#linuxcnc | Logs for 2012-10-01

[00:00:07] <Jymmm\AE> Hmmm
[00:01:21] <Jymmm\AE> thanks
[00:02:46] <djdelorie> good luck :-)
[00:45:18] <Loetmichel> mornin'
[01:13:20] <uw> has anyone ever tried to make a cnc shaper?
[01:26:43] <archivist> you only need to add a tool holder to a normal 3 axis mill and implement the shaper in gcode
[01:31:37] <archivist> the tool holder does not need the hinge because a cnc can lift the tool for the back stroke
[02:08:35] <DJ9DJ> moin
[05:30:42] <psha> logger[mah]: .
[05:30:42] <logger[mah]> psha: Log stored at http://linuxcnc.mah.priv.at/irc/%23linuxcnc/2012-10-01.html
[07:56:32] <pilla_> My terminal bug splats when I try to install OpenCNC...
[07:56:39] <pilla_> LinuxCNC*
[08:00:32] <jdh> OpenCNC might be a better name, but I don't really care for Open* names
[08:01:28] * jthornton can't find a translation for "terminal bug splats"
[08:01:59] <pilla_> :D My terminal gives a bug splat error.
[08:04:21] <jthornton> never seen a bug splat error before...
[08:04:35] <syyl_ws> my car has bug splats
[08:04:41] <syyl_ws> if they are terminal, who knows..
[08:05:13] <jthornton> in that case the bug don't have the guts to do that again
[08:05:43] <jthornton> ever seen parking lot birds hopping up and eating the bugs off of your bumper?
[08:09:59] <archivist> motor cyclist wear beards to soften the blow of insects at speed :)
[08:27:10] <pilla_> I get an error while trying to install LinuxCNC: Requires installation of untrusted packages. Packages are linuxcnc, linuxcnc-doc-en rtai-modules-2.... what to do?
[08:27:56] <cradek> are you installing from the buildbot, or one of the release repositories?
[08:28:02] <pilla_> repositories
[08:28:15] <cradek> what are your sources.list lines?
[08:28:57] <pilla_> linuxcnc.org/emc2/
[08:29:02] <pilla_> breezy
[08:29:04] <pilla_> and sources
[08:29:21] <cradek> paste the full lines please?
[08:29:38] <pilla_> http://www.linuxcnc.org/emc2/breezy emc2.1
[08:30:04] <pilla_> http://www.linuxcnc.org/emc2/breezy emc 2.1 (source code)
[08:30:10] <pilla_> is what I get in software center
[08:30:13] <cradek> those paths don't exist
[08:30:24] <pilla_> hmmmh
[08:30:25] <cradek> are you really running breezy?
[08:30:34] <cradek> what have you done to get to this point?
[08:30:35] <pilla_> lucid
[08:30:44] <pilla_> I followed the tutorial on the site xD
[08:30:45] <cradek> well then those lines are totally wrong.
[08:31:31] <cradek> go to www.linuxcnc.org, click download, click basic installation
[08:32:02] <cradek> you can use that cd image, or you can use the lucid-install script further down on the page
[08:32:33] <cradek> whatever you found is really outdated (breezy is from 2005). if you share the URL maybe someone can fix or remove it.
[08:32:45] <cradek> bbl.
[08:32:55] <jdh> cool.
[08:32:56] <pilla_> I did use that program actually
[08:33:16] <pilla_> the .sh file
[08:34:19] <pilla_> gpg: failed to create temporary file '-insert location-': permission denied
[08:34:34] <cradek> no, you didn't, because it doesn't point you to a breezy repository that doesn't exist
[08:35:11] <cradek> I just checked http://linuxcnc.org/install-scripts/lucid/linuxcnc-install.sh
[08:35:36] <cradek> there is nothing breezy-era on this page
[08:36:22] <pilla_> gpgkeys: http fetch error 6: couldn't resolve host 'pgpkeys.mit.edu'
[08:37:21] <pilla_> that's what I get when I run that .sh file, rest is now running
[08:37:38] <pilla_> t works :)
[08:37:39] <pilla_> thanks
[08:38:30] <pilla_> what machine lay-out suits a CNC 3020t?
[08:44:49] <pilla_> stepper_mm?
[08:45:15] <JT-Shop> what is a CNC 3020t?
[08:45:20] <pilla_> a miller
[08:45:44] <JT-Shop> a 3 axis mill?
[08:46:44] <pilla_> Yes
[08:47:24] <JT-Shop> well you could use the stepconf wizard to generate a configuration or pick one from the configuration picker
[08:49:01] <pilla_> Didn't know about the stepconf wizard. Thanks.
[08:49:11] <JT-Shop> welcom
[08:49:13] <JT-Shop> e
[09:15:52] <mazafaka> Shit. It looks like this spring something has happened while the mill was working and I was far. Foreman hsve cslled me and when I have come she looked onto my fingers. I think someone had been touching the mill and they simply hadn't told it to me. :/
[09:20:01] <mazafaka> this goddamn little plant is sucking whole the soul out of me.
[09:53:29] <pilla_> >.> how do I check the md5 checksum
[09:53:40] <jdh> md5sum
[09:55:11] <skunkworks> yikes http://www.cnczone.com/forums/showthread.php?p=1175454#post1175454
[09:55:15] <jdh> so, used boat
[09:56:59] <pilla_> isolinux disk error 32, AX = 4200, drive 9F <- when I try to start from the cd :(
[09:57:36] <JT-Shop> pilla_: have you read the Getting Started Guide yet?
[09:59:06] <pilla_> sorta
[09:59:34] <pilla_> My lector came over and said I'd better create a dualboot system, so I burned the iso on a dvd and tried starting it but meh
[10:00:26] <pilla_> Downloading the iso again, but I have no clue how to check the md5sum
[10:01:02] <pilla_> is it inside the iso? if yes, there's a text wall inside the md5sum, I have no idea which md5sum I have to compare
[10:05:04] <jdh> md5sum filename.iso, compare to the md5sum on the download page?
[10:05:28] <pilla_> under windows?
[10:06:54] <cradek> http://www.google.com/search?q=check+iso+md5sum+in+windows
[10:07:48] <JT-Shop> pilla_: that too is in the getting started guide
[10:07:52] <mazafaka> jdh: how do you know regular Linux distro shall have 'md5sum' included, or there are no regulations - everyone simply installs it by query?
[10:08:22] <jdh> I've never had to install it by itself.
