#linuxcnc | Logs for 2012-07-21

Back
[00:00:32] <Nick001-Shop> hoping to drive motors with H-Bridge cards
[00:02:30] <andypugh> Does this link work?
[00:02:31] <andypugh> http://tinyurl.com/d99uws8
[00:04:50] <Tom_itx> goes to gmail
[00:05:06] <andypugh> Nick001-Shop: Sounds like a job for 3 x 6-axis cards and a 4-port FPGA card, so probably 3x7i48, 1 x 7i37TA and a 5i22
[00:06:35] <andypugh> But I think Mesa might have a low-power version of the 8i20 in the pipeline, which might eliminate the analogue stages.
[00:07:14] <Nick001-Shop> any idea when it would be ready?
[00:07:47] <andypugh> None at all. This is a customer-rumour, not insider-information
[00:08:07] <andypugh> I just spotted the 7i30
[00:08:53] <Nick001-Shop> the 5i22 plugs into the PCI slot and the rest are daughter cards?
[00:09:06] <andypugh> 5i22 + 4 x 7i30 gives you PCI-slot-to-motor
[00:09:38] <andypugh> 7i30 is a quad H-bridge that plugs into the PCI cards.
[00:11:06] <andypugh> 5i22 + 4 x 7i30 is 16 of your motors + encoders. You would then need either a second PCI slot or you could consider a 7i43 on the parport for everything else.
[00:11:57] <micges> just mentioning: 7i40 have more amps than 7i30
[00:12:04] <Nick001-Shop> I have 5 PCI slots on these mb's
[00:12:49] <andypugh> micges: I think he only wanted 1.8A
[00:13:13] <andypugh> 7i30 is $90 for four channels, sounds a bargain
[00:13:14] <micges> ah ok
[00:14:15] <Nick001-Shop> how many amps on 7i40 card
[00:15:02] -!- toastydeath has quit [Ping timeout: 245 seconds]
[00:15:12] <andypugh> 2 x 5i20 probably makes more sense than 1 x 5i22 + something else then.
[00:15:16] -!- toastydeath [[email protected]] has joined #linuxcnc
[00:15:36] <andypugh> Nick001-Shop: www.mesanet.com
[00:16:47] <Nick001-Shop> have it - will look at it later almost time to go home -)
[00:17:00] <andypugh> 2 x 5i22 + 8 x 7i39 for a sweet 16 x brushless motor setup :-)
[00:17:28] <micges> heh
[00:17:45] <Nick001-Shop> what would be for brushed motors
[00:18:11] <andypugh> 7i30 is for brushed
[00:18:50] <andypugh> Be careful not to make PCW too rich, though. He might retire.
[00:20:41] <Nick001-Shop> I need a lot of other stuff - like resolver cards if I can find that fellow in Atlanta to see how he made out
[00:20:46] -!- WillenCMD [[email protected]] has joined #linuxcnc
[00:21:12] <Nick001-Shop> would be cost effective vs the pico way
[00:21:40] <andypugh> Resolver?
[00:21:44] <WillenCMD> how does one get he glade components to show up on a sim installed system
[00:22:13] <andypugh> Nick001-Shop: Which chap in Atlanta?
[00:23:02] <andypugh> WillenCMD: I am surprised that they don't
[00:23:28] <WillenCMD> me also
[00:23:46] <WillenCMD> im in the process of making a hal component gui
[00:23:48] <andypugh> FWIW I don't bother with sim systems, I run in VMS and simply ignore the RT errors when latency gets over 1 million. (and it does)
[00:24:07] <andypugh> Sorry, VMs not VMS
[00:24:20] <andypugh> ie, Virtual Machines
[00:24:21] <WillenCMD> thats what i ususally do
[00:24:35] <Nick001-Shop> I have other machines with resolvers that need to be converted
[00:24:58] <Jymmm> andypugh: Uh huh, you KNOW you really meant VMS =)
[00:26:07] -!- PCW has quit [Quit: ChatZilla 0.9.88.2 [Firefox 14.0.1/20120713134347]]
[00:26:33] <Nick001-Shop> andypugh - I think his name was ssl
[00:26:48] <andypugh> I have never knowingly used VMS. But who knows what I was using during my PhD when I was using fortran on a mainframe via a storage-screen terminal
[00:27:24] <andypugh> Nick001-Shop: What was he using? I know a fair bit about Resolvers and LinuxCNC
[00:28:53] <andypugh> I am using resolvers on my own machine.
[00:28:58] <skunkworks__> andypugh: have you done anything more with the arduino resolver interface?
[00:29:17] <Nick001-Shop> He redid a Hardinge HCNC with mesa cards
[00:29:27] <andypugh> skunkworks__: I converted it to fast serial last weekend, rather than emulated quadrature, which is a lot more elegant
[00:30:03] <Nick001-Shop> He was having issues of some kind with the cards
[00:30:18] <skunkworks__> neat - is it still 1000 counts per rev? also - when I played with it - it didn't seem to be linear throughout the rotation.
[00:30:26] <andypugh> Nick001-Shop: I also wrote the driver for the Mesa 7i49 Resolver/servo card, so if there are problems I am the one to talk to. If you get one I guarantee I will make it work.
[00:31:13] <skunkworks__> (and I still have not played with the accupins...)
[00:31:14] <andypugh> Nick001-Shop: He needs to tell me and/or PCW if there are problems.
[00:31:44] <Nick001-Shop> would be cost effective to replacing working resolvers with encoders that don't last as long
[00:31:59] <andypugh> skunkworks__: I haven't checked ilinearity. It is something that could be tweaked (for example by fudging the arctan table)
[00:32:29] <andypugh> Resolvers are much more satisying than incremental encoders.
[00:32:32] <Nick001-Shop> This was a couple of months ago and I haven't seen hin here lately
[00:33:15] <Nick001-Shop> I have resolvers that are still working after 30 years of use
[00:33:44] <skunkworks__> encoders shouldn't go bad... (as long as they don'
[00:33:45] <andypugh> Well, he has presumably either got it working and is making chips, or he has trashed the lot, fitted encoders and gecko step-dir drives and is using Mach
[00:33:50] <skunkworks__> t get crap in them)
[00:34:22] <andypugh> Yes, but you can use a resolver in chicken soup and it will still work perfectly.
[00:34:31] <Nick001-Shop> II rather think he got it working somehow - I hope
[00:34:49] <skunkworks__> yes
[00:35:13] <skunkworks__> (if the machine has resolvers - I would use them) ::)
[00:35:23] <Nick001-Shop> These machines throw out aan oil mist that gets into an encoder
[00:36:06] <Nick001-Shop> and I'v replaced 2 of them already in the 1 machine
[00:36:46] <andypugh> The trouble with Resolvers is that they are not the sort of thing you will buy to replace encoderss with, those things are expensive
[00:37:04] <Nick001-Shop> Put the resolver back in the spindle and used a pico adapter for it
[00:38:06] <andypugh> Going back to quadrature seems a little retrograde.
[00:38:10] <skunkworks__> heh
[00:38:12] <Nick001-Shop> machines came with them and they still work - unlike the Allen Bradley controls on them
[00:38:15] <skunkworks__> old school....
[00:39:47] <andypugh> And skunkworks already has an Arduino on the machine. And my Arduino code. And (IIRC) one of my Resolver boards. He could easily connect up a spindle resolver to the Arduino and use 2 channels for spindle and one for temperature.
[00:40:37] <skunkworks__> I think I have 3 - or maybe 2
[00:41:08] <andypugh> Then, with the serial mod, and a new firmware for his Mesa cards, he could measure spindle temperature at 1kHz, just in case :-)
[00:41:19] <skunkworks__> heh
[00:43:50] <Nick001-Shop> Time to shut the shop down and go home - be back later on the other computer
[00:44:01] -!- Nick001-Shop has quit [Quit: ChatZilla 0.9.88.2 [Firefox 14.0.1/20120713134347]]
[00:49:48] <andypugh> Is the Allegro AT667 really the only metal-detecting Hall sensor on the market? (ie, PCB-mount component with internal magnet and Hall device)
[00:50:37] <andypugh> <looks at djdelorie as somone who might know his components>
[00:50:43] <Jymmm> prox?
[00:50:56] <andypugh> Hmm, ATS667 I mean
[00:51:06] -!- geografa has quit [Quit: geografa]
[00:51:11] <Jymmm> are you looking for priximity sensor or ?
[00:51:43] <andypugh> Jymmm: Effectively, yes, but in a 4-pin SIP not a 2" long threaded housing
[00:52:05] <Jymmm> ah
[00:52:09] <Tom_itx> http://sensing.honeywell.com/products/magnetic_position_sensors?Ne=2308&N=3094
[00:52:14] * Tom_itx looks at andypugh
[00:52:32] <andypugh> I actually now have 4 of the ATS667, but couldn't find a UK source and imported from HK)
[00:53:45] <Tom_itx> http://www.hamlin.com/technical-detail-hall-sensor.cfm
[00:53:54] <Tom_itx> haven't seen any of theirs
[00:54:24] <andypugh> Tom_itx: I think that the top two detect magnetised targets. The ATS667 detects ferromagnetic targets.
[00:55:11] <Tom_itx> i wonder what i had
[00:55:14] <Tom_itx> it was allegro
[00:55:17] <andypugh> Tom_itx: Yes, the Hamlin one is the right sensor, wrong package
[00:55:33] <Tom_itx> they may have different packaging
[00:55:33] <andypugh> Tom_itx: Probably ATS667
[00:56:05] <skunkworks__> I have some ats667's that I need to play with too
[00:56:27] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/hall/HallSensor2.jpg
[00:56:46] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/hall/Hall_Sch.png
[00:56:54] <Tom_itx> mmm doesn't say which one
[00:57:05] <andypugh> There is nothing wrong with it, it's a god device, which makes it all the stranger that there appears to be no alternative on the market
[00:57:29] <Tom_itx> it was for current sense i believe
[00:57:42] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/hall/Hall2.jpg
[00:57:52] <Tom_itx> that was fun to cut
[00:58:28] -!- Nick001 [[email protected]] has joined #linuxcnc
[00:58:40] <andypugh> Tom_itx: That looks different. That's a normal Hall (detects magnetic field). The ATS667 has its own internal magnet and detects the change in field when metal passes by
[00:59:28] <Tom_itx> it was for a windmill current sense circuit
[00:59:39] -!- micges has quit [Quit: Leaving]
[01:05:14] -!- karavanjo has quit [Read error: Connection reset by peer]
[01:06:41] <WillenCMD> today i set up up another successfully machine with using only hal components, this one was by far more complicated than the first
[01:08:35] <WillenCMD> using the loadrt list basically has every logic component in it, i think im using like 6 mux2, 2 mux4, 6 mult, 5 sum2's, 15 nots, 12 ors and dozens more
[01:10:18] <andypugh> WillenCMD: At that point, you need to consider writing a custom comp
[01:10:31] <WillenCMD> i know, im trying right now
[01:10:44] <WillenCMD> i printed off every hal component
[01:11:46] <WillenCMD> i have almost completed my servo-stepper tuning interface, that allows you to change real time pin data & saves it to your hal file
[01:12:53] <Tom_itx> andypugh, see anything here: http://www.rohm.com/products/lsi/sensor/hall_sensor/ic/?gclid=CJG6k4nLqbECFWVvTAodR1EA_g
[01:13:02] <WillenCMD> it graphs the position command, encoder position, and velocity, makes tuning a breeze you can see oscillation in the graph and adjust accordingly
[01:13:15] <andypugh> Neat
[01:13:54] -!- gmagno has quit [Ping timeout: 264 seconds]
[01:14:11] <andypugh> Tom_itx: No, they all have an mT threshold.
