Back
[02:00:51] <evokanevo> Hello. Is this the right place to ask about writing custom M-codes in C/C++, or should I do that in linuxcnc-devel?
[02:01:32] <evokanevo> (I actually already asked over there but haven't heard back yet)
[02:03:04] <archivist> the trick in IRC is to lurk and wait for an answer as it depends on who is around
[02:03:38] <evokanevo> fair enough. i will repost every hour or so in case there are new users.
[02:04:30] <evokanevo> I hope this is the right place to ask - anyway, I'm trying to write a custom M-code in C and not finding any examples. I only see examples of how to do this in bash. I want to issue a simple command like "M64 P0" for example.
[02:04:30] <evokanevo> [23:46] <evokanevo> In the AXIS gui I can run the M-code with "M102 P... Q..." but I don't know how to get the MOTION or IO task to do anything as a result of what I'm doing in my M102 executable.
[02:04:32] * FinboySlick lurks and gives bad answers ;)
[02:04:55] <FinboySlick> archivist: If you're up, I'm up too late too.
[02:05:22] <archivist> 7:50 am here :)
[02:05:39] <FinboySlick> Heh, which means 2:50 here.
[02:05:44] <FinboySlick> OK, time to sleep.
[02:05:53] <DJ9DJ> moin
[02:06:05] <evokanevo> I should rephrase - I have an M102 executable and I want it to issue commands like "M64 P... Q..."
[02:06:07] <archivist> someone asked me a question long after I went to bed
[03:55:30] <Loetmichel> mornin'
[05:41:43] <skroon> hi all
[09:33:09] <andypugh> Everyone at church?
[09:37:56] <Tom_itx> naw
[09:38:18] <Tom_itx> trying to recover a hdd
[09:40:13] <Loetmichel> andypugh: church? whats that?
[09:42:28] <andypugh> It's where all decent folk go on a Sunday. So probably doesn't explain why the LinuxCNC folks are so quiet.
[09:45:58] <Tom_itx> still hungover?
[09:48:01] <andypugh> More probable, I guess
[09:57:21] <archivist> I only go to church during the week to fix clocks, never on a sunday
[09:59:20] <pcw_home> http://www.youtube.com/watch?v=28EAWlOXrYs
[09:59:21] <Tecan> (28EAWlOXrYs) "Melina Mercouri - Ta Paidia Tou Piraia (Never On Sunday)" by "Adam0792" is "Music" - Length: 0:03:35
[10:03:38] <archivist> looooong time since I have seen that :)
[10:20:38] <IchGuckLive> hi all B)
[12:19:11] <IchGuckLive> Crashdemon: bits are expensive so keep come
[12:24:55] <Roguish> hey, does anyone have experience with usini
[12:25:13] <Roguish> using an ultrasonic cleaner for small parts???
[12:25:48] <IchGuckLive> to make them oilfree for coding
[12:26:38] <archivist> Roguish, as a clockmaker, yes :)
[12:26:39] <Roguish> actually, i'm working on my bike and the derailluers are grimey
[12:27:58] <archivist> I often clean part I have made that way after making them, also when very grimy
[12:28:24] <Roguish> so, it works pretty good?
[12:29:14] <archivist> it work as well as it can, use the right solvents
[12:29:41] <Roguish> http://www.ebay.com/itm/DSA50SE-SK2-FULL-SET-INDUSTRIAL-GRADE-3L-3-17QT-250W-DIGITAL-ULTRASONIC-CLEANER-/290604862797?pt=LH_DefaultDomain_0&hash=item43a9664d4d
[12:29:52] <archivist> it may get lubrication you want to leave out too
[12:31:23] <archivist> bit small for the price
[12:33:08] <archivist> but they are never cheap except for the home jewelry
[12:35:42] <archivist> I have a couple of small ones similar to 360544098592 I abuse them with anything I want cleaning
[12:36:44] <archivist> there is a danger if the tank is not big enough and the item is not covered that you end up with a tide mark on your item
[12:37:26] <Roguish> ok, thats good info. i'm just checking now.
[12:37:27] <archivist> we had to rotate items during the clean to try to avoid
[12:38:04] <archivist> funny tingling in your bones of your finger ends
[12:40:02] <archivist> you can imagine how filthy this lot was after use before a dip
http://www.collection.archivist.info/archive/DJCPD/PD/2012/2012_11_07_Drive_gear/IMG_1400.JPG
[12:42:18] <andypugh> Did you ever get paid for those?
[12:42:20] <archivist> any mechanical cleaning loses edges on tiny stuff too
http://gears.archivist.info/gears/IMG_1214_hires.JPG
[12:42:26] <archivist> andypugh, nope
[12:43:14] <andypugh> I should have turned the nutter-filter up higher.
[12:43:30] <archivist> so they are still here as examples of work :)
[12:43:59] <archivist> one can never tell, some nutters do pay up
[12:44:56] <archivist> I think I may be on to a new winner, selling QR codes, download a generator...profit
[12:45:27] <archivist> some business people do get taken for a ride
[12:46:42] <skunkworks> https://www.youtube.com/watch?v=wslOMT6_e6k
[12:51:25] <IchGuckLive> archivist: did you use a Giant match for the photo
[12:51:39] <IchGuckLive> better 2 use coins
[12:55:23] <IchGuckLive> skunkworks: this is conz robo or
[12:59:01] <archivist> coins vary in size between countries
[13:00:31] <IchGuckLive> agree but you find matches as big as 2 inch also
[13:01:36] <archivist> never mind the length, feel the width
[13:01:43] <IchGuckLive> http://en.wikipedia.org/wiki/Penny_%28United_States_coin%29
[13:02:01] <IchGuckLive> this gives then the world a inication on how mesuments are
[13:02:47] <IchGuckLive> ok up to you thats gread work on miniture
[13:02:56] <IchGuckLive> By for me today
[13:04:33] <andypugh> Bah! I was just about to point him at:
http://petapixel.com/2011/08/18/large-objects-shot-as-miniatures-using-a-giant-coin-and-tilt-shift-effects/
[13:07:48] <archivist> I used a coin in another image and it fails hence the match on that one
[13:13:34] <archivist> this image
http://gears.archivist.info/P1010320_hires.JPG
[13:14:22] <andypugh> Did you see the pictures in the link though?