[10:08:33] <jdh> but, that is not an exhaustive answer :)
[10:10:27] <pilla_> thank you
[10:14:35] <pilla_> I'll now try to follow the Getting Started guide and not to bother you guys too much :) Thank again
[10:15:26] <JT-Shop> no bother here
[10:29:49] <psha> mazafaka: usually it's part of coreutils or something similar
[10:29:54] <psha> depends from distro to distro
[10:31:29] <mazafaka> psha: how do you know it? Did you learn some Lnux basics? Or just have to deal with this all and know?
[10:31:56] <JT-Shop> http://bbs.homeshopmachinist.net/threads/56108-Is-IMSRV-the-CNC-stuff-seller-still-around
[10:36:51] <psha> mazafaka: server side programming for more then 10 years gives at least some basic knowledge about linux...
[10:37:08] <jdh> anyone heard pricing range for 7i76e?
[10:39:05] <pilla_> sigh
[10:39:21] <JT-Shop> pilla_: still have problems?
[10:40:03] <pilla_> Checked checksum, was wrong, re-downloaded, correct checksum, made bootable usb
[10:40:18] <pilla_> now I have "error no configuration file found, no default or UI directive found!"
[10:40:28] <pilla_> so yep
[10:42:08] <jdh> did you use the stepconf wizard?
[10:44:40] <pilla_> didn't even got there :) still trying to install Ubuntu and LinuxCNC in a dualboot system
[10:44:51] <pilla_> teacher said I shouldn't be using VMware so derp
[10:45:10] <jdh> nope.
[10:48:59] <JT-Shop> pilla_: how did you burn the CD?
[10:56:36] <mazafaka> psha: yeah, experience
[11:14:10] <sliptonic> O.k. Cool. Thanks.
[11:19:23] <tjb1> Inventor blows.
[11:19:42] <JT-Shop-2> and Autocad just bought out HMS too
[11:21:02] <tjb1> HMS easystretch?
[11:21:35] <JT-Shop> Today Autodesk Inc. and HSMWorks ApS jointly announced that Autodesk Inc. has acquired all HSMWorks technology.
[11:22:03] <tjb1> Ah great
[11:22:11] <tjb1> I suppose that CAM for solidworks is gonna turn into a pile of crap
[11:23:02] <tjb1> Already have Autodesk all over the HSMworks website
[11:26:21] <skunkworks> I cannot believe autodesk is still in businell
[11:26:25] <skunkworks> I cannot believe autodesk is still in business
[11:29:55] <archivist> biggest steaming pile of .....
[11:37:46] <BHSPiMonkey> yay, my shapeoko is being built
[11:38:18] <BHSPiMonkey> or rather, the parts are being fabricated, so that I will eventually be able to build it.
[11:54:38] <gabe_w> has anyone had any luck using python-rtai?
[11:54:57] <gabe_w> when i import rtai i get an lxrt error
[11:56:35] <gabe_w> AttributeError: /usr/lib/liblxrt.so: undefined symbol: rt_task_delete
[12:15:57] <IchGuckLive> Hi all !
[12:23:18] <skunkworks> wow
[12:23:19] <skunkworks> http://www.cnczone.com/forums/mach_software_artsoft_software/163875-mach_4_unveiled_imts_2012_a.html
[12:23:52] <skunkworks> dual closed loop with any version of industrual controls
[12:25:03] <jdh> en englese por favor?
[12:25:38] <skunkworks> nfc
[12:26:00] <jdh> any version of Industry Controls
[12:26:20] <jdh> http://www.cnczone.com/forums/attachment.php?attachmentid=168267&d=1348899245
[12:26:43] <jdh> industry controls seems to make a control based on mach?
[12:31:05] <jdh> the dual closed loop is closed in the drive.
[12:32:28] <Connor> Sorry, Still not sure I want a controller running on top of Windows to handle my machine... for safety reasons.. :)
[12:32:45] <jdh> but, it runs in linux too!
[12:32:51] <jdh> which seems to be utterly pointless
[12:33:04] <Connor> huh?
[12:33:11] <Connor> They have a linux version too ?
[12:33:31] <jdh> yep.
[12:33:58] <RussianKid> still closed source, still a potential botnet
[12:34:08] <jdh> still non-realtime
[12:34:17] <jdh> still no closed loop in the control
[12:34:22] <RussianKid> the linux version might be?
[12:34:28] <jdh> nope
[12:34:29] <RussianKid> might be RT*
[12:34:40] <RussianKid> oh does it say somewhere in the article?
[12:34:45] <jdh> but, still a winner for a large class of users
[12:36:30] <IchGuckLive> it will find its costomers shure
[12:36:52] <RussianKid> I had a huge argument with lot of people on mycncuk about it a few moths back. i know exactly what that 'large class of users' comprises
[12:37:14] <jdh> it works for them... be happy for them and move on
[12:37:32] <Connor> RussianKid: What was the argument about?
[12:37:38] <jdh> if it doesn't work so well for htem, laugh at them, feel superior, move on :)
[12:38:38] <RussianKid> nono, there's no point having choice if you're not allowed to do it, by all means use what you want, but at least be civil abou tit
[12:39:17] <RussianKid> it was a thread about smoothstepper boards being available in the uk, and about how it was a huge improvement for a generic NEMA stepper DIY CNC
[12:39:41] <skunkworks> the linux verion is just a fancy gui that feed info to a motion box. (sam for the most part on the windows side)
[12:39:52] <skunkworks> they just can't do realtime very well.
[12:39:56] <RussianKid> and its not really, because the smoothstepper primarily provides much higher step rates, which almost never is a limiting factor
[12:40:29] <RussianKid> so you're spending like a lot of money on hardware which doesn't really give anything, and it was blatantly misleading, and they made handwaving statements like 'an improvement in accuracy was noted on all machines'
[12:40:48] <skunkworks> yikes
[12:41:08] <skunkworks> because the mach tp and step generation isn't perfect....
[12:41:25] <jdh> looks like the new stuff offloads all step gen to HW
[12:41:48] <jdh> much like many here do with mesa
[12:42:20] <skunkworks> jdh, not exactly
[12:42:54] <RussianKid> the thing that bothers me the most is that there is a large cluster of people who refuse to believe that mach's measure distance and divide by steps wizard makes your machine as accurate as the instrument you used to measure the distance
[12:43:35] <skunkworks> mach requires a outboard motion cotrol that gets fed high level command and the motion happens in the magic box
[12:44:11] <andypugh> Which does lead me to wonder what Mach is doing.