[01:15:09] <WillenCMD> my next goal is to make a graphical hal configure gui, that is similar to visual robotic programing. Putting hal comps as puzzle pieces allowing you to attach accordingly, it will then generate a hal file
[01:15:46] <andypugh> You will join a group of several in that quest.
[01:16:02] <WillenCMD> whats so hard about it?
[01:16:31] <cradek> the last 20%
[01:16:35] <Tom_itx> andypugh wasn't there an eagle thing that was supposed to do that?
[01:16:39] <andypugh> Folk do keep trying. Search the website for Graphviz, Crapahalic and Eagle
[01:17:17] <archivist> I should mod my database erd program to do it as well :)
[01:17:20] <andypugh> Can I change my answer to match cradek's?
[01:18:21] <WillenCMD> does graphiz work?
[01:19:14] <andypugh> Graphviz does, it is a well established package. Whether it works usefully for HAL is an open question
[01:19:28] <WillenCMD> gotcha
[01:19:31] <WillenCMD> ill check it out
[01:19:46] <WillenCMD> i did manage to get some of the glade components to show up
[01:20:21] <WillenCMD> in this ubuntu 12.04 sim instal
[01:21:22] <pcw_home> andypugh: I have not heard from anyone with a resolver/7I49 issue
[01:21:23] -!- cmorley1 [[email protected]] has joined #linuxcnc
[01:21:29] <andypugh> I think Graphviz is OK at HAL->diagram and Eagle is OK at diagram->HAL. And Crapahalic (which I hate the name of, by the way) is equally poor at both, but does try.
[01:22:12] <andypugh> pcw_home: Good. With LinuxCNC, or so you not know?
[01:22:36] -!- cmorley has quit [Ping timeout: 248 seconds]
[01:23:05] <andypugh> Making it look like a canonical encoder was the "not much thinking" approach, but seems like it might not ne optimum.
[01:23:31] -!- cmorley1 [[email protected]] has parted #linuxcnc
[01:24:05] -!- uw has quit [Remote host closed the connection]
[01:25:27] <Tom_itx> andypugh, ? http://www.melexis.com/Hall-Effect-Sensor-ICs/Special-Purpose-Hall-ICs/MLX90217-3.aspx
[01:26:49] <andypugh> Tom_itx: Aha! That's actually better than the Allegro one.
[01:27:19] <Tom_itx> there's not many of them out there
[01:27:24] <pcw_home> Well reading back to what Nick001-Shop said if someone had troubles with our hardware I would think we would hear about it
[01:27:41] <andypugh> But you have to add your own magnet, so I just changed my mind :-)
[01:27:47] <WillenCMD> i like to keep that stuff secret
[01:28:08] <Tom_itx> andypugh i thought it was on the back side
[01:28:27] <Tom_itx> maybe not
[01:29:32] <Tom_itx> so get a small magnet
[01:30:06] <andypugh> And I could have bought that from RS, after I would be finished the time machine: http://uk.rs-online.com/web/p/hall-effect-sensor-ics/6843301/
[01:31:29] <andypugh> Maybe the point I am missing is that they all work for my application if you add your own magnet?
[01:31:57] <Tom_itx> the allegro one has the magnet built in?
[01:32:12] <andypugh> Yes. Mine keep sticking together
[01:32:15] <Tom_itx> mmm
[01:32:50] <andypugh> I just picked up an AA battery with one
[01:33:28] <Tom_itx> http://unitednuclear.com/index.php?main_page=product_info&cPath=70_71&products_id=255
[01:33:33] <Tom_itx> stick a magnet on it
[01:33:48] <Nick001> andypugh - he was having to update something to got resolvers working properly and I lost track of what waas going on. He was also having tach problems which is common on that machine.
[01:35:26] <andypugh> There is good quailty velocity information from the resolvers at the 7i49, but getting that as a voltage to the drives is non-trivial
[01:35:52] -!- joe9 has quit [Quit: leaving]
[01:40:09] <djdelorie> andypugh: sorry, don't know. Was AFK too
[01:41:15] <andypugh> Sorry to drag you back :-
[01:42:01] <djdelorie> been an AFK day. made some fudge, worked on the southbend, cleaned the porch...
[01:42:21] <djdelorie> still haven't finished the inner control loop conversion for the controller boards, though...
[01:43:38] <andypugh> The good thing about a hobby is that you can work on the bits you fancy working on, or none at all. This is why I keep turning away paid work.
[01:44:40] -!- gmagno [[email protected]] has joined #linuxcnc
[01:44:40] <djdelorie> well, I have a special project that I need to do, and it requires a CNC machine
[01:44:57] <archivist> I dont get enough paid work :)
[01:45:16] <WillenCMD> i work for free but wish i was paid for it
[01:45:27] <WillenCMD> but i work for my dad still so...
[01:46:13] <Tom_itx> if i thought i could support it, i'd have a large cnc
[01:46:38] <Tom_itx> but keeping it a hobby is good
[01:46:38] <WillenCMD> actually funny but horrible story, last weekend i was submitting a resume to an employer and i must of some how printed it to the work printer and not realized it...
[01:46:53] <WillenCMD> lets just say that morning
[01:47:15] <WillenCMD> wasn't the best of mornings at work when they called me into the office holding my cover letter and resume to another employer
[01:47:28] -!- cmorley [[email protected]] has joined #linuxcnc
[01:47:30] <toastydeath> that sucks
[01:47:42] <toastydeath> everyplace I've worked so far I've given plenty of heads up i'm starting a job search
[01:47:44] -!- jpk has quit [Ping timeout: 260 seconds]
[01:47:45] <WillenCMD> i never make mistakes like that i was pissed
[01:47:49] <archivist> they should be offering a rise/raise
[01:47:51] <WillenCMD> oh i did also
[01:48:00] <WillenCMD> i told them last month
[01:48:08] <WillenCMD> but i guess they didn't take serious
[01:48:10] <toastydeath> ...then why did they call you into the office?
[01:48:14] <toastydeath> hah, morons
[01:48:20] -!- cmorley [[email protected]] has parted #linuxcnc
[01:48:38] <WillenCMD> they said they thought i was happy again, because the made me cnc supervisor(no pay raise though)
[01:48:59] <toastydeath> I've learned to never accept any additional responsibility without discussing pay firs
[01:49:02] <toastydeath> *first
[01:49:12] <WillenCMD> i laughed because i have been in charge un-officially for 3 years
[01:50:22] <WillenCMD> then they told me if i am going to quit they need to know now, and that i can't string them around like a yo-yo. Thats a direct quote
[01:50:22] -!- draig_ has quit [Read error: Connection reset by peer]
[01:50:30] -!- draig [[email protected]] has joined #linuxcnc
[01:50:35] <toastydeath> hahaha, did you laugh in their face?
[01:50:40] -!- cmorley1 has quit [Ping timeout: 260 seconds]
[01:50:41] <archivist> a label does not help feed you
[01:51:09] <toastydeath> "I'm not quitting, I'm giving you a heads up that it is in my near-term goals to resign."
[01:51:10] <WillenCMD> archvist: true but i did update the resume accordingly :)
[01:51:55] <WillenCMD> i told them, i am waiting for the right job and i can't give them a definite date.
[01:52:41] -!- uw [[email protected]] has joined #linuxcnc
[01:53:10] <WillenCMD> im actually putting a business plan together for my business in hopes of maybe receiving some investors
[01:56:38] <andypugh> it's always worth turning up at work in a suit and tie and then taking the afternoon off :-)
[01:57:18] <andypugh> I am being patronised on #sparkfun. I am quite enjoying it.
[02:00:16] <WillenCMD> do i just write a halcomp in gedit and save as a .comp?
[02:00:25] <WillenCMD> then install it
[02:02:34] <Tom_itx> andypugh have you met flyback yet?
[02:02:39] <Tom_itx> he's a real piece of work
[02:03:31] <archivist> flyback has earned a kline :)
[02:03:41] <andypugh> WillenCMD: Yes, save as .comp then sudo comp --install it
[02:04:32] <ReadError> Tom_itx,
[02:04:39] <ReadError> i think he just got banned on efnet
[02:05:10] <ReadError> trollin too hard
[02:05:58] -!- tjb1 [[email protected]] has joined #linuxcnc
[02:07:30] <tjb1> anyone here ever had to cut a laminate counter top to length
[02:07:50] -!- uw has quit [Quit: Ex-Chat]
[02:08:10] <Tom_itx> i've had them cut
[02:09:35] <tjb1> Not sure how to do it
[02:09:46] <tjb1> Im gonna have to make 4 cuts
[02:09:49] <andypugh> tjb1: I think I have
[02:10:11] <tjb1> Throw tape down the cut line and cut from the backside with a fine tooth circular saw?
[02:10:29] <Tom_itx> carbide if you have it
[02:11:04] <WillenCMD> then file it to fit :)
[02:11:34] <tjb1> Arent the teeth on circular saw blades carbide?
[02:11:37] <Tom_itx> they cap em and smooth that off with a router
[02:11:42] <Tom_itx> not all of them
[02:11:56] <Tom_itx> if they're silver soldered on they are
[02:12:01] <tjb1> One guy at lowes said about putting the blade on backwards
[02:12:17] <WillenCMD> it can help prevent it from fracturing
[02:12:22] <Tom_itx> keep the blade low
[02:12:30] <WillenCMD> do you have fine tooth saw blade?
[02:12:33] <tjb1> Yep
[02:13:01] -!- uw [[email protected]] has joined #linuxcnc
[02:13:32] <WillenCMD> do you have a vertical milling machine with a pallet table? thats an easy way :)
[02:14:09] <tjb1> Yeah in the shed!
[02:14:38] <WillenCMD> behind the lawnmower?
[02:15:03] <tjb1> no behind the 5 axis mill
[02:15:16] <WillenCMD> oh that has to be a tight fit
[02:15:25] <tjb1> Got a dewalt 6.5" 40 tooth fine finish carbide tooth blade
[02:15:33] <WillenCMD> did you put a cross on the shed and make it a church, to avoid taxes?
[02:15:45] <tjb1> no its not put in the ground
[02:15:46] <WillenCMD> go there to worship everynight
[02:16:06] <tjb1> we have 2 - asimos working the shop at night
[02:16:39] <WillenCMD> thats it, everyone knows you have to have 3 to get anything done...
[02:16:53] <WillenCMD> one has to be the tool bitch
[02:17:08] <tjb1> no we have a 5 axis robot do that
[02:17:19] <WillenCMD> is the shed hiring?