[13:16:25] <archivist> yes
[13:17:52] <andypugh> You could do something similar to make your gears look huge.
[13:19:22] <archivist> what! make a miniature coin!
[13:19:40] <archivist> probably a little hard
[13:20:17] <andypugh> Possibly illegal too.
[13:21:47] <archivist> ask the queen for some maundy money
[13:22:15] <andy1978> Hi at all
[13:22:43] <andypugh> Hi
[13:23:07] <archivist> an andy echo, is this recursion
[13:23:31] <andy1978> archivist: :-P
[13:23:43] <andy1978> beware of the inf loop
[13:24:29] <andy1978> archivist: Do you remember me? We talked about hershey fonts and ttt
[13:24:54] <andy1978> archivist: I made a repo here
https://github.com/Andy1978/hf2gcode if you are interested
[13:25:57] <Loetmichel> i have a lighter for size comparsion....
http://www.cyrom.org/palbum/main.php?g2_itemId=12265
[13:26:48] <andy1978> Loetmichel:
http://www.google.com/imgres?imgurl=http://carlighter.net/wp-content/uploads/2011/10/p_1600_1200_43ff9e05-d6ed-4f8e-a296-1f551a502edc.jpg&imgrefurl=http://carlighter.net/reasons-for-buying-a-big-lighter/&h=640&w=480&sz=93&tbnid=MzR7hoEorodmHM:&tbnh=109&tbnw=82&prev=/search%3Fq%3Dreally%2Bbig%2Blighter%26tbm%3Disch%26tbo%3Du&zoom=1&q=really+big+lighter&usg=__Clo9Dx3Vx8IALCW2SykRTWC1KPo=&docid=PLEvyOQNk5Zl
[13:26:53] <andy1978> uups, I'm sorry
[13:27:27] <Loetmichel> hrhr
[13:27:31] <Loetmichel> something like that
[13:27:42] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=12067
[13:27:55] <Loetmichel> thats a normal box of cigarettes on the left ;-)
[13:29:22] <andy1978> 5€ for 22 cigarettes?
[13:34:38] <Loetmichel> andy1978: jeah, tahts germany
[13:34:40] <Loetmichel> ;-)
[13:36:37] <Loetmichel> why? are they cheaper where you come from?
[13:36:41] <Loetmichel> or more expenive?
[13:36:45] <Loetmichel> +s
[13:43:18] <andy1978> Loetmichel: No, I stopped smoking some years ago before they got so expensive
[13:43:38] <andy1978> Loetmichel: die haben damals noch 5DM gekostet....
[13:44:01] <Loetmichel> hrhr
[13:59:14] <andypugh> A friend's drill table. Somebody either didn't care, or had no clue:
https://www.dropbox.com/s/o82rvdf63m1zkmm/drill15.jpg
[13:59:52] <pcw_home> Hardly a spot missed
[14:00:19] <andypugh> Indeed, it looks like they kept moving the table to find a new bit to drill into.
[14:02:08] <Loetmichel> andypugh: i thoght thats the factory condition of a drill table?
[14:02:18] <andypugh> :-)
[14:02:40] <Loetmichel> never seen any in (much) better condition
[14:02:40] <pcw_home> Break-away model?
[14:02:45] <Loetmichel> but some in worse ;-)
[14:03:00] <Loetmichel> ... one already broken away ;-)
[14:04:19] <pcw_home> Our circuit board vendor calls PCB breakaways done that way "mouse bites"
[14:11:44] <andy1978> can brackets () be escaped in g-code comments?
[14:12:55] <andy1978> ah, ; is also valid
[14:17:12] <andypugh> As is / I think.
[14:37:42] <Crashdemon> bye
[14:50:47] <GammaX> Afternoon all
[14:51:01] <GammaX> anyone aware of camworks post for a crusader II Lathe?!
[15:01:20] <Tom_itx> andypugh, some controls won't like ';' but will generally use '()'
[15:02:19] <Tom_itx> but lcnc may use either. i've always gotten used to '()'
[15:10:24] <andypugh> I just realised that "/" isn't a ocmment, it's a block-ignore isn't it?
[15:10:46] <micges> andypugh: yes
[15:15:18] <mrsun> hmm any info somewhere on how the signal from a jog wheel looks ?
[15:15:22] <mrsun> mpg wheel i guess its called
[15:15:31] <mrsun> got A B and ~A ~B ports on it
[15:15:36] <andypugh> Quadrature
[15:16:07] <andypugh> http://en.wikipedia.org/wiki/Quadrature_phase
[15:16:10] <mrsun> andypugh, yes but it only goes high for a sec then low again?
[15:16:50] <andypugh> That's not what I would expect.
[15:17:14] <andypugh> You don't have the scope on AC by some chance?
[15:17:21] <mrsun> andypugh, well measurments of it says it does that :P
[15:17:29] <mrsun> move one dot, and it goes high then low directly after
[15:17:48] <mrsun> might have to remeasure it, wasnt me who measured it :P
[15:17:48] <andypugh> It might have a full cycle per click.
[15:18:09] <mrsun> full cycle per click ?
[15:18:31] <andypugh> Yes
[15:18:49] <GammaX> anyone know if the linuxcnc forums have sponsors?
[15:18:52] <GammaX> or allow them?
[15:18:57] <mrsun> the whole four thingies in one go ?
[15:19:04] <mrsun> just to be able to count cycles and direction ?
[15:19:07] <andypugh> mrsun: Indeed
[15:19:26] <mrsun> hmm so a scope check it is then :P
[15:19:29] <andypugh> GammaX: We eschew any hint of commercilaity.
[15:19:43] <GammaX> understandable...
