#linuxcnc | Logs for 2015-03-03

Back
[00:00:06] <dirty_d> yea it definitely is
[00:00:27] <dirty_d> hmm, the holes measure about 0.124"
[00:00:35] <dirty_d> is 3.17mm used as a round metric number?
[00:00:57] <dirty_d> as opposed to 3.175
[00:00:58] <furrywolf> no, but 1/8" is...
[00:01:18] <dirty_d> so it probably really is 1/8" holes on a 23mm circle?
[00:01:29] <furrywolf> it's possibly. lol
[00:01:47] <dirty_d> when i saw "really" im thinking what the guy that designed it wrote down for dimensions
[00:04:10] <furrywolf> or "we want to put a M3 screw in, but want a little wiggle room, and 1/8" bits are really cheap..."
[00:05:57] -!- dramz has quit [Ping timeout: 276 seconds]
[00:06:34] <SpeedEvil> It's clearly 3.14mm
[00:06:44] <SpeedEvil> It's meant to be 10mm circumference
[00:07:46] <tjtr33> just got this: cnc rotary with pneumatic brake and servo motor/tacho and .001degree encoder & sliding sleeve stud pull i told zee about http://ibin.co/1tUPUHeJpc3v
[00:08:20] <tjtr33> i'm amazed what these surplus guys just _cut_ off equipment
[00:09:59] -!- Valen has quit [Quit: Leaving.]
[00:10:49] <SpeedEvil> ow
[00:12:28] <MacGalempsy> sliptonic: do you have a make and model?
[00:12:50] <MacGalempsy> try the waybackmachine?
[00:13:23] <MacGalempsy> tjtr33: that looks pretty cool
[00:13:42] <furrywolf> the sensors I got for my spindle encoder like 5V... but annoyingly you have to drive the LED separately, with your own resistor. since the sensor end requires regulated 5v, I have no idea why they didn't just put a resistor and drive the led off that, other than to make me have a bunch of extra wires to hook up.
[00:16:52] <dirty_d> SpeedEvil, i do need a little clearance, so why not pi?
[00:17:25] kb1bdw is now known as mwortley
[00:18:06] <tjtr33> sliptonic, i would not cnx it to 24Vdc, i'd series a 1.5 or 2K res to it ( swag that 20mA would be ok ) dont trust my numbers till YOU do the math
[00:19:55] -!- syyl_ has quit [Ping timeout: 265 seconds]
[00:27:37] <sliptonic> MacGalempsy, They're made by Sharp circa 1986. There's various numbers but no idea what is model.
[00:29:09] <sliptonic> They're mounted on a little board with a 'blister' on the back so I think they have built-in electronics.
[00:30:16] <tjtr33> sliptonic, does this help? and does the board have a slot for a flat head screwdriver to adjust the position? http://goo.gl/4fFzTl
[00:31:17] <tjtr33> now way to tell whats inside a ptted blister.
[00:31:23] -!- swingley has quit [Remote host closed the connection]
[00:31:25] <tjtr33> potted
[00:33:25] <sliptonic> no adjuster. The gate looks like most of those images but I don't see any that look like the complete assembly. I'll add a res to bring it down.
[00:34:02] <MacGalempsy> sliptonic: looks like the waybackmachine only goes to 1996
[00:34:38] <tjtr33> might be of use ( a gnrc opto interupttor from sharp ) http://sharp-world.com/products/device/lineup/data/pdf/datasheet/gp1a52hr_e.pdf
[00:37:06] -!- mhaberler has quit [Read error: Connection reset by peer]
[00:44:08] <sliptonic> FWIW, they look like this: https://plus.google.com/u/0/photos/yourphotos?pid=6121802918993722386&oid=107717094180262901848
[00:47:20] <tjtr33> looks like a black screen? :) i logged in 2x now and its always a black screen
[00:47:25] <dirty_d> do you guys know if aluminum round stock is usually like +0.005" -0" or just +/- 0.005"?
[00:48:41] -!- mhaberler [mhaberler!~mhaberler@extern-171.stiwoll.mah.priv.at] has joined #linuxcnc
[00:49:36] <dirty_d> i think i wanna use an aluminum shaft instead of steel
[00:49:43] <dirty_d> its overkill either way
[00:49:51] <dirty_d> has to fit in bearings though
[00:54:21] <sliptonic> tjtr33, try this one https://lh4.googleusercontent.com/-zk7PcOwnNDM/VPUDUfOR_BI/AAAAAAAAE6M/pMD9aTE5-2U/w958-h719-no/IMG_20150302_184006.jpg
[00:55:23] -!- rob_h has quit [Ping timeout: 246 seconds]
[00:59:59] <Rab> tjtr33, you were having latency problems with your DC7800? I got mine today, and latency numbers aren't terrible: 41070/38801 after a few hours of glxgears, etc.
[01:00:22] amnesic_away is now known as amnesic
[01:00:47] <Rab> Worth noting that the numbers were actually a lot better when I booted from a USB live image, before installing to an internal SATA drive.
[01:04:22] <dirty_d> using the debian live cd?
[01:04:59] -!- Nick001-shop has quit [Quit: ChatZilla 0.9.91.1 [Firefox 34.0.5/20141126041045]]
[01:05:05] <PCW> Ive got one on order, will try preemt-rt also
[01:05:48] <PCW> replacing the anemic CPU with a e8500 would also probably help
[01:05:52] <Rab> dirty_d, yeah
[01:06:39] <dirty_d> Rab, i noticed that the terminal program on there made my latency go really high when i move the mouse over it
[01:06:41] <dirty_d> very wierd
[01:06:49] <dirty_d> up to 80,000ns
[01:06:51] -!- AR_ [AR_!~AR@24.238.81.234.res-cmts.sth2.ptd.net] has joined #linuxcnc
[01:06:58] <dirty_d> 15,000 with it not running
[01:07:10] <dirty_d> maybe something to do with the framebuffer?
[01:07:30] <dirty_d> xterm doesnt have the saem problem
[01:08:58] <tjtr33> Rab thx, look at what i saw if you like https://www.dropbox.com/sh/n95e1h9qlj09cp2/AAA4g5Td3eGlicFJWh4cRWf4a?dl=0
[01:09:44] <tjtr33> sliptonic, i dont recognize it the opto-interupttor there
[01:12:25] <tjtr33> outside of the bios settings, i turned off some services too, bluetooth, cups and... its all recorded in a log file in that dropbox dir
[01:13:46] <tjtr33> i can try it again w/o the WD SATA drive
[01:14:40] -!- dirty_d has quit [Quit: Leaving]
[01:14:56] -!- patrickarlt has quit [Quit: Leaving...]
[01:14:56] <Rab> tjtr33, cool, when I have time I'l try to reproduce your exact configuration.
[01:15:01] <tjtr33> yeah, i saw 150uS latency at worse
[01:15:30] <tjtr33> i really hope you have better luck and i just goofed something. every bios page is screenshot in that dir
[01:15:41] -!- sumpfralle has quit [Ping timeout: 246 seconds]
[01:15:50] <Rab> Booted from USB flash drive again and numbers are still better. 34559/14441 even after mounting the SATA drive and hitting it pretty hard. I'm not sure what's different.
[01:16:20] <tjtr33> time :) and chance :)
[01:17:11] <tjtr33> i saw wildly diff vlaues using latency-test and latency-histogram, same system same run, same load, just end one and try the other
[01:18:02] <Rab> I'm running 4 instances of glxgears, a bunch of heavy stuff like Gimp/OpenOffice/Glade, and my torture test:
[01:19:05] <Rab> $ for file in $(find /) ; do md5sum $file ; done
[01:19:33] <Rab> Latency numbers jump noticeably when that starts. ^_^
[01:20:38] <Rab> Really puzzled why booting off USB gives better figures, even when doing the find/md5sum thing on the mounted SATA drive.
[01:21:30] <tjtr33> try 1 glxgears from a terminal , the terminal wndo itself no chg to latency, i had 65273/31243 then jiggle the window , i got 139665/90118
[01:22:00] <tjtr33> ps2 mouse and built in video
[01:22:27] -!- sumpfralle [sumpfralle!~lars@erina.opennet-initiative.de] has joined #linuxcnc
[01:23:56] <tjtr33> for you, jiggling a window does not double or triple the latency?
[01:25:43] <Rab> It didn't double or triple, but it did go up to: 35223/19213
[01:26:18] <tjtr33> i concluded that it was video ( my hammering the sata didnt matter, usb mouse vs ps2, audio/no audio none of the had the drama of a window resize or reposition )
[01:26:37] <tjtr33> so, what video rezolution? ( i assume you have vga driver? )
[01:27:13] -!- Connor has quit [Quit: Leaving.]
[01:27:24] <tjtr33> and thanks for looking
[01:27:39] -!- Connor [Connor!~Connor@24.214.127.194] has joined #linuxcnc
[01:28:07] <Rab> 1280x1024@60
[01:28:35] <tjtr33> thx, i think i may have a widescreen format...
[01:29:38] <Rab> Looks like just vesa/fbdev...
[01:30:20] <tjtr33> before i puts with it, i had it running 2 days no with NO TOUCHY , no strees, just the latncy-test. its at 10000/11000 !
[01:31:31] <tjtr33> vid is 1440x900(16:10)@60hz
[01:31:50] <tjtr33> now chg to yours ( i hope i hope i hope )
[01:33:53] -!- mhaberler has quit [Quit: mhaberler]
[01:35:20] <tjtr33> 10871/13282 wont let me get the 4th glxgears but mouse & kbd are snappy.
[01:35:33] <tjtr33> gonna let it cook a while BIG THANKS!
[01:36:13] <tjtr33> wont let me means... i typed it into the terminal, and it did not happen
[01:36:38] <Rab> Wait, so you changed resolution and everything's better now?
[01:37:00] <tjtr33> well its only a few minutes but yes wildly crazy huge differnce
[01:37:47] <tjtr33> i have see it go south before, dont get excited
[01:38:46] -!- deanputney has quit [Remote host closed the connection]
[01:39:21] <tjtr33> 1140x900@60 ->1280x1024@60 is only diff so far
[01:40:15] <tjtr33> 12k/13k now ( and maxd one glxgears then restired it already )
[01:40:49] <tjtr33> ask me tomorrow morning :)
[01:41:42] <tjtr33> you may want to try some of the bios things i did ( and isolcpus=1 and turn off unneeded services )
[01:43:42] -!- fablab has quit [Ping timeout: 272 seconds]
[01:45:15] <tjtr33> adding your md5sum stress now
[01:47:02] -!- Loetmichel [Loetmichel!~cylly@p54B11914.dip0.t-ipconnect.de] has joined #linuxcnc
[01:48:00] -!- Loetmichel2 has quit [Ping timeout: 264 seconds]
[01:54:40] -!- PetefromTn_ [PetefromTn_!~IceChat9@75-136-60-251.dhcp.jcsn.tn.charter.com] has joined #linuxcnc
[01:56:03] -!- fablab [fablab!~fablab@2001:a60:1334:2701:715e:dfbf:5a8a:69d9] has joined #linuxcnc
[01:58:38] <tjtr33> Rab took a long time to get the terminal to appear, then ran your md5sum loop. now at 11310 servo and 13154 base
[01:59:55] <tjtr33> the latency-test window is alive and the data is constantly changing for 'last' values. one of the glxgears (5th) is barely moving
[02:00:03] <Rab> tjtr33, weird. Terminals open almost instantly for me. The system seems pretty snappy in all respects.
[02:00:33] <tjtr33> yep, the kbd and mouse are fine, but opening another window is dead slowok, dont stare at the monkeys
[02:00:53] <tjtr33> i intend to... not stare at the monkeys now :)
[02:01:23] <tjtr33> but thats good enuf for parport stepping right now
[02:01:46] <zeeshan> what are you guys testing
[02:03:02] <tjtr33> the video rez on the 39$ hp dc7800 has big impact, 1440x900 was 150uS latency, 1280x1024 is 10uS huge diff ( >>>so far <<< )
[02:05:10] -!- Servos4ever [Servos4ever!~chatzilla@173-87-48-244.dr01.hnvr.mi.frontiernet.net] has joined #linuxcnc
[02:07:42] -!- fablab has quit [Ping timeout: 265 seconds]
[02:11:22] <Rab> tjtr33, how did you change the resolution BTW?
[02:11:50] <tjtr33> menu|prefs|monitors
[02:12:21] <tjtr33> on the fly
[02:12:52] <tjtr33> maybe you could try 1140x900? highly depends on monitor
[02:13:00] <tjtr33> 1440x900
[02:13:51] -!- Anonymous [Anonymous!bdaf0c76@gateway/web/freenode/ip.189.175.12.118] has joined #linuxcnc
[02:13:51] <the_wench> anonymous: archivist_herron said about http://linuxcnc.org/docs/html/common/User_Concepts.html
[02:13:56] <tjtr33> System|Preferences|monitors
[02:14:51] <Anonymous> Hello
[02:15:15] <Anonymous> I have a question related to the use of Delta's ASDA servo drives with linuxcnc
[02:15:54] <Anonymous> it says in the page that it has been tested
[02:16:27] <Anonymous> but i wonder how does the computer interface with said servos
[02:16:47] <Anonymous> over rs232 or will it need an pcie ethercat card
[02:17:58] <Anonymous> i was wondering if anyone has experience with the drives
[02:19:44] -!- unfy [unfy!~Miranda@wsip-184-185-82-30.om.om.cox.net] has joined #linuxcnc
[02:20:29] -!- fablab [fablab!~fablab@2001:a60:1345:2301:715e:dfbf:5a8a:69d9] has joined #linuxcnc
[02:20:59] <tjtr33> Anonymous, i do not know ASDA but i see that it has been used by the German linuxcnc groups
[02:22:22] <Anonymous> ok, ill do some googling, thanks man
[02:23:04] <tjtr33> sorry ich nicht spechen zie Deutsche
[02:23:53] <tjtr33> and cant type either
[02:24:18] <Anonymous> haha =)
[02:24:46] <tjtr33> you will find the German group does some very sophistcated etercat work
[02:26:02] <tjtr33> bbl
[02:28:51] <zeeshan> anyone here use g73?
[02:29:01] <zeeshan> or even g83
[02:29:11] <zeeshan> what if your spindle axis isn't the Z axis..?
[02:29:25] <zeeshan> like in my case, the vertical spindle mode is Z axis
[02:29:29] <zeeshan> but horiziontal spindle is Y axis
[02:29:46] <zeeshan> same for tapping cycles :(
[02:30:08] -!- gambakufu has quit [Read error: Connection reset by peer]
[02:31:07] -!- stustev [stustev!~stustev@ip68-102-48-177.ks.ok.cox.net] has joined #linuxcnc
[02:31:24] -!- gambakufu [gambakufu!~ah@bzq-79-183-58-115.red.bezeqint.net] has joined #linuxcnc
[02:31:43] amnesic is now known as amnesic_away
[02:33:04] -!- Anonymous has quit [Quit: Page closed]
[02:36:04] <MacGalempsy> when dealing with the z axis brake, should I link this pin to axis.2.amp-enable-out ?
[02:38:55] <PetefromTn_> probably going to have to create dual configurations of linuxCNC one for vert and one for horiz I would think to be able to use the linked movements
[02:39:05] <zeeshan> =/
[02:39:28] <zeeshan> i wonder if theres a way to change axis_1 to axis_2
[02:39:37] <zeeshan> with the click of the button inside linuxcnc
[02:39:39] <zeeshan> without having to reset
[02:42:09] <tjtr33> MacGalempsy, beware of race conditions. on other systems, i wait for inposition, then clamp brake, then dis-able servo. similar concerns the other way.
[02:43:38] <PetefromTn_> on my teco drives they handle the Z brake timing internally and there is adjustment for delay.
[02:44:43] <MacGalempsy> tjtr33: ok. that sounds like a good method. thanks
[02:44:48] <MacGalempsy> PetefromTn_: lucky you ;P
[02:44:52] <tjtr33> the idea of waiting for in-position is so there is no servo resistance if there is no position error ( when using position mode )
[02:46:16] <tjtr33> on my machines some axis ar eonly for positioning heavy loads, so i brake and unbrake between poses
[02:46:22] <PetefromTn_> I wasn't telling you that to brag more to maybe have you check to ensure your drive is not equipped similarly...
[02:46:53] -!- Mr_Sheesh has quit [Ping timeout: 256 seconds]
[02:47:30] <MacGalempsy> didn't take it the wrong way :)
[02:47:44] <PetefromTn_> ok
[02:48:18] <MacGalempsy> PetefromTn_: did all your snow melt yet?
[02:49:00] <PetefromTn_> yes indeed
[02:49:22] <PetefromTn_> altho apparently we are getting blessed with more on thursday possibly
[02:49:29] <MacGalempsy> here too. finally got to go out and rampage town again
[02:49:54] <tjtr33> ice storm tomorrow in Chicago yay
[02:49:58] <MacGalempsy> need to make a run to OKC this week, but I fear its too cold to paint
[02:50:18] <PetefromTn_> are you in Oklahoma?
[02:50:42] -!- mablae__ has quit [Ping timeout: 256 seconds]
[02:50:49] <MacGalempsy> Fayetteville.
[02:50:54] <tjtr33> mmmmm Nelsons Buffeteria in Tulsa
[02:50:58] -!- Mr_Sheesh [Mr_Sheesh!~mr_s@unaffiliated/mr-sheesh/x-0757054] has joined #linuxcnc
[02:51:04] <MacGalempsy> trying to sell our house in OKC, and no one likes the kitchen colors
[02:51:20] <PetefromTn_> I feel your pain
[02:51:46] <PetefromTn_> too bad you are not in south Florida we could trade heh
[02:51:50] <tjtr33> http://www.nelsonsbuffeteria.com/Nelsons_0816/Welcome.html best food i knows of
[02:51:55] <MacGalempsy> I want to take my pcar to the shop, but it is too low for the trailer, so trying to figure out a plan to get two vehicles there
[02:52:09] <PetefromTn_> pcar?