[02:17:26] <tjb1> lol
[02:17:29] <tjb1> there is no shed
[02:17:33] <WillenCMD> i know
[02:17:40] <WillenCMD> we did good though
[02:17:47] <WillenCMD> built a hell of a fake shop
[02:18:15] <tjb1> yep yep
[02:18:26] <tjb1> http://www.amazon.com/DEWALT-DW9196-2-Inch-Precision-Framing/dp/B00697YQIU
[02:18:28] <tjb1> There she be
[02:19:13] <tjb1> Heh bought that blade a week ago and according to the dewalt website it has been discontinued
[02:19:14] <andypugh> Oh, who was asking about cutting Alu with a wood blade?
[02:19:48] <tjb1> Dunno
[02:20:26] <andypugh> I routinely cut through 1" Alu with my hand-held, cordless, circular saw using the original blade. No problems so far other than underpants full of swarf.
[02:20:28] -!- cmorley1 [[email protected]] has joined #linuxcnc
[02:20:34] -!- cmorley1 [[email protected]] has parted #linuxcnc
[02:20:49] <Tom_itx> that's gotta be a good feeling
[02:20:55] <tjb1> Going to be interesting cutting the backsplash
[02:21:05] <tjb1> Blade isnt big enough to do it without rotating the saw
[02:21:06] <WillenCMD> yes, but Andy you seem to accomplish more than the every day do-it-yourselfer
[02:21:24] <WillenCMD> you drill holes through carbide with your cordless drill
[02:21:39] <WillenCMD> :)
[02:22:43] -!- cmorley has quit [Ping timeout: 252 seconds]
[02:23:02] <jdh> backsplash cuts fine from the back 7" circular saw
[02:30:39] -!- GH-1234 has quit [Quit: Leaving.]
[02:39:19] <r00t4rd3d> painters tape
[02:39:27] <r00t4rd3d> no chipping
[02:39:55] <jdh> score with a knife
[02:40:04] <r00t4rd3d> score with your wife
[02:42:17] -!- karavanjo has quit [Read error: Connection reset by peer]
[02:44:41] -!- draig has quit [Ping timeout: 246 seconds]
[02:51:47] -!- gmagno has quit [Remote host closed the connection]
[02:56:43] -!- cmorley1 [[email protected]] has joined #linuxcnc
[02:56:48] -!- cmorley1 [[email protected]] has parted #linuxcnc
[03:03:23] <r00t4rd3d> sketchup rocks
[03:03:52] <r00t4rd3d> with the phlatscript
[03:04:07] -!- Thetawaves [[email protected]] has joined #linuxcnc
[03:04:13] <r00t4rd3d> http://www.phlatforum.com/
[03:04:57] <r00t4rd3d> i can simply make accurate parts
[03:05:31] <andypugh> You can define actual sizes?
[03:05:36] <r00t4rd3d> yeah
[03:05:57] <r00t4rd3d> anally so
[03:05:57] <andypugh> I still find it unlikely that it actually "rocks"
[03:06:18] <r00t4rd3d> i can make parts with it and i cant with anything else so for me it rocks
[03:08:14] <andypugh> Ah, you reminded me of AutoCAD Mechanical 1998 vintage, where you would create a sketch to extrude, and it would tell you it lacked 7 constraints, and give you no clue at all which constraints were missing (and you couldn't use a sketch/profile unless it was fully constrained). It was like the world's least interesting logic puzzle, and mandatory.
[03:09:02] <toastydeath> hahahahah
[03:09:09] <WillenCMD> lol
[03:09:41] <andypugh> You laugh now. I had to design parts.
[03:10:34] <WillenCMD> im laughing because i can imagine how much of a pain that was
[03:10:35] <toastydeath> no, I laugh because I have had to use plenty of shitty modeling packages
[03:10:44] <toastydeath> i'm glad autocad mechanical wasn't one of them
[03:12:22] -!- draig [[email protected]] has joined #linuxcnc
[03:13:10] <andypugh> They threw it all away and made Inventor, and that is lovely. Then they dragged it back out of the skip and carried on with it, but in parallel to Inventor. (and I am sure that AutoCAD is better now)
[03:13:59] <toastydeath> i hated inventor
[03:14:08] <toastydeath> the packages i've liked so far are pro/e and solidworks
[03:14:14] <toastydeath> but i haven't used pro/e in a hot minute
[03:14:34] <Tom_itx> i think alot of packages beat acad for cad cam use
[03:15:00] <Tom_itx> i knew a guy way back when that used it along with nc polaris. i never understood why he liked it but he did
[03:15:07] <andypugh> I have almost no complaints about Inventor, other than that it is a nuiscance if you accidentally create an Imperial drawing and wish it had been in Metric. But it's no worse there than Aibre.
[03:15:22] <Tom_itx> http://www.ncpolaris.de/html/ncp_english.html
[03:15:33] <Tom_itx> i'm sure it's come a long ways since then
[03:16:19] <andypugh> That seems to be a CAM system, and claims to work with Inventor
[03:16:41] <Tom_itx> it was an add on for acad back then
[03:17:20] <Tom_itx> it worked but i sure didn't like it
[03:17:47] <andypugh> It is worth stressing that Inventor and AutoCAD are rather different (and Maya and Fusion are different again). They all do 3D modelling, but with rather different paradigms
[03:19:45] -!- tjb1 has quit [Quit: tjb1]
[03:20:55] <toastydeath> i didn't like inventor because at the point i was using it, I was most practiced in pro/e
[03:21:06] <toastydeath> which beats the everliving shit out of inventor for features and speed of modeling
[03:21:33] <toastydeath> i've never used autocad for anything except drafting
[03:28:47] <andypugh> I have never used ProE.
[03:29:45] <andypugh> I have found something Inventor can't do, though, and it might be a fun test for ProE. I could _not_ get it to assemble a hexapod.
[03:29:46] <Tom_itx> how does it compare to rhino?
[03:30:19] <andypugh> Tom_itx: It doesn't. Very different applications
[03:30:56] <Tom_itx> i've never used either one
[03:30:58] -!- uwe_ has quit [Ping timeout: 245 seconds]
[03:31:08] <toastydeath> castings is a good place to go for shit that doesn't work well in most modeling packages, that only pro/e and catia can handle
[03:31:35] -!- uwe_ [[email protected]] has joined #linuxcnc
[03:31:41] <toastydeath> but the speed and flexibility of pro/e was far greater than inventor
[03:31:56] <andypugh> The hexapod had hookes-roints at the top, and spherical feet with sliding square legs. It seemed unable to rotate the legs to solve the assembly.
[03:31:56] <toastydeath> *castings are
[03:32:21] <andypugh> What about castings?
[03:32:37] <andypugh> ie, what does ProE do that Inventor doesn't?
[03:33:54] <andypugh> It has a mould-design feature (which I have never used) which appears to go an awful lot further than split-plane definition, it seems to be able to design sliding parts automatically too,
[03:33:56] <toastydeath> radiusing on complicated revolved surfaces
[03:34:17] <toastydeath> some drilled holes on the same type of surface
[03:34:55] <andypugh> I have never had it fail to drill a hole, though sometimes it can be hard to define the hole.
[03:35:05] <toastydeath> I had a casting with 250+ features
[03:35:15] <toastydeath> in solidworks, and it couldn't place some of the things
[03:35:26] <toastydeath> and inventor is worse than solidworks for that
[03:36:05] <andypugh> You might end up having to create a web of points axes and planes to locate it in some situations, but I suspect that is fundamental geometry, not the package
[03:36:15] <toastydeath> ot
[03:36:18] <toastydeath> *it's the package.
[03:36:38] <toastydeath> both inventor and solidworks have types of features that they acknowledge cannot be made
[03:36:53] <toastydeath> which is why pro/e and catia are used in things like automotive and aerospace over solidworks and inventor
[03:37:09] <toastydeath> they don't handle importing geometry as well
[03:37:23] <Tom_itx> yeah, my friend went with catia in the end
[03:37:35] <Tom_itx> mostly i think because his vendors used it
[03:37:41] <toastydeath> moulds aren't that hard to make, they rarely have bizarre features
[03:38:10] <Tom_itx> but even that required an additional cam package
[03:38:41] <andypugh> Well, I work in Automotive, we appear to use Inventor to design factories, and Unigraphics to design cars.
[03:39:44] <toastydeath> unigraphics is on the same level as pro/e and catia
[03:40:15] <Tom_itx> boeing did alot of customization on catia for their purposes
[03:42:16] <andypugh> Anyway, time to sleep before the sun gets all the way up,
[03:42:17] -!- geo01005 [[email protected]] has joined #linuxcnc
[03:42:30] <toastydeath> gn
[03:55:03] <Jymmm> Is there another alcohol base clear coat other than shellac?
[03:57:36] -!- andypugh has quit [Quit: andypugh]
[03:58:46] <WillenCMD> anybody know how to uninstall a halcomp
[04:11:31] <cpresser> WillenCMD: just delete the file. but most likely you just want to unload it :)
[04:12:17] <WillenCMD> cpresser: no i made the component
[04:12:21] <WillenCMD> just testing it out
[04:12:55] <cpresser> then just delete the file.
[04:14:03] -!- ve7it has quit [Remote host closed the connection]
[04:20:16] -!- mhaberler [[email protected]] has joined #linuxcnc
[04:40:34] -!- Valen [[email protected]] has joined #linuxcnc
[05:03:32] -!- Fox_Muldr has quit [Ping timeout: 265 seconds]
[05:04:54] -!- Fox_Muldr [[email protected]] has joined #linuxcnc
[05:07:15] -!- Valen has quit [Quit: Leaving.]
[05:15:29] -!- dew_addict [[email protected]] has joined #linuxcnc
[05:41:58] -!- dhoovie [[email protected]] has joined #linuxcnc
[05:43:24] -!- dew_addict has quit [Quit: Ex-Chat]
[06:19:14] -!- geo01005 has quit [Ping timeout: 246 seconds]
[06:25:17] -!- WalterN [[email protected]] has joined #linuxcnc
[06:30:13] -!- WillenCMD has quit [Quit: WillenCMD]
[06:57:15] -!- DJ9DJ [DJ9DJ!~Deejay@unaffiliated/dj9dj] has joined #linuxcnc
[06:58:05] -!- DJ9DJ has quit [Client Quit]
[06:59:11] -!- DJ9DJ [DJ9DJ!~Deejay@unaffiliated/dj9dj] has joined #linuxcnc
[06:59:17] <DJ9DJ> moin
[07:09:32] -!- mhaberler has quit [Quit: mhaberler]
[07:11:17] -!- mhaberler [[email protected]] has joined #linuxcnc
[07:29:00] -!- Thetawaves has quit [Quit: This computer has gone to sleep]
[08:10:33] -!- rob_h [[email protected]] has joined #linuxcnc
[08:15:21] -!- draig_ [[email protected]] has joined #linuxcnc
[08:15:22] -!- draig has quit [Read error: Connection reset by peer]
[08:15:22] draig_ is now known as draig
[08:42:56] -!- bootnecklad [[email protected]] has joined #linuxcnc
[08:43:06] -!- bootnecklad [[email protected]] has parted #linuxcnc
[08:47:33] -!- mhaberler has quit [Quit: mhaberler]
[08:48:21] -!- karavanjo has quit [Quit: Leaving]
[09:04:13] -!- odogono has quit [Ping timeout: 252 seconds]
[10:14:15] -!- mhaberler [[email protected]] has joined #linuxcnc
[10:38:22] -!- syyl [[email protected]] has joined #linuxcnc
[10:54:31] -!- vladimirek [[email protected]] has joined #linuxcnc
[11:01:22] -!- Guthur [[email protected]] has joined #linuxcnc
[11:04:38] <Guthur> pjb: I possibly could use CL to implement this CL, though I'd still need to decide where to start
[11:04:52] <Guthur> oops wrong channel, sorry
[11:10:38] -!- syyl_ [[email protected]] has joined #linuxcnc
[11:13:49] -!- syyl has quit [Ping timeout: 255 seconds]
[11:18:44] -!- syyl_ws [[email protected]] has joined #linuxcnc
[11:23:45] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust639.basl.cable.virginmedia.com] has joined #linuxcnc
[11:24:42] <archivist> andypugh, fleabay toy 271019007935
[11:30:18] <syyl_ws> i see 1000 uses for that, spontanious
[11:30:50] <syyl_ws> but my milling machines will tip over, if i put it on the table...