[15:40:15] <pcw_home> Most 100 count MPGs Ive seen are 100 PPR = 400 CPR quadrature
[15:42:35] <mrsun> pcw_home, hmm ok so it will have a full cycle of 01 11 10 00 then for each step
[15:43:01] <pcw_home> So if you want then to count exactly one count per detent, you need the encoder counter set to 1X mode
[15:43:03] <pcw_home> (and the correct phase/polarity chosen so it does not count in the 'valley' of the detent)
[15:43:08] <pcw_home> them
[15:43:32] <mrsun> pcw_home, gonna read it directly into a uC
[15:43:59] <pcw_home> then you want to implement 1X mode
[15:45:29] <pcw_home> (count up on rising edge of A when B is high,count down on falling edge of A when B is high)
[15:47:48] <L84Supper2> what dates are the meetup?
[15:49:25] <L84Supper2> http://linuxgizmos.com/beaglebone-black-climbs-the-device-tree-with-linux-3-8/ we are trying to get RTAI to catch up to 3.8
[15:56:36] <mhaberler> mrsun: this is some robust encoder code:
http://www.mikrocontroller.net/articles/Drehgeber#Signalauswertung
[15:57:19] <mrsun> thanks =)
[16:00:56] <mhaberler> (nb there are lots of broken 'quadrature decoder' code fragments out there - seems thats the #1 project for arduino n00bs - "lets ignore the CV and redo this!" ;)
[16:01:24] <GammaX-Tower> im lookin at arduino tutorials right now
[16:04:40] <mrsun> mhaberler, "ignore the CV" ?
[16:04:44] <mrsun> what is CV?
[16:04:54] <mhaberler> conventional wisdom
[16:05:10] <DJ9DJ> v like wisdom? ;)
[16:06:44] <mhaberler> vizdom
[16:07:17] <mhaberler> duh..
[16:07:34] <DJ9DJ> hrhr
[16:07:52] <L84Supper2> vich is vat one must alvays do vit dis code :)
[16:09:58] <L84Supper2> doesn't the Arduino IDE have the quadrature decoder as a macro?
[16:10:04] <mhaberler> zerman tspelling
[16:10:33] <mhaberler> tick a) macro b) robust ;)
[16:11:12] <mhaberler> pcw probably has a few stories on what separates the men from the boys in this space
[16:11:38] <L84Supper2> sorry 4am here
[16:11:51] <mhaberler> just tick b)
[16:11:53] <GammaX-Tower> L84Supper, where u live? london?
[16:12:02] <L84Supper2> back in China
[16:12:03] <GammaX-Tower> no that would be wrong...
[16:12:05] <GammaX-Tower> ahhhh
[16:12:10] <GammaX-Tower> that where u live?
[16:12:25] <L84Supper2> >50% of the time now
[16:12:43] <L84Supper2> need anything?
[16:13:17] <GammaX-Tower> pm?
[16:13:32] <L84Supper2> shopping for parts is fun here
[16:14:01] <Loetmichel> ... if you are in shenzen ;-)
[16:15:23] <L84Supper2> was at the Beijing electronics market yesterday, you have to screen the vendors by how they store their parts....
[16:15:46] <L84Supper2> in tubes and trays vs plastic bucket :)
[16:16:07] <GammaX-Tower> L84Supper, I actually am interested in a few items... but a bulk order....
[16:17:04] <GammaX-Tower> can you recomend any tool manufacturers for custom collet chucks?
[16:17:40] <L84Supper2> no lead or ballscrew vendors I've found yet, Hiwin in Taiwan is where they still turn to
[16:17:58] <L84Supper2> how many are you looking for?
[16:18:05] <GammaX-Tower> 100+
[16:18:33] <DJ9DJ> gn8
[16:18:43] <GammaX-Tower> later DJ9DJ
[16:18:53] <L84Supper2> so < 1 hours work for a big shop here
[16:18:57] <andypugh> GammaX-Tower: Try
http://www.ystool.com.cn/en/index.asp
[16:19:09] <L84Supper2> what size and type collets?
[16:20:05] <GammaX-Tower> andypugh, you know marbo?
[16:20:19] <GammaX-Tower> L84Supper, can you check your pm please?
[16:21:59] <andypugh> GammaX-Tower: No, I hardly know anyone.
[16:22:27] <GammaX-Tower> haha I spoke with sktool for a bit they have helped alot im just trying to get price down more...
[16:26:20] <andypugh> Some people live in a different world to me (and, I suspect, you lot). Please forgive this extended quote
[16:26:45] <andypugh> "I've got this really cool app called Bubble on my phone, it's two
[16:26:45] <andypugh> concentric rings with a little circle in the middle that moves around
[16:26:47] <andypugh> inside the rings, it's a game or something, you have to get the circle
[16:26:48] <andypugh> bang in the middle of the two rings. If you hold the phone vertically
[16:26:50] <andypugh> the rings turn into a little rectangle with a dip out of the top of
[16:26:51] <andypugh> it, and you have to get the circle in the middle from side to side
[16:26:53] <andypugh> instead. It's quite cool. I can imagine it'd be useful in a situation
[16:26:54] <andypugh> like this. I spose you could actually use it as a tool. I don't know
[16:26:54] <andypugh> what you'd do if the battery ran out. I wonder if they do manual ones?"
[16:27:31] <Loetmichel> hrhr
[16:27:50] <Loetmichel> someone who doent know what a level is... ok...
[16:28:39] <GammaX-Tower> lol
[16:28:47] <Loetmichel> said that: i have an app like this on my phone.. its more or less useless because the sensors are not good enough in the Samsung GTN-7000
[16:29:17] <L84Supper2> sktool is down near Shenzhen, the more western type quality machine tool makers are up north
[16:29:45] <L84Supper2> Suzhou is ground zero for high tech
[16:31:02] <andypugh> I just sent some PCB artwork to Shenzhen
[16:32:14] <L84Supper2> thats where all the high volume manufacturing goes on
[16:32:48] <L84Supper2> the R&D is up in Suzhou
[16:33:59] <Tom_itx> L84Supper, an Aussie in #avr is in Shenzhen peddling parts for us and commutes from Au to Ch
[16:34:12] <L84Supper2> if you need lots and lots of PCB's (100k+) I have contacts
[16:34:20] <andypugh> Not, presumably, daily.