[02:52:14] <MacGalempsy> porsche
[02:52:43] <MacGalempsy> tjtr33: Tulsa is not the best way to get to OKC from NW Ark
[02:52:47] <PetefromTn_> I loaded my Fiero GT's onto trailers by using multiple boards to get it high enough
[02:53:01] <MacGalempsy> I cant even get a jack under it! lol
[02:53:16] <PetefromTn_> they do sell low pro jacks
[02:53:32] <PetefromTn_> but I used to drive it onto taller and taller boards
[02:53:32] <MacGalempsy> the shop is going to put it back to factory height
[02:55:05] <PetefromTn_> my second one was slammed down on 18" rims and was a bitch to get over curbs LOL
[02:55:23] <furrywolf> that's why you own better vehicles. :P
[02:56:18] -!- swingley has quit [Remote host closed the connection]
[02:56:30] -!- Servos4ever has quit [Quit: ChatZilla 0.9.91.1 [SeaMonkey 2.26.1/20140612173529]]
[02:56:34] <MacGalempsy> the other day, I went to lunch with my college mentor and we went over a low speedbump and I heard a crunch. then a piece of black plastic shot out from under the car. I thought, oh crap there goes part of the front bumper. luckily it was a piece of someone elses car!
[02:57:28] <MacGalempsy> furrywolf: what is a "better car"?
[02:57:45] <furrywolf> anything that doesn't scrape on speedbumps?
[02:58:04] <MacGalempsy> a 78 pinto doesnt scrape, but its not better
[02:58:09] <furrywolf> you know, one designed with enough ground clearance to handle everyday life.
[02:59:04] <MacGalempsy> I got a couple of those, for everyday life
[02:59:20] <MacGalempsy> but there is something about busting out the supercar
[03:00:14] <PetefromTn_> supercar? WTF kind of Porsche is it?
[03:01:26] -!- swingley has quit [Ping timeout: 252 seconds]
[03:02:05] <tjtr33> well, its funny to see lowriders sitting on top of 6"of snow, the wheels off the ground after turning onto an unplowed street.
[03:03:23] <furrywolf> fastest thing I've driven was a DeTomaso Pantera... 550hp in the back seat to 14" wide tires.
[03:03:33] <MacGalempsy> how fast you get it?
[03:04:10] <furrywolf> not very.
[03:04:30] <MacGalempsy> then it may have not been the fastest thing you driven :\
[03:04:37] <furrywolf> was selling it for a relative who inherited it... I only got to drive it to take photos and get it to the shipping company to the new owner.
[03:04:56] -!- adam3999_ [adam3999_!~adam3999@pool-173-60-57-4.lsanca.fios.verizon.net] has joined #linuxcnc
[03:07:17] <furrywolf> fastest I've driven a vehicle is my '84 subaru, around 90. 83hp, baby!
[03:08:44] -!- grummund_ [grummund_!~user@unaffiliated/grummund] has joined #linuxcnc
[03:09:35] <Tom_itx> we built a 240z for a detective here once... 3 webbers, cam, pistons.. the whole bit... he kept bringing it back in complaining about a miss in it.. come to find out it wouldn't start missing until it hit ~140
[03:09:35] -!- grummund has quit [Ping timeout: 250 seconds]
[03:09:57] <MacGalempsy> Tom_itx: lol
[03:10:12] <MacGalempsy> got to love that
[03:11:57] <MacGalempsy> furrywolf: I will tell you that a 2014 Dodge Challenger rev limits out at 155.
[03:11:57] <Tom_itx> true story...
[03:12:55] <MacGalempsy> someone told me that, so I assume it is right. he couldnt pull 175
[03:14:33] <furrywolf> the pantera was geared for off-the-line performance, not speed... but with those tires and that much hp, it sure could get to its top speed awfully quickly.
[03:14:53] <furrywolf> the one I drove was built for more than 200hp over stock
[03:15:00] <Tom_itx> i swear you can't spin the tires on a pantera
[03:15:03] <MacGalempsy> those had the same engine and driveline as the original gt40, right?
[03:15:04] -!- pcw_home has quit [Read error: Connection reset by peer]
[03:15:51] <Tom_itx> not a stock one anyway..
[03:17:12] <MacGalempsy> oh. hmm.
[03:17:18] <furrywolf> similar, I think...
[03:17:32] <Tom_itx> ford winsor block
[03:18:18] <furrywolf> small-block ford and a zf5, mid-engined.
[03:18:47] -!- pcw_home [pcw_home!~chatzilla@c-50-143-148-115.hsd1.ca.comcast.net] has joined #linuxcnc
[03:18:56] <furrywolf> the fastest thing I own right now is my '84 subaru... it no longer has 83hp.
[03:19:05] -!- bobo_ [bobo_!63283ad9@gateway/web/freenode/ip.99.40.58.217] has joined #linuxcnc
[03:21:10] <furrywolf> I still need to deal with radiator fans. going to have to re-engineer the hood latch to make room in front of the radiator, as there sure ain't any behind it.
[03:22:01] <furrywolf> I managed to cram two 8" pusher fans in diagonal corners, but cooling is marginal when crawling.
[03:22:33] <MacGalempsy> hood pins?
[03:24:02] <furrywolf> nah, I think I can keep the stock mechanism, I just need to make a much more compact mount for it, with diagonal braces to the brackets I'll be building for the fan, instead of the stock vertical one.
[03:25:41] <furrywolf> only thing I've worked on with hood pins is a relative's xj12 with a chevy in it.
[03:26:16] -!- johtso has quit [Quit: Connection closed for inactivity]
[03:31:37] <furrywolf> also, it needs another inch of lift. :)
[03:33:32] <furrywolf> I put on 2", but it needs 3".
[03:34:15] <renesis> thats she said
[03:34:18] -!- swingley has quit [Remote host closed the connection]
[03:34:25] <renesis> get it because you put it on
[03:34:32] <tjtr33> Rab, pcw_home, i should not have looked http://ibin.co/1tVQsAjq8w10 http://ibin.co/1tVR9el94Jdu
[03:34:55] <furrywolf> ... huh?
[03:35:34] <renesis> needs more words but if you squint its rly funny
[03:35:41] <furrywolf> tjtr33: may I suggest a mesa board instead of hardware step generation?
[03:35:59] <tjtr33> cant even use mesa boards with that kind of latency
[03:36:13] <furrywolf> sure you can
[03:36:16] <tjtr33> the traj planner wont get to the mesa fast enif
[03:36:45] <tjtr33> the servo update wont happen, then servo sample wont....
[03:36:53] <furrywolf> I seem to remember it only needs 1ms if you don't use hardware step generation...
[03:36:56] <pcw_home> YOu can use mesa cards with 500 usec or latency
[03:37:02] <pcw_home> of
[03:37:19] <pcw_home> or maybe more on a a 500 Hz thread
[03:37:35] <tjtr33> well im wrong fer sure, didn think that was possible ( or reasonable )
[03:37:55] <Cromaglious> OK unbricked my Bufalo, now unbricked my TP-Link and reinstalling dd-wrt
[03:38:38] <pcw_home> looks like you disabled one core in your BIOS setup, might be better with both
[03:39:08] <tjtr33> ok, try that next, just no isolcpus at all in grub
[03:39:18] <Tom_itx> pcw_home, are you running those J1800 1900 boards on single core?
[03:39:34] <tjtr33> thx ( a 500hz thread ,, wow )
[03:39:37] <pcw_home> no
[03:39:48] <pcw_home> all cores, better performance
[03:39:50] -!- TTN has quit [Ping timeout: 246 seconds]
[03:40:22] <Tom_itx> i recall running D525 with single core seemed to do better...
[03:40:38] <furrywolf> tjtr33: to put it simply, how far do you expect your mill to move in 1ms? unless it's more than your tolerance, you won't have any problem updating at that speed...
[03:41:29] <pcw_home> its less critical than that with the correct hardware
[03:42:06] <furrywolf> yep, since it knows in advance what's happening
[03:43:29] <pcw_home> The primary error from latency spikes with a servo or stepper system is the apparent position error cause by late sampling of the position
[03:45:15] <pcw_home> with recent Hostmot2 configs, this sampling can be done via a DPLL thats phase locked to Linuxcnc 's periodic accesses
[03:45:49] -!- AR_ has quit [Ping timeout: 256 seconds]
[03:45:57] <Tom_itx> how recent?
[03:46:18] <pcw_home> 2.7
[03:47:12] * furrywolf wishes to be able to afford to get a mesa setup to really learn all these things
[03:47:31] <pcw_home> the DPLL reduces the sampling jitter from 10 to 200 times (depending on settings)
[03:48:16] -!- Cromaglious has quit [Ping timeout: 255 seconds]
[03:49:02] <Tom_itx> are other hardware vendors keeping up with these changes?
[03:49:08] <tjtr33> ok, its 80k servo & 90k base now after a couple minutes. and i read thats not a problem. very much news to me.
[03:49:15] <Tom_itx> i only seem to hear about mesa here
[03:49:30] <Tom_itx> not sure why i'm asking, just curious mostly
[03:50:23] <pcw_home> well I think we are the main vendor that doesn't make Mach stuff so concentrate on linuxcnc
[03:50:48] <Tom_itx> i wasn't sure about the others
[03:51:33] <pcw_home> Theres Pico Systems (Jon Elson)
[03:51:34] <pcw_home> General Mechatronics
[03:51:46] <pcw_home> (in Hungary I think)
[03:52:00] <Tom_itx> i've heard of pico
[03:52:00] <pcw_home> Vital systems
[03:52:19] <Tom_itx> the others not so much
[03:52:56] <pcw_home> Of course theres all the MachineKit BBB stuff
[03:58:17] <tjtr33> I re-read the latency-test page. its true then that 100000 nS is ok for hardware stepping. thx pcw_home.
[03:58:27] <pcw_home> for 2.7 stepgen the DPLL is setup like this hal:
[03:58:29] <pcw_home> setp hm2_[HOSTMOT2](BOARD).0.dpll.01.timer-us -50
[03:58:30] <pcw_home> setp hm2_[HOSTMOT2](BOARD).0.stepgen.timer-number 1
[03:59:20] <pcw_home> 100000 may require setting up the DPLL and-or using a PID stepgen setup
[04:00:33] -!- Cromaglious [Cromaglious!~robi@pool-108-23-106-180.lsanca.fios.verizon.net] has joined #linuxcnc
[04:01:17] <Tom_itx> and if you don't set those?
[04:01:25] <Tom_itx> or is that a requirement now for 2.7?
[04:01:57] <pcw_home> you will get the non-dpll sampling of the stepgen feedback position
[04:02:10] <pcw_home> (the old behaviour)
[04:02:26] <tjtr33> which exhibits the jitter becuz it isnt sampled at the right time
[04:02:33] <tjtr33> yes?
[04:02:39] <pcw_home> Yes
[04:03:00] <tjtr33> so, do i have servo threads with this dpll method?
[04:03:07] <pcw_home> yes
[04:03:11] <tjtr33> ( do i set a thread rate )
[04:03:19] <pcw_home> same as normal
[04:03:41] <tjtr33> ok, and what do i base it on, the latency-test result?
[04:04:39] <pcw_home> just set it to 1KHz if thats your normal servo thread rate
[04:04:48] <tjtr33> arrrrgh!
[04:04:52] <tjtr33> whats normal?
[04:05:00] <tjtr33> sorry'
[04:06:00] <pcw_home> 1 KHz is the most common as its adequate for velocity mode servos unless you have very high accelerations or accuracies
[04:06:28] <tjtr33> ok, this is just like religion, you gotta have faith and set it to 1khz, ok
[04:07:21] <tjtr33> i have not high acel and have 1um resolution to encoders
[04:07:28] <pcw_home> No you can go through the exercise and figure out the chord errors tha occure because of the velocity steps at the servo thread rate
[04:07:36] <tjtr33> sorry & thx
[04:08:55] <tjtr33> chord errors? something about circular interp? all new terms in ref to linuxcnc for me.
[04:09:05] <tjtr33> i'll try the 5i20 at 1khz
[04:11:10] -!- mwortley has quit [Ping timeout: 256 seconds]
[04:12:27] <tjtr33> pcw_home, thx, you do a lot for this community. maybe the community doesnt understand this well, maybe its just me.
[04:19:56] -!- Cromaglious has quit [Read error: Connection reset by peer]
[04:23:45] <tjtr33> chord error velocity = jerk i think
[04:27:40] <pcw_home> With a velocity mode servo or stepgen, the velocity is stepped every servo thread resulting in a new straight line segment every servo thread
[04:27:41] <pcw_home> The error between these straight like segments and the desired path is proportional to the velocity and the curvature
[04:28:16] <pcw_home> (I think this is the same as centripetal acceleration)
[04:29:19] <cradek> I bet the path a real machine actually follows is quite circular but maybe a tiny bit undersized
[04:29:49] -!- Mr_Sheesh has quit [Ping timeout: 264 seconds]
[04:30:20] <tjtr33> yes, thx reading this now http://www.mmrc.iss.ac.cn/~xgao/papernc/29-09.pdf beginning to see influence of servo thread period on chord len at given velocity, and complication of turning a corner in one or a few such periods
[04:30:28] <pcw_home> unless you have very high accelerations, the deviation is quite small relative to common machine/servo tolerances
[04:30:37] <cradek> if you were moving laser beams instead of cast iron maybe you could actually make a sharp corner every millisecond
[04:31:15] <pcw_home> yeah its low pass filtered by the mechanics
[04:32:51] -!- sumpfralle has quit [Ping timeout: 256 seconds]
[04:34:34] <tjtr33> ok, so where does the 1ms come from? isnt it more related to the possible response of the mechanical system? (like my system cant really respond to a 400hz sine wave, or some number )
[04:35:09] <tjtr33> we used to test machine to find that value and limit the servo update rate to less
[04:35:30] <cradek> it's a nice round number that's normally fast enough, and easy to reach with any old hardware
[04:35:43] <furrywolf> it was pulled out of someone's ass, and works.
[04:35:50] <cradek> some old CNCs would update at 100Hz
[04:36:31] <tjtr33> ok, so its a useful place to start. ok. i'd need to do bode plots to find better.
[04:36:46] -!- PetefromTn_ has quit [Quit: I'm Outta here!!]
[04:36:47] <cradek> are you working on a special machine?
[04:37:05] -!- Mr_Sheesh [Mr_Sheesh!~mr_s@unaffiliated/mr-sheesh/x-0757054] has joined #linuxcnc
[04:38:33] <tjtr33> yes and the response it not the special part. as usual i only work on EDM. and the fast bit is a reciprocating head with high speed. 5" retract and return in 800mS. i have videos
[04:38:42] <tjtr33> will go look for them now
[04:39:57] <cradek> sounds neat
[04:43:15] <pcw_home> if your servo system/PC is up to it you can sometimes improve performance by raising the servo thread rate
[04:44:18] <pcw_home> and a higher than 1 KHz servo thread is usually required when you have torque mode drives
[04:47:00] <pcw_home> did you see this latency with preemt-rt:
[04:47:01] <pcw_home> http://ibin.co/1tUeHa2Voo4A
[04:47:03] <pcw_home> it definitely is getting better
[04:48:39] -!- TTN [TTN!~TTN@unaffiliated/ttn] has joined #linuxcnc
[04:49:21] <furrywolf> bbl, wolfy bedtime
[04:51:34] <tjtr33> https://videobin.org/+8mk/bhp.html still being transoded
[04:58:12] -!- furrywolf has quit [Ping timeout: 256 seconds]
[04:58:36] <tjtr33> i dont know how long video bin wlll take to transcode that video
[05:01:07] <tjtr33> pcw_home, what value is there to the low latency? is it only for software stepping? ( only for non mesa setups ) the picture makes me think that is a virtue i want to pursue.
[05:01:23] <tjtr33> sorry to beat a dying horse, but its confusing
[05:04:29] <pcw_home> If you are pushing performance you still want reasonable latency
[05:07:13] <tjtr33> heres the vid from a competitor at 0:29 https://www.youtube.com/watch?v=uc8BUu61LvI
[05:07:40] <tjtr33> pcw_home, thx, is reasonable 5-15uS?
[05:07:51] <pcw_home> if a 1 Khz servo thread is good enough, the right hardware/setup can tolerate 100 usec latency or more
[05:07:52] <pcw_home> (but 1 KHz is probably _not_ good enough for that linear motor if linuxcnc is closing the servo loop)
[05:08:50] <tjtr33> i ended up saying i could not do it with linuxcnc and use a standalone control ( amplifier with built in positions ) and let linuxcnc just triggger it
[05:09:23] <tjtr33> linuxcnc says 'jump', it says 'done'
[05:10:39] <pcw_home> really depends on your drives (assuming velocity mode) , no so much linuxcnc
[05:12:18] <tjtr33> i use a 12mm pitch KSS ballscrew and Teco 5000rpm motor/drive, yes linuxcnc doesnt know its there, its sort of a plc driven W axis ( but not an axis, a device )
[05:13:13] <tjtr33> it takes the tool tip away from the path and returns it to the path, just very quickly
[05:14:16] <pcw_home> no reason step/dir position mode would not work as well
[05:14:31] -!- stustev [stustev!~stustev@ip68-102-48-177.ks.ok.cox.net] has parted #linuxcnc
[05:15:34] -!- Crom_ [Crom_!~Robi@pool-108-23-106-180.lsanca.fios.verizon.net] has joined #linuxcnc
[05:16:29] <tjtr33> well i can try it now, the system ( and several others ) are good enuf to use with hardware stepgens, and no need to use it as an axis, a pid in position mode can be hal controlled.
[05:16:29] <Crom_> Need to get a longer extension cord for new router in attic
[05:16:51] <tjtr33> very good info thanks for bearing with me.