[12:44:13] <jdh> yeah, still, all the step gen is done in hardware (and some more)
[12:44:18] <skunkworks> So you are stuck with the 'realtime' things that that motion device manufacuter builds in.
[12:44:31] <jdh> it's drip feeding a hw control
[12:45:10] <skunkworks> while emc does all the motion internally. the interface cards are just dumb (sorry pcw) encoder counter, pwm generators and i/o.
[12:45:51] <skunkworks> so any realtime things (like rigid tapping or such) added to linuxcnc are available to all devices. (that is why you can rigid tap with just printer port hardware if you want to)
[12:46:08] <RussianKid> wait will the HW motion box be compulsary, or will there still be the option to output via parallel port?
[12:46:46] <andypugh> p-port is only going to be available in the Pro version
[12:46:50] <skunkworks> sounds like pro will have the printer port still (although limited to 32bit)
[12:47:03] <skunkworks> and art doesn't know how long...
[12:47:17] <RussianKid> so theyve split not just windows/linux but also normal/pro? that's a lot of versions to maintain
[12:47:35] <skunkworks> but the lite version will require a external motion device.. (which puts it in the 299 price range so far)
[12:47:39] <andypugh> tbf it isn't Mach's fault if 64-bit wondows hides the paarport
[12:47:55] <FinboySlick> RussianKid: Maybe they wrote it in .Net ;)
[12:48:05] <skunkworks> which you can get a mesa setup that is a lot lot more flexible for $200
[12:50:04] <RussianKid> tl;dr mach failed to get better
[12:50:10] <gmouer> mesa boards and linuxcnc is a fair amount cheaper than mach and a smoothstepper or similar, and much more powerfull , done both
[12:50:12] <skunkworks> RussianKid, http://groups.yahoo.com/group/mach1mach2cnc/message/136901
[12:51:06] <skunkworks> gmouer, glad you converted!
[12:51:17] <gmouer> me too!!!
[12:51:28] <skunkworks> (I think I have read your post on the mach yahoo groups) correct?
[12:52:02] <gmouer> yes, and home shop machinist too just recently (Sparky_NY)
[12:52:12] <skunkworks> ah - Hi
[12:52:15] <skunkworks> :)
[12:52:28] <skunkworks> so when is the mill conversion starting?
[12:53:07] <RussianKid> >lowering the complications
[12:53:08] <RussianKid> >all about simple
[12:53:08] <RussianKid> it reminds me of that line on apple's site 'effortlessly adjust exposure, saturation and contrast with your mouse. You don't even need to know what those things are. Just click and drag until everything looks picture perfect'
[12:53:09] <gmouer> well, the mill actually works decent under mach so there is no big hurry, but its going to happen for sure
[12:53:19] <skunkworks> heh
[12:53:35] <gmouer> I am spoiled now with linuxcnc
[12:54:54] <gmouer> my 14x40 lathe gets reretrofitted from mach first, thats the next project, a simple one but the reason I went linuxcnc
[12:55:05] <gmouer> got fed up LOL
[12:55:25] <andypugh> RussianKid: Actually, that Apple idea sounds quite reasonable.
[12:55:59] <skunkworks> gmouer, Once you get buy the 'oh my god it is linux!' it isn't so bad :)
[12:56:08] <RussianKid> it's the way they word it. I mean you're editing a picture, why WOULD you want to know what exposure, saturation and contrast are?
[12:56:16] <andypugh> I didn't have that barrier, I was on a Mac.
[12:56:48] <RussianKid> i mean to each their own, but in my opinion, knowing what those things are is a prerequisite to image editing, not an optional extra
[12:57:10] <gmouer> the linux part wasn't all that bad, Ubuntu is not that different than windows to a new operator Linuxcnc is where the hairpulling began
[12:57:14] <FinboySlick> andypugh: But Macs drain away your testosterone...
[12:57:19] <andypugh> I slightly disagree. You dopn't need to know what they are called to be able to judge the effect.
[12:57:25] <Connor> One issue I have with Mach is it's interface is just dog ugly and overly complicated
[12:57:36] <andypugh> FinboySlick: I have no practiacl use for testosterone anyway.
[12:57:48] <RussianKid> there are lots of screensets for it, and its not a huge deal to write your own either
[12:58:05] <gmouer> yea, but that is also machs biggest strength, it interface gui and the ability to customize it easily
[12:58:05] <Jymmm\AE> andypugh: You are woman, we hear you roar! ;)
[12:58:27] <FinboySlick> andypugh: wait... which Andy Pugh are you? Switched to being the babe in the picture again?
[12:58:30] <andypugh> But you can create a custom GUI for LinuxCNC with only a few weeks solid work :-)
[12:59:10] <RussianKid> i've been wanting to change a fair bit of axis for ages, it's on my to do list
[12:59:11] <Jymmm\AE> FinboySlick: Did you see the bald guy in the photo?
[12:59:35] <FinboySlick> Jymmm\AE: No, he only showed me a picture of a babe.
[13:00:13] <Jymmm\AE> FinboySlick: The bald guy was NOT using a Mac and pulled out all his hair in frustration =)
[13:00:17] <FinboySlick> I had a question for the egghead version of Andy last night though. I hope he'll switch back sometime soon.
[13:00:35] <IchGuckLive> im off
[13:01:34] <jthornton> there is even a tutorial for creating your own GUI for linuxcnc
[13:01:44] <andypugh> Actually, there is supposedly a link between testosterone and baldness. SO if there _is_ a Mac/baldness anti-correlation, perhaps FinboySlick has a point.
[13:01:49] <RussianKid> there are the source files though
[13:02:37] * FinboySlick amused.
[13:03:09] <andypugh> Anyway, I am not babe-Andy (she's some sort of artist) so what was the question?
[13:04:57] <FinboySlick> andypugh: Metalurgical genius Andy... I was sort of trying to wrap my head around what the so called cold-tempering process does to steel. That is the technique of supercooling steel to supposedly remove internal stresses. I don't see how freezing material would do that.
[13:05:27] <FinboySlick> Or is it all gimicky pseudo-science?
[13:06:04] <andypugh> I am not 100% convinced that it does what they say it does.
[13:06:05] <RussianKid> I got a sheet of alu recently that said stress relieved by stretching. I've not read up on it but it doesn't make sense to me how stretching can remove stress either.
[13:06:23] <andypugh> It _will_ tend to reduce retained austenite.
[13:06:49] <andypugh> But that retained austenite will convert to martensite on transformation.