[16:34:22] <Tom_itx> he also does PCB's
[16:34:37] <L84Supper2> Tom_itx: what kinds of parts?
[16:34:41] <andypugh> I thought I was pushing the boat out ordering 20
[16:34:48] <Tom_itx> he will try to find anything you need
[16:34:54] <Tom_itx> andypugh, no :)
[16:34:56] <andypugh> (And I expect to give most of them away)
[16:35:09] <Tom_itx> he send me 100 free samples
[16:35:29] <L84Supper2> electronic components are sold in markets here, similar to fish markets :)
[16:35:41] <Tom_itx> yes
[16:36:05] <andypugh> I don't like taking free samples when I know that there won't be a follow up order. Though even when I explained this to Lee Spring, they still sent me the spring I needed :-)
[16:36:39] <evokanevo> If anyone here is familiar with path following tolerance, i.e. path blending, would you please take a look at this? I'm reposting here becuase it's not getting many views on the forum.
http://www.linuxcnc.org/index.php/english/forum/20-g-code/26580-g64-path-blending-not-working-as-expected
[16:37:06] <L84Supper2> we can also get lots of parts out the backdoor from high volume buyers
[16:37:26] <L84Supper2> usually a 1-2% premium over their 1M piece price
[16:37:27] <Tom_itx> andypugh, he wanted me to try them to replace one i was currently using
[16:37:46] <Tom_itx> i've made several orders with him
[16:38:39] <L84Supper2> we can ask for 1M piece pricing and the seller will still expect us to pay $.22 for each sample
[16:39:24] <L84Supper2> sample are not free here, just sold at the 1-10 pc price
[16:39:43] <Tom_itx> http://www.evilmadscientist.com/2009/a-visit-to-the-electronics-markets-of-shenzhen/
[16:40:10] <L84Supper2> there goes my grammar :)
[16:42:29] <L84Supper2> the south is electronics and small equipment, the farther north you go the larger the machines get, forging and smelting is way up north along with large machinery
[16:43:13] <L84Supper2> the offices up north are not even heated, we were asked if our printers work at -10C
[16:43:51] <Tom_itx> http://www.youtube.com/watch?v=BF0Qaxzp2nQ
[16:43:51] <Tecan> (BF0Qaxzp2nQ) "SED Electronics Market (Tablets Market) in Shenzhen walk-through" by "Charbax" is "Tech" - Length: 0:20:55
[16:43:55] <Tom_itx> video walk thru
[16:44:05] <L84Supper2> I said office temps and they said in Dec-March their office is +5 to -10C
[16:44:27] <andypugh> evokanevo: Has me baffled too
[16:45:26] <evokanevo> Thanks for looking.
[16:45:30] <Loetmichel> shouldnt be a problem for a laser printer ;-)
[16:45:42] <andypugh> evokanevo: This does not answer your question, I realise, but if you want to follw paths at that speed on a machine with those accel levels, then you will need to use exact-path ot exact-stop.
[16:47:11] <L84Supper2> walked passed the German embassy here yesterday, beer garden next door :)
[16:47:12] <evokanevo> andypugh: I actually want smooth, curvy moves. It's a weird CNC machine.
[16:52:56] <pcw_home> Is it related to the A axis? I tried that gcode in sim (minus A) and the curves were symmetrical
[16:53:53] <andypugh> There are no A-moves, to it shouldn't matter.
[16:55:05] <pcw_home> so whats different I wonder
[16:55:31] <andypugh> Is it possible for inverse time mode to make a short move happen faster than a long one?
[16:56:26] <pcw_home> Was that in inverse time mode?
[16:56:58] <andypugh> Probably not. I am grasping at straws
[16:59:09] <evokanevo> pcw_home: It could be A-axis related - I am not using trivkins. However, the A axis is the same in every part of the path so it seems like it shouldn't matter.
[16:59:17] <pcw_home> anyway I didn't see any difference in the curves (of course the sim accel values are not the same as evokanevo's)
[17:01:17] <pcw_home> probably have to divide/conquer to find where its gone wrong
[17:01:22] <evokanevo> Nope. Just checked with my trivkins version and it's the same. So kinematics/inverse-kinematics aren't messing with it.
[17:05:01] <pcw_home> Well looks OK in 2.5 sim
[17:05:17] <evokanevo> I'm using 2.5.2
[17:08:25] <andypugh> There are unlikely to be any differences between 2.5 and 2.5.2
[17:08:49] <andypugh> Do the three moves all happen at the same speed?
[17:10:00] <evokanevo> Not sure exactly what you mean - they have the same feedrate set, however the Z is slower.
[17:10:23] <evokanevo> So far it looks like moving from slow segment to fast segment results in large arc.
[17:10:32] <evokanevo> And fast to slow results in a small arc.
[17:12:41] <evokanevo> Yep, that's it. Just switched the axis to move only in the XY-plane (same max acceleration/velocity) and all arcs are the same.
[17:12:55] <evokanevo> Thanks for your help.
[17:13:16] <andypugh> That probably makes sense, the motion planner always guarantees that it can stop before the end of the next segment.
[17:13:45] <andypugh> I think you might see this in action if you made the last Z move longer.
[17:16:42] <tjtr33> ?? "thats it" meaning one axis had different acc? while x&y had identical acc's & macvel? looks symmetric to me on 2.6.0~pre at F200 2000 and 20000 resp.
[17:16:45] <evokanevo> I tried it but it gave the same behavior. I think it's that the Z-acceleration is lower than the X and Y acceleration.
[17:17:07] <tjtr33> k
[17:17:12] <evokanevo> tjtr33: yes, Z has different acceleration
[17:17:31] <tjtr33> thx
[17:17:37] <pcw_home> Doing the best it can with the Z contraint
[17:18:43] <evokanevo> I think the idea is it starts accelerating to the next line direction as soon as it is P-distance away from the endpoint.