[05:17:37] <pcw_home> I guess the retract could overshoot without a problem
[05:18:23] <tjtr33> right, i planned on that, only the start posn is important ( i even tried pneumatics with hard stops )
[05:19:23] <pcw_home> jerk limited plunge might improve performance
[05:19:31] -!- unfy has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
[05:19:48] <tjtr33> haha yes, that is just a very tight corner ( 0 radius )
[05:20:31] -!- Cromaglious [Cromaglious!~robi@pool-108-23-106-180.lsanca.fios.verizon.net] has joined #linuxcnc
[05:20:55] <tjtr33> i put air dampers at bottom ( airpots )
[05:21:02] <pcw_home> actually you might want trapezoidal start and jerk limited stop on the plunge
[05:21:09] <tjtr33> but they dont help ( i think they cant refill )
[05:21:37] <tjtr33> you ar eway past me on the possibilites in 2.7
[05:21:38] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[05:22:32] <tjtr33> is that in 2.7 or are we talking control theory
[05:23:08] <pcw_home> maybe possible with the limit3 comp
[05:23:47] -!- Crom_ has quit [Ping timeout: 246 seconds]
[05:24:14] <pcw_home> since the limits are all pins so might easily be changed on the fly
[05:24:47] <tjtr33> thx, i got lots to do again ( i have a voice coil motor to try too from SMAC )
[05:25:14] <pcw_home> Sounds like fun
[05:25:51] Cromaglious is now known as Cromaglious_
[05:26:08] <pcw_home> bbl
[05:27:22] <bobo_> tjtr33 what is the travel on a voice coil motor ?
[05:27:30] <tjtr33> https://www.youtube.com/watch?v=tW5-fI2ElPc the kss screw look near 1:00 m thanks pcw_home
[05:27:46] <tjtr33> bobo only 15mm ( cheap but proof of concept )
[05:28:21] <bobo_> can they be stacked ?
[05:29:35] <tjtr33> https://www.youtube.com/watch?v=tbPmrMSrVY4 end to end? i doubt it, the mass is accumulated that way
[05:30:39] -!- Cromaglious_ has quit [Ping timeout: 256 seconds]
[05:31:01] -!- anth0ny has quit [Quit: anth0ny]
[05:31:39] <tjtr33> i gotta thin about what he said... jerk limited plunge.. but its trapezoidal, the de-acell kicks in at least i got some ideas
[05:32:15] -!- AnxiousGarlic [AnxiousGarlic!~Spider@130.225.244.206] has joined #linuxcnc
[05:33:05] -!- AnxiousGarlic [AnxiousGarlic!~Spider@130.225.244.206] has parted #linuxcnc
[05:34:58] <tjtr33> hokay , had to copy the session to text file for future. gnite & thx all
[05:35:06] -!- tjtr33 has quit [Quit: Leaving]
[05:39:39] -!- maximilian_h [maximilian_h!~bonsai@dslb-088-066-232-059.088.066.pools.vodafone-ip.de] has joined #linuxcnc
[05:40:26] -!- Izakey [Izakey!~Izakey@41.205.22.18] has joined #linuxcnc
[05:40:31] -!- Cromaglious [Cromaglious!~robi@pool-108-23-106-180.lsanca.fios.verizon.net] has joined #linuxcnc
[05:40:57] <Cromaglious> ok Lets try WiFi only no wired network
[05:42:28] -!- swingley has quit [Remote host closed the connection]
[05:42:37] -!- maximilian_h1 has quit [Ping timeout: 264 seconds]
[05:45:14] -!- jerryitt has quit [Quit: Connection closed for inactivity]
[05:56:02] -!- deanputney has quit [Ping timeout: 246 seconds]
[05:59:11] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[06:01:18] -!- deanputney has quit [Read error: Connection reset by peer]
[06:01:33] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[06:02:18] -!- Vq has quit [Ping timeout: 250 seconds]
[06:07:29] -!- deanputney has quit [Ping timeout: 256 seconds]
[06:08:08] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[06:08:18] -!- anth0ny has quit [Quit: anth0ny]
[06:12:01] -!- Praesmeodymium has quit [Ping timeout: 256 seconds]
[06:15:21] -!- Praesmeodymium [Praesmeodymium!~kvirc@c-24-21-129-95.hsd1.or.comcast.net] has joined #linuxcnc
[06:21:55] -!- mhaberler [mhaberler!~mhaberler@extern-171.stiwoll.mah.priv.at] has joined #linuxcnc
[06:25:08] -!- mk0 [mk0!~Orr@plasmoteg-1.bas-net.by] has joined #linuxcnc
[06:26:29] -!- Camaban has quit [Ping timeout: 246 seconds]
[06:36:07] -!- ve7it has quit [Remote host closed the connection]
[06:38:12] -!- mk0 has quit [Ping timeout: 272 seconds]
[06:38:58] -!- Tanz [Tanz!~chatzilla@conr-adsl-209-169-105-33.consolidated.net] has joined #linuxcnc
[06:43:37] -!- Cromaglious has quit [Ping timeout: 245 seconds]
[06:56:44] -!- cornmander has quit [Ping timeout: 265 seconds]
[06:57:03] -!- cornmander [cornmander!~cornmande@98.116.247.205] has joined #linuxcnc
[06:58:24] -!- Cromaglious [Cromaglious!~robi@pool-108-23-106-180.lsanca.fios.verizon.net] has joined #linuxcnc
[07:02:54] -!- syyl_ [syyl_!~sg@p579F77D8.dip0.t-ipconnect.de] has joined #linuxcnc
[07:03:35] -!- deanputney has quit [Remote host closed the connection]
[07:04:02] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[07:04:10] -!- deanputney has quit [Remote host closed the connection]
[07:06:42] -!- kwallace2 has quit [Ping timeout: 272 seconds]
[07:08:34] -!- _nexxus_ has quit [Ping timeout: 252 seconds]
[07:11:56] -!- _nexxus_ [_nexxus_!~bwg@ragnar.generalamalgamated.com] has joined #linuxcnc
[07:16:12] -!- karavanjo has quit [Ping timeout: 272 seconds]
[07:35:12] -!- dan2k3k4 has quit [Ping timeout: 272 seconds]
[07:36:12] -!- dan2k3k4k5 has quit [Client Quit]
[07:42:14] <MacGalempsy> deadsville!
[07:42:25] -!- mttr has quit [Ping timeout: 264 seconds]
[07:42:55] <MacGalempsy> Cromaglious: what have you got going on tonight?
[07:43:00] <archivist> do something then
[07:44:24] <MacGalempsy> just got done playing some xbox.
[07:44:35] <MacGalempsy> work has nothing going on at all
[07:44:50] <MacGalempsy> but still have to be up for another 4 hours
[07:44:59] <MacGalempsy> archivist: what are you up to?
[07:45:17] <archivist> faking some scanning dates....
[07:46:03] <MacGalempsy> sounds mischivious
[07:46:55] <archivist> just want a doc scan to look recent, done specially for the user
[07:48:23] <archivist> I dont ask for money for scanning just hope for donations
[07:50:16] <Connor> zeeshan: You get your shim? You get the backlash fixed?
[07:57:25] -!- Izakey has quit [Ping timeout: 256 seconds]
[07:57:34] <MacGalempsy> archivist: what kind of stuff do people ask you to scan?
[07:57:41] <MacGalempsy> manuals, mags?
[07:57:43] -!- Deejay [Deejay!~Deejay@unaffiliated/dj9dj] has joined #linuxcnc
[07:57:48] <MacGalempsy> moin
[07:58:08] <Deejay> moin
[07:58:13] <MacGalempsy> ;)
[07:58:21] <archivist> today a 1960s bench power supply manual
[07:58:28] <Deejay> :)
[07:59:42] <archivist> an unusual design for the time as it has a SCR switching regulator before the linear regulator, ran very cool
[08:00:15] -!- adam3999_ has quit [Ping timeout: 256 seconds]
[08:01:34] <Swapper> anyone have a recommendation for a small 3-phase 400v rated plug with matching chassie connector ?
[08:02:37] <archivist> I only use CEE for 3 phase or direct connection
[08:02:47] <Swapper> Cee is huge
[08:02:56] <MacGalempsy> I saw the oil tank today at the thriftstore. it was a 1 gallon steel pail with a small pump in the bottom that was connected to a garden spicket
[08:03:22] <Swapper> and its for a servo motor so dont feel confident that would meet best practice if anyone connects some other stuff to the CEE connector
[08:04:09] <archivist> I direct wire the motors
[08:04:41] <Swapper> thats plan B i guess
[08:04:48] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[08:05:11] <archivist> not often disconnecting, easy to label
[08:05:49] -!- c_morley has quit [Ping timeout: 244 seconds]
[08:06:28] -!- adam3999_ [adam3999_!~adam3999@pool-173-60-57-4.lsanca.fios.verizon.net] has joined #linuxcnc
[08:09:26] -!- deanputney has quit [Ping timeout: 244 seconds]
[08:35:14] -!- adam3999_ has quit [Ping timeout: 245 seconds]
[08:54:43] -!- mhaberler has quit [Quit: mhaberler]
[08:57:14] -!- mhaberler [mhaberler!~mhaberler@bigmac.stiwoll.mah.priv.at] has joined #linuxcnc
[08:58:58] -!- georgenz [georgenz!~georgenz@203-211-101-15.ue.woosh.co.nz] has joined #linuxcnc
[09:01:09] <georgenz> Hey guys... I have just fitted some teco servos, drives and 7i77 to my machine, z axis had some chatter in it. I tuned it today, but the chatter is still there. Any ideas?
[09:02:04] <georgenz> It's not hunting, jst random chatter sounds like electrical noise or something?
[09:05:59] <archivist> is it sliding down and correcting
[09:08:13] -!- georgenz has quit [Ping timeout: 255 seconds]
[09:15:22] -!- Tanz has quit [Ping timeout: 240 seconds]
[09:18:20] -!- rob_h [rob_h!~robh@90.203.32.28] has joined #linuxcnc
[09:24:13] -!- Tanz [Tanz!~chatzilla@conr-adsl-209-169-105-33.consolidated.net] has joined #linuxcnc
[09:45:54] -!- sumpfralle [sumpfralle!~lars@p5DCFD1FE.dip0.t-ipconnect.de] has joined #linuxcnc
[09:46:42] -!- SolarNRG [SolarNRG!~SolarNRG@213.207.181.14] has joined #linuxcnc
[09:47:00] <SolarNRG> does anybody have a photo of the PCB of a gecko stepper motor controller?
[09:49:36] -!- PCW_ [PCW_!~chatzilla@99.88.10.65] has joined #linuxcnc
[09:50:58] -!- PCW has quit [Ping timeout: 255 seconds]
[09:51:00] PCW_ is now known as PCW
[09:52:40] -!- sumpfralle has quit [Ping timeout: 265 seconds]
[10:00:15] -!- bobo_ has quit [Quit: Page closed]
[10:02:16] <SolarNRG> preferable the g540
[10:10:16] -!- dhoovie [dhoovie!~kvirc@182.69.237.109] has joined #linuxcnc
[10:16:09] -!- dhoovie has quit [Ping timeout: 252 seconds]
[10:20:22] -!- Tanz has quit [Quit: ChatZilla 0.9.91.1 [Firefox 36.0/20150222232811]]
[10:21:31] <XXCoder> work was interesting heh
[10:21:38] <XXCoder> getting ready to bed its 2:21 am lol
[10:22:29] -!- danylevskyi has quit [Ping timeout: 250 seconds]
[10:23:40] -!- sumpfralle [sumpfralle!~lars@p5DCFD1FE.dip0.t-ipconnect.de] has joined #linuxcnc
[10:24:11] <SolarNRG> xxcoder what controller are you using?
[10:24:36] <XXCoder> some el cheapo chinese controller
[10:24:41] <XXCoder> my hobby router is tiny
[10:24:50] <SolarNRG> what motors are you using?
[10:24:57] <XXCoder> nemo23
[10:25:05] -!- moorbo [moorbo!~moorbo@dyn-21-50.mdm.its.uni-kassel.de] has joined #linuxcnc
[10:25:06] <SolarNRG> will it handle nema34's?
[10:25:35] <XXCoder> depends on power requirements. by time youre using 34s I would suggest seperate umm whats it called... drivers?
[10:26:00] <SolarNRG> any chance you can take the driver apart for me and take two photos of each side of the PCB?
[10:26:11] <XXCoder> I dont use any
[10:26:20] <SolarNRG> don't use any what?
[10:26:24] <XXCoder> driver
[10:26:31] <XXCoder> mine is controller direct drive
[10:26:41] <XXCoder> it isnt fancy project after all
[10:26:59] <XXCoder> well ready to sleep now so later
[10:27:13] <SolarNRG> I can't afford a gecko g540 but there are tonnes of electronic parts about, transistors and the like and I want to make my own controller to plug into my pc
[10:27:27] -!- moorbo_ [moorbo_!~moorbo@dyn-21-50.mdm.its.uni-kassel.de] has joined #linuxcnc
[10:29:55] -!- moorbo__ [moorbo__!~moorbo@dyn-21-50.mdm.its.uni-kassel.de] has joined #linuxcnc
[10:30:51] -!- moorbo has quit [Ping timeout: 265 seconds]
[10:31:21] -!- moorbo_ has quit [Read error: Connection reset by peer]
[10:47:32] -!- tiwake has quit [Ping timeout: 244 seconds]
[10:47:37] <MacGalempsy> finally in the home stretch of finishing up with work!
[10:50:05] -!- c_morley [c_morley!~chris@S0106204e7f8c229b.no.shawcable.net] has joined #linuxcnc
[10:51:36] -!- gonzo_nb [gonzo_nb!~gonzo@host-92-14-209-118.as43234.net] has joined #linuxcnc
[10:52:20] -!- monkeyisl has quit [Read error: Connection reset by peer]
[10:52:59] -!- monkeyisl [monkeyisl!~monkeyisl@115.95.237.2] has joined #linuxcnc
[10:53:31] -!- sumpfralle has quit [Ping timeout: 255 seconds]
[10:54:35] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[10:59:22] -!- deanputney has quit [Ping timeout: 255 seconds]
[11:01:37] -!- karavanjo has quit [Ping timeout: 255 seconds]
[11:05:39] -!- [cube] has quit [Ping timeout: 245 seconds]
[11:12:07] -!- [cube] [[cube]!~cube@bas17-ottawa23-1279418485.dsl.bell.ca] has joined #linuxcnc
[11:19:31] -!- mk0 [mk0!~Orr@plasmoteg-1.bas-net.by] has joined #linuxcnc
[11:24:20] -!- AnimalFarmPig_ has quit [Ping timeout: 246 seconds]
[11:29:47] amnesic_away is now known as amnesic
[11:33:35] amnesic is now known as amnesic_away
[11:39:17] -!- moorbo__ has quit []
[11:45:55] -!- sivu has quit [Ping timeout: 244 seconds]
[11:53:41] -!- mk0 has quit [Ping timeout: 252 seconds]
[12:00:25] -!- TTN has quit [Ping timeout: 264 seconds]
[12:08:27] -!- syyl_ has quit [Ping timeout: 256 seconds]
[12:17:53] <Swapper> anyone have a example of a way to stop the "at speed" and speed input of the spindle to stop blinking ?
[12:18:22] <Swapper> i guess its cuse the motor i have is a servo and it drifts slooowly when stoped.
[12:18:33] <Swapper> Maybe some sort of deadband setting ?
[12:26:10] -!- WalterN [WalterN!~Octavia@71-84-132-89.dhcp.knwc.wa.charter.com] has joined #linuxcnc
[12:26:42] -!- MacGalempsy has quit [Remote host closed the connection]
[12:37:55] <jthornton> 0 is at speed
[12:39:03] <jthornton> it should not drift when under the command of LinuxCNC
[12:39:16] <Swapper> its not pid controlled
[12:39:49] <Swapper> i asked for that before and since the driver is very linear there has not been any need to close it in linuxcnc
[12:40:16] <Swapper> its +-10v controlled servo as spindle
[12:41:58] <Swapper> i have similar behaviur on the axies that it varies between encoder counts and 1.0 becomes 0.999 and back several times
[12:42:01] <Swapper> its not static
[12:43:24] <jthornton> yea, it will only be plus minus an encoder count on an axis
[12:43:52] <Swapper> is that "ok" or should i bother fixing it?
[12:43:54] <jthornton> is the at speed light one you added?
[12:44:13] <Swapper> yea and it shows the RPM
[12:44:49] -!- WalterN has quit [Ping timeout: 244 seconds]
[12:44:50] <jthornton> put something in there to ignore at speed when the spindle command is off
[12:44:57] <Swapper> and since the driver is enabled and keeping the spindle locked when stopped but drifting slowly it flashes
[12:45:38] <Swapper> i where thinking some typ of deadband that ignores values in the range of 8rpm
[12:46:10] <Swapper> and probably only filter the value to the gmocapy "led" pin
[12:46:12] <jthornton> it drifts at 8rpm?
[12:46:48] <Swapper> not that much
[12:46:53] <Swapper> but to get rid of it
[12:47:34] <jthornton> gmocapy should be smart enough to ignore the at speed when the spindle is not running
[12:47:43] <Swapper> its not :)
[12:48:36] <Tom_itx> you can use 'near' with it to come 'close'
[12:48:59] <Tom_itx> value or percent
[12:49:45] <Swapper> i do have that allready
[12:49:54] <Swapper> probably only wrong lowpass setting ?
[12:50:12] -!- FreezingCold [FreezingCold!~FreezingC@135.0.41.14] has joined #linuxcnc
[12:50:18] <Swapper> net spindle-vel-cmd-rps => near.0.in1
[12:50:18] <Swapper> net spindle-vel-fb => near.0.in2
[12:50:18] <Swapper> net spindle-at-speed <= near.0.out
[12:50:18] <Swapper> setp near.0.scale 1.750000
[12:50:18] <Swapper> #setp near.0.difference 3.333333
[12:50:21] <Swapper> #setp near.0.difference 5
[12:50:27] <Swapper> the last 2 where in there...