[11:34:42] <anonimas1> i found a rotary table like that
[11:34:51] <anonimas1> only problem was it weighed 200kg..
[11:35:13] <anonimas1> the big issue was the physical size.. not actual weight tho..
[11:37:47] <andypugh> That one isn't huge, if the table is 8"
[11:38:20] <andypugh> And I have a friend who lives in Dudley
[11:44:17] <archivist> the only thing about 8" is getting up close with the spindle at 90 degrees
[11:44:40] <archivist> I have enough problems with a 6"
[11:45:18] <archivist> but that does look a reasonable lump of iron so start with
[11:45:34] <anonimas1> well, the top of the table is above my spindle..
[11:45:50] <anonimas1> only 650mm of Z travel :)
[11:46:21] * anonimas1 is pondering if 2 hours to 3d contour a part is a long time
[11:46:32] <archivist> I had to jack up my column as I ran out of height too
[11:49:35] -!- psha [[email protected]] has joined #linuxcnc
[12:19:09] -!- sumpfralle [[email protected]] has joined #linuxcnc
[12:19:12] -!- gmagno [[email protected]] has joined #linuxcnc
[12:22:20] -!- The_Ball has quit [Ping timeout: 248 seconds]
[12:27:54] -!- The_Ball [[email protected]] has joined #linuxcnc
[12:32:08] -!- WillenCMD [[email protected]] has joined #linuxcnc
[12:33:06] <WillenCMD> Top of the mornin
[12:34:00] <WillenCMD> nothing like making som halcomps, while listening to pandora "big band orchestral"
[12:34:10] <WillenCMD> some*
[12:34:28] <andypugh> No, you are right, there is very little like that.
[12:35:24] <WillenCMD> what i would give to go back 50 or 60 years and see some of these bands in person
[12:39:56] <WillenCMD> whats the advantage of parameters over pins?
[12:47:56] <andypugh> Not a lot
[12:48:27] <WillenCMD> both can be set in real time correct?
[12:48:53] <andypugh> In fact there was a project to change most parameters to pins a while ago, but it faltered when it became less than obvious that it was a good idea
[12:49:15] <andypugh> No. Parameters can only be set from userspace, because they can't be netted.
[12:49:39] <andypugh> In fact I don't think parameters even need to be in shared memory.
[12:50:12] <WillenCMD> gotcha, by they way not to sound creepy buti had a dream i was buying a tread mill from you lol
[12:50:18] <andypugh> I suspect that their existence is due to a desire to keep the shared memory area small.
[12:50:18] <WillenCMD> you wanted 800 for it
[12:50:33] <andypugh> Bizarre.
[12:50:35] <WillenCMD> i had to take over the financing
[12:50:47] <WillenCMD> we where working out shipping lol im not making this up
[12:50:56] <WillenCMD> maybe i need abreak from this place
[12:51:25] <andypugh> The only treadmill I would be likely to be getting rid of would be devoid of motor
[12:52:19] <andypugh> It's an interesting thing about dreams though, that you "just know" stuff. Like, how did you know it was me? I bet the dream-me was nothing like the real-me.
[12:52:21] <WillenCMD> true ever since picked up electronics as a hobby a few years ago, i can't let anything go without first stripping it of its usefull components, transofrmer, heavy duty diodes
[12:52:48] <WillenCMD> you where britsh thats all i remember talking to you on the phone
[12:53:07] <WillenCMD> i guess from watching your youtube videos
[13:20:27] -!- syyl_ws has quit [Quit: Verlassend]
[13:21:50] <archivist> hmm that monitor failure was expensive :) fleabay 330763624079
[13:22:41] <andypugh> Bargain.
[13:23:20] <archivist> going to cost as much to fetch almost!
[13:29:42] -!- mhaberler has quit [Quit: mhaberler]
[13:33:48] -!- syyl_ has quit [Ping timeout: 248 seconds]
[13:34:09] <WillenCMD> i keep getting an lvalue to left of assignment, on this http://pastebin.com/heNxXUuF
[13:34:22] <WillenCMD> its hard going back and forth from python to c
[13:35:38] <WillenCMD> im just trying to get the flow of this down
[13:36:14] <archivist> I have some cutters that need sharpening
[13:36:45] <WillenCMD> i just converted our hob sharpener into a cnc
[13:37:39] <WillenCMD> but i want to clean the hal file up so im attempting to make my own component
[13:39:38] <archivist> I was reviewing my hob collection yesterday while I restore/convert my barber colman
[13:40:32] <WillenCMD> what model?
[13:41:35] <archivist> S
[13:41:53] <archivist> seems to be a rare one
[13:42:29] <archivist> a before pic http://www.collection.archivist.info/archive/DJCPD/PD/2012/2012_07_10_Barber_Colman/IMG_1261.JPG
[13:43:15] <archivist> it did need to be rebuilt http://www.collection.archivist.info/archive/DJCPD/PD/2012/2012_07_12_Barber_Colman/IMG_1266.JPG
[13:43:51] <WillenCMD> we have 2 of those
[13:44:01] <archivist> you can just make out the broken gib adjustment screw
[13:44:28] <archivist> you have a manual ? and a scanner ?
[13:44:35] <WillenCMD> i bet we do
[13:44:49] <WillenCMD> im heading in there in a bit
[13:45:02] <WillenCMD> i'll check if we have it i'll make a pdf
[13:45:23] <WillenCMD> we have the auto loader for them
[13:45:51] <archivist> I dont have all the autoloader or any change wheels
[13:46:12] <WillenCMD> gears? we have a ton of the gear's
[13:46:41] <archivist> and there is not a lot of room to get a spindle sensor in the head for cnc
[13:47:59] <archivist> the one change wheel on the machine turned out to be a mod 1 not 24 dp so that is not from the machine
[13:48:20] -!- syyl [[email protected]] has joined #linuxcnc
[13:49:13] -!- sumpfralle has quit [Quit: Leaving.]
[13:49:17] -!- sumpfralle1 [[email protected]] has joined #linuxcnc
[13:49:41] <archivist> also thinking of going without the auto so I can get a little bigger wheels on it
[13:49:55] <WillenCMD> i would
[13:49:57] <WillenCMD> brb
[13:51:50] -!- geo01005 [[email protected]] has joined #linuxcnc
[13:53:53] -!- jpk has quit [Ping timeout: 252 seconds]
[14:08:28] -!- dhoovie has quit [Ping timeout: 248 seconds]
[14:13:11] -!- joe9 [[email protected]] has joined #linuxcnc
[14:14:57] <joe9> alex4nder: ReadError: how do you lubricate the x-axis ways? I use way-oil and am painting the ways with a brush. not sure if that is the correct method.
[14:15:30] <joe9> with the z and y axis, it is easy to see the oil on the ways.
[14:15:55] <joe9> but, with the x-axis it is not that easy as it is on the underside. any thoughts, please?
[14:17:05] <r00t4rd3d> i smear grease on with my finger
[14:20:07] <Tom_itx> think auto oiler
[14:21:14] <archivist> each machine has its quirks with lubrication
[14:23:03] <Loetmichel> i've done it with some brass tube... so i can regrease the ways wit a grease press...
[14:23:17] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=4660
[14:23:29] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=4663
[14:23:33] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=4669
[14:27:05] -!- bedah [[email protected]] has joined #linuxcnc
[14:28:46] <joe9> Loetmichel: that is some serious stuff.
[14:29:02] <joe9> r00t4rd3d: yes, that is what I am doing for the lead screw.
[14:29:17] <Loetmichel> joe9: in which way?
[14:29:26] <joe9> but, for the ways, the way oil just flows, it does not stay as grease does.
[14:29:42] <joe9> Loetmichel: looks pretty complicated there. atleast to my amateur eyes.
[14:29:51] <Loetmichel> isnt really
[14:30:02] <Loetmichel> just a bit of brass tube bent and soldered
[14:30:21] <Loetmichel> the whole machine was build on my home table
[14:30:46] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #linuxcnc
[14:34:02] <joe9> that is interesting.
[14:34:10] <Loetmichel> thats the whole machine as i had left it for the now owner of the model company i build it for: http://youtu.be/BUqo3zxv5Vo
[14:35:49] <Loetmichel> and yes, its HUGE
[14:36:08] <Loetmichel> about 1500*1020*160mm work area
[14:36:18] <anonimas1> http://ah.io23.net/ul/files/IMAG2590.jpg
[14:36:21] <anonimas1> time for new spindle bearings.
[14:36:38] -!- syyl_ws [[email protected]] has joined #linuxcnc
[14:37:01] <Loetmichel> looks more like a machine missing som rigidity ;)
[14:37:25] <anonimas1> I'm regularily having 200kg on the table.. :/
[14:37:27] -!- WalterN has quit [Read error: Connection reset by peer]
[14:37:39] <anonimas1> it's not the machine
[14:37:48] <Loetmichel> ok, than you are right
[14:37:53] -!- WalterN [[email protected]] has joined #linuxcnc
[14:37:53] <Loetmichel> than the bearings are gone
[14:38:03] <Loetmichel> or hte head is tilted slightly
[14:38:23] <Loetmichel> but then the marks should disappear in one direction
[14:38:33] <anonimas1> they dont that's why i assume it's the spindle bearings..
[14:38:34] <anonimas1> :)
[14:38:48] <anonimas1> equally uniform looking like crap
[14:39:21] <anonimas1> i dont know why they stuck standard tapered roller bearings in the spindle tho.
[14:39:56] <Loetmichel> anonimas1: because the usual fault for this "crap looking" is a bending (hobby)machine
[14:40:03] <Loetmichel> and not the spindle bearings
[14:40:09] <anonimas1> skf 32013x/q
[14:40:16] <anonimas1> indeed :)
[14:40:44] <anonimas1> well it'll serve it's purpose for getting the ballnut stuck on the hobby mill(actually measuring shaft so i can center it all.