[17:18:57] <evokanevo> (endpoint of the current line segment)
[17:19:41] <evokanevo> And the XY is so fast in comparison to the change in Z-velocity that the Z-displacement is minor.
[17:19:52] <evokanevo> Anyway, problem solved. Thanks again.
[17:31:02] <tjtr33> yes making the sim axis z acc 1/2 that of x&y produces the asymmetry.
[17:31:04] <tjtr33> but it seems wrong,and is certainly unexpected. A caveat about equal axis 'tuning' might be a good idea
[17:32:57] <evokanevo> Yes I am still not quite comfortable with it, but I need to think about it more.
[17:35:03] <tjtr33> and this is just the predicted path based on the given accelerations ( ignore the mechanics for now ) and some people are guessing thier acc values :(
[18:34:17] <andypugh> L84Supper: Can you read the spec of this?
http://www.ebay.co.uk/itm/40mm-Outer-Diameter-Incremental-Type-Photoelectric-Encoder-Rotary-Encoder-/281099904847
[20:39:07] <RyanS> What's the point of three turrets on a turning centre? Is it simply for faster production?
[20:40:19] <toastydeath> faster production, and you can often cut harder with two turrets
[20:40:28] <toastydeath> third turret is either for boring or 2nd chuck work
[20:40:45] <andypugh> Probably. Or maybe to make it easier to get long tools out of the way. Or possibly to perform operations on the other spindle simulaneously.
[20:42:11] <RyanS> I think the other thing that I have seen is pinch milling thin parts I guess you would need 2
[20:42:46] <toastydeath> pinch turning works just as well
[20:43:21] <toastydeath> consider that a swiss screw machine has what... seven or eight tool slides that all work simultaneously?
[20:44:22] <RyanS> But if it only the has one spindle?
[20:44:44] <toastydeath> yep
[20:45:32] <RyanS> Swiss machines confuse me I think because they are so fast it's difficult to see what's going on
[20:46:22] <toastydeath> the chuck is also hidden
[20:46:52] <RyanS> yeah
[20:48:41] <RyanS> Is it impossible for a fixed head turning centre to achieve the same level of precision as swiss.. Or its more about the speed
[20:49:17] <toastydeath> considering the most accurate turning machines in the world are not swiss machines, i'd say "yes"
[20:49:41] <toastydeath> it is difficult to beat a swiss screw machine (cnc or cam driven) for accuracy AND speed
[20:49:58] <RyanS> ah ah
[20:50:22] <toastydeath> the cam driven machines are a good margin faster
[20:50:37] <toastydeath> but cannot handle the same range of work, and are a little less accurate than their cnc counterparts
[20:50:52] <velcrow> I am making my own CNC mill, but am not sure what to use as the cutting tool/spindle. I was thinking of using my Dremel, but am not sure if that's good. Any suggestions? I want to mill PCBs, <6mm MDF/plywood/acrylic (is this possible?). Thanks.
[20:52:35] <RyanS> cam systems are Windows-based? What happens if the computer crashes halfway through machining
[20:53:18] <toastydeath> no, cam, as in camshaft
[20:53:48] <toastydeath> non-cnc screw machines are driven by a set of cams
[20:56:52] <RyanS> ah....nevertheless computer-aided machining..... solidcam is a well-known one. If it crashes as much as solidworks
[20:57:05] <Tom_itx> ie brown & sharp, davenport, new brittan
[20:58:39] <RyanS> I just find it strange you have a a few million dollars worth of machinery being run by a desktop computer
[20:59:07] <toastydeath> ?
[20:59:34] <toastydeath> DNC machining is pretty uncommon nowdays
[20:59:59] <toastydeath> the machines are networked and can pull/push nc files across the network, but actual drip feeding machines is rare
[21:00:09] <toastydeath> (for exactly the reason you're pointing out)
[21:00:11] <RyanS> DNC?
[21:00:18] <toastydeath> direct numeric control
[21:00:26] <toastydeath> or is it distributed
[21:00:27] <toastydeath> i forget
[21:00:59] <toastydeath> you throw the machine into DNC mode, and hit start, and it sits there and waits for commands to come in over ethernet or rs232
[21:01:24] <RyanS> "Each SolidCAM post-processor is customized to the machinist's individual needs and generates ready-to-run CNC-programs" ?
[21:01:30] <toastydeath> was used back in the day of tape storage for long ass programs
[21:01:50] <toastydeath> that's just CAM; it's a finished program
[21:02:06] <toastydeath> you take the gcode the postprocessor spits out, send it to the machine, and the machine runs it locally
[21:02:55] <Tom_itx> yeah most machines probably have alot more memory now
[21:03:15] <Tom_itx> we had one we had to DNC all the time
[21:03:17] <RyanS> ah – so you never have a machine that is simply like a dumb terminal
[21:03:26] <toastydeath> the only place i've ever heard of DNC being used is OLD school 5 axis machines in the early days of aerospace machining
[21:03:34] <toastydeath> like, 60's and 70's
[21:03:35] <andypugh> velcrow: The Bosch Colt router is popular
[21:03:44] <toastydeath> RyanS, yeah, exactly
[21:03:54] <toastydeath> they can theoretically be used that way, but they aren't
[21:04:07] <velcrow> andypugh, Thanks, looking into it.
[21:04:39] <velcrow> andypugh, A specific model?
[21:04:50] <pcw_home> A dremel is entirely too wobbly for (fine) PCB routing
[21:05:47] <RyanS> hmm I wonder what sort of CPU in the machines controller probably RISC and not all that high clock speed?
[21:06:34] <toastydeath> my understanding is that most industrial controls use several processors at a low clock speed, yeah
[21:06:58] <andypugh> velcrow: Have you seen this sort of thing on eBay?
http://www.ebay.com/itm/1-5KW-WATER-COOLE-MILLING-AND-GRINDING-SPINDLE-MOTOR-WITH-INVERTER-DRIVE-VFD-b2-/170745848157?