[12:50:38] <Swapper> setp scale.spindle.gain 60
[12:50:39] <Swapper> setp lowpass.spindle.gain 1.000000
[12:52:01] <jthornton> motion.spindle-on OUT BIT TRUE when spindle should rotate
[12:53:18] <jthornton> you should mention this on the never ending gmocapy thread on the forum
[12:53:52] <Swapper> yea guess i do that
[12:53:57] <Swapper> ty for the input
[12:54:19] <Swapper> i think it where mentioned a while back
[12:54:21] <Swapper> ill check
[12:58:18] <jthornton> as busy as that thread is I'd assume a lot gets forgotten unless brought up again
[12:58:32] <Swapper> yep
[13:09:53] -!- Deejay__ [Deejay__!~Deejay@p54B6757F.dip0.t-ipconnect.de] has joined #linuxcnc
[13:10:30] -!- Deejay has quit [Disconnected by services]
[13:10:34] Deejay__ is now known as Deejay
[13:10:40] -!- Deejay has quit [Changing host]
[13:10:40] -!- Deejay [Deejay!~Deejay@unaffiliated/dj9dj] has joined #linuxcnc
[13:10:40] -!- akex2 [akex2!uid58281@gateway/web/irccloud.com/x-hzdvgdlpovkcvwaw] has joined #linuxcnc
[13:12:13] -!- WalterN [WalterN!~Octavia@71-84-132-89.dhcp.knwc.wa.charter.com] has joined #linuxcnc
[13:13:24] -!- dan2k3k4 has quit [Ping timeout: 272 seconds]
[13:14:22] -!- amiri has quit [Ping timeout: 240 seconds]
[13:20:33] -!- pandeiro has quit [Remote host closed the connection]
[13:27:51] <SolarNRG> has anybody here made their own stepper motor drivers?
[13:30:30] <skunkworks> yes. isn't worth it. buy them.
[13:30:55] <skunkworks> Same goes for servo drives...
[13:31:56] <Swapper> yea, and small ones for mini projects are cheap on ebay
[13:32:03] <malcom2073> +1. Is the $10 you save on a small stepper driver, or $100 you save on a large one, worth the amount of engineering time you'll spend makign your own? :)
[13:32:07] <Swapper> cant even get the parts for the price some times
[13:35:14] <SolarNRG> why is it not worth making your own ones?
[13:35:27] <_methods> http://www.ebay.com/itm/5pcs-Geeetech-Stepper-Driver-DRV8825-and-heatsink-RepRap-Prusa-Mendel-3D-Printer-/231180196237
[13:35:31] <_methods> $3/each
[13:35:46] <malcom2073> SolarNRG: Because they're silly inexpensive.
[13:35:47] <_methods> you can't even get the components that cheap
[13:36:22] <SolarNRG> will this handle the current and voltage of a nema 34?
[13:36:30] <_methods> not for nema34
[13:36:36] <malcom2073> SolarNRG: You didn't specify that.
[13:36:46] <malcom2073> What current and voltage are you looking for?
[13:37:06] <SolarNRG> well I asked everywhere yesterday and no shops sell stepper motor drivers
[13:37:22] <malcom2073> Where's everywhere, the internet?
[13:37:40] <SolarNRG> but they had some power fets and I seen some open source stepper motor drivers ppl have made, a lot of the time they are slow and shite, but some people have used dedicated PIC's or ATMEGA's and got a lot higher speeds out of them
[13:37:41] <_methods> http://www.ebay.com/itm/4Axis-Stepper-motor-driver-PEAK-7-8A-256-micsteps-DM860A-24-80V-for-Nema-34motor-/301319682760
[13:38:00] -!- mablae has quit [Read error: Connection reset by peer]
[13:38:28] <_methods> pretty sure you can find those dm860a's cheaper
[13:38:54] <_methods> around $50
[13:39:36] <_methods> once again the voltage and current requirements would help
[13:40:47] <SolarNRG> can I put a volt meter across the terminals of my motor and find that out?
[13:41:24] <malcom2073> No, but you can look at the datasheet for your motor to find that out
[13:51:24] <SolarNRG> ok i got the 60BYGH301B which states no. phases 4, voltage 3.9, current 3, inductance 3.2, resistance 1.3, hold trq 22, rotor inertia 860, weight 1.4, length 88, lead wires 8
[13:52:02] <malcom2073> Those are nema 23, not nema 34
[13:52:10] -!- skunkworks has quit [Ping timeout: 255 seconds]
[13:52:30] <SolarNRG> right so the chinese dude on ebay ripped me off then
[13:52:34] <SolarNRG> great
[13:53:06] <SolarNRG> shall i continue working with what i got or should i get some bigger motors?
[13:53:32] <malcom2073> That depends, what are you working on? If your motor moutns are nema34, nema23's won't fit, and if nema34 torque levels are required, then nema23 won't work
[13:53:51] <SolarNRG> I'm drilling my own mounts out of steel plate
[13:54:05] <SolarNRG> using sand and glass to make them nice and flat
[13:55:52] <SolarNRG> but if i just want to make a very basic diy system on the cheap with some fets so i can at least get the motors turning, what do?
[13:58:36] <malcom2073> If you want to just get the motors turning just ot see them turn, spend $15 on one of these: http://www.ebay.com/itm/CNC-Single-Axis-TB6600-4-5A-Two-Phase-Stepper-Motor-Driver-Controller-/261244221548?pt=LH_DefaultDomain_0&hash=item3cd35e786c
[13:58:46] * jthornton got the lathe g code generator to generate arcs with GTk YEA!
[13:58:46] <malcom2073> I have a couple of them, they work fairly well for bench testing and the like
[13:59:13] <malcom2073> If you want to design your own, that's out of my field, maybe someone else can help you...but I'm fairly certain it'll cost you more than $15 to design/build one :P
[14:00:50] <SolarNRG> I was thinking find a pcb file for one that works pretty good take it down the uni and getting it printed then soldering on my own fets etc.
[14:01:44] <_methods> http://www.3ders.org/articles/20130107-powerlolu-board-powerful-open-source-stepper-motor-driver.html
[14:01:56] <_methods> if you can make your pcb's there you go
[14:02:08] <_methods> https://github.com/fluidfred/powerlolu
[14:02:19] <_methods> .brd and .sch there
[14:02:35] <SolarNRG> sweet thank you
[14:02:43] <SolarNRG> seems a bit slow but works
[14:05:56] <SolarNRG> any reason why you wouldn't use this?
[14:06:30] <_methods> because i have better things to do and i would just buy an off the shelf driver?
[14:08:49] <SolarNRG> good answer
[14:09:25] -!- SolarNRG has quit [Quit: Leaving]
[14:09:27] <malcom2073> Time really, drivers are cheap enough. If you spend 5 hours getting it made and soldering it together (including drive time to go get it done), and you could buy one and have it at your door for $50, then your time is worth $10 an hour, ignoring any cost of components.
[14:09:41] <malcom2073> Heh
[14:11:53] -!- FinboySlick [FinboySlick!~shark@74.117.40.10] has joined #linuxcnc
[14:12:52] -!- SolarNRG [SolarNRG!~SolarNRG@213.207.181.14] has joined #linuxcnc
[14:14:54] <SolarNRG> can that open source driver handle 8 wire stepper motors?
[14:15:22] <_methods> i have no idea
[14:15:38] <_methods> you'd probably need to read about it's capabilities
[14:15:59] <SolarNRG> how many wires do your stepper motors have?
[14:29:12] -!- stustev1 [stustev1!~stustev@wsip-184-176-200-171.ks.ks.cox.net] has joined #linuxcnc
[14:32:22] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[14:36:46] -!- deanputney has quit [Ping timeout: 256 seconds]
[14:38:17] <sliptonic> Anyone know a part# or link for the edge connectors to connect to servo amps like this: http://www.grmolds.com/wp-content/uploads/2014/05/amps-1024x765.jpeg
[14:40:49] <_methods> i think that's a jst connector
[14:40:54] <_methods> how many pins is that
[14:41:00] <_methods> 16 pin jst?
[14:41:07] <sliptonic> yes 16
[14:43:06] <pcw_home> 8 wire stepper motors have 2 windings per phase that allow you to connect them in series or parallel
[14:43:08] <pcw_home> when in parallel the inductance is lower but twice the current must be supplied for the same torque
[14:46:52] <SolarNRG> the datasheet said this motor uses 3.6v does that mean I can get away with 5V?
[14:47:20] <SolarNRG> and most power supplies won't handle anything above 500mA where do I get a power supply from that will handle the current?
[14:48:54] <_methods> sliptonic: i think that's a jst ph connector
[14:49:04] <_methods> but you'd have to measure it and check the specs to make sure
[14:49:10] <_methods> http://www.jst-mfg.com/product/pageview/connector_e/pageview.html#page_num=60
[14:49:36] <pcw_home> No 3.6V is the DC voltage drop at rated current, you typically need 10 to 20 times this for good performance (36 to 72V)
[14:49:39] <SpeedEvil> In general, steppers don't care at all about voltage until you hit 200V or so
[14:49:48] <SpeedEvil> Then the insulation breaks down.
[14:50:00] <SpeedEvil> You need to keep the current to the nominal amount
[14:50:25] <sliptonic> Thanks _methods
[14:50:36] <_methods> np but that's probably not it
[14:50:38] <pcw_home> I have some 360V rated steppers (for rectified 220 line operation)
[14:50:43] <_methods> it might be a chinese knockoff of jst
[14:50:48] <SpeedEvil> Current limiting drivers work well.
[14:50:55] <SolarNRG> so would an old pc power supply work?
[14:51:21] <pcw_home> only 12v so slow but would work
[14:51:29] -!- karavanjo has quit [Quit: karavanjo]
[14:51:35] <sliptonic> I think it's 2.54 pitch. An ideal solution would be a female connector that adapts to screw terminals but this gets me in the ballpark.
[14:51:41] <_methods> it might actually be a jst xh
[14:51:46] <_methods> http://www.jst-mfg.com/product/pageview/connector_e/pageview.html#page_num=92
[14:51:47] <pcw_home> but for 3.6v step motors you need a current source drive
[14:52:00] <_methods> pitch is 2.5 on xh
[14:52:17] <SolarNRG> what sort of shop would sell an appropriate power supply?
[14:52:28] <pcw_home> ebay?
[14:52:48] <pcw_home> where you can also get the drivers
[14:53:34] <pcw_home> note that the critical rating or matching step motors with drives is current
[14:53:43] <pcw_home> for matching
[14:57:36] -!- dan2k3k4 has quit [Ping timeout: 244 seconds]
[14:58:48] -!- skunkworks [skunkworks!~skunkwork@68-115-41-210.static.eucl.wi.charter.com] has joined #linuxcnc
[15:14:08] amnesic_away is now known as amnesic
[15:38:57] -!- akex2 has quit [Quit: Connection closed for inactivity]
[15:44:30] -!- miss0r [miss0r!~miss0r@87-59-179-253-dynamic.dk.customer.tdc.net] has joined #linuxcnc
[15:44:31] <Rab> sliptonic, that's a Molex KK 254 series connector. http://www.molex.com/molex/products/family?key=kk_254_rpc_connector_system&channel=products&chanName=family&pageTitle=Introduction&parentKey=kk_interconnect_solutions
[15:44:36] -!- Vq [Vq!~vq@90-227-168-203-no123.tbcn.telia.com] has joined #linuxcnc
[15:45:40] <CaptHindsight> just FYI http://www.pchub.com/uph/ carries lots of the hard to find replacement parts for laptops, PC's and LCD monitors
[15:50:58] <Rab> sliptonic, this is probably the part you want: http://www.molex.com/molex/products/datasheet.jsp?part=active/0022012165_CRIMP_HOUSINGS.xml
[15:54:09] WalterN is now known as tiwake
[15:55:25] <Rab> Selecting the proper connector is a little tricky because Molex have several different feature variants. The one I linked to has a "locking ramp"; you'll want to make sure your connector uses that.
[15:58:07] <archivist_herron> some manuals are following me home today http://www.ebay.co.uk/itm/121579154539
[15:58:14] <Rab> Note that this is only the housing. You also need contacts. That housing is compatible with 2759 and 4809 series contacts; I strongly recommend the 4809 "cat ear" contacts for reliability: http://www.molex.com/molex/products/listview.jsp?query=4809&path=cHome%23%23-1%23%23-1~~ncCRIMPTERMINALS%23%230%23%23b&offset=0&autoNav=1&sType=s&filter=&fs=&channel=Products
[15:59:04] -!- kwallace has quit [Read error: Connection reset by peer]
[15:59:11] <Rab> The standard traditional KK pin everyone uses only contacts one side of the pin. Not a good design IMO.
[15:59:55] <Rab> Armed with those part numbers you can go to a distributor like Digi-Key or Mouser for pricing and availability.
[16:00:26] <archivist_herron> seen them burnt up when too much current put though them
[16:01:20] <Rab> Yeah, overcurrent or resistance heats the contact and it loses its spring temper. Less tension = worse connection = escalating situation.
[16:01:21] <pcw_home> My experience is use tin connectors only to about 30% of rated current
[16:01:36] -!- Loetmichel has quit [Remote host closed the connection]
[16:01:46] -!- Loetmichel [Loetmichel!~cylly@p54B11914.dip0.t-ipconnect.de] has joined #linuxcnc
[16:02:10] <archivist_herron> parallel the contacts for higher current
[16:03:23] <FinboySlick> archivist_herron: http://www.ebay.co.uk/itm/Edgwick-16-20-Phase-3-Shaping-Machine/111607730406 was a couple related links from your auction. It's a lovely machine.
[16:05:15] <archivist_herron> way over my budget though...grmbl
[16:05:42] <archivist_herron> locking up here be at home soon
[16:06:18] -!- dan2k3k4 has quit [Ping timeout: 272 seconds]
[16:06:50] <FinboySlick> I guess if you found a busted one, sturdy as those are built, you could convert it to a pretty darn solid vertical mill.
[16:23:39] -!- deanputney [deanputney!~mustardha@104.244.221.171] has joined #linuxcnc
[16:26:35] <CaptHindsight> anyone know the going hourly rate is the US for cnc machining? Say a part has a 1 hour cycle time and you're making 50 parts and setup is trivial or extra.
[16:27:22] -!- sumpfralle [sumpfralle!~lars@erina.opennet-initiative.de] has joined #linuxcnc
[16:28:04] <FinboySlick> CaptHindsight: I think it'll depend on if you're doing *only* 50 or batches of 50 with a commitment for a higher number.
[16:29:01] <_methods> we usually charge $120/hour
[16:29:10] <CaptHindsight> just a 50 - 100 run of something milled on it's own
[16:29:13] <_methods> if it's an insert killer we'll charge a tooling charge
[16:29:48] <_methods> if you want to send a print i'll get you a quote
[16:29:51] <CaptHindsight> rockwell hardness 100.000 :)
[16:30:04] <_methods> um nm don't send me that print lol
[16:30:07] -!- PetefromTn_ [PetefromTn_!~IceChat9@75-136-60-251.dhcp.jcsn.tn.charter.com] has joined #linuxcnc
[16:30:28] <CaptHindsight> _methods: no, I was just wondering what US rates are in the range of
[16:30:51] -!- miss0r has quit []
[16:31:37] -!- quiqua has quit [Quit: quiqua]
[16:31:37] -!- sumpfralle1 [sumpfralle1!~lars@subaru.opennet-initiative.de] has joined #linuxcnc
[16:32:27] <_methods> yeah we charge $120/hr for cnc, i think $85/hr for manual, laser is $220/hr and waterjet is like $150/hr
[16:32:42] <_methods> plasma is $100/hr
[16:33:09] <_methods> unless it's got a ton of holes then we might up it because of electrode wear
[16:33:34] <_methods> fab is random price the owner makes up at will
[16:33:50] <_methods> pretty sure he has a dart board hidden somewhere
[16:34:23] -!- sumpfralle has quit [Ping timeout: 246 seconds]
[16:34:38] <alex4nder> _methods: are you guys a one-stop shop?
[16:34:45] <_methods> we're a job shop
[16:34:47] <alex4nder> werd
[16:34:50] <_methods> we'll whores
[16:35:02] <_methods> s/we'll/we're
[16:35:12] <alex4nder> do you TIG weld as well?
[16:35:15] <_methods> yeah
[16:35:17] <alex4nder> nice
[16:35:18] <alex4nder> that's fun
[16:35:29] <_methods> we have a robot welder too for med production welding jobs
[16:36:11] <CaptHindsight> was just considering the cost for people to machine laptop cases in USA
[16:36:45] <alex4nder> you doing a product?
[16:36:47] <_methods> well aluminum chassis should be around the $120 range
[16:36:53] <_methods> $120/hr that is
[16:37:17] <_methods> if you were to supply material you might be able to beat them down on price a bit too
[16:37:42] <CaptHindsight> people want to know the DIY price since the files will be open
[16:38:01] <_methods> ah say to get a local shop to machine a one off part like that
[16:38:04] <_methods> will be expensive
[16:38:12] <_methods> you'll be eating all the setup
[16:38:30] <CaptHindsight> no, for a shops in the US that might do a run once a week
[16:38:35] <_methods> ahh
[16:38:41] <_methods> yeah that would be the way to do it
[16:38:48] <CaptHindsight> http://openlunchbox.com/smf/index.php?topic=35.0
[16:38:51] <_methods> sounds like a prime job for PetefromTn_
[16:39:22] <alex4nder> CaptHindsight: that looks a lot like an X-series thinkpad chassis
[16:39:34] <CaptHindsight> yeah, they were made well
[16:39:42] -!- deanputney has quit [Remote host closed the connection]
[16:39:43] -!- furrywolf [furrywolf!~randyg@184.194.30.56] has joined #linuxcnc
[16:39:53] <CaptHindsight> these are for the PC enthusiast or pro user
[16:39:56] <alex4nder> yah
[16:39:59] <alex4nder> I loved that hardware
[16:40:35] -!- micges_ [micges_!~micges@dag189.neoplus.adsl.tpnet.pl] has joined #linuxcnc
[16:40:36] <alex4nder> CaptHindsight: have you ever looked at/used the novena?