[13:08:35] <andypugh> Another possible effect is that if there are thermal stresses caused by shrinkage from the high forming temperaure, then going much colder will increase those stresses, possibly past the plastic limit. Then when you warm up again, they are gone.
[13:09:17] <FinboySlick> OK, so it could potentially be useful in already tempered material then.
[13:10:24] <andypugh> Look at it another way. If the part has thermal streses at room temperature, then they ought to be less at a higher temperature. Cryo-cooling makes -192C the new "room temperature" and the operating temperature is now relatively hot.
[13:11:19] <FinboySlick> andypugh: Yeah, that last bit I had understood.
[13:11:33] <FinboySlick> I had to wiki austentite and martensite though ;)
[13:13:15] <andypugh> Austenite is what steel is when it is hot. It then turns to Ferrite (tough, soft) and Martensite (brittle, hard) when you cool it. How you cool it, and the metal composition, affect the relative ratios of ferrite and martensite.
[13:16:46] <FinboySlick> I'm smarter now :)
[13:16:50] <FinboySlick> Thanks.
[13:17:47] <Jymmm\AE> Anyone familiar with thin (=<0.015") SS sheet grades? I need to BOW a 6"x12" sheet around a 4" diameter pipe and have it spring/flex/return back to it's flat state as much as possible. I'm tried 316, but it seems to bend more than it bows. 412 seems too stiff.
[13:19:11] <Jymmm\AE> I have some 312/314 hybrid samples on the way. but still not sure that's going to do it.
[13:21:12] <Jymmm\AE> The 0.010" aluminum roof flashing you can get from home depot/lowes seems to have the right characteristics, but need stainless instead.
[13:21:30] <andypugh> You probably need something cold-rolled. A spring steel grade
[13:22:06] <andypugh> Look for something like "hard temper"
[13:25:12] <Jymmm\AE> andypugh: Says "nickel" is the key component... http://en.wikipedia.org/wiki/Spring_steel
[13:25:48] <andypugh> Low nicket 316 is what Lee Steel Strip used.
[13:26:08] <Jymmm\AE> Lee Steel?
[13:26:24] <andypugh> Not the same as Lee Spring
[13:27:20] <andypugh> Merged with Avesta some time after I worked with them: http://www.cisionwire.com/avesta-sheffield/r/avesta-sheffield-purchase-of-lee-steel-strip-approved,e14095
[13:27:55] <andypugh> But, they used deliberately low-stability stainless steel and rolled it to massive hardness.
[13:28:04] <andypugh> (presumably they still do)
[13:28:18] <Jymmm\AE> andypugh: And what does that process accomplish?
[13:29:20] <andypugh> It gives the material a much harder yield stress. Effectively you need to bend it further than the rolling process did to reach the elastic limit.
[13:30:33] <Jymmm\AE> and the yield stress makes it want to return to it's original flat self?
[13:31:51] <BHSPiMonkey> Could someone advise on the pros/cons between two stepper motors? (One has 1.8deg steps and max speed of 600rpm, other has 0.9deg and 300rpm)
[13:32:14] <BHSPiMonkey> In other words it looks like I could trade off speed for finer step resolution
[13:32:46] <BHSPiMonkey> Torque is pretty identical between the two
[13:38:23] <Jymmm\AE> andypugh: I think I found a "plain english" SS Grade chart... http://www.mastainless.com/grades/index.html
[13:39:13] <Jymmm\AE> andypugh: See also: http://www.mastainless.com/tempered/index.html
[13:41:03] <FinboySlick> BHSPiMonkey: How fast do you need the machine to move?
[13:42:22] <FinboySlick> If torque is similar and it'll move as fast as you need, I'd say go for the 0.9° version and save some of the disadvantages of microstepping.
[13:43:32] <pcw_home> Static torque may be the same but torque vs speed at a given drive power supply
[13:43:33] <pcw_home> voltage will drop off twice as fast for the .9 degree stepmotor.
[13:43:35] <pcw_home> so what FinboySlicksaid : how fast do you need to move?
[13:44:09] <BHSPiMonkey> I have no real needs; I'm diving into the CNC world by building a Shapeoko
[13:44:17] <BHSPiMonkey> Just for fun
[13:45:32] <BHSPiMonkey> Realistically, based on past discussion, it doesn't seem like the precision difference is *that* substantial in terms of actual results (feel free to challenge that claim)
[13:46:20] <BHSPiMonkey> So my impression is that the increased movement speed will be the "better" advantage to have
[13:46:54] <pcw_home> a 1.8 degree stepmotor with double the ustep ration may well perform better than a .9 degree motor (in terms of torque vs speed)
[13:49:00] <BHSPiMonkey> I appreciate the expertise
[13:49:46] <skunkworks> andypugh, I think mach is trying for the industrial cnc market...
[13:50:44] <pcw_home> This also depends on how close your loads are to stall torque. If you are running close to stall torque, more poles will beat a higher Ustep ratio (.9 degrees to peak torque vs 1.8 degrees)
[13:51:30] <BHSPiMonkey> pcw_home: does that basically come down to how tough the material is to chew through?
[13:51:56] <pcw_home> and the gearing and mass and accell
[13:52:02] <andypugh> Well, that is where the real money is. But can they offfer industrial levels of support?
[13:53:30] <BHSPiMonkey> pcw_home: I'm probably just going to be doing simple tasks with wood/mdf/acrylic
[13:54:31] <BHSPiMonkey> So far it feels like the speed is probably more immediately useful than the step resolution
[13:54:36] <skunkworks> andypugh, if all they are is a gui running galil or other motion boards.. maybe?
[13:54:55] <BHSPiMonkey> (Plus I could always just buy different motors later on if I NEED them)
[14:01:26] <JT-Shop> pcw_home: do you have a sample 7i76 config like http://freeby.mesanet.com/7i77.zip ?
[14:04:12] <JT-Shop> mhaberler: is he talking about you? http://linuxcnc.org/index.php/english/component/kunena/?func=view&catid=48&id=24816#24816
[14:04:41] <mhaberler> hm?
[14:05:07] <mhaberler> aja, will answer
[14:05:13] <mhaberler> thanks!
