe 10 encoder counts
[15:59:27] <cradek> proper tuning should be within about 5 counts at all speeds
[16:00:19] <aventtini6> if its linair ?
[16:00:58] <cradek> if there is any looseness between motor and scale this setup can be hard to tune
[16:01:17] <cradek> check for backlash everywhere, ballnut, ballscrew mounts, pulleys/belts
[16:01:23] <cradek> gibs
[16:01:36] <aventtini6> that i was afraid
[16:01:40] <aventtini6> in fact
[16:02:50] <aventtini6> motor is on drive belt
[16:03:18] <aventtini6> and if the ballscrew has some 0,04mm backslash
[16:03:23] <aventtini6> then its a problem
[16:03:37] <aventtini6> i cant understand why the P is so small
[16:07:25] <cradek> backlash between motor and feedback will cause oscillation
[16:07:48] <cradek> feedback directly on the motor is easier to tune
[16:08:07] <aventtini6> yes i know
[16:08:23] <aventtini6> this is the first problem
[16:08:58] <aventtini6> let me show you some UK made milling machine with fiber optics on a win95
[16:10:03] <cradek> bbl
[16:10:26] <aventtini6> http://imgur.com/a/SLftB
[16:11:03] <aventtini6> how do you rettrofit this
[16:11:05] <aventtini6> :))))
[16:11:30] <aventtini6> its a bridgeport type milling machine with 3 axis servo
[16:11:48] <aventtini6> i was thinking 1 week and i convert it
[16:12:05] <aventtini6> but surprise surprise
[16:12:07] <aventtini6> :))
[16:12:15] <aventtini6> no linuxcnc
[16:27:45] <MacGalempsy> do you guys recommend an optional stop button?
[16:29:02] <XXCoder> not e-stop?
[16:29:27] <MacGalempsy> estop too
[16:29:33] <MacGalempsy> this is for a m98
[16:29:38] <XXCoder> if not emergacy then dunno you do have time to just work pc itself
[16:29:50] <XXCoder> isnt optional stop m01
[16:29:52] <MacGalempsy> not it is for trial maching
[16:30:05] <MacGalempsy> well optional stop m01 is only at the end of the program
[16:30:24] <XXCoder> lol ok
[16:30:27] <MacGalempsy> but if you use a / with fanuc controls the optional stop button will skip the code
[16:31:04] <MacGalempsy> so say you set 4 trys to trial machine and you get it after one, you can hit the optional stop button to skip the other thee tries
[16:31:22] <XXCoder> interesting, never used m98
[16:32:40] <MacGalempsy> doesnt look like linuxcnc has m98
[16:52:05] <Tom_itx> it's got subroutine calls instead
[16:53:01] <Tom_itx> MacGalempsy, i think pause/continue would server the same purpose as op stop
[16:54:19] <MacGalempsy> Tom_itx: they seem similar, but with the op stop switch it will automatically skip instead of pause
[16:55:05] <Tom_itx> i thought you meant to continue after a programmed op stop
[16:55:38] <Deejay> gn8
[16:55:41] <Tom_itx> i guess if your style of programming requires one i'd add one
[16:56:01] <Tom_itx> switch instead of button..
[16:56:14] <MacGalempsy> the instructor last night was mentioning it was pretty handy for trial machinging
[16:56:27] <Tom_itx> meh.. i suppose
[16:56:29] <MacGalempsy> but surely there are ways around
[16:56:41] <Tom_itx> axis has it on the screen
[16:56:44] <MacGalempsy> mostly for boring bars i guess
[16:57:01] <Tom_itx> go to exciting bars instead
[16:57:12] <MacGalempsy> good plan
[16:57:37] <Tom_itx> Jymmm,
[17:08:42] <MacGalempsy> timmmeh
[17:39:10] <zeeshan|2> tgif :D
[17:44:22] <SpeedEvil> tjpeg
[17:45:09] <Wolf_> tpng
[17:45:41] <zeeshan|2> nurds
[17:46:03] <XXCoder> PNG
[17:46:14] <XXCoder> png is king of all
[17:46:21] <Wolf_> tiff
[17:46:31] <XXCoder> raw bitmap
[17:46:45] <Wolf_> i was going to say raw next lol
[17:48:00] <XXCoder> :)
[17:50:44] <renesis> 22:22 < XXCoder> png is king of all
[17:50:48] <renesis> truth.