[14:41:32] <Tom_itx> run it thru a tumbler and you'll never know it
[14:42:03] <Loetmichel> hrhr
[14:42:21] <anonimas1> i doubt aluminium for a ballnut mount :)
[14:43:40] <anonimas1> tho I dont get why they used standard bearings for a spindle.
[14:43:45] <Loetmichel> i lately had the "opertunity" to repair the electronics of a earlier copy of THIS: http://www.ebay.de/itm/120946203621
[14:43:54] <Loetmichel> ... WHAT a pile of crap
[14:43:58] <anonimas1> http://www.lathes.co.uk/abene/page4.html <- that kind of mill but with a heidenhain control
[14:44:39] <Loetmichel> and the user who is usung this has no clue at all for CNC, so he got a steep learning curve that evening ;-)
[14:46:05] <Loetmichel> first after changeing the TB6560 ws to grease the acme muts/spindles.
[14:46:08] <Loetmichel> nuts
[14:46:13] <anonimas1> Loetmichel: other alu work :) http://ah.io23.net/ul/files/IMAG1041.jpg
[14:46:49] <anonimas1> http://ah.io23.net/ul/files/IMAG1056.jpg
[14:46:55] <Loetmichel> result: G0 of F2400 instead of F400 before grease...
[14:47:33] <Loetmichel> anonimas1: is that a stamp/press tool?
[14:47:34] <andypugh> Loetmichel: Interesting vacuum hose design
[14:47:39] <Loetmichel> because of the mirror
[14:48:08] <anonimas1> Loetmichel: insert for a fibreglass mold
[14:48:17] <Loetmichel> andypugh: i got some problems with vacuum cleaner hose because this machine does 90% CF and GF plastics
[14:48:20] <anonimas1> Loetmichel: http://ah.io23.net/ul/files/IMG_6290%5B1%5D.JPG <- before my bearings were dying
[14:48:52] <Loetmichel> so the dust is quite abrasive and eats away the hose in the corners in about two months
[14:48:58] <anonimas1> minor difference? :D
[14:49:15] <Loetmichel> so i used 50mm PE pipe for waste water
[14:49:48] <Loetmichel> which has 2.5 mm wall thickness and is much more durable ;-)
[14:49:57] <Loetmichel> anonimas1: MUCH better ;-)
[14:50:34] <anonimas1> and i were suspecting they were bad even back then.. because of minor imperfections that you see in that picture(but hey the bearings are like 20 years old)
[14:51:05] <anonimas1> they used precision bearings, but they changed to normal one's probably because the precision ones from 196x something are less accurate then todays standard bearings
[14:51:22] <Loetmichel> anonimas1: hrhr, i used a kress FM6990E in that big machine before i got the isel MA2.05
[14:51:37] <anonimas1> I have a isel 0.75kw for my hobby machine
[14:51:37] <Loetmichel> and was going thru 2 sets of bearings a year ;-)
[14:51:57] <anonimas1> but a ufm one
[14:52:01] -!- Tom_L has quit []
[14:52:18] <Loetmichel> now i have a cheap chinese watercooled spindle
[14:52:20] <anonimas1> Loetmichel: http://www.spline.nl/photos/isel-750w-brushless-motor.png
[14:52:29] <Loetmichel> and i THINK the bearings will give in soon
[14:52:31] <anonimas1> 5-36k rpm
[14:52:44] <anonimas1> will run all the stuff I want to do at home fine i think
[14:53:17] <Loetmichel> because you can hear them like a "ciiiirrrrrpp" grasshopper at certain speeds/temperatures
[14:53:20] <anonimas1> err isa-750 is the spindle..
[14:53:41] <Loetmichel> nice
[14:53:47] <Loetmichel> but expensive i presume
[14:53:55] <anonimas1> yeah :/
[14:54:06] <anonimas1> the spindle itself isnt so expensive as the inverter for it
[14:54:21] <anonimas1> err 26k rpm sorry.
[14:54:50] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=12569 <- have payed about 300 Eur for spindle AND VFD
[14:54:57] <anonimas1> we'll see if i can run stainless on it.
[14:54:59] <Loetmichel> with all taxes and shipping
[14:55:08] <Loetmichel> as far as now i am satisfied
[14:55:41] <anonimas1> i were thinking of the cheaper isel router like one, but it felt like it was pretty stupid to stick that when im sticking good serovs and all onto the machine
[14:55:43] -!- gallenat0r [[email protected]] has joined #linuxcnc
[14:55:52] <Loetmichel> http://www.cyrom.org/palbum/main.php?g2_itemId=12506 <- got this in the package
[14:55:59] <anonimas1> as usual it starts like hobby, and in a year i'lll be running 12 hours a day :D
[14:56:35] <anonimas1> somehow, my big mill was like that (we have like 20 parts to make this year and outsourcing them is expensive)
[14:57:40] <anonimas1> and now somehow parts magically appear to be made every second
[14:57:53] <Loetmichel> hrhr
[14:57:58] <Loetmichel> better than me
[14:58:20] <Loetmichel> my little machine has to do stuff for the company approx. onece a month
[14:58:23] <Loetmichel> once
[14:58:36] <Loetmichel> ... AFTER regular work....
[14:58:41] <Loetmichel> and on weekends
[14:58:47] <Tom_itx> anonimas1 did you get those plastic parts to run?
[14:59:03] <anonimas1> yeah, I ran 240 of the small spacers last weekend
[14:59:26] <anonimas1> and the boxed i did like 60 of so far
[14:59:28] <anonimas1> boxes..
[14:59:40] <Tom_itx> no time to spend all that money you're making
[14:59:56] <anonimas1> i dont make a whole lot 50*3min of cycletime
[14:59:58] <Loetmichel> anonimas1: but not with a little selfbuild cnc made of aluminim?
[15:00:15] <anonimas1> no
[15:00:24] <Loetmichel> the little one is to small, to slow, to weak for high throuput work-
[15:01:02] <Loetmichel> i still have to engrave and drill about 400 slot plates for the next run of pcs for the company in the next two weeks...
[15:01:07] <jthornton> Tom_itx, I've added to the Glade3 tutorial
[15:01:18] <Tom_itx> good stuff
[15:01:19] <anonimas1> my limit is a slow control, but i end up programming alot in the shop so the heidenhain is really convenient for that
[15:01:26] <jdh> Loetmichel: I'd do that on company time and charge 10x
[15:01:38] <Loetmichel> ... you see me exited... especially as one plate takes about 6 min to make
[15:02:39] <Loetmichel> jdh: i am still hoping that with this order the company has enough money to buy the LONG promised CNC fot the company workshop
[15:03:13] <Loetmichel> i meand: 5kEur isnt THAT much... thats about one month of my salary (cost for the Boss)
[15:03:22] <Loetmichel> -d
[15:03:23] * anonimas1 nods
[15:03:42] <anonimas1> im still looking for a deal like that for a decent cnc
[15:04:00] <anonimas1> with a decent controller.(without mucking around to re-build it)
[15:04:41] <Tom_itx> jthornton at the end of each one, add a 'back to main' for us lazy folk
[15:04:54] <jthornton> back to index?
[15:05:07] <Tom_itx> yeah, after tutorial one part b
[15:05:10] <Tom_itx> for example
[15:05:54] <Tom_itx> or go on to 2
[15:06:17] <Loetmichel> anonimas1: i had this machine in mind for the company, that will be sufficient, because most work is modifying existing Printer/monitor/PC parts -> http://www.team-haase.de/maschinen/al-serie
[15:06:21] <jthornton> ok
[15:06:41] <Loetmichel> its about 5000 Eur in 1000mm*650mm and ballsrews
[15:06:48] <Loetmichel> and its plug and play
[15:07:04] <Loetmichel> get a computer, install LinuxCNC, plufg it in and go ;-)
[15:09:51] <anonimas1> I were thinking a dmg.. of some kind
[15:10:41] <Loetmichel> anonimas1: we are a company which makes Computers for military and services
[15:10:56] <Loetmichel> so most of our metalwork is outsourced
[15:11:06] * anonimas1 nods
[15:11:19] <Loetmichel> only the small parts/prototypes are build inhouse
[15:11:37] <Loetmichel> and for that this cheap CNC is sufficient
[15:11:53] <anonimas1> true
[15:12:05] <anonimas1> we do alot of prototyping inhouse, if we were to send that off we'd pay alot for it
[15:12:07] <Tom_itx> jthornton, looks good
[15:12:16] <anonimas1> and mainly steel
[15:12:19] <jthornton> Tom_itx, do a refresh
[15:12:32] <Loetmichel> i dont do steel
[15:12:53] <Loetmichel> more likely aluminium and occasionally plastics
[15:12:53] <anonimas1> http://www.dmgecoline.com/en-SE/28-dmc-1035-v-ecoline <- that one i would like
[15:13:06] <Loetmichel> harhar
[15:13:12] <Loetmichel> if you can pay for it ;-)
[15:13:41] <anonimas1> but it's kindof too small.
[15:13:45] <Tom_itx> jthornton, yeah.. i was thinking on the part b because on the first one you can just hit the back button once
[15:14:56] <Tom_itx> it's all good
[15:15:04] <anonimas1> haas seems like a better deal but why would i buy a machine that will hold 0.02mm on a good day.
[15:15:07] <anonimas1> :D
[15:18:34] <anonimas1> not so impressed by forum posts about them
[15:18:50] <jthornton> Tom_itx, do another refresh and see if it is a bit better
[15:18:59] <Tom_itx> aren't they more midrange machines anonimas1?
[15:19:04] <anonimas1> yeah
[15:19:21] <Tom_itx> high production of shit parts...
[15:19:30] <mazafaka> http://www.youtube.com/watch?v=2xbetIm4cvA&feature=related
[15:19:42] <anonimas1> it's weird that todays new machines like hurco cant stand up to a 1988 mazak. in terms of precision
[15:19:44] <Tom_itx> fadal is kinda like that too
[15:19:46] <anonimas1> and accuracy
[15:20:04] <mazafaka> this vehicle beside moto and Gelendevagen is UAZ, Loetmichel
[15:20:31] <syyl_ws> mh
[15:20:35] <anonimas1> i looked at a hurco somewhile back absolute positioning for the lathes were like 0.05mm(what do you do if you get a quote for a 0.01 part)?
[15:20:41] <anonimas1> err 0.005.. sorry
[15:20:42] <syyl_ws> friend of mine is runing a haas tl-2 lathe
[15:20:59] <anonimas1> syyl_ws: how many 0.001mm does it change between two parts?
[15:21:00] <syyl_ws> he does one-off jobs most the time
[15:21:46] <syyl_ws> he said, he has problems to measure the difference between two parts
[15:21:55] <syyl_ws> most error comes from the toolchange
[15:22:14] <anonimas1> what does he measure the parts with?
[15:22:34] <syyl_ws> mitutoyo digital micrometers
[15:22:36] <syyl_ws> most the time
[15:23:17] <anonimas1> that's pretty good
[15:23:23] <syyl_ws> off course they show a little deviation on the display
[15:23:32] <anonimas1> little?