[14:05:19] <JT-Shop> np
[14:05:43] <pcw_home> JT-Shop: There's a 5i25-7I76.ini
[14:05:44] <pcw_home> It just uses the standard hm2-stepper.hal file
[14:05:49] <mhaberler> I have so much crap on youtube.. need to look which one he could mean
[14:06:41] <JT-Shop> pcw_home: ok, thanks mine will be close to that then
[14:07:28] <pcw_home> freeby.mesanet.com/5i25-7i76.ini
[14:07:30] <pcw_home> I think its only a very minor change from hm2-stepper.ini
[14:07:48] <JT-Shop> yea, I think I just changed one or two lines for mine
[14:08:14] <pcw_home> or I should say 5i20.ini or whatever it is
[14:09:48] <JT-Shop> yea, I know what you mean
[14:30:22] <pcw_home> so a 5i25-7i76.ini could be added to the configs/hm-stepper dir and just work
[14:30:24] <pcw_home> the 7I77 required hal file changes because the analog outputs
[14:30:26] <pcw_home> and enables are different
[14:41:06] <JT-Shop> ok
[14:48:56] <JT-Shop> I think I got my mind right boss on the 7i47TA pin names http://imagebin.org/230562
[14:57:49] <pcw_home> JT-Shop yeah except you need to have the odd numbered Ins and outs on TB2
[14:58:09] <JT-Shop> oh I forgot to change them after copying lol
[16:22:41] <DJ9DJ> gn8
[16:26:05] <andypugh> Backlash? http://linuxcnc.org/index.php/english/component/kunena/?func=view&catid=10&id=24623&limit=6&start=6#24836
[16:28:17] <andypugh> I do like the warning "warning: suggest parentheses around assignment used as truth value", So much more polite than "This is C, you idiot, comparison is =="
[16:37:18] <BHSPiMonkey> if only I had a clue what was going on in that post
[16:40:58] <andypugh> You can ignore the G-code
[16:49:47] <tjb1> Tomorrow needs to get here.
[16:50:27] <tjb1> I need to finish my Z
[17:03:58] <tjb1> JT-Shop: Do I need to apply G92.1 before re-doing the probe? or is it fine to leave it.
[17:15:46] <andypugh> Did I ever post this engine rebuild animation link here: http://youtu.be/daVDrGsaDME ?
[17:15:59] <andypugh> I think it's one of the best things I hve seen on Youtube.
[17:17:59] <tjb1> andypugh: Dont hunt me down please.
[17:18:34] <andypugh> Huh?
[17:18:47] <tjb1> I applied to the forum :)
[17:20:49] <andypugh> OK, you're in
[17:21:23] * tjb1 goes into underground bunker
[17:22:38] <tjb1> Java is so confusing...
[17:25:48] <WillenCMD> what kind of Java? Carmel Machiatto or Hazlenut..
[17:26:56] <tjb1> the fantasatic programming kind.
[17:27:17] <djdelorie> not the island?
[17:27:32] <tjb1> I wish
[17:28:08] <andypugh> "Java" simply doesn't mean "coffee" at all, in the slightest, over here.
[17:31:43] <WillenCMD> really im an american and i don't have time to learn ther cultures
[17:31:53] <WillenCMD> i expect you to know mine though
[17:32:02] <WillenCMD> :)
[17:49:06] <JT-Shop> tjb1: can you elaborate a bit? I'm not following you...
[17:49:29] <tjb1> JT-Shop: http://linuxcnc.org/index.php/english/component/kunena/?func=view&catid=49&id=21954&start=30#22069
[17:51:06] <JT-Shop> the G92.1 is at the end of the file, after your done cutting, a second touch off simply resets the Z offset
[18:00:05] <PCW> Ha CNCzone is now listed as a suspicious site (by google)
[18:00:22] <andypugh> LinuxCNC frequently is.
[18:01:02] <andypugh> What bugs me is that the filter at work considers "hackaday" to be a "related to hacking" and won't let me view it.
[18:01:12] <PCW> Not for a while though
[18:02:14] <PCW> yeah two different meanings of hack
[18:03:30] <andypugh> Though today it is quite strong on lockpicking
[18:09:47] <andypugh> halrun
[18:09:49] <andypugh> Doh
[18:10:14] <JT-Shop> run hal run
[18:15:52] <jdh> proxy through an ssh tunnel
[18:20:07] <andypugh> With JT as the proxy?
[18:21:51] <jdh> sure
[18:22:00] <jdh> I have friends with colo boxes I use
[18:23:09] <andypugh> Sorry, I think I might be confused
[18:26:48] <jdh> I ssh -D, or use putty to proxy web pages through ssh tunnels.
[18:26:59] <r00t4rd3d> lol
[18:27:03] <r00t4rd3d> newb
[18:27:56] <jdh> ?
[18:28:12] <andypugh> Ah, right, I thought you were referring to my "halrun" and JT typing it back (JT as a proxy, see)
[18:28:24] <jdh> ahh... nah, the hackaday thing
[18:57:40] <Valen> ey pcw able to chat on work time for a minute?
[18:57:59] <PCW> Yes
[18:58:41] <Valen> I'm the guy who wants to use half an encoder again, any suggestions for how it'd work (other than badly)
[18:59:02] <Valen> IE how i'd get a count out of the card
[18:59:38] <Valen> or would it require changes to the firmware in the card?
[19:00:20] <PCW> I can hard wire a up/dpwn counter mode that takes the drive directions as motion direction (yes a firmware change)
[19:00:54] <PCW> But I honestly dont have a lot of hope it would work well enough
[19:01:40] <Valen> its for a prototype for a science experiment, people are going to be touching an array of little wheels and the wheels get driven a few mm
[19:02:00] <Valen> (6mm motors with 1000:1 gearboxes on them driving 8mm wheels)
[19:02:08] <andypugh> Valen: Use the software counters?
[19:02:37] <Valen> what are these software counters of which you speak?
[19:02:45] <andypugh> Sounds like there won't be a huge count rate?
[19:02:57] <andypugh> The HAL encoder component
[19:03:46] <andypugh> (I am not certain that it will help)
[19:04:25] <Valen> I spose i could look at doing it all in software
[19:04:33] <Valen> just poll a pin
[19:04:43] <Valen> I want it all to be as off the shelf as possible
[19:04:46] <JT-Shop> could this be similar to my winder example in hal?
[19:05:27] <JT-Shop> or is this just a MPG?
[19:05:33] <PCW> if you need them to hold position I dont think it will work If you only need to drive them approximately N counts
[19:05:35] <PCW> this way or that a non quadrature thing should work
[19:06:08] <andypugh> Can you describe the problem?