[21:07:37] <velcrow> andypugh, Not on ebay, but I have seen it elsewhere during my research. Why?
[21:07:40] <eric_unterhausen> the Haas at work uses the biggest circuit board I have ever seen
[21:07:49] <eric_unterhausen> even in old-school electronics
[21:08:16] <andypugh> Well, they are actually designed for the job (then built down to a price, which may undermine things a little)
[21:09:03] <andypugh> I am sure we can find you spindle motors up to any budget you care to name.
[21:09:46] <velcrow> andypugh, A little out of my price range though? Are there cheaper solutions? I'd go up to about $100. I can make the entire machine except for the spindle mechanism stuff because I have no experience with milling.
[21:09:48] <eric_unterhausen> wish I could figure out how to run the Moore tool spindles I have
[21:10:06] <toastydeath> eric_unterhausen, you mean you have jig grinder spindles?
[21:10:26] <eric_unterhausen> don't think so, they are small, router sized
[21:10:36] <toastydeath> bout 6" long?
[21:10:40] <toastydeath> rectangular base?
[21:10:41] <eric_unterhausen> yeah
[21:10:51] <toastydeath> what rpm
[21:10:55] <toastydeath> 40k, 60k, 120k?
[21:10:58] <eric_unterhausen> 40k, 60k,
[21:11:02] <andypugh> There are people making spindles from brushless RC motors.
[21:11:13] <toastydeath> lol those are worth several thousand dollars if they're in good condition
[21:11:22] <toastydeath> they are not router spindles, they're ultraprecision grinding spindles
[21:11:22] <velcrow> i use bl motors for my rc hobby :)
[21:11:25] <velcrow> same ones?
[21:11:30] <andypugh> Yes.
[21:11:40] <velcrow> And they'd work for my purposes?
[21:11:40] <eric_unterhausen> toastydeath, aware of that
[21:11:50] <toastydeath> ah, sorry
[21:11:58] <eric_unterhausen> but I'm not sure what to feed them
[21:12:22] <eric_unterhausen> have the drive
[21:12:40] <toastydeath> i wish i didn't throw away all my business cards, I had a guy from moore nanotech who could get all the manuals for those machines
[21:12:44] <RyanS> What machines are typically used in aerospace (or any requiring ultra-precision), just high-end models of the major brands? Or are we talking about some exotic machines from a manufacturer most have never heard of
[21:13:13] <toastydeath> RyanS, the former
[21:13:23] <toastydeath> aerospace isn't ultraprecision, they're just precision
[21:13:27] <andypugh> velcrow: You take a brushless motor, one of these collets:
http://www.ebay.co.uk/itm/C16-ER16A-150L-Straight-Collet-Milling-Chuck-Holder-for-ER16-Collet-/290913553085 and some bearings. Make a housing (which needs a lathe, you are wasting components unless you get the housing right) and then either drive with a coupling or a belt drive.
[21:14:00] <toastydeath> Kitamura, DSG, Mori Seiki, Ikeagi, Okuma
[21:14:12] <toastydeath> and more i can't remember now, but all fairly expensive
[21:14:14] <velcrow> andypugh, Thanks, that helps a lot.
[21:14:35] <eric_unterhausen> moore nanotech and some other related company are ultra-precision
[21:14:46] <andypugh> There is somebody making them semi-commercially, but I can't find him.
[21:14:49] <pcw_home> Sodick makes some tiny ones
[21:15:13] <pcw_home> all linear
[21:15:15] <andypugh> toastydeath: Which DSG?
[21:15:24] <toastydeath> dean, smith, and grace
[21:15:25] <RyanS> What's the Swiss brand, starts with a k..
[21:15:45] <toastydeath> either way a lot of companies make "aerospace" quality machines now
[21:16:21] <toastydeath> because hitting .0001" tolerances anywhere in a machine's envelope isn't as hard as it used to be
[21:16:57] <toastydeath> provided you're willing to pay the cost of doing that, that is
[21:17:37] <andypugh> toastydeath: I think you might have meant DMG? Though DSG made this lovely thing:
http://www.pdbengineering.com/pdf/Travelling%20gantry%20machine%20feature.pdf
[21:18:17] <toastydeath> I thought DSG was bought by somebody and was still an active division, but DMG is obviously super-high quality
[21:18:33] <toastydeath> much like Leblonde was bought, though their quality went to shit and are virtually unknown now
[21:18:58] <RyanS> Kern..
[21:19:14] <andypugh> DSG are stilll the place to go for _really_ good lathes.
http://www.deansmithandgrace.co.uk
[21:19:14] <RyanS> Odd looking machines
[21:20:53] <toastydeath> damn that's a big goddamn gantry mill
[21:21:44] <toastydeath> i was wondering if it was worth waiting for the load time
[21:21:49] <toastydeath> but the answer is yes
[21:22:50] <RyanS> nanotech.. looks exotic
[21:23:21] <toastydeath> ?
[21:23:42] <toastydeath> the diamond turning machine company?
[21:24:05] <RyanS> yeh you mentioned that b4
[21:24:24] <toastydeath> oh, moore nanotech
[21:24:42] <toastydeath> they're two different companies
[21:25:06] <eric_unterhausen> http://www.youtube.com/watch?v=vAvfrrlMZg4&list=UUvR6oZZrgw7dz50ANYv2gQA&index=26
[21:25:07] <Tecan> (vAvfrrlMZg4) "Diamond Turning Copper Sample with Measurements.wmv" by "mdrlpsu" is "Tech" - Length: 0:01:30
[21:26:13] <andypugh> toastydeath: What I like about that gantry is that it is conceptually identical to a DIY CNC gantry, even in the construction and bolt-on linear slides. But it is a lot bigger.
[21:26:23] <toastydeath> i used to work for a company that made air bearings and one of our major products were linear stages for diamond turning machines
[21:26:54] <eric_unterhausen> air bearing linear stages?
[21:27:02] <eric_unterhausen> which company?
[21:27:10] <RyanS> Ive seen that video very impressive
[21:27:12] <toastydeath> new way air bearings
[21:27:22] <eric_unterhausen> how was Drew to work for?