[17:51:12] <XXCoder> yeah lossless compression rocks
[17:59:56] <Wolf_> so, whats everyones opinion on audible edge finders? http://amzn.com/B0006J3DOA I’ve never used one…
[18:00:22] <Tom_itx> they're noisy
[18:00:32] <Tom_itx> and take batteries
[18:00:56] <Wolf_> not electronic, spinning style
[18:01:40] <Wolf_> I have a SPI led edge finder, I’m not so impressed with it
[18:01:46] <Tom_itx> i have one of those but it's noiseless
[18:01:58] <Tom_itx> works just fine
[18:03:11] <Tom_itx> if you're gonna get fancy, get a renishaw probe and find your edge automatically
[18:03:20] <Wolf_> yeah, I have a double ended mitutoyo one, but it doesn’t seem to jump off the center very much
[18:03:49] <Wolf_> s much/easy
[18:04:53] <Wolf_> working on the renishaw approach too, just trying to not spend $1000 to find a edge lol
[18:05:14] <XXCoder> just ran 1993 scoast.scr screensaver on my pc
[18:05:18] <XXCoder> works great lol
[18:05:27] <XXCoder> too bad I couldnt ever find registered version
[18:05:43] <XXCoder> linux rocks as I woulnt be able to run it on windows.
[18:06:26] <XXCoder> http://cd.textfiles.com/desktopworks/SCRNSAVE/SCR_FILE/SCOAST.TXT
[18:06:32] <XXCoder> maybe you guys will have better luck.
[18:08:08] <Tom_L> https://www.youtube.com/watch?v=SGuV86Su430&feature=youtu.be
[18:08:13] <Tom_itx> Wolf_, make one ^^
[18:09:03] <Wolf_> conductive?
[18:09:27] <Tom_L> http://tom-itx.no-ip.biz:81/~webpage/emc/probe_index.php
[18:10:12] <Tom_itx> jt TN!!
[18:10:20] <JT-Tn> Hi from East Tn
[18:10:22] <Tom_itx> how's the ride?
[18:10:40] <JT-Tn> 7 hrs at 80+
[18:10:58] <JT-Tn> I was so glad to get off of 40
[18:11:00] <Tom_itx> go to W texas and you'd get passed
[18:11:12] <JT-Tn> Lol
[18:11:37] <Tom_itx> been down 40 a few times..
[18:12:55] <JT-Tn> At least the road up to the cabin is paved
[18:13:09] <Wolf_> hmm neat, no spring on the probe/center thing?
[18:13:43] <Tom_itx> yes
[18:14:23] <JT-Tn> Heading back to the cabin now. Bellies full.
[18:14:40] <Tom_itx> yeah, every trip i've made down 40 was 19-20 hrs.. don't care for those much
[18:15:11] <Tom_itx> have fun
[18:15:57] <Tom_itx> Wolf_, yeah i guess i didn't put a pic up showing the spring
[18:16:34] <Tom_itx> http://fadedbits.com/2011/02/touchprobe/
[18:16:53] <Wolf_> I’ve been trying to catch a renishaw setup but no luck on eBay yet, missed one setup for $150
[18:17:58] <Tom_itx> bbl
[18:19:51] <andypugh> Well, I expected to spend this evening routing out more pattern parts, but then I found I couldn’t fit the blank under the spindle, so I had to make a specially short long tool holder. https://picasaweb.google.com/lh/photo/2gXzbJaTCpJDrK9-t8NXjtMTjNZETYmyPJy0liipFm0?feat=directlink
[18:21:21] <andypugh> Wolf_: My experience is that “Renshaw” probes are a lot cheaper than “Renishaw” ones on eBay.