[19:06:09] <Valen> there should be no back drive, and .1 of a mm precision would be heaps
[19:06:22] <Valen> gimme a sec i'll stick some images up
[19:07:54] <JT-Shop> I'm amazed that I can still stand up after pedeling 12 miles through the woods
[19:08:01] <JT-Shop> pedaling
[19:08:20] <Valen> this is the overall thing
[19:08:24] <Valen> http://www.vapourforge.com/jake/temp/pin/CCW%20body.jpg
[19:08:33] <Valen> the insides http://www.vapourforge.com/jake/temp/pin/CCW%203%20backs%20removed%20labeled.jpg
[19:08:41] <Valen> detail of the top http://www.vapourforge.com/jake/temp/pin/CCW%20top%20back%20removed%20labled.jpg
[19:09:01] <Valen> for scale the purple motors are 6mm diamiter
[19:09:28] <Valen> there is going to be an array of 49 of these things that you can rest your hand on
[19:10:46] <PCW> all they do is rotate a little?
[19:10:47] <andypugh> Is there a really good reason not to use a quadrature sensor?
[19:11:02] <Valen> the science people want to be able to drive each individual pin up into contact with your hand, (set the position) then drive back down, and up to that point again. then rotate the wheel when its in contact with the skin
[19:11:21] <JT-Shop> 149 inputs for quad?
[19:11:43] <Valen> andypugh: there is just barely enough room for a pair of 805 packages
[19:12:18] <Valen> I want to use the mesa toolchain for the prototypes, I plan on rolling a custom avr per pin setup for the 49 motor package
[19:13:25] <andypugh> Can you squeeze this in? It's a 2-channel photo-interrupter: http://uk.rs-online.com/web/p/photointerrupter/7103825/
[19:14:31] <Valen> sure thats 2 ch?
[19:14:48] <PCW> yep its 2
[19:14:50] <andypugh> Yes
[19:15:01] <Valen> ahh right you are
[19:15:19] <andypugh> There are even smaller reflective style ones, but they seem fussy about target.
[19:15:41] <Valen> I'm not sure, but i'll find out, I'd much rather off the shelf it
[19:15:54] <Valen> don't spose they come with code wheels?
[19:16:33] <andypugh> Can you use the magnetic sensors on the end axis?
[19:16:50] <Valen> no, the motors are all single ended
[19:17:03] <Valen> and theres not enough room on the driven end
[19:17:18] <Valen> i would have done that in a heartbeat lol
[19:17:28] <andypugh> DC motors, or brushless?
[19:17:31] <Valen> dc
[19:17:52] <andypugh> But I was meaning embed a little magnet in the end wheel.
[19:18:05] <Valen> hmm interesting
[19:18:28] <andypugh> http://docs-europe.electrocomponents.com/webdocs/0d59/0900766b80d59dc6.pdf
[19:18:30] <PCW> with 1000/1 gearing 1 PPR may be enough
[19:18:32] <Valen> I can get a DC gearhead motor in this size for $20, for brushless its like $200 at least from my regular components
[19:19:19] <andypugh> That ams chip is a bit big.
[19:19:39] <Valen> it also needs to be on center and fairly close as i recall
[19:20:47] <Valen> hmm, open loop would suffice for all this, if i could get a cheap brushless or super mini stepper, then all this goes away
[19:21:42] <andypugh> I am using these, and they are proper tiny, but fussy about target http://uk.rs-online.com/web/p/rotary-encoders/7160604/
[19:22:10] <andypugh> Ideally you would put a magnetic pattern in the edge of the wheel and sense it with a magenetic hall encoder sensor.
[19:22:22] <andypugh> But...
[19:22:26] <Valen> bit much work for what we are doing
[19:22:46] <Valen> although with 50 of them i spose heating a bit of iron wouldn't be insane to do
[19:23:21] <andypugh> If you take the output of an encoder in counter mode, pass it through a ddt component, then multiply the output by the direction, then you probably get what you need.
[19:24:00] <Valen> well that one you just posted looks pretty good
[19:24:16] <Valen> everything we are doing is CnCed so pointing shouldn't be too much of an issue
[19:24:26] <Valen> if it solves the other problems then i'm for it
[19:24:46] <Valen> where are you getting the codewheels from?
[19:25:06] <andypugh> Laser cut by a stencil maker
[19:25:31] <Valen> what are you making out of curiosity?
[19:25:45] <andypugh> A hollow-shaft servo
[19:26:03] <Valen> for what?
[19:26:23] <Valen> i mean hobby or 5Kw milling spindle ;->
[19:26:31] <andypugh> These guys made my encoder wheels: http://www.sparkslaser.com/Sparks_Laser_Services.html
[19:26:39] <andypugh> Valen: Hobby
[19:26:45] <Valen> Al?
[19:27:40] <r00t-Shed> :D
[19:27:48] <andypugh> https://picasaweb.google.com/108164504656404380542/CNCUnsorted#5761456633570778562
[19:27:56] <andypugh> You can see the detector there
[19:28:08] <Valen> how fussy is fussy?
[19:28:28] <andypugh> It needs to be really really shiny, and have holes in it
[19:29:08] <andypugh> You need holes through the wheel, I think.
[19:29:15] <Valen> polished Al or i spose stainless
[19:29:30] <andypugh> And any sensor you get will limit your wheel width
[19:31:04] <Valen> I'm still liking that reflective encoder you have there
[19:32:17] <andypugh> I don't see how you can use it on your wheels though.
[19:32:30] <Valen> make new wheels
[19:32:34] <Valen> oh hang on
[19:32:56] <Valen> see http://www.vapourforge.com/jake/temp/pin/CCW%203%20backs%20removed%20labeled.jpg
[19:33:00] <andypugh> The wheels have to interface with people as wellas with the sensor
[19:33:14] <Valen> the sensor is on the motor shaft itself
[19:33:34] <andypugh> Ah, OK.
[19:33:36] <Valen> you can see the "code wheel" as the red disk on the end of the motor
[19:33:47] <Valen> the same setup is used at the bottom motor
[19:33:57] <andypugh> How clever do you want to get?
[19:34:09] <Valen> simple and reliable
[19:34:17] <andypugh> There is this tiny encoder that I invented..
[19:34:34] <Valen> uh huh...... ;->
[19:34:58] <andypugh> You take a shaft and machine half away in 2 places with a 90 degree rotation.
[19:35:19] <andypugh> You then shine leds past the shafts.
[19:35:59] <andypugh> You get two varying outputs, with a 90 degree shift, just like a resolver. And you solve for position using the same method.
[19:36:28] <andypugh> I had that working on a 3mm shaft in a 5mm housing.