[21:27:34] <toastydeath> he was an interesting guy
[21:28:12] <eric_unterhausen> he's making some interesting shaped air bearings
[21:28:14] <toastydeath> didn't work directly under him except for a handful of times
[21:28:22] <RyanS> How about this
https://www.youtube.com/watch?v=epjtsPc58Gc
[21:29:18] <toastydeath> RyanS, that's actually engraving microscopic patterns on the drum, not just a regular diffraction grating
[21:29:19] <eric_unterhausen> they have some company that uses theirs to make reflective material
[21:29:36] <eric_unterhausen> takes more than a shift to do a part
[21:29:57] <toastydeath> and if you jump or drop something, it ruins the part
[21:30:11] <eric_unterhausen> probably
[21:30:21] <toastydeath> nice big white line shows up
[21:30:25] <RyanS> Looks like a granite bed
[21:30:37] <eric_unterhausen> fancy concrete
[21:30:37] <toastydeath> at least it's easy for inspection to catch
[21:30:50] <andypugh> I find myself wanting to make a lathe, with an air-bearing, direct drive spindle, and a granite bed.
[21:31:25] <eric_unterhausen> andypugh, and do what with it?
[21:31:50] <andypugh> Oh, nothing. I am only interested in makine machines, not using them :-)
[21:32:33] <RyanS> It wouldnt be too hard to pour a granite bed level.... no maybe those machines use actual machineed rock
[21:32:48] <eric_unterhausen> easy to get ruined spindles on ebay :)
[21:32:49] <andypugh> I expect to die with the capability to machine absolutely anything, but having never actuallt done so.
[21:33:38] <toastydeath> you can get pretty good results with an air bearing spindle and not a whole lot of knowledge
[21:34:05] <andypugh> Indeed, they aren't actually complicated.
[21:34:27] <andypugh> I use to work for a company that used air bearings partly because they were easy.
[21:34:43] <eric_unterhausen> depends on the precision you need
[21:35:19] <toastydeath> naturally, but for a homebuilt machine a homebuilt air bearing is going to be more accurate than most commercial bearings of any type that is in the price range
[21:36:00] <eric_unterhausen> I have a private vid that has been watched 17 times
[21:36:07] <andypugh> I designed this (well, the high-speed parts, the base machine was a standard frame).
http://www.nordson.com/en-us/divisions/dage/products/bondtesters/Pages/4000HSHighSpeedBondtester.aspx
[21:36:14] <toastydeath> we got away with stuuuupid laziness on experimental projects for things that went into cmms and interferometers
[21:36:53] <eric_unterhausen> I was working with a cmm, surprised how crappy it is
[21:37:02] <eric_unterhausen> they move approximately where you tell them to go
[21:37:10] <toastydeath> "CLOSE ENOUGH"
[21:37:37] <andypugh> They achieve very good accuracy without actually using expensive components. The air-bearing lets you land on the substrate of a chip, back off three microns, then push through a track and measure the force.
[21:37:41] <eric_unterhausen> I did this to work with cmm
http://www.youtube.com/watch?v=azv9_aDfg_E&list=UUvR6oZZrgw7dz50ANYv2gQA
[21:37:43] <Tecan> (azv9_aDfg_E) "Cylindricity Spindle TEST.mp4" by "mdrlpsu" is "Tech" - Length: 0:00:16
[21:38:18] <toastydeath> was that using the moore method
[21:38:33] <eric_unterhausen> what moore method?
[21:39:12] <toastydeath> measure deviation for 360 degrees, rotate test artifact 180, measure again, the differential error is the error in the spindle
[21:39:12] <RyanS> Do people use epoxy granite for diy because it doesn't require foundry stuff? Most commercial machines are cast-iron
[21:39:44] <andypugh> RyanS: Basically, yes. Though some manufacturers apparently fill cast iron framws with it too.
[21:39:46] <toastydeath> moore came up with it in the 70s as the first direct measurement of spindle error
[21:39:57] <eric_unterhausen> toastydeath, no
[21:40:00] <toastydeath> without relying on an artifact's accuracy
[21:40:03] <andypugh> However, my lathe idea was to use a granite kitchen worktop :-)
[21:40:06] <eric_unterhausen> we have better methods anyway
[21:40:22] <toastydeath> yeah, that's why i was curious to see only one probe
[21:40:33] <toastydeath> i haven't seen a one probe spindle test in a long long time
[21:40:42] <eric_unterhausen> we do single probe multiprobe :)
[21:40:46] <RyanS> What about a cheap surface plate?
[21:41:05] <toastydeath> how does that work?
[21:41:15] <eric_unterhausen> put the bearing on a rotary table
[21:41:29] <eric_unterhausen> saves you 20k in cap probes
[21:41:40] <toastydeath> ah, so you're indexing it by 120 every time?
[21:41:48] <eric_unterhausen> no, magic angles
[21:41:51] <eric_unterhausen> 120 doesn't work
[21:42:26] <eric_unterhausen> angles are published, just magic
[21:42:29] <toastydeath> hahaha
[21:42:55] <toastydeath> i believe it
[21:43:08] <toastydeath> this conversation is making me sad i got out of ultraprecision
[21:43:22] <eric_unterhausen> it's a very small world
[21:43:37] <toastydeath> hindsight i should have stuck with drew
[21:43:54] <eric_unterhausen> drew really seems like he could have a bright future
[21:44:07] <toastydeath> the manufacturing team and the r&d team were not best friends, and the mfg side basically told me "pick us or him"
[21:44:19] <toastydeath> since i was trying to become a machinist at the time, i picked them
[21:44:50] <Tom_itx> heh
[21:44:56] <toastydeath> looking back, that was a stupid thing to do
[21:45:38] <Tom_itx> too bad we weren't born with hind sight
[21:45:53] <Tom_itx> i'm sure we'd all change a few things
[21:45:54] <toastydeath> it ended any chances I had of going out to the precision machining conferences
[21:46:05] <toastydeath> only got to go to one
[21:48:14] <toastydeath> maybe after i get my math degree i'll send my resume around
[21:49:21] <andypugh> I did a project on machining 100um dia glass cylinders to 3um thick. :-)
[21:49:34] <toastydeath> hahahah
[21:49:39] <toastydeath> how did that work out
[21:49:46] <toastydeath> i wouldn't want to be the one to have to handle those
[21:50:06] <eric_unterhausen> no
[21:50:49] <andypugh> Pretty easy with the right machine:
http://www.gatan.com/pdf/addl_pdf/Dimple%20Grinder.pdf
[21:53:22] <toastydeath> haha
[21:53:44] <andypugh> Looks like the paper isn't available without subscription:
http://link.springer.com/article/10.1007%2FBF00275452
[21:54:49] <andypugh> I got quite good at engraving the specimen number in the wall-thickness of sections of 3mm brass tube.