[18:22:52] <andypugh> Though today, that is not the case. http://www.ebay.com/itm/USED-RENSHAW-RE3-PROBE-WITH-HPRA-ARM-AND-BASE-WOODEN-CASE-INCLUDED-N7-/252078232170?hash=item3ab108926a:g:x54AAOSwyQtV50WH
[18:22:52] <Wolf_> well, I have been eyeing a MP9 renishaw listed at for parts/not working, I might get it as a parts donor
[18:23:12] <andypugh> There are very few ways for one to break.
[18:23:42] <Wolf_> well, mp9 is optical linked head
[18:24:14] <andypugh> I have an MP3. I just took off the inductive stuff and fitted a physical wire connection.
[18:24:31] <andypugh> (Partly because I needed it to be a lot shorter)
[18:24:34] <Wolf_> yeah, that was pretty much the same idea I had
[18:25:10] <Wolf_> or just take the probe end off the TX and mount it on something else
[18:26:17] <andypugh> I did get a bit elaborate, I made a DIY “Magsafe” connector in case I turned on the spindle with it plugged in. https://picasaweb.google.com/lh/photo/8hEPrpkf9lYayE3O0-evI9MTjNZETYmyPJy0liipFm0?feat=directlink
[18:26:43] <Wolf_> nice
[18:27:13] <andypugh> If you scroll forwards and back, you will see that I did take off the transmitter
[18:31:23] <Wolf_> you sure about that? :P
[18:33:36] <XXCoder> magnet connector eh
[18:35:55] <Wolf_> oh, wait yeah pic of just the probe half
[18:43:39] <Wolf_> I have been looking for info on the Renishaw Mi4 box but haven’t been able to find anything
[18:47:38] <andypugh> Unless your probe is going to live in your toolchanger none of the fancy interfaces are relevant.
[18:48:23] <Wolf_> well, if I can keep it wireless for under $150 total its would be neat to do
[18:48:49] <Wolf_> but I can’t find any docs for the Mi4 box
[19:31:35] <JT-Tn> Made it back to the cabin. Dang tough at night.
[19:32:05] <XXCoder> fun
[19:32:58] <JT-Tn> It is now lol sitting in the swing sipping on a cervesa
[19:34:25] <XXCoder> lol
[20:19:02] <trentster> Hi all - I want to build a solid workbench today - I found 12 beams (two-by-four) and I want to make all the parts on a cnc router with a limited work area and wondering if anyone has some suggestions on fixturing or creating some type of template to be able to do this accurately? https://monosnap.com/file/W57SFI0QxESBMNFxDa6DIuA95ON017.png
[20:20:36] <XXCoder> trentster: one way is features in common
[20:20:47] <XXCoder> for example dowel holes that align for flipping
[20:21:28] <XXCoder> that grooves can be used that way
[20:25:40] <jesseg> This hasn't anything really to do with CNC or Linux but I did a crude experiment to machine titanium using salt water and electricity: http://videoflier.com/movies/1445045895890083517272
[20:25:52] <trentster> XXCoder: the problem is the beam will need to be moved down to get to the point where the cut in half point is
[20:26:45] <trentster> it can not be fixtured in place where router head can reach the point to cut in two and the edge where the pocket has to be cut
[20:27:08] <trentster> so it will need to be moved
[20:28:23] <trentster> I guess having a circular saw to cut them first would be handy in this situation. I need to buy one
[20:48:03] <Tom_itx> jesseg, how'd that work out?
[20:53:07] <jesseg> Tom_itx, it worked. As you can see from the video, it wasn't real clean looking because I didn't control my active area at all. However it is very promising because it did effect the removal of quite a bit of titanium with very little effort.