[19:36:31] <Valen> I'm trying to stay away from the analog domain if i can
[19:36:33] <PCW> with a 1000/1 gear box 4 counts/rev is likely OK
[19:36:46] <andypugh> (but the software guys couldn't figure it out at the time)
[19:36:58] <PCW> thats less than .1 degree
[19:36:59] <Valen> i don't actually have access to the motor shaft, i'm coming off the end of the gearbox
[19:37:12] <PCW> Oh
[19:37:31] <Valen> nothing is "good" about this whole thing lol
[19:37:48] <Valen> unless somebody knows a source for tiny dual shaft motors?
[19:37:58] <andypugh> I would say sense commutation spikes.
[19:38:02] <PCW> are any of the magnetic sensors small enough?
[19:38:27] <Valen> the problem is they generally need to be on axis
[19:38:37] <andypugh> Codestrip.
[19:38:57] <Valen> magnetic code strip?
[19:39:02] <andypugh> yes
[19:39:24] <Valen> keep in mind this whole thing fits in something the size of your thumb
[19:39:34] <andypugh> Aye.
[19:39:58] <PCW> the 3 mm sensor is to big?
[19:40:10] <PCW> too
[19:40:12] <andypugh> Wrap the wheels in casette tape. Record a code pattern. Sense with {magic}
[19:40:29] <Valen> which one PCW? the photo interrupter or the reflective encoder?
[19:40:42] <PCW> the interrupter
[19:40:56] <Valen> it could be, I'll have to pass it to the mechanical guy
[19:41:11] <Valen> but worst case the reflective one looks good
[19:41:18] <Valen> and both give quadrature outputs
[19:41:26] <Valen> just need to find somebody to make a code wheel
[19:41:33] <andypugh> Tape is dirt cheap: http://www.machine-dro.co.uk/high-accuracy-magnetic-linear-tape.html
[19:41:45] <andypugh> Then you just need two tiny little hall sensors.
[19:43:20] <Valen> more screwing about than using either of those optical sensors methinks
[19:44:13] <andypugh> How does the wheel motor rotate the wheel? Fricton drive?
[19:44:22] <Valen> bevel gear
[19:44:48] <andypugh> What's the bevel gear made of?
[19:44:52] <Valen> plastic
[19:45:32] <Valen> you can see it http://www.vapourforge.com/jake/temp/pin/CCW%20top%20back%20removed%20labled.jpg
[19:46:04] <andypugh> What diameter do you think you have for a codewheel>
[19:46:05] <andypugh> ?
[19:46:25] <Valen> I'll be estimating from the picture lol
[19:46:42] <Valen> looks like a bit over 6mm ;->
[19:47:38] <andypugh> That's really rather tight for the reflective sensor. But it looks like you could slot the body out for clearance?
[19:48:07] <Valen> its all offset to one side to get the drive wheel back on center
[19:48:48] <andypugh> I keep coming back to measuring commutation pulses.
[19:49:43] <andypugh> Anyway, I should be in bed
[19:50:17] <Valen> with ~100 other motors hanging off the same power supply and precious metal brushes it sounds like a tight game
[19:50:25] <PCW> what resolution do you figure you needed?
[19:50:44] <Valen> thanks for your help, I'll pass those onto the mechanical guy and see if he can fit something
[19:51:04] <PCW> 'nite Andy
[19:51:08] <andypugh> Maxxon motors?
[19:51:09] <Valen> PCW one of the problems we have with this project is there are no requirements, its all "as good as we can get"
[19:51:20] <Valen> actually a different group
[19:51:24] <andypugh> I always found them really helpful, they might have ideas.
[19:51:39] <Valen> they don't have sensors at 6mm from the factory
[19:51:42] <Valen> i have asked ;->
[19:51:56] <PCW> I notice you can get hall sensors in sot23s
[19:52:00] <andypugh> Piezo crawler onto the wheel edge?
[19:52:39] <Valen> sounds rather expensive
[19:54:22] <PCW> The linear magnetic sensors would give you fairly good resolution but I think the chips are big
[19:54:44] <PCW> and ic-haus is always offline
[19:55:33] <Valen> the problem i have with magnetics is the codewheel
[19:55:37] <andypugh> http://www.physikinstrumente.com/en/news/fullnews.php?newsid=106
[19:56:10] <andypugh> Valen: I would look at magentising a pattern into the actual top wheel.
[19:56:21] <PCW> http://www.ams.com/eng/Products/Magnetic-Encoders/Linear-Encoders/NSE-5310
[19:56:31] <Valen> its all getting further from "the shelf" if you will
[19:56:44] <PCW> AMS is off the shelf
[19:57:42] <Valen> looking at them now
[19:57:50] <PCW> thats about 4000 counts/rev with 6mm codewheel
[19:57:59] <Valen> yeah i was just working that out
[19:58:05] <andypugh> I suspect that there is some form of piezo motor available off the shelf that would let you shring even smaller, and that would work as an open loop stepper.
[19:58:28] <Valen> probably cost more than $20 though I'd wager
[19:58:35] <Valen> the piezo motor
[19:58:43] <PCW> sorry 2000 (~9 1mm pitch pole pairs needed)
[19:59:07] <andypugh> I wouldn't bet on it, piezo is really quite cheap stuff.
[19:59:25] <andypugh> Buzzers cost pennies. Motors are rather more specialised though.
[19:59:51] <PCW> Nitinol!
[20:00:02] <andypugh> http://en.wikipedia.org/wiki/Piezoelectric_motor
[20:00:05] <Valen> we were looking at using piezo stuff (actuators) for this in the past and they were like $100 an actuator
[20:00:12] <Valen> pcw is that muscle wire?
[20:00:29] <andypugh> USM motors are standard on camera lenses now, that has to have brought the price down.
[20:00:37] <PCW> Yeah (assumoing i spelled it right)
[20:00:44] <Valen> we also looked at that
[20:00:50] <Valen> had a whole design for it too
[20:01:10] <Valen> but to get the response speed, the wires (even in an oil bath) were like .1mm
[20:01:35] <Valen> then they decided they liked the wheel system because it didn't actually have a displacement
[20:02:17] <PCW> piezo electric (20 feet long)
[20:02:35] <Valen> looked at those too
[20:03:01] <Valen> needed something like 50:1 amplification to get enough motion from a small enough package ;->
[20:03:23] <PCW> I have a little linear piezo motor here (a inchworm)
[20:03:35] <Valen> even putting a lump of iron over the top of a grid of 6 coils but worked out we would melt the coils
[20:03:46] <andypugh> Inchworm directly on the wheel is a really attractive idea.