[16:44:14] -!- micges has quit [Ping timeout: 252 seconds]
[16:44:37] <CaptHindsight> alex4nder: yes, they almost made a laptop
[16:45:03] <alex4nder> haha
[16:45:08] <alex4nder> they might disagree with you. ;)
[16:46:19] -!- sumpfralle [sumpfralle!~lars@erina.opennet-initiative.de] has joined #linuxcnc
[16:46:47] <CaptHindsight> alex4nder: I have no idea why they decided to make tooling for injection molding
[16:47:45] <alex4nder> I only ever saw their aluminum case, and their limited-run wood and aluminum
[16:47:47] -!- sumpfralle1 has quit [Ping timeout: 245 seconds]
[16:48:21] <CaptHindsight> we are going to SLA these cases
[16:48:33] <CaptHindsight> tooless production
[16:48:41] <alex4nder> cool
[16:49:29] <CaptHindsight> but you're free to mod the design and machine them
[16:55:54] <CaptHindsight> when you add up all the parts costs for a current high end laptop you find that they are making 100-200% margins vs 10-20% or less on the low end
[16:56:47] <CaptHindsight> so it looks like they just don't want to make good laptops anymore
[16:57:03] <CaptHindsight> just cheap laptops at high prices
[16:57:14] -!- SpeedEvil has quit [Quit: No Ping reply in 180 seconds.]
[16:57:59] -!- SpeedEvil [SpeedEvil!~quassel@tor/regular/SpeedEvil] has joined #linuxcnc
[16:59:23] -!- Loetmichel has quit []
[17:02:58] -!- toast-work [toast-work!88a0a9f3@gateway/web/freenode/ip.136.160.169.243] has joined #linuxcnc
[17:04:14] -!- Loetmichel [Loetmichel!~cylly@p54B11914.dip0.t-ipconnect.de] has joined #linuxcnc
[17:06:20] <ssi> need an opinion
[17:06:36] <ssi> last laser was 100W reci Z4
[17:06:51] <ssi> and honestly I felt like I never could really fully utilize all the power it could make
[17:06:54] <renesis> capthindsight: laptops have huge development cost and time compared to pc
[17:07:11] <ssi> 100Ws aren't in stock in the US at the moment, but 80Ws are... should I just do 80W this time?
[17:07:13] <renesis> youre not just paying for parts, even before you add the marketing costs
[17:07:47] <renesis> higher end laptops tend to suck less, in terms of reliability, mechanicals, and gap and step in the finish
[17:07:57] <renesis> that shit costs $$$
[17:08:32] <renesis> its not a desktop, that shit is one injection mold for the snap in face, maybe get fancy with some fan ducts, everything else off the shelf
[17:09:19] <renesis> you could have a decent tech shop for desktop parts from oem vendors and dev that shit in a few days
[17:09:58] -!- md-2 has quit [Quit: Leaving...]
[17:10:45] -!- deanputney [deanputney!~mustardha@2601:8:a581:a700:3ca7:5dee:703c:483d] has joined #linuxcnc
[17:11:27] <renesis> also how do you assign a price to a custom motherboard to calculate a margin
[17:11:59] <SpeedEvil> NRE cost / volume * profit
[17:12:08] <SpeedEvil> Or something
[17:12:54] <renesis> yeah but you would have to know the profit
[17:13:28] <renesis> how do you know nre cost without like, full teardown bom and then industrial esionage to find out how much they paid for all of the things?
[17:13:56] <SpeedEvil> Err - what?
[17:14:07] <SpeedEvil> If you're making a laptop, you know how much it costs to make a laptop
[17:14:37] <SpeedEvil> If you're making a laptop, and you don't know how much it's going to cost within narrow margins going in - you shouldn't be making a laptop
[17:14:50] <renesis> im saying if youre not making it, like you are capthindsight, how do you know they are making 100-200% margins
[17:14:57] <SpeedEvil> Ah
[17:15:11] <renesis> based on parts cost
[17:15:20] <CaptHindsight> I know the NRE costs
[17:15:31] <CaptHindsight> and the typical run volumes
[17:15:33] <renesis> so my argument is profit is probably the same or less than for low end
[17:15:38] <renesis> but nre is way more
[17:16:03] <renesis> capthindsight: do you work in laptop manufacturing?
[17:16:27] <CaptHindsight> renesis: and yes, the corporate overhead is all in the marketing, support, branding etc etc
[17:16:29] <renesis> or do you have recent teardown pricing link or something
[17:16:44] <SpeedEvil> https://technology.ihs.com/414163/
[17:16:46] <SpeedEvil> for example
[17:16:57] <renesis> when you add up all the parts costs for a current high end laptop
[17:17:04] <CaptHindsight> renesis: been designing PC's since the early 80's
[17:17:15] <SpeedEvil> You can get IhS to commission an analysis for a particular laptop
[17:17:24] <renesis> right i just explained why a laptop and a pc have nothing to do with each other in terms of development
[17:17:45] <alex4nder> renesis: I just bought a keyboard kit for $300, because it had an aluminum case, and a trackpoint. there is someone who will pay $N for anything.
[17:17:53] <renesis> unless you mean you design mobo since the early 80s
[17:17:58] <CaptHindsight> pc's laptops, servers, boards etc
[17:18:03] <SpeedEvil> https://technology.ihs.com/344375/teardown-analysis-samsung-series-9-np900x3a-notebook-computer - for example
[17:18:12] <SpeedEvil> CaptHindsight: laptop from component level? Neat
[17:18:28] <renesis> alex4nder: right but a company will spend more time making sure it gets manufacturered correctly to a higher tolerance
[17:18:38] <alex4nder> that's fair
[17:18:38] <XXCoder> morning
[17:18:52] <alex4nder> renesis: unless you've got boutique customers with cash to burn
[17:18:59] <renesis> laptops are crazy custom, high density
[17:19:14] <renesis> so in a lot of models, you break even or lose money on high end
[17:19:21] <renesis> because itll sell more of your low end
[17:19:36] <CaptHindsight> I don't leave anything up to the actual assembler
[17:19:37] <renesis> which you are pumping out in high volume, and getting lots of deals because of the volume
[17:19:46] -!- JLuc69 [JLuc69!JLuc69@2a01:e35:2e88:cdf0:b9c1:241:38a0:8780] has joined #linuxcnc
[17:19:46] -!- JLuc69 has quit [Client Quit]
[17:19:48] <XXCoder> ran this machine today http://twentywheels.com/imgs/a/a/m/u/y/fadal_vmc_4020_cnc_vertical_machining_center___1988_2_lgw.jpg
[17:19:52] <XXCoder> ancient. lol
[17:20:05] <CaptHindsight> well tooling is >$250K for a laptop case
[17:20:16] <alex4nder> XXCoder: awesome
[17:20:41] <SpeedEvil> https://technology.ihs.com/Search?f=261:2588&q=laptop&so=MostRelevant
[17:20:42] <SpeedEvil> hmm
[17:20:52] <SpeedEvil> alas all of their laptop teardowns and analysis are subscribe only
[17:25:12] <CaptHindsight> most designs take the 1,000 monkeys approach and end up working (almost)
[17:26:30] <CaptHindsight> AMD and Intel do much of the motherboard work for them
[17:29:48] <CaptHindsight> cut and paste gerber
[17:29:48] -!- capricorn_one has quit [Quit: Konversation terminated!]
[17:29:48] <alex4nder> CaptHindsight: I always wondered how those second-tier motherboard manufacturers churned out boards that worked at all
[17:29:48] <CaptHindsight> and the large laptop houses have internal people from both the brand they are making and the major suppliers
[17:29:48] <CaptHindsight> so foxconn has Apple and Intel people in house
[17:29:48] <renesis> apple and intel would be insane not to have engineers at the CM
[17:30:23] <renesis> i think at this point the only people not camping out at asian CM are small-medium companies who are new at CM manufacturing who are about to have the asses handed to them
[17:30:44] <alex4nder> yah :/
[17:30:58] <alex4nder> another leg up for local production
[17:31:00] <renesis> at which point theyll die or become another company training china how to make all the things
[17:31:12] <renesis> well, soon maybe
[17:31:34] * alex4nder looks at his Mac Pro
[17:31:57] <renesis> we dont have infrastructure like before so economy maybe has to get pretty meh here before accountants decide its cheaper to make things here
[17:32:00] -!- akex2 [akex2!uid58281@gateway/web/irccloud.com/x-gbcpssfakgmeqxqr] has joined #linuxcnc
[17:32:03] <renesis> gonna go slow at first =\
[17:32:07] <alex4nder> yah
[17:32:15] <CaptHindsight> Apple has a sea of mini cnc mills churning out cellphone and laptop cases
[17:32:40] <CaptHindsight> they don't even call them mills since they are stripped down to just make cases
[17:33:28] <renesis> most of their line workers just load QA and assembly robot jigs
[17:33:51] <renesis> like short distance fedex tasks
[17:34:45] <CaptHindsight> Fosconn was installing over a million robots in their plants
[17:35:12] <CaptHindsight> but since they bought Chinese they aren't accurate enough to make iphones
[17:35:38] <CaptHindsight> http://www.macrumors.com/2014/12/05/foxconn-robots-iphone-assembly/
[17:36:05] <CaptHindsight> the people that run these co's aren't really that bright unfortunately
[17:37:33] <renesis> depends
[17:37:58] <renesis> a lot of times theyre pretty smart but theyre not going to do more work than what they feel they were paid for
[17:39:51] <renesis> if you deliver great work for low cost, you cant charge extra for great work
[17:40:33] <alex4nder> assuming our cultural priorities match theirs is a costly mistake
[17:40:36] <renesis> thats not a chinese thing, thats just how you run a shop and hit deadlines
[17:40:43] * LeelooMinai points out that people in Asia have highest average IQ overall
[17:41:29] * ssi points out that IQ is mostly bogus
[17:41:44] <renesis> in this context yeah
[17:42:11] <renesis> communication and team building actually go a long way, high iq doesnt necessarily lend itself to that
[17:42:19] -!- jvrousseau has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
[17:42:29] <CaptHindsight> it's not IQ, it's more narcissism, ego and cunning
[17:42:45] <renesis> sounds like corporate america to me
[17:42:47] <alex4nder> renesis: dealing with the chinese, I don't think it's entirely apples-to-apples with how shops are run.
[17:43:04] <alex4nder> the chinese are interesting, because stuff that they think is totally acceptable socially, is considered horrible in the US
[17:43:32] <alex4nder> but the US went through its own struggles on that front
[17:43:47] <renesis> we chopped up kids in machines during our industrial revolution
[17:43:51] <alex4nder> that's what I'm saying
[17:43:59] <renesis> just because we had no concept of servicability
[17:44:05] <alex4nder> but now that's not part of the american narrative
[17:44:09] <alex4nder> and we tell ourselves that we are better
[17:44:26] <renesis> china is doing fine, they the reason we have such a high standard of living
[17:44:32] <alex4nder> eh
[17:44:40] <alex4nder> that remains to be seen
[17:45:03] <renesis> shit will get worse here and better there
[17:45:18] <renesis> im okay with that, its fair
[17:45:25] <alex4nder> you ever been to china?
[17:45:31] amnesic is now known as amnesic_away
[17:45:42] <LeelooMinai> "yes, 3 times"
[17:45:47] <renesis> nope, but i work closely, hear tons of stories
[17:46:35] <renesis> my point is, we drove them towards the state theyre in now, theyre dealing with issues but overall theyre probably in a more positive place than before
[17:46:39] <CaptHindsight> it's like blade runner, without the flying cars
[17:46:58] <alex4nder> at least the noodles are good
[17:47:08] <LeelooMinai> I heard they make new BladeRunner, with Harrison Ford
[17:47:17] <renesis> ?
[17:47:47] <LeelooMinai> He is kind of old now... maybe they will call it BladeWalker:)
[17:47:53] <CaptHindsight> yes, I hope Riddley does a good job vs just make a few $
[17:48:04] <toast-work> side question: anyone have a favorite book on camshaft design with a focus on the theory/application to heat engines
[17:48:30] <CaptHindsight> flying wheel chairs
[17:51:44] -!- sirdancealot [sirdancealot!~koo5@236.152.broadband3.iol.cz] has joined #linuxcnc
[17:51:54] <sliptonic> Rab: Thanks for the info!
[17:52:14] <CaptHindsight> LeelooMinai: people were wondering if Sean Young were to also return as an older replicant
[17:52:36] <LeelooMinai> As Sean Old:)
[17:53:21] <CaptHindsight> and she should just be herself, run around with martinis, get tossed out of places etc
[17:53:37] <alex4nder> haha
[18:00:18] -!- postaL has quit [Read error: Connection reset by peer]
[18:01:14] -!- postaL [postaL!~postaL@c-107-3-170-109.hsd1.ca.comcast.net] has joined #linuxcnc
[18:06:20] <archivist> toast-work, I have some docs but rather old probably
[18:07:39] <XXCoder> CaptHindsight: if I recall, replinants dont live very long?
[18:09:07] <CaptHindsight> XXCoder: except for Rachel, she had no expiration date
[18:09:19] <XXCoder> oh yea. been a long while
[18:10:49] <CaptHindsight> http://en.wikipedia.org/wiki/Blade_Runner_2:_The_Edge_of_Human they kill Rachael :(
[18:11:04] -!- rob_h has quit [Ping timeout: 272 seconds]
[18:12:51] <XXCoder> interesting
[18:13:15] <CaptHindsight> it would be nice of the movie just doesn't sux
[18:13:31] <alex4nder> the whole movie is going to just be a loop of this: http://www.reactiongifs.us/wp-content/uploads/2013/08/who_gives_a_shit.gif
[18:13:36] <_methods> that's asking a lot of a movie that never needed a sequel
[18:14:03] <XXCoder> alex4nder: lol
[18:19:58] postaL is now known as postaL_offline
[18:24:48] -!- pandeiro has quit [Remote host closed the connection]
[18:25:53] <XXCoder> http://cdn.smosh.com/sites/default/files/ftpuploads/bloguploads/pug-man-swap.jpg
[18:25:55] <XXCoder> enjoy.
[18:26:15] <PetefromTn_> Personally speaking as a HUGE fan of Blade Runner and everything it was I think that anything that takes away from it would be a crime. Having said that and witnessing the new Star Trek movies by JJ Abrams being pretty damn amazing I would HOPE that any sequel or prequel would be of similar quality.
[18:27:15] -!- adb [adb!~IonMoldov@2a02:120b:c3d1:5780:baac:6fff:fe67:305f] has joined #linuxcnc
[18:30:34] <PetefromTn_> I do think that reading the story line on the wikipedia page there is adequate story there to make it work if done with enough effort and passion for the original.
[18:30:34] <CaptHindsight> back in the 80's I used it a a test tape so it was on 12 hours a day, I must have watched it 200 times
[18:30:34] <XXCoder> if. yeah
[18:30:34] <XXCoder> sequels pretty rarely improve on orginial
[18:30:34] <PetefromTn_> it really was a masterpiecce especially the soundtrack.
[18:30:34] <CaptHindsight> most the characters are now in the late 60's or early 70's
[18:30:34] <PetefromTn_> yeah which is a difficult hurdle to overcome..
[18:30:34] <XXCoder> Capt whats likely is only few will repeat
[18:30:34] <XXCoder> certainly main charactor
[18:30:34] <CaptHindsight> or lots of CGI like that film with Bruce Willis a few years ago
[18:30:34] <PetefromTn_> I thought it would be impossible to recreate the original Star Trek cast adequately but I think the new Star Trek cast is amazing really...
[18:30:34] <CaptHindsight> make him and the costar look 20 years younger
[18:30:34] postaL_offline is now known as postaL
[18:30:34] <XXCoder> pete few changes but it works
[18:30:34] <XXCoder> it can be explained away that its alternate universe
[18:30:57] <CaptHindsight> http://en.wikipedia.org/wiki/Surrogates
[18:31:03] <PetefromTn_> I think the reality is that everyone ages and people are forgiving of new actors taking famous roles AS_LONG_AS it is done with the correct care and passion for what made it famous in the first place
[18:31:09] -!- danylevskyi_ has quit [Ping timeout: 252 seconds]
[18:31:15] <XXCoder> indeed
[18:32:37] <PetefromTn_> The new Star Wars movies are coming right around the corner. I am THRILLED at the possibilities they bring and honestly it is an amazing time to be a fan of SCI FI with all of the comics heroes and recreations of the most amazing movies of our times being redone with better tech and skills.
[18:33:26] <PetefromTn_> maybe I am just a geek at heart but I can't wait to see what they do with it.
[18:33:35] <Rab> They should turn the premise on its head, 30 years later genetic engineering is prevalent and unmodified humans are considered a threat.
[18:33:50] <CaptHindsight> PetefromTn_: things have come a long way from https://www.youtube.com/watch?v=pFnLirXjjto Clutch Cargo
[18:34:38] <PetefromTn_> hehe thats cool
[18:35:44] <PetefromTn_> I remember one of my favorite movies was the old Flash Gordon from the 80's...