[15:23:34] <syyl_ws> but there also plays the measuring error with
[15:23:41] <anonimas1> over 20 parts i have about +/0.001 because of ambient temp in the shop on my old lathe.. :/
[15:23:42] <syyl_ws> 2..4/1000 mm
[15:23:55] <Tom_itx> jthornton you forgot to rename at least one link
[15:24:01] <Tom_itx> part2b to 3
[15:24:07] <WillenCMD> brb
[15:24:28] <jthornton> ok thanks
[15:24:39] <Tom_itx> or it's going in circles :)
[15:24:49] <syyl_ws> thats pretty good, i think, when you have no climated measuring room ;)
[15:24:59] <anonimas1> syyl_ws: so basically between two parts you are about half out of tolerance for a 0.01 dimension
[15:25:21] <syyl_ws> depends on the tolerance ;)
[15:25:44] <anonimas1> +/- 0.005 ;)
[15:25:50] <syyl_ws> *runs*
[15:26:18] <anonimas1> yeah it's good, but not exceptional for a new machine :p
[15:26:20] <jthornton> Tom_itx, fixed the copy and paste error
[15:26:38] <syyl_ws> its a cheap machine, thats what to keep in mind :D
[15:26:46] <syyl_ws> even if it says "tool room" lathe ;)
[15:27:24] <syyl_ws> he bought it used, with only 75hours spindle time
[15:27:27] <anonimas1> we make some shafts for some bearings in one of our assemblies they need to be wiped down because if they get dusty you have to beat them into the bearings..
[15:27:55] <syyl_ws> good ol' hammer ;)
[15:28:29] <Tom_itx> jthornton, better :)
[15:28:51] <anonimas1> tho it werent by intention, i forgot to spec them smaller when i drew them up..
[15:28:51] <jthornton> thanks for proof reading it for me
[15:29:15] <Tom_itx> i'll study it more if i ever load glade
[15:29:20] <syyl_ws> ah, error of the engineer :D
[15:29:25] <anonimas1> the micrometer reads (75.000)
[15:30:18] <anonimas1> tho we dont have temp control either so i guess they are small when they get cold.
[15:30:21] <anonimas1> ;)
[15:30:41] <WillenCMD> is there an easy way to apply global permissions to all files in ubuntu
[15:30:52] <syyl_ws> sounds like tolerances that get the guy crazy, that has to assemble the stuff :D
[15:31:14] <anonimas1> they are a nice sliding fit, _if_ you wipe them first.
[15:31:34] <anonimas1> I were mounting myself the first time.
[15:31:37] <syyl_ws> what are the bearings bores? 75.000 plus?
[15:31:53] <anonimas1> i dont have that tolerance, and they measure 75.00?
[15:32:28] <syyl_ws> ok
[15:32:44] <anonimas1> I guess h6 or h5
[15:32:52] <anonimas1> H6/H5
[15:32:55] <syyl_ws> i like that, if the engineer tries his stuff by himself :)
[15:34:16] <anonimas1> we are imposing qc on all the stuff now instead of measuring by caliper and test fitting as we make small series of machines so we can keep spare parts on the shelve
[15:35:27] <anonimas1> we did alot of precise one/off stuff but they are usually trial fit, now it's all iso toleranced instead
[15:35:35] <anonimas1> much more sane.
[15:35:39] <pcw_home> WillenCMD: man chmod (but you will be quite unhappy if you change permissions on _all_ files)
[15:35:49] <syyl_ws> :)
[15:35:54] <syyl_ws> sounds reasonable
[15:36:08] <anonimas1> try to remember you made that shaft 0.03 under until next time ok?
[15:36:18] <anonimas1> because the hole is a bit tight..
[15:36:25] <anonimas1> for one of the machines in a series..
[15:36:32] <syyl_ws> of course we will remember
[15:36:40] <syyl_ws> when we try to assemble it ;)
[15:36:54] <anonimas1> worse, customer calls I need shaft with part number XF312312
[15:37:23] <anonimas1> ah, yeah no problem and you pull up the drawing, and the shaft is too big.. by 0.01mm.
[15:37:34] <syyl_ws> "ah, i remember, can you please measure your bore, so we can make you the shaft?"
[15:38:00] <anonimas1> it never works good
[15:38:20] <anonimas1> we still tried to keep to iso before but nowdays it's a self imposed thing out of nescessity
[15:39:29] <anonimas1> and it gives the impression that we are actually professional :D
[15:39:44] <syyl_ws> :D
[15:39:47] <syyl_ws> absolutely
[15:40:09] <syyl_ws> i have a colleague that gives a shit on the tolerances pointed out in a drawing
[15:40:10] <Tom_itx> anonimas1 now write an iso9000 qc manual for it
[15:40:25] <syyl_ws> he just fits everything together so it works kind a...
[15:40:26] <Tom_itx> we went thru all that once
[15:40:31] <anonimas1> Tom_itx: I am getting there someday
[15:41:01] <syyl_ws> its a pain to work with such a guy...
[15:41:07] <anonimas1> Tom_itx: (never iso certified tho)
[15:41:47] <WillenCMD> <= machinist + narcolepsy = bad news
[15:41:54] <anonimas1> I consulted for some iso certified company
[15:42:27] <anonimas1> they werent allowed to make any parts without a full drawing
[15:43:16] <anonimas1> Tom_itx: it seems like too much of a pain to buy 10 calipers and retire them after 2 years and re-calibrate them constantly
[15:43:57] <syyl_ws> calipers with a vernier are not calibrated in our company
[15:43:57] <anonimas1> also, the part where you cant have a non conforming measurement device in a iso certified setting
[15:44:00] <syyl_ws> only the digital one..
[15:44:29] <syyl_ws> and that every year...
[15:44:42] <anonimas1> ie, i cant give a guy at the shop floor a caliper that's not accurate enough for machining precision stuff but precise enough to measure the id of a hole down to 0.05mm..
[15:45:27] <anonimas1> (most stuff is mesured with verniers so even a old mitutoyo caliper would be twice as good)
[15:45:36] <anonimas1> measured.
[15:45:38] <syyl_ws> you can put a sticker on it "not for quality relevant measurement"
[15:45:51] <syyl_ws> normal thats ok with iso
[15:46:02] <syyl_ws> at least we do it ;)
[15:46:11] <anonimas1> well, as im the machinist and i do qc for the parts and draw the machines up, that's not really a problem
[15:47:07] <anonimas1> i have measuring gear that dosent leave the shop and i use my own caliper for anything accurate that i check against the micrometer gauge blocks when im in doubt
[15:47:52] <anonimas1> calipers are pretty crap anyway for anything accurate(measurement pressure is so damn critical)
[15:48:00] <syyl_ws> yeah
[15:48:05] <syyl_ws> every machinist should know that
[15:48:42] <syyl_ws> but there are still people that trust them for 1/100mm accuracy..
[15:48:54] <syyl_ws> and for bores...
[15:48:56] <anonimas1> tho, I dont lend micrometers out to anyone I'll happily come measure but please dont take them out of the shop on your own
[15:49:09] <anonimas1> (not to the other parts of the shop either)
[15:52:21] <anonimas1> mainly because I doubt most people would admit dropping one of them. so it can be sent for checking or fixing
[15:52:42] <anonimas1> or replacing.
[15:52:52] <JT-Shop> you can ony borrow my micrometers if my hand is attached to it
[15:53:26] <syyl_ws> "you have to take it from my dead, cold hand"
[15:53:27] <syyl_ws> ;)
[15:53:43] <JT-Shop> I have the other children that play in my shop trained to not touch them and clean up after themself :)
[15:53:49] <anonimas1> (um i dropped your ~300$ micrometer.., sorry...")
[15:54:34] <syyl_ws> "ok, run."
[15:54:42] <anonimas1> im slowly learning that machining stuff is never a problem, it's measuring it after..
[15:54:56] <anonimas1> :)
[15:54:57] -!- factor has quit [Read error: Connection reset by peer]
[15:55:28] <JT-Shop> and making the correct judgment on when to take a corrective action
[15:55:35] <anonimas1> yep
[15:56:15] <anonimas1> worst part is assuming that something is ok, and ending up trying to fit stuff by force. :)
[15:57:04] <anonimas1> and remaking parts because you were really sloppy the first time :D
[15:57:34] <syyl_ws> "the hydraulic press will make it fit"
[15:57:41] <syyl_ws> not the best way ;)
[15:58:12] <syyl_ws> but still better than parts with hammermarks..
[15:58:38] <Tom_itx> anonimas1 we had a lab come in every so often and recert everything
[15:59:15] <anonimas1> syyl_ws: standard fitting procedure :D
[15:59:24] <archivist> I use the first "few" to set the machine
[15:59:24] <Tom_itx> i'll admit the iso thing is more a paper pushing thing than anything
[15:59:36] <anonimas1> we started making holes for sliding hammers so you can mount stuff in field easier
[15:59:53] <anonimas1> even if things are a bit worn and beat
[16:00:19] <Tom_itx> JT-Shop did you train them by smashing their fingers a couple times?
[16:00:53] -!- RagingComputer has quit [Ping timeout: 248 seconds]
[16:01:10] <anonimas1> Tom_itx: agreed
[16:01:58] <anonimas1> why throwing out good measurement devices because of wear if it dosent happen(you mean you use that 24" micrometer everyday? 24/7 for 3 years
[16:02:15] <Tom_itx> anonimas1 but if a part fails they can follow back and know who to fire :)
[16:02:31] <anonimas1> yeah, for a big production op i understand it
[16:02:42] <anonimas1> it's cheap replacing stuff compared to failing parts
[16:02:52] -!- mazafaka [mazafaka!~mazafaka_@unaffiliated/mazafaka] has parted #linuxcnc
[16:02:53] <Tom_itx> boeing required it for all their vendors
[16:02:57] -!- mazafaka [mazafaka!~mazafaka_@unaffiliated/mazafaka] has joined #linuxcnc
[16:03:47] <Tom_itx> once it was in place, they never looked at it
[16:04:31] -!- FinboySlick [[email protected]] has joined #linuxcnc
[16:05:40] -!- RagingComputer [[email protected]] has joined #linuxcnc
[16:08:46] -!- factor [[email protected]] has joined #linuxcnc
[16:20:45] -!- servos4ever [servos4ever!~chatzilla@74-47-247-180.dr01.hnvr.mi.frontiernet.net] has joined #linuxcnc
[16:30:21] -!- heathmanc [[email protected]] has joined #linuxcnc
[16:32:59] <JT-Shop> Tom_itx: no, they are easy to train they are doctors
[16:33:53] -!- heathmanc has quit [Client Quit]
[16:34:25] <JT-Shop> and one is a retired phone guy and one installs medical equipment and they all know not to bite the hand that makes parts for them LOL
[16:35:32] <Tom_itx> been around the block a few times then
[16:36:18] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #linuxcnc
[16:36:26] -!- atom1 [[email protected]] has joined #linuxcnc
[16:36:26] -!- atom1 has quit [Changing host]
[16:36:26] -!- atom1 [atom1!~tom@unaffiliated/toml/x-013812] has joined #linuxcnc
[16:37:23] -!- Tom_L has quit [Client Quit]
[16:38:20] <WillenCMD> was wondering if anyone could explain why my halcomp i made never loads when i loadrt it
[16:38:32] <WillenCMD> it just says waiting for halcomponent
[16:39:00] <JT-Shop> yea, the old fart is 75 and still practicing
[16:40:33] <Tom_itx> that gets me about doctors. when do they quit practicing and actually get it right?