[20:04:11] <andypugh> How much force do you need?
[20:04:19] <Valen> "about that much"
[20:04:50] <PCW> certainly lighter and potentially cheaper (a slippy step drive)
[20:04:54] <andypugh> You could probably make a sector of a brushless motor and drive the wheel direct (embedded magents, all cast in epoxy). But you might not want to make 49. of the,
[20:05:22] <Valen> the force on that would be drastically drastically lower
[20:05:37] <Valen> they have played with a motor with a wheel on it and they liked it
[20:06:47] <PCW> the linear sensor might be usable directly for the linear axis
[20:07:02] <Valen> thats a good idea pcw
[20:07:21] <PCW> (the COB one is 2.5x3.9x.6 mm)
[20:08:20] <andypugh> If I think of anything, I will get back to you
[20:12:34] <PCW> oops COB package is custom
[20:13:43] <PCW> how far does the wheel have to turn?
[20:14:52] <Valen> continious rotation is desired
[20:18:18] <PCW> about to head home but Ill keep this in mind in case I get any inspiration
[20:18:24] <Valen> thanks
[20:18:37] <Valen> I'll see about stuffing those optical sensors in for the rotation
[20:19:17] <PCW> Those are cheap and should be good if they fit and have enough resolution
[20:19:50] <Valen> yeah, just need to see if they will work with that diamiter
[20:21:31] <Valen> those magnetic encoders all seem to use strontium magnets at a minimum, so magnetising up some iron looks doubtfull
[20:34:05] <andypugh-iPhone> Valen?
[20:34:57] <andypugh-iPhone> How about an optical mouse sensor looking at the wheel?
[21:38:14] <jp_> PCW, will the 7i73 eventually support 8bit parallel?
[21:39:54] <pcw_home> Maybe though it seems unnecessary as almost all displays seem to support nibble mode
[21:40:28] <pcw_home> (its just something the driver chips do)
[21:42:02] <pcw_home> Its a trivial change though it would probably need an EEPROM option to select the mode
[21:42:48] <jp_> looking at the HD61830 data sheet from hitachi i dont see any mention of 4bit mode vs 8bit
[21:43:14] <jp_> should i just try it and see?
[21:45:06] <jp_> i was going to try and use an old robot teach pendant with it the HD61830 is the driver for the LCD
[21:45:11] <jp_> http://store.industrialcontrolrepair.com/product/fanuc-a05b2020c141-teach-pendant-rf-29120.cfm
[21:50:13] <pcw_home> Its looks close but it does not say it supports nibble mode
[21:50:15] <pcw_home> I would have to see if the protocol is identical as well (control signals look the same)
[21:50:16] <pcw_home> It may well be the the 7I73 has the 8 bit code in it as we started that way but found that all the current
[21:50:18] <pcw_home> character displays we could find worked in both modes
[21:50:37] <pcw_home> (so saved the extra pins)
[21:51:45] <pcw_home> I think Andy found his display on Ebay and the suggested one is nice but expensive ($20 or so) from Digikey
[21:53:18] <jp_> ill try the LCD with my teensysy if it works then i see about getting the 7i73 to play with it. I have a few 4x20's kicking around as well that support 4bit.
[21:53:27] <pcw_home> (just fixed the manual since Optrex was bought out by Kyocera and the PN changed)
[21:53:53] <pcw_home> and the price went up :-(
[21:53:57] <jp_> just thought it would be neat to breath some life into these old pendants i have
[21:54:44] <pcw_home> I'll ask Gilbert about 8 bit mode its probably lurking in the source
[21:54:56] <pcw_home> (on Wednesday)
[21:55:01] <jp_> thx
[21:57:27] <jp_> pcw_home, oh one more thing is the 4x20 a limit on the 7i73 or just common size for reference?
[21:58:39] <pcw_home> Its a common (and useful) size
[21:59:44] <jp_> so larger sizes are not an issue?
[22:02:25] <pcw_home> They may be because the interfaces change (some larger panels are actually 2 controllers)
[22:03:33] <pcw_home> I think the 40X4 panels are 2x 40x2 panels for example
[22:03:50] <pcw_home> (logically)
[22:09:02] <r00t4rd3d> http://i.imgur.com/h1eIy.jpg
[22:09:09] <r00t4rd3d> a sign for a friend
[22:10:37] <r00t4rd3d> 12x3.5
[22:11:02] <pcw_home> Moose country?
[22:11:20] <r00t4rd3d> no, he is fat
[22:11:25] <r00t4rd3d> :D
[22:11:41] <r00t4rd3d> but yeah moose country too
[22:13:05] <pcw_home> looks nice
[22:14:22] <r00t4rd3d> thx
[22:15:51] <r00t4rd3d> you can see the lines, I only did a 8% stepover with a vbit.
[22:16:05] <r00t4rd3d> but i hate waiting for 2-3%
[22:20:59] <panasync> that's why you walk away and goto sleep :)
[22:22:12] <r00t4rd3d> last time i walked away i came back to a broken motor shaft and garage full of smoke
[22:23:03] <r00t4rd3d> http://i.imgur.com/gJLX8.jpg
[22:24:33] <r00t4rd3d> For Sale: One used stepper, slightly used.
[23:01:47] <panasync> well I've got 4 used steppers with 5/16" shafts for sale. odd ball shaft size. I needed eiher 3/8" or 1/4". too bad I didn't know to specify when I bought these
[23:23:14] <r00t4rd3d> they make 5/16 couplers
[23:23:45] <r00t4rd3d> http://www.ebay.com/sch/m.html?_nkw=5%2F16&_sacat=0&_odkw=&item=330423833395&_osacat=0&_ssn=carolbrent
[23:27:09] <panasync> I needed 5/16" pulley's
[23:27:20] <panasync> 20 tooth or 22 tooth
[23:27:59] <panasync> I ended up using some copper pipe and made some split sleeve adapters for a 3/8" pulley
[23:28:29] <panasync> seems that 3/8" copper pipe is exactly 5/16" inside diameter
[23:29:28] <panasync> but I've since replaced the motors and gotten 1/4" shaft ones from good american companies instead of longs_motor in china
[23:39:40] <r00t4rd3d> "This web page at www.cnczone.com has been reported as an attack page and has been blocked based on your security preferences."
[23:39:49] <r00t4rd3d> lol