[21:54:51] <toastydeath> my school is pretty crappy but we have a lot of subscriptions
[21:54:54] <toastydeath> i may have access
[21:55:20] <toastydeath> from all the money they get from like, dance students, they spend it on physics and engineering journals
[21:55:25] <toastydeath> ...there's no engineering department
[21:56:12] <andypugh> You can't dance without physics, but you can do physics without dancing :-)
[21:56:45] <toastydeath> there's like ten people in the physics dept
[21:57:15] <andypugh> <digression> It is possible to plug an Rpi into a Beaglebone. I wonder if that is useful in any way?
[21:59:48] <pcw_home> With Ethernet or serial of course, probably SPI as well whether its a useful or not I dont know
[22:00:04] <pcw_home> useful combination
[22:00:52] <andypugh> I was mainly meaning that the Beagle has a female header, and the RPi a male one ;-)
[22:02:49] <Willburrrr2003> hello, i hope everyone is doing well
[22:04:02] <Willburrrr2003> i finally upgraded from emc 2.4.6 to linuxcnc 2.5.0 today
[22:04:51] <pcw_home> as long as it does not cause a I/O-->GND or I/O -->PWR short there are probably many possible connections
[22:05:13] <Willburrrr2003> i was able to get my config files edited for the linuxcnc changes. i had to comment out my ngcgui in my ini file, as it keeps crashing my linux cnc
[22:06:30] <pcw_home> crashing is such a harsh word perhaps program control anomaly?
[22:06:40] <Willburrrr2003> does anyone know if there was any changes in the way linuxcnc sets up and calls ngcgui vs emc?
[22:07:02] <Willburrrr2003> ok, i can go with that
[22:07:36] <andypugh> Typically it isn't a "crash", it is a controlled exit with an error report.
[22:08:05] <andypugh> ie, the software finds something it isn't happy with, and gives up.
[22:08:19] <Willburrrr2003> it closes linuxcnc as soon as it opens, and if i comment out all the tab setup lines it doesnt close linuxcnc as soon as it loads up
[22:08:39] <andypugh> Now, the code I write, that can crash. In fact I crashed an 8i20 so badly it had to go back to Mesa.
[22:08:42] <Willburrrr2003> yes andy it does spit out an error report
[22:10:17] <andypugh> So, what is the error report?
[22:11:35] <Willburrrr2003> let me fire up the computer and i will tell you what it says
[22:12:01] <tjtr33> the blue leds on the bbb are blinding, really long term persistance of vision.
[22:12:03] <tjtr33> but pretty cool it just shows up on desktop and has its own web page
[22:12:28] <eric_unterhausen> yeah, they are lowering the led leves
[22:12:31] <eric_unterhausen> levels
[22:13:07] <andypugh> It was rather annoying blinking there next to me.
[22:13:18] <eric_unterhausen> too much
[22:13:40] <eric_unterhausen> I'm going to put mine in a case and hope not too much light gets out
[22:13:46] <tjtr33> type left handed, right cupped over annoying leds :|
[22:14:19] <tjtr33> wheres that cardboard box :)
[22:14:32] <tjtr33> ahhh!
[22:15:12] <andypugh> Lay it face-down? I don't think that the photon pressure is enough to launch it accross the room. Though it's probably close.
[22:15:42] <Willburrrr2003> andy, in the error window under debug it says tkinter.tclerror: expected integer but got "normal"
[22:15:46] <Connor> http://www.cnczone.com/forums/benchtop_machines/173025-zachs_g0704-6.html#post1281782
[22:15:54] <Connor> someone may want to help him with that.
[22:17:06] <andypugh> I just properly crashed my Linuxcnc box. Like, need to walk accross the room and pull the power plug.
[22:17:33] <eric_unterhausen> I've seen 0-5v and direction before
[22:18:08] <Willburrrr2003> andy, also at begin of debug section says can't find -sec Mot - var Mot -num 1, & same can't find for IO,Linuxcnc,Emc
[22:18:15] <andypugh> Willburrrr2003: Sounds like a question for Dewey on the forum.
[22:18:24] <andypugh> You can ignore those.
[22:18:31] <Willburrrr2003> wow andy, that does sound like a crash
[22:19:11] <Willburrrr2003> ok andy, i will post for dewey in the forums and see what he says
[22:19:12] <andypugh> And, annoyingly, no clue what is wrong. But I am doing hariy things with void pointers in kernel modules.
[22:26:51] <andypugh> Connor: I wasn't quite quick enough to beat some prat suggesting throwing out the Mesa kit and using a kflop.
[22:27:20] <tjtr33> for the bbb, should i use raring or quantal ubuntu image on uSD card?
[22:28:25] <Connor> andypugh: K flop won't work on linuxcnc will it
[22:28:40] <andypugh> Not as far as I know.
[22:37:10] * ssi fpgas up the world
[22:40:01] <andypugh> Night all
[22:40:52] <tjtr33> gnite
[22:49:17] <zultron> GammaX-Tower, I have folks in Suzhou who can do some legwork if you need.
[23:26:10] <ssi> my verilog-fu is exceedingly weak :(
[23:50:51] <The_informer> !HY-TB3DV-M setting