[16:41:25] -!- gmagno has quit [Ping timeout: 248 seconds]
[16:41:50] <Tom_itx> some of those guys are the best to have around though
[16:42:08] <JT-Shop> he is good at what he does...
[16:42:09] <Tom_itx> i had an old friend that loved fixing coo coo clocks of all things
[16:42:15] <Tom_itx> dr
[16:42:23] <JT-Shop> WillenCMD: not enough info to guess
[16:45:00] -!- atom1 has quit [Quit: Leaving]
[16:45:53] <WillenCMD> http://pastebin.com/RdREajir
[16:48:50] -!- karavanjo has quit [Remote host closed the connection]
[16:49:44] -!- IchGuckLive [[email protected]] has joined #linuxcnc
[16:49:52] <IchGuckLive> hi all
[16:51:06] <Tom_itx> does 10.04 use SysV style init process?
[16:51:33] <Tom_itx> or upstart
[16:52:26] <IchGuckLive> Tom_itx: https://help.ubuntu.com/community/UbuntuBootupHowto
[16:53:31] <andypugh> WillenCMD: I suspect that you can't use "index" as a pin name, because it is used in the HAL comp code, so you are macro-substituting important bits of the generated code...
[16:55:18] <andypugh> try using index_ instead
[16:55:29] -!- gmagno [[email protected]] has joined #linuxcnc
[16:56:49] -!- mhaberler [[email protected]] has joined #linuxcnc
[16:58:24] <WillenCMD> okay
[17:00:04] <WillenCMD> no luck "waiting for component 'hob-sharpener' to become ready.....................
[17:00:35] <WillenCMD> i have to hit 'ctrl +c ' to break from it
[17:00:57] <WillenCMD> frustrating
[17:01:06] <IchGuckLive> in customhal?
[17:01:55] <WillenCMD> yes
[17:03:21] <IchGuckLive> did you check the standart pin state positiv or negative
[17:04:05] <JT-Shop> WillenCMD: the component hobindex must be the same as the name iirc
[17:04:07] <WillenCMD> when i make changes to i have to remove and reinstall or just comp --install again
[17:04:59] <IchGuckLive> the ini of the mashine does not need to be recompiled
[17:05:12] <IchGuckLive> also the AXIS module if a change in python
[17:05:36] <WillenCMD> jt-shop: your the man
[17:05:46] <WillenCMD> :)
[17:06:12] <WillenCMD> the file name has to match the component name
[17:06:47] <JT-Shop> yes
[17:07:12] <WillenCMD> thanks
[17:07:17] <JT-Shop> just comp hobindex --install
[17:09:43] <andypugh> I was just about to say that it appears to work fine here
[17:09:57] <andypugh> But I chose the right filename...
[17:11:05] <WillenCMD> that solves one problem
[17:11:08] <WillenCMD> thanks
[17:11:14] <WillenCMD> now to just get it to work
[17:11:27] <WillenCMD> if i change gash to 11, enable isn't set to true
[17:11:46] <WillenCMD> oops
[17:11:48] <WillenCMD> n/m
[17:14:38] <WillenCMD> works great
[17:14:45] <WillenCMD> had to add it to the thread
[17:14:54] <WillenCMD> pretty damn easy
[17:16:27] <JT-Shop> yea
[17:21:28] <WillenCMD> i remember a 2 years ago hen i first starting using linuxcnc
[17:21:32] <WillenCMD> reading through the hal manual
[17:21:49] <WillenCMD> i laughed when i saw you could run it through command line thinking who the hell would do that
[17:22:09] <WillenCMD> now, i do it all the time
[17:22:59] <WillenCMD> hal is the best part of linuxcnc in my opionion
[17:26:43] <andypugh> Yes, I didn't see the point, but now I do nearly all my testing using halrun, as you can change things on the fly, and use a very minimal config to save confusion.
[17:27:37] <WillenCMD> yes beats the hell out of waiting for the debug box to popup realizing you typed a pin name wrong or set a parameter wrong, going down the list one by one
[17:27:56] <WillenCMD> its instantanious debugging on the fly
[17:31:11] <WillenCMD> man i should of wrote hal comps for the first machine andy
[17:31:20] <WillenCMD> would of been better than calbacks
[17:31:33] <WillenCMD> nothing beats realtime command handeling
[17:32:23] <WillenCMD> this hobsharpener it has to index so fast i don't have the luxury of using callbacks, with the first gladevcp app i used i could see delays in setting hal pins
[17:33:27] <archivist> production of new hobs?
[17:33:32] <WillenCMD> so for this i let hal take care of indexing, and do limit the gash error i pick up the index pulse every revolution
[17:33:43] <WillenCMD> to*
[17:34:15] <WillenCMD> no sharpening
[17:34:45] <WillenCMD> i have the stepper motor holding the hob directly
[17:34:57] <WillenCMD> well i should say driving the hob not holding
[17:36:55] <WillenCMD> to eliminate any backlash or wiggle i set the Pgain on the stepper as high as possible using an encoder i almost doubled the holding torque of the stepper motor
[17:37:25] <WillenCMD> it starts fighting back as soon as you turn the torque wrench
[17:39:39] <IchGuckLive> someone is using deskproto CAM
[17:41:25] <WillenCMD> how long can my halcomp program be are there any restrictions?
[17:42:15] <andypugh> I don't know of any restrictions, and there are some rather long ones.
[17:42:28] -!- IchGuckLive has quit [Quit: ChatZilla 0.9.87 [Firefox 13.0/20120601201853]]
[17:42:36] <WillenCMD> okay just making sure before i go to town on it
[17:53:59] -!- mazafaka has quit [Ping timeout: 246 seconds]
[17:57:14] -!- geo01005 has quit [Ping timeout: 260 seconds]
[18:03:05] -!- Nick001 has quit [Ping timeout: 246 seconds]
[18:04:06] -!- Thetawaves [[email protected]] has joined #linuxcnc
[18:07:52] -!- CyberGabber [[email protected]] has joined #linuxcnc
[18:21:42] -!- joe9 has quit [Quit: leaving]
[18:24:46] -!- vladimirek has quit [Remote host closed the connection]
[18:26:57] -!- GH-1234 [[email protected]] has joined #linuxcnc
[18:27:57] <WillenCMD> does the halcomp loop over and over again?
[18:29:57] <pcw_home> No its typically invoked in a thread and must "fall through" each time
[18:29:59] <pcw_home> that is you cannot "busywait" for things in a HAL comp
[18:30:36] <WillenCMD> but if a pin changes signal does it get called back?
[18:30:46] <pcw_home> No
[18:31:26] <pcw_home> the comp is a bit of code executed every thread invocation (say at 1KHz servo thread)
[18:31:49] <WillenCMD> oh okay so it is constantly updated
[18:32:08] <pcw_home> thats what the addf statement is about
[18:34:21] -!- mazafaka [mazafaka!~mazafaka_@unaffiliated/mazafaka] has joined #linuxcnc
[18:34:31] <andypugh> WillenCMD: Just imagine that some outer-loop is calling your code every millisecond, and all the pin values get updated in that outer loop. And if you don't return to that outer loop before your code is due to be called again, the Universe ends.
[18:34:59] <WillenCMD> is morgan freeman there?
[18:35:04] <andypugh> You _can_ busywait in the EXTRA_SETUP code
[18:35:39] <WillenCMD> as long as morgan freeman is there to explain when the Universe ends i'll be fine
[18:36:16] <Jymmm> WillenCMD: Ok, Mr Smith.
[18:36:46] <pcw_home> andypugh: so the extra setup stuff is called before the threads are started?
[18:37:05] <andypugh> Yes
[18:37:14] <pcw_home> OK good to know
[18:37:58] <WillenCMD> do you watch "in through the worm hole" ?
[18:38:08] <andypugh> In fact, working at the command line, the threads only start after you type "start" and I think in LinuxCNC the threads start after all the HAL files have been processed.
[18:39:39] <WillenCMD> okay im so mad, i spent so much time using all those components to make that thing index and i could of just made a simple hal component
[18:39:58] -!- geo01005 [[email protected]] has joined #linuxcnc
[18:40:01] <WillenCMD> i had a piece of paper mapping all the ands,or,and mux2 pins lol
[18:41:03] <WillenCMD> from now on, i'll have to be like a child and talk every idea over on here first to make sure im taking the best approach
[18:41:07] -!- WalterN has quit [Ping timeout: 240 seconds]
[18:46:01] <andypugh> Well, there are occasionally good reasons to use standard HAL blocks, in a sample config, or one for someone else to use in a normal LinuxCNC installation.
[18:46:51] <andypugh> I am doing a rather complex HAL mess today, and thinking how much easier it would be as a comp
[18:52:16] -!- syyl_ws has quit [Quit: Verlassend]
[18:55:20] <andypugh> Such elegant structures as conv_foat_s32 straight into conv_s32_float..
[19:03:28] <WillenCMD> andy do you have a great resource where i can learn about serial communication
[19:04:02] <WillenCMD> how fast could i send 7 bits using the standard serial port at its highest baudrate?
[19:06:30] <andypugh> Try Wikipedia
[19:06:41] <WillenCMD> whats that? ;)
[19:07:02] <andypugh> Seriously, it is likely to be a good starting point
[19:07:05] <archivist> and note what mode your serial is set up as, what handshake etc
[19:07:27] <WillenCMD> i did start there, its merely half a page
[19:07:37] <archivist> dont go overrunning your rx
[19:08:00] <andypugh> But 115,000bps is typical, so that's 90uS (about)
[19:08:42] <archivist> dont forget the start and stop bits for real length
[19:09:00] <andypugh> I assumed 10 bit backets
[19:09:38] <WillenCMD> i use serial now, just pyserial makes it easy and its only for sending an encoder position at startup
[19:11:30] -!- gmagno has quit [Ping timeout: 264 seconds]
[19:14:31] <WillenCMD> just want to run this by you guys, would it be better to:
[19:15:29] <WillenCMD> pick up the index position of the encoder, then grab the command position , subract it from the current position. Reset encoder to zero and set the position command to the difference. Or
[19:15:39] -!- Nick001-Shop [Nick001-Shop!~chatzilla@clsm-74-212-2-169-pppoe.dsl.clsm.epix.net] has joined #linuxcnc
[19:15:42] <WillenCMD> go 360 degree's one direction
[19:15:46] <WillenCMD> then 360 degree's back
[19:16:18] <WillenCMD> this is to of course eliminate accumulation error
[19:16:25] <WillenCMD> or as much as possible
[19:17:33] <WillenCMD> so if you have a 16 gash hob, it would be 15 degree's per gash, so it increments 15 every idex
[19:17:44] <WillenCMD> then when it hits 360, it decrements 15
[19:18:45] -!- ve7it [[email protected]] has joined #linuxcnc
[19:20:52] <WillenCMD> i am concerned the index position option might not be able to update all the parameters fast enough
[19:22:21] -!- pfred1 [pfred1!~pfred1@unaffiliated/pfred1] has joined #linuxcnc
[19:22:27] <andypugh> You can use index-enable to reset the encoder on index. Then when you see index-enable low again in your component, you know that the current encoder counts represent how far past index it has moved in the time between index and your code being called.