[18:36:25] -!- rob_h [rob_h!~robh@90.203.32.28] has joined #linuxcnc
[18:36:50] <PetefromTn_> https://www.youtube.com/watch?v=x4mBGEdB-XE
[18:36:55] -!- deanputney has quit [Read error: Connection reset by peer]
[18:37:17] <CaptHindsight> he did a guest scene in Ted
[18:37:29] -!- deanputney [deanputney!~mustardha@2601:8:a581:a700:3ca7:5dee:703c:483d] has joined #linuxcnc
[18:38:26] <toast-work> archivist: yeah, i'm looking at fairly modern stuff directly as it relates to engine performance
[18:41:07] -!- deanputney has quit [Read error: Connection reset by peer]
[18:41:41] -!- deanputney [deanputney!~mustardha@2601:8:a581:a700:3ca7:5dee:703c:483d] has joined #linuxcnc
[18:41:53] <PetefromTn_> Oh yeah I didn't know that.
[18:42:23] <PetefromTn_> while it was really hokey it was sure a fun movie I thought. Would make for a great modern movie I think..
[18:42:37] -!- moorbo [moorbo!~moorbo@dyn-21-121.mdm.its.uni-kassel.de] has joined #linuxcnc
[18:46:00] <PetefromTn_> https://www.youtube.com/watch?v=miU6ruUPZHI LOL I never saw that before... hilarious!
[18:47:55] <archivist> toast-work, zeeshan may have some resources as he play with engines
[18:48:15] <toast-work> definitely, I plan on asking him
[18:48:19] <PetefromTn_> https://www.youtube.com/watch?v=fcTPhAp3Ax8
[18:49:37] <archivist> toast-work, I got books and data from Lucas who made early injection as well as ignition systems
[18:50:21] <toast-work> paper-books or digital books? I'd love to take a look if they're available
[18:50:24] <toast-work> if not, no worries
[18:50:28] <_methods> wut i might have to watch ted now that's freakin hilarious
[18:50:56] <moorbo> it's a funny movie ted
[18:52:27] <PetefromTn_> I know right... thats some funny stuff.
[18:53:50] -!- phantoneD [phantoneD!destroy@a89-152-21-144.cpe.netcabo.pt] has joined #linuxcnc
[18:54:52] <archivist> toast-work, my stuff is paper, 99% not scanned but a reasonable amount catalogued http://www.collection.archivist.info/searchv13.php?searchstr=lucas
[18:54:58] <XXCoder> lol floating city
[18:55:18] <toast-work> that's awesome, ty sir
[18:55:51] <XXCoder> Oh that movie wherwe that talking teddy bear comes from - I hate movie since it so balantly stole from webcomic
[18:56:18] <XXCoder> http://www.cartoonbrew.com/comics/seeing-double-macfarlanes-ted-vs-imagine-this-60531.html
[18:57:22] -!- phantoxeD has quit [Ping timeout: 245 seconds]
[18:58:03] <XXCoder> "Imange THIS" webcomic is gone because of it
[18:58:10] <XXCoder> I loved that webcomic
[18:58:50] -!- TTN [TTN!~TTN@unaffiliated/ttn] has joined #linuxcnc
[19:03:30] -!- Rickta59 has quit [Quit: leaving]
[19:12:35] -!- pandeiro has quit [Remote host closed the connection]
[19:13:12] -!- karavanjo has quit [Read error: No route to host]
[19:14:14] -!- acdha has quit [Quit: Textual IRC Client: www.textualapp.com]
[19:14:33] -!- georgenz [georgenz!~georgenz@118.148.252.141] has joined #linuxcnc
[19:15:42] -!- Rickta59 [Rickta59!~Rickta59@cpe-174-106-151-175.ec.res.rr.com] has joined #linuxcnc
[19:17:42] <CaptHindsight> who make a rugged motorized screw jack with limit switches?
[19:21:04] -!- deanputney has quit [Read error: Connection reset by peer]
[19:21:37] -!- deanputney [deanputney!~mustardha@c-67-183-112-251.hsd1.wa.comcast.net] has joined #linuxcnc
[19:22:31] <georgenz> I cant seem to figure out how to get jitter out of my z axis. Teco 2kW servo, just tuned it yesty. But its always had a small amount of jitter, not visible on dro.. but audible. Im concerned it will wear ballscrews etc.
[19:22:42] <georgenz> Anyone got any ideas?
[19:22:47] <JT-Shop> data:image/jpeg;base64,/9j/4AAQSkZJRgABAQAAAQABAAD/2wCEAAkGBxQQEhQUEBISFhUVFBgVFBUYFRYYFBUVFhYWFxgVFxUZHCggGBwlHxcUITEiJSorLi4uFyAzODMsNygtLysBCgoKDg0OGhAQGiwkHSUsLCwsLC0sLSwsLC8tLCwsLCwsNCwsLCwsLCwsLCwsLCwsLCwsLCwsLCwsLCwsLCwsLP/AABEIAOEA4QMBEQACEQEDEQH/xAAcAAEAAQUBAQAAAAAAAAAAAAAAAwECBAUGBwj/xABJEAACAgECAwQHAwYLBgcAAAABAgADEQQSBSExBhNBUQciMmFxgZEjQsEUM1KSodEkJUNTYnJzoqOxshWT0uHw8Rc1VGN0tML/xAAZAQEBAQEBAQAAAAAAAAAAAAAAAQIDBAX/xAAyEQEAAgIABAUCBAYDAQEAAAAA
[19:22:47] <JT-Shop> AQIDEQQSITETIkFRcRRhMoGx8AWRocHR8TNC4SMV/9oADAMBAAIRAxEAPwD3GAgIGPfrqkOHtrU9cM6g488EzM3rHeViJlLVarAFSCD0IIIPwIliYnsi+UICAgICAgICAgICAgICAgICAgICAgICAgQ6vUCtGdgSFBJAGSceQmbWisTMrEbnTQ0cR1AA1DhWpfrWuC1SjkGB+8fP8J4q58n/ACTHln+cOs0r+H1XaGuvVX3WFNybEVCyEAkA5IDDMVimbLM63GiZmlYhDwnjAq06VhWa8eqKsEEkknJJHJfM/jJhzxTHy+vst6TNt+jP4ZxGwP3OpCiwjcrL7DL5e4jmPl9e2LNabcmSNSxaka3Xs3E9TmrAQEBAQEBAQEBAQEBAQEBAQEBAQEBAQNdruMVVP3b7y20NhUZuRJA5gY8DOOTPSs8s924pMxuGD2fXNlzVqyUkjajA
[19:22:48] <JT-Shop> jLEDcwB9kdRicOHjz25fwtZO0b7ttr9Wmnqe2w4StSzfBRnlPdSm5isONp1G5cCfShT3YuOksBLGtMsm47VDOc+Q3Vj37vcZ6/8A86ebvG9PP9VGuzM9HnaWvW2XZVhfgOzMQdyZxhAPZVfVGPeJ5r8BbBPPedzLtTiIyeWI0rx70n6XS2tWq2WlCQ7JtCBh1UMx5kdOXKenFwOS9ebs5X4itZ13Yus9LFFdj1nT3EoxUkNXg4OPObr/AA+8xvcMzxVYnWmRqPSfp66K7Wqt32bilXq7tisV3sc4UEhgPE4mI4G82msT29Wp4isRtdpPSbprKLbQlgeoKzUnbuZWdU3I2cMAWGfGLcFki0R7+pHEVmNrk9J+kaiy1RZvrCk0sFV2DOqZU52tjdk884ETwOSLRHv6n1FNbdJ2a4yuu01eoRWRX3gK2Nw2OyHOOXVSZ58uOcd5pPo7
[19:22:53] <JT-Shop> UtzV22k5tEBAQEBAQEBAQEBAQEBAQEBA0+t0l4uNtBq51qhDhj0LH7pHnPLfHkjJz11206VtXl1KThOvd2sruCCysjkucEMAQefxmsOW1rTW0amEtWIiJjs4b0tdo1apNLQ6u1j5tCMDgIfVQ4PIlscv6Pvn2OBwzzc9o6Q8PEX3HLDiNNw9tTqk0+nVLBpkOQSQj92d1xJAJwzlgMeBWey+SuPHOS3r+4ca1m1orDMGrfh/EK9TYK61uZnYVsWQJYStq5IBypO7Hh6vunHFkpxXDeSd69/s3etsWTqv7P8AGV4LdqK9Zpu8ZsbX9XJUE4ZS3JkbIOR/23lxzxFazSzNLximeaHX+l0J+QVOqqu65DnAB9ZHOMiebgd+LMTLtxGuTbhNOG0T6LW2099QawRn2MrvQpk5AYHDDPX649tpjJF8cTqdvPHl5bTHRDxLdrG1mtqp7qgB
[19:22:54] <XXCoder> holy crap
[19:23:00] <JT-Shop> c49jczVoFB5AsTljjp9M2kxjiuOZ3JO7TNtdFLezf8Wpr0fI7wpahxhRvKKyn44BB/Sl8f8A+3hynh+Tnet+ik/xXp/jd/8AYtnyuM/57fl+j24P+OHWzyuxAQEBAQEBAQEBAQEBAQEBAQINbv2N3W3fg7d3QH3zF98s8vdY1vq5vhuqZF7qitjqGYm57ByU+Ls33s+AH/fw4skx
[19:23:02] <JT-Shop> sorry
[19:23:06] <JT-Shop> http://image.tradett.com/images/newaylights/20130922124442896104321.jpg
[19:24:11] <Rab> CaptHindsight, are you looking for an industrial-grade jackscrew? Or something you might find on an RV?
[19:25:08] <CaptHindsight> Rab: http://www.northerntool.com/shop/tools/product_200381584_200381584
[19:25:45] <CaptHindsight> some customers break whatever you make for them
[19:25:57] <archivist> light weaight ones used to be available for satelite pointing
[19:26:23] -!- MrSunshine has quit [Quit: Leaving]
[19:26:41] <CaptHindsight> I used some lighter ones with a 100kh rating and they crashed them lifting 4KG loads
[19:26:50] <Cromaglious> heh
[19:26:51] <CaptHindsight> 100Kg rating
[19:26:52] <Cromaglious> nice
[19:27:09] <CaptHindsight> it's mounted on gantry
[19:27:33] -!- moska1 [moska1!~chatzilla@ip-78-94-79-59.hsi02.unitymediagroup.de] has joined #linuxcnc
[19:27:37] <CaptHindsight> so of course they move the X before the Z and bend them
[19:27:48] <Rab> Warner Linear make actuators like that: http://www.warnerlinear.com/Default.asp
[19:28:30] -!- moorbo has quit [Remote host closed the connection]
[19:28:30] <Cromaglious> so there is no interlock switch to prevent Z from moving until X is unfolded?
[19:28:58] <CaptHindsight> controlled by Linuxcnc and several PhD's :)
[19:30:20] <jdh> I'm playing with a bunch of LinMot's right now
[19:30:26] -!- MrSunshine [MrSunshine!~mrsun@c-c825e255.4551536--62697410.cust.bredbandsbolaget.se] has joined #linuxcnc
[19:30:38] <PCW> georgenz: post a halscope plot of the jitter, otherwise its not really possible to guess whats going on
[19:30:57] -!- theorbtwo has quit [Remote host closed the connection]
[19:31:12] <Cromaglious> HAHAHA like action mouths
[19:31:21] <Cromaglious> live
[19:31:40] <Cromaglious> VERY bad animation
[19:31:43] <CaptHindsight> Cromaglious: Clutch Cargo?
[19:31:54] <CaptHindsight> was from ~ 1960
[19:32:04] <Cromaglious> CaptHindsight, yeah
[19:32:14] <CaptHindsight> almost creepy
[19:32:31] <georgenz> Sure... whats the best way to grab a screen capture onl linux
[19:32:34] -!- nerdfiles has quit [Ping timeout: 256 seconds]
[19:32:38] <Cromaglious> almost? Definately REALLY creepy
[19:33:12] <Rab> georgenz, I use xv, but only because it's familiar.
[19:33:38] <archivist> georgenz, it is it applications/accessories/take screenshot in this old linux box
[19:33:49] -!- cnc1 [cnc1!~chatzilla@ip-78-94-79-59.hsi02.unitymediagroup.de] has joined #linuxcnc
[19:33:59] <georgenz> It not hunting... jst random jitter
[19:34:16] <georgenz> Sweet on my way into workshop... will do it shortly
[19:34:35] <Rab> Apparently The Gimp can take screenshots.
[19:34:39] <georgenz> Just halscope of following error?
[19:35:08] <Rab> There was some really basic screenshot that was in most Linux distros, but the name escapes me.
[19:35:39] <archivist> I use the pre installed default
[19:36:32] <alex4nder> Rab: import?
[19:36:49] <alex4nder> (worst name ever)
[19:37:03] <Rab> Looks plausible. Also xwd.
[19:37:11] <alex4nder> werd
[19:37:25] <Rab> Sorry, this is worse: "escrotum-git screen capture using pygtk, inspired by scrot"
[19:37:33] <alex4nder> : |
[19:38:04] <Rab> https://wiki.archlinux.org/index.php/taking_a_screenshot
[19:40:26] <Rab> Ooh yeah, import wins. Part of ImageMagick, outputs to common formats, fast usage.
[19:41:48] <Rab> Funny thing, the man page doesn't actually describe how to use it (invoke from command line, then click on window to capture).
[19:42:21] <Rab> (The root window also being a valid selection, of course.)
[19:43:37] <PCW> alt printscreen works on ubuntu not sure about debian
[19:45:20] <Rab> Doesn't seem to work with LinuxCNC boot disk Xfce environment.
[19:46:25] <_methods> apt-get install xfce4-screenshooter-plugin
[19:46:45] <alex4nder> "how many machinists does it take to make a screenshot?"
[19:46:48] <cradek> Rab: I happened to just answer this for someone else, here's a transcript: http://pastie.org/9996791
[19:47:05] <CaptHindsight> heh, reminds me of the tech support calls where people pressed their credit card or paperwork on the screen and pressed "print screen"
[19:47:46] micges_ is now known as micges
[19:47:51] <Rab> cradek, cool, thanks!
[19:49:00] -!- akex2 has quit [Quit: Connection closed for inactivity]
[19:52:04] -!- swingley has quit [Remote host closed the connection]
[19:59:28] -!- tmcw has quit [Remote host closed the connection]
[20:09:17] -!- tmcw has quit [Remote host closed the connection]
[20:12:17] -!- akex2 [akex2!uid58281@gateway/web/irccloud.com/x-dbnpxnhrqabhmtqw] has joined #linuxcnc
[20:26:24] <furrywolf> yay, shopmaster shipped my mt3 extension
[20:26:50] <XXCoder> finally
[20:29:25] <furrywolf> what's the cheapest place to find an AGM U1 battery?
[20:31:47] <furrywolf> also, can we please kill every ebay seller who lists the same identical item 6000 times with a different "for <random product>" in the title?
[20:32:07] <XXCoder> well it makes it more findable, but bit annoying yes
[20:33:36] <furrywolf> http://www.ebay.com/itm/SigmasTek-12V-35AH-Group-U1-Deep-Cycle-Sealed-Battery-/321661603397 is the cheapest I've found so far. there's a few listed for a couple bucks less, but they appear to be SLA, not AGM.
[20:35:03] <XXCoder> http://www.batterymart.com/p-YTX4L-BS-AGM-Maintenance-Free-Battery.html?utm_source=google&utm_medium=cpc&utm_term=YTX4L-BS&utm_campaign=PLA&gclid=Cj0KEQiA1NWnBRDchObfnYrbo78BEiQA-2jqBdby__WHf29w5uSkpi3KAUwooruoVk5xUOlLY7ZmU4caAjR38P8HAQ
[20:35:09] <XXCoder> anywhere near what you want?
[20:35:42] <XXCoder> free shipping special
[20:35:42] <furrywolf> that's not U1 size.
[20:35:51] <XXCoder> ok
[20:36:10] <furrywolf> that's a fucking TINY battery. 3Ah! U1 is usually 35Ah...
[20:36:28] <XXCoder> yea saw agm wondered if what you wanted. didnt see u1
[20:36:33] <furrywolf> http://shop.batterymart.com/search#w=U1
[20:37:05] <XXCoder> $77 cheapest
[20:37:11] <furrywolf> the ebay one is $63...
[20:37:17] <XXCoder> yeah
[20:37:25] <XXCoder> check shipping though is ebay one free too?
[20:37:29] <furrywolf> I need the type with lead posts on top, not the type with screw holes.
[20:37:30] <furrywolf> yes
[20:38:29] <furrywolf> argh! another webdesigner who felt compelled to make something utterly brain-damagedly simple, like going to the next page of search results, require tons of javascript.
[20:38:43] <furrywolf> and slow
[20:38:45] <furrywolf> and broken.
[20:38:54] <XXCoder> lets see how well aliexpress price em lol
[20:39:08] <XXCoder> not that I would trust em with battery
[20:39:38] <XXCoder> no result probably too large for cheap shipping
[20:40:05] <XXCoder> oh im wrong
[20:40:11] <furrywolf> and air shipping is questionable. (yes, you can ship sealed batteries air... no, they won't always take them)
[20:40:48] <roycroft> just label them as "energy containment apparatus" when you declare them on the shipping manifest
[20:41:11] <furrywolf> the guy at the local interstate battery suggested a Concorde if I want it to last the longest... he says every time he gets one in as a core, it's 10+ years old.
[20:41:46] <XXCoder> $100 for 100 ah deep cycle
[20:41:53] <XXCoder> 12v agm
[20:41:59] <XXCoder> big battery.
[20:42:12] <XXCoder> $471 shipping though lol
[20:42:16] <furrywolf> lol
[20:42:59] <XXCoder> 2v 1000AH agm battery lol
[20:43:05] <XXCoder> deadly.
[20:43:10] -!- tmcw has quit []
[20:43:28] <XXCoder> yet anohter at 2v 2000AH
[20:43:40] <XXCoder> its worth 5 of my vans
[20:43:47] <XXCoder> $10k basically
[20:44:28] -!- deanputney has quit [Remote host closed the connection]
[20:45:02] <XXCoder> yea nothing worth it from aliexpress, not surpised. heavy things tend not to
[20:45:38] <furrywolf> batteries are one of the few things still made in the US.