[20:53:23] <jesseg> I think my water flow was also insufficient
[20:53:37] <Tom_itx> how much time did it take?
[20:54:27] <jesseg> I'd have to guess around half an hour of cutting time. More time in all for setup, stopping and inspecting, adjusting, etc. (I just had it set up as a jig screwed down to a block of wood.)
[20:54:38] <jesseg> Didn't want to use salt water as a coolant on my regular mill :P
[20:55:53] <jesseg> Now I want to make a little sherline style mill all out of nylon or HDPE so it's rust proof :P
[21:20:19] <Tom_itx> crickets
[21:25:18] <C_P-Away> Sup al...
[21:26:00] <Tom_itx> crickets
[21:26:12] <MacGalempsy> chirp .... chirp ...
[21:27:30] <MacGalempsy> Tom_itx: you machining crickets? thats some high resolution
[21:43:07] <wzl> hi from Argentina, anybody has experience with spindle orientation?
[21:47:14] <MacGalempsy> sure
[21:47:17] <MacGalempsy> m19?
[21:47:53] <Tom_itx> they make me dizzy watching them
[21:48:16] <Tom_itx> i did it on my sherline for rigid tapping
[21:48:23] <Tom_itx> but not for orienting for tool change
[21:48:46] <Tom_itx> andy p wrote a tool change comp for that that includes orient i think
[21:48:55] <Tom_itx> it's in 2.7 as a comp
[21:49:22] <Tom_itx> even has a cute little demo for it
[21:52:56] <wzl> not really m19. I just purchased a new to me hyundai bt30 tap drill with a broken yaskawa regenerative spindle drive. in short I manage to buy a modern , closed loop, vfd from yaskawa that is capable of drive the motor and there is a option for spindle orientation firmware but is not available at the local dostributor
[21:54:12] <Tom_itx> hmm
[21:54:33] <Tom_itx> well all the smart guys took the night off
[21:54:36] <wzl> the controller is a yasnac x3 and I want to retrofit with lcnc in steps. first using lcnc to simulate the spindle orientation pins for the yasnac controller
[21:55:45] <wzl> so, the question is which is the best way to address this, hal, ladder ...
[21:55:51] <Tom_itx> http://linuxcnc.org/docs/2.7/html/man/man9/carousel.9.html
[21:56:35] <Tom_itx> probably doesn't apply to you
[21:57:00] <Tom_itx> http://linuxcnc.org/docs/2.7/html/man/man9/orient.9.html
[21:57:02] <Tom_itx> might
[21:57:17] <wzl> not at this time
[21:58:58] <wzl> if I'm right I need to run orient by a input pin and set an output at the finish
[21:59:56] <wzl> is as simple or I miss something
[22:00:56] <Tom_itx> not that savvy on orient using external hardware
[22:02:52] <wzl> ok, in your experience, where is the best place to ask this kind of questions
[22:03:21] <Tom_itx> here
[22:03:31] <Tom_itx> but like i said, the smart ones are off tonight
[22:04:38] <Tom_itx> where are you?
[22:04:51] <Tom_itx> mmm argentina.. nm
[22:05:12] <Tom_itx> give it some time.. someone will help you
[22:05:38] <wzl> great! I think that a night out is a better option that a dirty machine anytime :)
[22:05:39] <Tom_itx> 10pm on a friday may not be the best time to ask
[22:08:15] <wzl> thanks a lot, I will install an irc client right now.
[22:39:12] <XXCoder> trentster: HMM
[22:39:51] <XXCoder> Fixture thats wider than machine for step 2, the middle?
[22:46:25] <Sync_> wzl: what's the error with the yawkawa? input overcurrent?
[23:08:56] <wzl> yes f300 input overcurrent
[23:58:46] <XXCoder> LOL http://i.crackedcdn.com/phpimages/photoshop/9/7/0/460970_v1.gif