[20:46:35] -!- Valen [Valen!~Valen@c211-30-128-202.blktn6.nsw.optusnet.com.au] has joined #linuxcnc
[20:47:41] <XXCoder> http://www.aliexpress.com/item/CNC-router-kit-metal-milling-machine-metal-engraving-machine-3050116C/32262332104.html
[20:47:44] <XXCoder> interesting
[20:48:09] <ssi> argh I'm stymied
[20:48:30] -!- adoyle88 [adoyle88!cef82b35@gateway/web/freenode/ip.206.248.43.53] has joined #linuxcnc
[20:50:33] -!- stustev1 [stustev1!~stustev@wsip-184-176-200-171.ks.ks.cox.net] has parted #linuxcnc
[20:51:06] <XXCoder> interesting. http://www.aliexpress.com/item/PCB-2020B-engraving-machine-Mini-DIY-2020B-CNC-router-without-the-control-box-2020B-CNC-frame/1129455336.html
[20:51:13] <XXCoder> cheap but used ehh
[20:51:52] <XXCoder> looks like it is made from cheap plastic
[20:54:35] <XXCoder> furrywolf: think this is worth it http://www.aliexpress.com/item/3040-CNC-router-milling-machine-mechanical-kit-CNC-aluminium-alloy-Frame-ball-screw-for-DIY-user/2051273724.html
[20:54:48] <XXCoder> because I now has job that earns enough for me to afford it.
[20:54:53] <XXCoder> skip the crappy wood stage
[20:55:05] <XXCoder> go stright to crappy alum stage lol
[20:55:38] * furrywolf thinks very few things on aliexpress are worth it
[20:56:10] -!- TTN has quit [Ping timeout: 244 seconds]
[20:57:12] <furrywolf> that does look like a lot of aluminum for the price, however
[20:57:22] <XXCoder> shipping isnt too bad
[20:57:24] <XXCoder> $166
[20:58:55] <XXCoder> lol theres one $300 kit with $800 shipping. but then it has b5 inch block of something
[20:59:03] <XXCoder> looks like conscerete
[20:59:47] <furrywolf> lol
[21:00:02] <XXCoder> I would pour my own thanks LOL
[21:00:47] <furrywolf> I guess that's one way to add some mass to your machine, make it a bit more rigid...
[21:00:54] <XXCoder> jeez. one seller forgot one slight detail
[21:01:02] <XXCoder> how large working space is. :P
[21:01:17] <adoyle88> Hello everyone. My dad and I are attempting to make a CNC tubing bender and we have a few questions about using linuxcnc. Is it ok for me to ask them here?
[21:01:20] <XXCoder> for all I know it's 1"x1"x0.01"
[21:01:34] <XXCoder> ask away and see if someone knows
[21:01:41] * furrywolf waits for someone to build a large mill entirely from prestressed concrete
[21:01:43] <XXCoder> and wait a bit, that expert may take a while
[21:01:56] <cradek> adoyle88: absolutely. always go ahead and just ask
[21:02:00] <XXCoder> furry someone made small router cnc with epoxy granite
[21:02:34] <adoyle88> The bender we've built bends in a single plane so it just needs to feed the tube, make a bend and then repeat. We have a stepper motor that feeds the tube. That was easy enough for us to figure out.
[21:03:07] <adoyle88> What we need help with is using linuxcnc to trigger a solenoid operated hydraulic valve to pressurize the hydraulic cylinder. We would like to use an optical encoder wheel to count degrees of travel and reverse the valve when the bend reaches the target angle.
[21:03:08] -!- skunkworks has quit [Read error: Connection reset by peer]
[21:03:29] <adoyle88> We are using a break out board from probotix. For the encoder, we planned to machine out our own slotted wheel and use a simple optical sensor.
[21:03:45] <adoyle88> First question: Can I output a signal to trigger the solenoid operated valve? Second: Can I use linuxcnc to count pulses from the optical sensor? Third: Can I send another signal when the target angle is reached?
[21:05:05] <XXCoder> furrywolf: another not bad one. larger and free shipping http://www.aliexpress.com/item/Free-shipping-3020-CNC-router-DIY-CNC-frame-with-trapezoidal-screw-for-small-engraving-machine/32286449808.html
[21:05:22] <_methods> adoyle88: i would say yes to all 3
[21:05:48] <XXCoder> encoder probably would work good for angle?
[21:05:54] <XXCoder> _methods: what ya think
[21:06:06] <_methods> he's making his own encoder
[21:06:21] <XXCoder> yea
[21:06:38] -!- arek_at_work has quit [Quit: Konversation terminated!]
[21:06:42] <_methods> but you could probably skip the encoder and just use a limit switch on a protractor base
[21:06:56] <furrywolf> adoyle88: yes on all three. but I don't know nearly enough to answer how.
[21:07:22] <XXCoder> adoyle88: is it always same angle?
[21:07:23] <furrywolf> _methods: that wouldn't be programmable
[21:07:50] <adoyle88> the angle varies
[21:08:18] <furrywolf> XXCoder: that one has normal screws, not ball.
[21:08:28] <XXCoder> yea and smaller than first obne
[21:08:33] <XXCoder> so forget that
[21:08:48] <_methods> then you'll need some sort of method to bend to set angles like you have planned with your encoder
[21:08:48] <adoyle88> for example, on of our pieces is a roll hoop for a car. It will bend 60 degrees, feed 15 inches, bend 90 degrees, feed 40 inches, bend 90, feed 15, bend 60
[21:08:49] <XXCoder> Engraving Area:390×285mm
[21:08:56] <XXCoder> other is only 300x200
[21:09:49] <adoyle88> those values were arbitrary, they might make a ridiculous shape if you actually draw them
[21:09:52] -!- nerdfiles has quit [Ping timeout: 245 seconds]
[21:09:56] <_methods> yeah
[21:10:10] <_methods> i understand you need to make multiple bends of diff angles on 1 piece
[21:10:20] <adoyle88> yep
[21:10:29] <_methods> an encoder should work
[21:11:07] <XXCoder> would guess - tolence level of say 0.1 degree?
[21:11:18] <adoyle88> more like .25 degree
[21:11:19] <_methods> that depends on the resolution of his encoder
[21:11:19] <Valen> morning chaps
[21:11:20] <XXCoder> I cant imange car frames being that precise
[21:11:33] <adoyle88> maybe even .5 degree. we have quite a bit of wiggle room
[21:11:50] <XXCoder> _methods: yeah was thinking about bounds but encoder probably can easily do wat better
[21:11:58] <furrywolf> my experience with prebent tube things is prepare for some prying anyway. :)
[21:12:19] <Valen> I'm trying to make some nice 7mm shaft for a bearing mount out of 4140 and I'm not having much luck
[21:13:02] <adoyle88> we have jigs that will massage pieces into place for welding
[21:13:18] <Valen> the surface finish is all yicky like its tearing the metal, if i use a carbide or one of those diamond shape HSS toolholders
[21:13:58] <_methods> i think the tubing bender would be a perfect match for classic ladder
[21:14:10] <furrywolf> adoyle88: linuxcnc should be able to do what you need, but I'm afraid I have no clue how to implement it.
[21:14:28] <_methods> but i don't know i've never tried to set somethign like that up with classicladder
[21:15:00] <furrywolf> you might be able to define the bending as an axis in torque servo mode, in which case no manual setup would be needed at all.
[21:15:27] <adoyle88> Valen: I actually work at a machine shop (a part is running right now while I get help here). What RPM are you turning?
[21:15:49] <Valen> thats a good question lol
[21:15:58] <XXCoder> that would work I would guess furry
[21:16:15] <Valen> we have a VFD on the lathe, but I don't know the relationship between motor and spindle RPM
[21:16:19] <XXCoder> drawing on screen would look really funny
[21:16:23] <Valen> it is going pretty fast
[21:16:26] -!- Computer_Barf has quit [Quit: Leaving.]
[21:16:38] <Valen> I found taking a larger cut seemed to help, so I was doing a .7mm finish pass
[21:16:52] <Valen> taking less seemed to make it more furry
[21:17:19] <adoyle88> Valen: When it looks like it is tearing, you generally need to speed up the spindle.
[21:17:38] <Valen> I'll give it a go
[21:17:55] <_methods> adoyle88: i'd look at some encoder feedback examples
[21:18:04] <furrywolf> XXCoder: I don't think it'll draw an accurate representation without significant coding, no matter how you implement it. as far as I know (which isn't much, I'll admit), none of the guis have any concept of bending. heh.
[21:18:07] <adoyle88> that finish pass is going to need a really high spindle speed.
[21:18:17] <XXCoder> indeed!
[21:18:25] <XXCoder> but if it works it works
[21:18:27] <Valen> its at a stage where the bearing won't fit on it, then if i rub it with sand paper for a few seconds its a kinda loose fit
[21:18:33] <XXCoder> pipes get bent and adoyle gets happy
[21:18:48] <Valen> what about the depth of cut?
[21:19:00] <adoyle88> XXCoder: It'll be alright if the drawing doesn't look right. We've used our mills as lathes by putting the part in the spindle and the tools on the table. Those never look right on the screen
[21:19:16] <adoyle88> that was for furrywolf*
[21:19:18] <XXCoder> adoyle88: damn
[21:19:25] <XXCoder> thats heck of hack
[21:19:51] <Valen> hack is when you do that but put a dremel on the table ;-P
[21:19:59] <furrywolf> lol
[21:20:08] <XXCoder> Valen: I once saw one shop that used dermel on spidle
[21:20:14] <XXCoder> so it could do 90 degree drills
[21:20:15] <cradek> I've sure put a bench grinder on the table
[21:20:23] <adoyle88> haha
[21:20:24] <XXCoder> spidle rotates then moves in so dermel drills
[21:20:34] <cradek> http://timeguy.com/cradek-files/01188770083/DSCN6327.JPG
[21:20:50] <cradek> grinding tooling for itself, actually
[21:20:51] <XXCoder> grinder lol
[21:21:04] <Valen> cradek: hmm, is it bad I'm thinking about doing that now for my shaft?
[21:21:18] <furrywolf> lol
[21:21:25] <cradek> what's bad about a clever setup to make a part you need with what you have?
[21:21:36] <cradek> that's called being a machinist
[21:22:07] <Valen> I should just be able to lathe a shaft to 7mm lol
[21:22:38] <furrywolf> I still want to use my mill to cnc sharpen drill bits. will need to do that one of these days.
[21:22:51] <Swapper> anyone seen any cheap thread mills (the cutting sort) on ebay ?
[21:22:55] <Swapper> cant find that realy
[21:22:59] <Swapper> would like to try it out
[21:22:59] <XXCoder> wouldnt lathe be better (or mill used as lathe)
[21:23:06] <cradek> the passable but not beautiful result: http://timeguy.com/cradek-files/01188770083/DSCN6329.JPG
[21:23:30] <XXCoder> cradek: you made own umm chuck?
[21:24:06] <cradek> changed a NMTB30 jacobs to QC30
[21:25:14] <_Sync_> Valen: try a hss bit, they are less likely to tear the metal
[21:25:30] <_Sync_> if it is particularily annoying, try a shear tool
[21:26:05] <cradek> Valen: yes, hss, and have it SHARP
[21:26:06] <adoyle88> can I share an imgur link here of my current abstract setup?
[21:26:10] <furrywolf> LOL! reading the local news... every year, on 4/20, one of the local parks gets filled with useless potheads. the cops told them last year they weren't allowed to do that. now they're claiming it's a "Free speech event" and can't be told no. the cops solution: reserve the park in advance for officer training on that day. :)
[21:26:10] <_methods> adoyle88: you'll probably need to do somethign like this
[21:26:10] <Valen> I have one of these http://www.eccentricengineering.com.au/index.php?option=com_content&view=article&id=2&Itemid=18
[21:26:13] <_methods> http://www.mogi.bme.hu/TAMOP/robot_applications/math-ch10.html
[21:26:23] <cradek> adoyle88: always
[21:27:02] <_Sync_> try it, those are said to be working fine
[21:27:17] <Valen> I did, no joy
[21:27:21] <Valen> well dad did anyway
[21:27:33] <Valen> though he was taking a fine cut for his finish pass
[21:27:35] <_Sync_> well, then something was wrong
[21:27:44] <_Sync_> those usually like a fine cut when sharp
[21:27:44] -!- deanputney [deanputney!~mustardha@2601:8:a581:a700:2d2b:90f8:e6d7:ae4f] has joined #linuxcnc
[21:27:51] <_Sync_> but you need non suck HSS for your steel
[21:28:02] <adoyle88> we've made a gear box to turn multiple saws to cut grooves in stainless steel turbine backers: http://i.imgur.com/ZBsBj9w.jpg
[21:28:05] <_Sync_> preferrably cobalt containing
[21:28:33] <Valen> yep that is the one we have
[21:28:42] <cradek> adoyle88: neato
[21:28:52] <_Sync_> sharpened it properly?
[21:28:56] <Valen> yep
[21:29:00] <_Sync_> also do you know the state of your steel?
[21:29:03] <_methods> it's a vertical horizontal lol
[21:29:11] <Valen> comes pre-hardened and tempered
[21:29:24] <_Sync_> hm prehard, as expected
[21:29:25] <Valen> I believe its the 4140 steel, it doesn't like fine cuts for finishing, at least thats what everybody seems to say
[21:29:32] <_Sync_> yeah
[21:29:35] <_Sync_> because it is pre hard
[21:30:11] <Cromaglious> swapper I've been looking and haven't found any yet either
[21:30:12] <Valen> The CnC stuff yesterday came out ok ;-> http://imgur.com/a/ccQ3N
[21:30:20] -!- skunkworks [skunkworks!~chatzilla@str-broadband-ccmts-ws-26.dsl.airstreamcomm.net] has joined #linuxcnc
[21:30:27] <_Sync_> just get the grinder out
[21:30:31] <Swapper> Cromaglious: ok, weerd
[21:31:22] <Cromaglious> I found some not ebay but it's like $56 for one to do 1/8" screw hole
[21:31:33] <_Sync_> yeah
[21:31:37] <_Sync_> sounds reasonable
[21:31:46] <Valen> don't have a toolpost grinder :-<
[21:32:20] <Cromaglious> I don't have a grinder stable enough to use as a tool post grinder.
[21:32:39] <_Sync_> a regular good quality die grinder "works"
[21:32:59] <Valen> suggestions for how to hang on to the bastard?
[21:34:09] <Cromaglious> My dad has a nice potrer cable die grinder, but he's 4500 miles away
[21:34:12] <Cromaglious> My dad has a nice potrer cable die grinder, but he's 450 miles away
[21:34:59] -!- syyl_ [syyl_!~sg@p579F77D8.dip0.t-ipconnect.de] has joined #linuxcnc
[21:35:06] <_Sync_> rough it out, and polish it
[21:35:11] <_Sync_> or use some non suck steel
[21:35:11] <adoyle88> anyone know of a good hal tutorial? I feel like a deer in headlights reading through some of the stuff
[21:35:18] <cradek> adoyle88: yes, the hal tutorial
[21:35:25] <Cromaglious> harveytool.com
[21:36:03] <adoyle88> advanced was the first result on google. this oughta be fun ;)
[21:36:07] <cradek> adoyle88: http://www.linuxcnc.org/docs/html/
[21:36:21] <cradek> this is the index to all the docs
[21:36:47] <adoyle88> awesome, thanks
[21:36:50] <cradek> looks like there are two hal tutorials now
[21:39:58] <Cromaglious> http://www.harveytool.com/cat/Thread-Mill-Cutters/Specialty-Profiles/Browse-Our-Products_203.aspx
[21:40:32] <Cromaglious> decent prices, but still to steep for my budget
[21:40:57] -!- nerdfiles has quit [Read error: Connection reset by peer]
[21:41:19] <adoyle88> I just remembered another interesting way we used a CNC mill. We put a rotary table facing upward on the table and a lathe tool holder in the spindle with a carbide insert. The spindle did not spin. Instead, it moved up and down cutting splines into a part in the rotary table.
[21:42:01] <adoyle88> once the spline was to depth, the table rotated and the machine cut the next spline
[21:44:23] <MrSunshine> Valen: thats freakin pornographic =)
[21:44:47] <Valen> heh as one of my friends said "you can tell who doesn't pay for machine time by the hour"
[21:46:00] <furrywolf> adoyle88: I've seen that done
[21:47:08] <adoyle88> we were doing 200 super high tolerance splined fittings. time was also crucial to make money. that was the best way to do it
[21:47:19] <adoyle88> best way for us*
[21:47:33] <Cromaglious> Swapper, Most of my sizes could be cut with a #8, #12, 1/4,5/16,1/2" cutters thread 8-32 to 7/8"-10
[21:47:46] -!- jthornton has quit [Read error: Connection reset by peer]
[21:47:56] -!- jthornton [jthornton!~john@184.21.239.59] has joined #linuxcnc
[21:47:57] -!- JT-Shop has quit [Read error: Connection reset by peer]
[21:48:09] -!- georgenz has quit [Ping timeout: 256 seconds]
[21:48:57] * furrywolf wonders what a "shredder2" is, because despite the name, it sure doesn't look like a heavyweight robot wars entry...
[21:49:10] -!- JT-Shop [JT-Shop!~john@184.21.239.59] has joined #linuxcnc
[21:49:22] <Swapper> Cromaglious: i where looking for 1 of these https://www.youtube.com/watch?v=T4ZVoHei1uE
[21:49:29] -!- micges has quit [Quit: Ex-Chat]
[21:49:39] <MrSunshine> Valen: hehe .. =) my friends are lucky like that with my machine also ... 12 hours of work in cad/routing in the router ... payed for 1hour :P
[21:49:51] <Swapper> 60degree v cutter seems to be able to cut some metric threads (im all metric =:)
[21:50:25] <MrSunshine> well ... threads are 60 degrees? :P
[21:50:28] <Valen> furrywolf: its a 1.3kg robotwars robot
[21:50:41] <Deejay> gn8
[21:50:50] <Cromaglious> Swapper, yeah that's what the Harvey tool is
[21:50:56] <Valen> furrywolf: http://www.robowars.org/forum/viewtopic.php?t=1751
[21:51:14] <furrywolf> I don't know if anything that small counts. :P
[21:51:19] -!- Deejay has quit [Quit: bye]
[21:51:41] <furrywolf> (yes, I'm aware they have classes for tiny robots.... but it just doesn't quite seem in the spirit)
[21:51:51] -!- moorbo [moorbo!~moorbo@dyn-21-121.mdm.its.uni-kassel.de] has joined #linuxcnc
[21:52:40] <Valen> well you give me the mill and the cheque book that will fit turning 400kg of aluminium into 10kg and I'll build one of those
[21:52:44] * furrywolf was planning on a SHW walker, but decided useful things were a much better use of time and money
[21:52:51] <furrywolf> lol
[21:52:56] <Valen> we only run 13.6kg bots here as our heaviest
[21:53:12] <Valen> but they typically have as much KE as a .50cal armour peircing round in their spinners
[21:54:04] <furrywolf> I was planning on a crusher... nothing like a nice fat hydraulic ram to make other things smaller.
[21:54:42] <Valen> yeah, you just need the other guy to hold still while you are doing that
[21:54:44] <Cromaglious> ok back to work figuring out DMX board
[21:55:01] <Valen> rather than hitting you with super high KE spinning weapons
[21:55:44] <furrywolf> nah. you just need jaws strong enough to take the blow, and fast enough closure that they don't get away...
[21:56:13] <furrywolf> the same hunk of steel that can take tends of thousands of pounds of force can also take a weapon hit
[21:56:24] <Valen> see you have all this fancy jaw mechanism, I have the same weight of steel as just a plain bar and its doing 5000RPM
[21:56:34] <furrywolf> the problem with hydraulics is most people who used them couldn't figure out speed, and they were SLOW. pathetically slow.
[21:56:48] <Valen> accumulators are explicitly banned
[21:56:51] <furrywolf> let's roll around at 1mph and take 30 seconds to actuate our weapon!
[21:57:01] <furrywolf> not in all arenas.
[21:57:06] <Valen> (mostly because I think the safety people are idiots)
[21:57:46] <Valen> put it this way, in the 13.6kg weight class, we punch holes in 6mm high hardness armour steel plate
[21:57:54] <Valen> like tear chunks of it off
[21:58:21] -!- gonzo_nb has quit [Read error: Connection reset by peer]
[21:58:23] <Valen> crushers and the like will work, but in the "no spinners" class
[21:58:40] * furrywolf thinks no one has made a good crusher yet
[21:59:18] <furrywolf> there was one that did well in robot wars, can't remember its name right now, but otherwise every one I've seen sucked in some way.
[21:59:21] <Valen> because everybody who has competed knows they will get eaten in the first 4 seconds against a big spinner
[21:59:25] <Valen> razor
[21:59:38] <Valen> it stopped competing when people started entering spinners
[22:00:28] -!- toast-work has quit [Quit: Page closed]
[22:00:33] -!- FinboySlick has quit [Quit: Leaving.]
[22:01:25] <Valen> razer sorry
[22:01:28] <furrywolf> my plan was a crusher with enough jaw strength to absorb a spinner hit... and a lot of those spinners are nice and squishy in the middle.
[22:01:30] <Valen> http://en.wikipedia.org/wiki/Razer_(robot)
[22:01:53] <furrywolf> yep, that's the one.
[22:03:25] -!- logger[mah] has quit [Ping timeout: 264 seconds]
[22:03:30] <Valen> right, this is what is winning these days http://files.tested.com/photos/2012/11/28/41692-lastrites_teaser.jpg
[22:03:32] <Valen> or similar
[22:03:56] <furrywolf> didn't that lose to a simple rammer recently?
[22:04:08] -!- karavanjo1 has quit [Ping timeout: 272 seconds]
[22:04:13] <Valen> see the 75lb lump of steel hanging off the front?
[22:04:16] <furrywolf> sewer snake or something like that
[22:04:33] <Valen> entirely feasible
[22:04:41] <Cromaglious> that looks like a Hard edge off a loader bucket
[22:04:52] <Valen> the two robot styles are brick and the grenade
[22:04:55] <Valen> mortal enemies
[22:05:43] <Valen> the heavyweights aren't as exciting as the feathers anyway
[22:05:54] <Valen> its too expensive to go as nuts with them
[22:06:11] <Valen> so they are always slow and pound for pound under performing
[22:06:20] amnesic_away is now known as amnesic
[22:06:59] -!- logger[mah] [logger[mah]!~loggermah@mah2.mah.priv.at] has joined #linuxcnc
[22:07:14] <furrywolf> I've been doing some playing with high-efficiency hydraulics... some prototype motors are one of the first things on the list when I get my new machine fully functional...
[22:13:40] -!- mttr [mttr!~mttr@0016019a1697.cpe.westmancom.com] has joined #linuxcnc
[22:14:37] <furrywolf> your thread mentions ring deforming... what about a multi-part ring with some form of crushable material between the layers?
[22:14:53] <Valen> I have been thinking a hydraulic crusher for the sportsman (no crusher) class
[22:15:09] <Valen> I thought about it, but I don't have any facility to balance stuff
[22:15:11] <furrywolf> or mill a honeycomb shape into the middle of the ring
[22:15:49] <Valen> I might wind up doing that on this one, I've taken at least 100 grams out of the ring for the upgraded shell
[22:16:41] <furrywolf> that is, solid in the middle where it runs on the bearings, a layer of honeycomb, strong on the outside for the teeth... if you design the honeycomb right, it should transfer torque well, but crush without deforming the inside too much.
[22:17:38] <furrywolf> no crusher class?
[22:17:46] <Valen> yeah, I had the same idea
[22:17:54] <Valen> sorry, no spinner class
[22:18:22] <Valen> I'm just loath to do it because it runs the risk of unbalancing the ring and I don't have any good way of balancing it
[22:18:51] <furrywolf> hydraulics are hard to make both fast and strong, and that's where most designs fail... they make them strong, and they're so slow they spend the whole time getting killed without their weapon ever doing anything.
[22:18:58] <furrywolf> bring spare rings? heh
[22:19:22] <Valen> if the ring bends at speed it'll just destroy the robot ;->
[22:19:37] <furrywolf> speed and maneuerability are at least as important as weapon effectiveness.
[22:19:41] -!- MacGalempsy [MacGalempsy!~quassel@ip72-204-45-74.fv.ks.cox.net] has joined #linuxcnc
[22:19:53] <Valen> actually driving ability is one of the most important things
[22:19:58] <Valen> and you can't buy it
[22:20:05] <furrywolf> yep
[22:20:13] <MacGalempsy> hello
[22:20:14] <Valen> the guy here who builds the scariest robots, is also the crappest driver
[22:20:16] <_Sync_> Valen: a balancing rig is easy to make
[22:20:18] <furrywolf> lol
[22:20:46] <Valen> _Sync_: balancing a ring with no center?
[22:21:01] <_Sync_> yeah you have to make an adapter for that
[22:21:05] <_Sync_> but the rest is simple
[22:21:05] <furrywolf> what about nested leaf springs?
[22:22:04] <furrywolf> like very curved spokes... able to transfer rotation well, but just springing on side impacts
[22:22:48] -!- mhaberler has quit [Quit: mhaberler]
[22:28:09] -!- cnc1 has quit [Quit: ChatZilla 0.9.91.1 [Firefox 11.0/20120310193829]]
[22:29:45] -!- MacGalempsy has quit [Read error: Connection reset by peer]
[22:31:17] -!- mhaberler [mhaberler!~mhaberler@extern-171.stiwoll.mah.priv.at] has joined #linuxcnc
[22:32:17] <furrywolf> bbl
[22:33:31] -!- MacGalempsy [MacGalempsy!~quassel@ip72-204-45-74.fv.ks.cox.net] has joined #linuxcnc
[22:35:32] -!- Nick001-shop [Nick001-shop!~chatzilla@50.107.167.196] has joined #linuxcnc
[22:35:37] <_methods> do you need to feed those chinese db25 breakout boards with 5v from the pc you're using to control them?
[22:35:49] <_methods> or can you just power them with 5v from anywhere?
[22:38:01] -!- patrickarlt has quit [Remote host closed the connection]
[22:38:20] <furrywolf> you can power them with 5v from anywhere, but keep in mind it'll be fed straight to the pc, so it needs to be regulated and clean, and won't be isolated.
[22:38:31] <furrywolf> I got an isolated 24v to 5v dcdc for mine
[22:39:29] <_methods> wasn't sure if the 5v needed to come from signal side
[22:40:14] <furrywolf> well, that's how the one I have is built. yours could be different. :P
[22:40:52] <_methods> it can be fed from both and it has jumpers to split for isolation
[22:41:30] -!- TTN [TTN!~TTN@27.110.122.139] has joined #linuxcnc
[22:41:30] -!- TTN has quit [Changing host]
[22:41:30] -!- TTN [TTN!~TTN@unaffiliated/ttn] has joined #linuxcnc
[22:42:22] -!- unfy [unfy!~Miranda@wsip-184-185-82-30.om.om.cox.net] has joined #linuxcnc
[22:42:58] <MarkusBec> where can i find the current information about installing the 7i92 ethernet mesa card
[22:43:35] <MarkusBec> the wiki infos seems to be outdated
[22:45:18] -!- Camaban has quit [Ping timeout: 272 seconds]
[22:48:37] -!- georgenz [georgenz!~georgenz@118.148.181.28] has joined #linuxcnc
[22:49:12] <georgenz> whats a good picture sharing website?
[22:49:41] <_methods> imgur
[22:50:01] <georgenz> sweet - I'll just sign up
[22:50:46] <_methods> you don't have to sign up
[22:50:53] <_methods> you can just upload the pics then share
[22:52:52] <georgenz> oh - i've jst signed up
[22:52:56] <georgenz> http://imgur.com/0yRUlA0
[22:53:12] <georgenz> this is my z axis chattering
[22:54:38] <MarkusBec> PCW: pcw_home any tipps for the 7i92
[22:54:45] <georgenz> teco 2kW motor and drive - not sure how to fix, sometimes it's not his bad, I'll move it to a position where it is less angry.
[22:54:55] -!- stustev [stustev!~stustev@ip68-102-48-177.ks.ok.cox.net] has joined #linuxcnc
[22:54:58] -!- jvrousseau has quit [Quit: Textual IRC Client: www.textualapp.com]
[22:55:36] <PCW> what are the units?
[22:56:52] <georgenz> http://imgur.com/c3BsHqJ
[22:57:19] <PCW> units?
[22:57:20] <georgenz> less angry, jitter noise never goes away, but it does get better or worse depending on position
[22:58:01] <georgenz> 1m (not sure)/div vertical
[22:58:13] <PCW> inches?
[22:58:18] <georgenz> metric
[22:59:22] <georgenz> 2msec/div horizontally
[23:00:12] <PCW> most of that is due to the encoder resolution limits
[23:00:32] -!- deanputney has quit [Read error: Connection reset by peer]
[23:00:38] <PCW> you can see that its quantized to +-2 counts
[23:01:08] -!- deanputney [deanputney!~mustardha@2601:8:a581:a700:2d2b:90f8:e6d7:ae4f] has joined #linuxcnc
[23:02:10] <Tom_itx> maybe run a pico psu on it
[23:02:17] <PCW> a little I term may help get rid of it
[23:02:32] <Tom_itx> err wrong channel but you're here too :D
[23:02:44] <PCW> yeah a pico PS should run a J1900 fine
[23:02:53] <Tom_itx> i've got a 120 on this
[23:03:23] <PCW> much more than my J1900 HTPC (50W wall wart)
[23:03:25] -!- MacGalempsy has quit [Ping timeout: 255 seconds]
[23:04:25] <Tom_itx> i think i'll put an order together tonight
[23:04:35] <PCW> MarkusBec tips?
[23:04:40] <MarkusBec> PCW: is this the current how to vor hm2_eth and uspace installtion http://linuxcnc.org/index.php/english/forum/27-driver-boards/27253-7i80-and-7i77?start=50#51638
[23:04:53] -!- Miner_48er has quit [Quit: Leaving]
[23:04:53] <MarkusBec> for
[23:05:41] -!- nerdfiles has quit [Read error: Connection reset by peer]
[23:05:50] <PCW> should be OK (or you can download the packages from the buildbot if you dont want to build from source)
[23:06:35] <MarkusBec> how can i install it form the buildbot
[23:07:21] <PCW> I think the buildbot has some (rather terse) instructions
[23:07:52] <_methods> PCW: the 5i24 is the same as the 5i25 but low profile?
[23:09:20] <PCW> The 5I24 is a 72 I/O FPGA card (with 3x 50 pin connectors)
[23:09:56] <PCW> 5I25 is a 34 I/O FPGA card with (dual) parallel port type connectors
[23:10:12] <PCW> (also 5I24 has a bigger FPGA)
[23:10:23] <_methods> is there low profile version of the 5i25?
[23:10:25] <PCW> both are low profile
[23:10:31] <_methods> ah ok
[23:12:51] <_methods> you have the 5i25 7i76 plug and go kits in stock?
[23:13:01] <PCW> Yes
[23:13:12] <_methods> cool
[23:13:37] -!- adoyle88 has quit [Ping timeout: 246 seconds]
[23:13:56] -!- MacGalempsy [MacGalempsy!~quassel@ip72-204-45-74.fv.ks.cox.net] has joined #linuxcnc
[23:15:53] <georgenz> PCW, FYI and not sure if this makes any difference to your answer, it does this jittering whether P is at 1 or at 12, makes little difference
[23:16:07] <PCW> MarkusBec: other 7i92 (hm2_eth) hints:
[23:16:09] <PCW> Dont use a slow CPU (D525)
[23:16:11] <PCW> Preemt-RT 3.18.7 is a better RT kernel than the stock one with wheezy
[23:16:13] <PCW> Intel MACs need interrupt coalescing turned off
[23:16:14] <PCW> Set Ethernet device to an oddball IP address (say the EEPROM default of 10.10.10.10)
[23:17:15] <PCW> georgenz: that means its mainly in the drives velocity loop (assuming you just have P and FF1 in linuxcnc's PID loop)
[23:17:53] <georgenz> Yeah - I have P and FF1 set in linuxcnc
[23:18:42] <PCW> So you may need to tune the drives velocity loop
[23:20:09] <PCW> (I assume the drive is connected to the machine, bare drives are nearly impossible to tune)
[23:20:33] <georgenz> ok, that sortof makes sense too as I shorted the drive analog input to ground and it still did it
[23:21:07] <georgenz> yeah - it has maybe a 300kg spindle head hanging off it
[23:21:11] <PCW> may have to lower the drives velocity loop gain a bit
[23:21:11] -!- eFuchs [eFuchs!~home_swee@dyn-31-25-157-250.ewacom.ropa.net] has joined #linuxcnc
[23:21:15] <georgenz> jittering up and down
[23:21:35] <georgenz> ok, thanks i'll give it a go
[23:21:44] -!- nerdfiles has quit [Read error: Connection reset by peer]
[23:21:46] <georgenz> what about turning the drive onto auto tune?
[23:22:20] <PCW> you can try...
[23:23:05] <PCW> you could also just adjust the velocity loop gain
[23:23:36] <PCW> (assuming you have a easy way to do this)
[23:23:49] -!- moska1 has quit [Ping timeout: 264 seconds]
[23:25:18] <georgenz> do you know what the load-inertia ratio does?
[23:25:38] -!- deanputney has quit [Read error: Connection reset by peer]
[23:25:51] <georgenz> the manual also refers to that under speed loop gain
[23:26:11] -!- deanputney [deanputney!~mustardha@2601:8:a581:a700:2d2b:90f8:e6d7:ae4f] has joined #linuxcnc
[23:28:56] -!- akex2 has quit [Quit: Connection closed for inactivity]
[23:32:50] -!- ve7it [ve7it!~LawrenceG@S01060014d19d0b68.pk.shawcable.net] has joined #linuxcnc
[23:33:48] -!- deanputney has quit [Read error: Connection reset by peer]
[23:34:00] -!- Valen has quit [Remote host closed the connection]
[23:34:24] -!- deanputney [deanputney!~mustardha@c-67-183-112-251.hsd1.wa.comcast.net] has joined #linuxcnc
[23:34:42] -!- asah [asah!~asah@173-228-112-206.dsl.dynamic.fusionbroadband.com] has joined #linuxcnc
[23:34:57] <PCW> you probably have to lower loop gain if you have a large load inertia ratio so they are related
[23:36:52] <asah> pcw: having issues in commutation on a 7i39 using an absolute encoder
[23:37:24] <asah> what is the best method to ensure I have correctly aligned the encoder to the electrical phase of the motor
[23:40:02] -!- adb has quit [Ping timeout: 252 seconds]
[23:42:29] -!- nerdfiles has quit [Ping timeout: 246 seconds]
[23:43:08] <MarkusBec> PCW: with the documatation and the buildbot version
[23:43:37] <MarkusBec> I get these error ERROR could not retrieve mac adress
[23:45:20] <Nick001-shop> <PCW> Is there info on using 5i25/7i76 cards to convert a Bridgeport Series 1 with a boss 5 control to Linuxcnc? Also keeping the large steppers.
[23:45:28] -!- sirdancealot has quit [Ping timeout: 272 seconds]
[23:45:56] <renesis> when you do his nick like that its weird because it looks like you are quoting him
[23:56:22] -!- patrickarlt has quit [Remote host closed the connection]