[19:22:27] <andypugh> Does that help at all?
[19:23:16] <Tom_itx> when do you decide to use hal over glade over classic ladder etc?
[19:24:01] <Tom_itx> halui*
[19:24:54] <andypugh> Your component sees a snapshot. Nothing else will update your pin values while your code is running. In general. Your servo-thread code can be interrupted by base-thread functions, but you will typically be looking at the servo-thread pins from the companion functions of the base-thread functions.
[19:25:11] <andypugh> Tom_itx: Whim.
[19:25:22] <Tom_itx> i wondered
[19:25:44] <Tom_itx> i didn't know if one was better suited to do certain things
[19:26:05] <WillenCMD> andypugh: yes i thought of that, right now on startup the machine is homed, then the hob is placed on the spindle, the setup tech touches off to the grinding wheel. I could save the difference between the touch off and when the endex signal goes high that offset value is the poistion command
[19:26:07] <andypugh> Horses for courses. You can't do a UI in a realtime hal component, and you can't do realtime stuff in a UI
[19:26:25] <pfred1> how come?
[19:27:18] <Tom_itx> i notice axis doesn't get updated as quick with halui doing some things
[19:27:21] <andypugh> Many reasons. You can't even try to a large extent, because realtime code runs in kernel space.
[19:28:15] <pfred1> what machine access doesn't?
[19:28:55] <WillenCMD> pfred1: i have seen first hand what the delay between the ui and hal can cause, i basically now only use the ui for toggling hal components and feedback. but all of the data is sent to hal on cycle start & hal does the rest.
[19:29:04] <pfred1> just because it runs in kernel space doesn't restrict it to there
[19:29:39] <WillenCMD> of course thats using gladevcp
[19:31:44] <pfred1> it could be that glade exacts too high a performance penalty
[19:31:54] <WillenCMD> probably
[19:32:29] <pfred1> all I know is a lot of the power of computing is pissed away today
[19:33:04] <WillenCMD> its because we are lazy, who wants to write in assembly?
[19:33:19] <pfred1> it is the only language I've ever had much success in
[19:33:42] <WillenCMD> C does pretty good
[19:33:43] <pfred1> the higher level languages are too abstract for me
[19:34:03] <pfred1> they say C compiles as well as assembler
[19:34:34] <WillenCMD> python is great for fast developement, and you can write modules in C for resource hungry operations
[19:35:18] <WillenCMD> python+glade i can make a gui pretty fast and im still new to it
[19:36:28] <pfred1> ever tried gambas?
[19:36:30] <WillenCMD> how is gambas going?
[19:36:47] <pfred1> well I haven't fooled with it much lately been working on cleaning something else up
[19:36:47] <WillenCMD> is your memory fading? lol
[19:37:01] <pfred1> I don't commit much to memory
[19:38:03] <WillenCMD> oh, ever since i saw this bright light outside my house. I remember everything i see, i can read a book in a few minutes learn languages in as little as an hour.
[19:38:10] <pfred1> I wanted to get something knocked off my todo list so I set my other project aside
[19:38:35] <WillenCMD> im working on moving a pencil with my mind
[19:38:41] <pfred1> turned out it has taken me a little longer than I initially thought it would though
[19:39:09] <pfred1> I went to school with a kid that said he had a blackboard in his mind
[19:39:27] <pfred1> after he got a masters out of stanford he told me he upgraded to a calculator
[19:39:32] <WillenCMD> was it john travolta?
[19:39:53] <pfred1> no it wasn't john travolta
[19:41:13] <WillenCMD> i was just kidding earlier, don't want people thinking im crazy because they haven't seen the movie phenomenon
[19:41:28] <pfred1> I've seen it 3 times
[19:41:41] <WillenCMD> have you seen the second one or the lifetime remake?
[19:41:54] <WillenCMD> much better ending
[19:41:54] <pfred1> didn't travolta have a brain embilism?
[19:42:04] -!- cylly [[email protected]] has joined #linuxcnc
[19:42:07] <WillenCMD> brain tumor
[19:42:19] -!- sumpfralle [[email protected]] has joined #linuxcnc
[19:42:19] -!- sumpfralle1 has quit [Read error: Connection reset by peer]
[19:42:24] <pfred1> no that was arnie it's not a tumor!
[19:43:06] -!- demacus [[email protected]] has joined #linuxcnc
[19:43:18] -!- Loetmichel has quit [Ping timeout: 264 seconds]
[19:44:20] -!- mhaberler has quit [Ping timeout: 248 seconds]
[19:44:42] -!- demacus has quit [Remote host closed the connection]
[19:47:55] <r00t4rd3d> http://i.imgur.com/8PazS.jpg
[19:51:23] -!- toastydeath has quit [Read error: Connection reset by peer]
[19:52:05] -!- toastydeath [[email protected]] has joined #linuxcnc
[19:53:37] -!- mhaberler [[email protected]] has joined #linuxcnc
[19:56:21] -!- demacus [[email protected]] has joined #linuxcnc
[20:00:08] -!- demacus has quit [Remote host closed the connection]
[20:00:42] cylly is now known as Loetmichel
[20:11:16] -!- mhaberler has quit [Read error: Operation timed out]
[20:13:22] -!- mhaberler [[email protected]] has joined #linuxcnc
[20:14:11] -!- psha has quit [Quit: Lost terminal]
[20:28:45] -!- syyl_ [[email protected]] has joined #linuxcnc
[20:30:04] -!- syyl has quit [Ping timeout: 260 seconds]
[20:47:07] -!- Thetawaves has quit [Ping timeout: 252 seconds]
[20:49:50] <DJ9DJ> gn8
[20:50:09] -!- DJ9DJ has quit [Quit: bye]
[20:53:13] <syyl_> ich werf jetzt mal ne trovaltrommel mit auf die bau-liste
[20:53:14] <syyl_> Oo
[20:56:20] -!- syyl [[email protected]] has joined #linuxcnc
[20:56:54] -!- mhaberler has quit [Quit: mhaberler]
[20:57:37] -!- syyl_ has quit [Ping timeout: 240 seconds]
[21:02:39] -!- syyl has quit [Ping timeout: 255 seconds]
[21:04:19] -!- syyl [[email protected]] has joined #linuxcnc
[21:05:47] -!- the_wench has quit [Ping timeout: 246 seconds]
[21:07:18] -!- archivist has quit [Ping timeout: 265 seconds]
[21:19:55] -!- archivist [[email protected]] has joined #linuxcnc
[21:29:59] <pfred1> The BBC is reporting today that Microsoft has swiftly removed a bunch of code labelled 0xB16B00B5 which can be read, as 'Big Boobs'
[21:34:16] <WillenCMD> why won't halcomp let you set in pins
[21:34:53] <WillenCMD> i want a pin thats toggled true immediatly forced false after the function runs
[21:36:00] <r00t4rd3d> why do you have to get so technical?
[21:36:21] <cradek> if you need either "end" of a pin to write to it, you need to use a bidirectional pin
[21:36:50] <WillenCMD> cradek: io pins do let me write to it, but then i can't link an out signal to it
[21:37:02] <cradek> this is how index-enable works (one end sets it high to request wait for index, when index passes the other end sets it back low)
[21:37:17] <cradek> yes that is correct
[21:37:54] -!- geo01005 has quit [Ping timeout: 264 seconds]
[21:38:05] <WillenCMD> thats similart to what i want, a limit switch comes on sets index true, function runs but in case it returns and the switch is still true i dont' want it running again
[21:38:09] <WillenCMD> similar*
[21:38:31] <WillenCMD> i could put a count in the function i suppose
[21:38:42] <Jymmm> or a flag
[21:38:44] <cradek> so only run on the rising edge
[21:38:48] <WillenCMD> yes
[21:39:06] <WillenCMD> rotates, in this case the hob the required degree's
[21:39:16] <WillenCMD> and waits for the table to go back and forth
[21:39:22] <cradek> good that you said what your goal was, because it's easy
[21:39:53] <WillenCMD> halcomp is easy, i would like to give the maker a hug
[21:40:10] <WillenCMD> or a handshake if he or she isn't into hugging
[21:40:13] -!- odogono has quit [Quit: odogono]
[21:40:23] -!- gmagno [[email protected]] has joined #linuxcnc
[21:40:37] <Jymmm> Give em a big fat wet sloppy kiss dab smack on the lips!
[21:40:37] <cradek> you mean comp, the thing that builds and installs hal components?
[21:40:47] <WillenCMD> yes
[21:41:08] <cradek> ah, that would be jepler
[21:41:13] -!- bedah has quit [Quit: bye]
[21:41:44] <WillenCMD> where you getting ready to explain how it is easy or where you just implying it would be easy for you?
[21:41:59] <cradek> well actually neither
[21:42:17] <WillenCMD> should i just implement a count in the function
[21:42:29] <cradek> to detect rising edge, you just check "is it high NOW but was low last time I checked"
[21:42:34] <Jymmm> WillenCMD: Don't let cradek fool you, he had part to do with it too! So you can give him a semi sloppy kiss too!
[21:42:42] <cradek> so you just have to remember last time's value
[21:42:49] <cradek> it's not a matter of counting at all
[21:42:57] <WillenCMD> okay i follow you
[21:43:14] <WillenCMD> i was just making sure it wasn't something i was doing wrong with in pins
[21:43:20] <andypugh> Normally you have a variable to store the old value, and then "so stuff" when the new value is true and the old value is false. Look at oneshot for example.
[21:43:26] <cradek> nope, if they're IN they're IN only
[21:43:39] <WillenCMD> thanks!
[21:43:49] <Jymmm> if ($now==HIGH && $was ==LOW){}
[21:44:11] <cradek> beware that you only read the input once per run
[21:45:02] <Jymmm> if ($now==HIGH && $was ==LOW && $run_once==TRUE){}
[21:45:03] <cradek> an easy bug to write: if (READPIN and !old_pin) {do it}; old_pin = READPIN;
[21:45:04] <WillenCMD> r00t4rd3d: I think what im doing with linuxcnc is somewhat new, branching out past what most people use it for
[21:45:22] <cradek> ... becase the pin's value can change between the first and second reads
[21:45:29] <WillenCMD> true
[21:45:34] <WillenCMD> very detailed answer
[21:45:37] <WillenCMD> thank you
[21:45:52] <cradek> welcome
[21:46:41] <WillenCMD> should i share components like this? i presume it would be usefull if you had to do drilling on say hexagon stock, or no?
[21:46:52] <WillenCMD> on a rotary table
[21:47:06] <WillenCMD> actually it will do any number of sides
[21:47:32] <cradek> there is a spot on the wiki for user-contributed comps; it's great if you share when you think it's useful for others
[21:51:27] <andypugh> WillenCMD: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?ContributedComponents
[21:52:01] <Jymmm> couldn't he just check it in to the repository?
[21:52:38] <Jymmm> ...for inclusion into the next release?
[21:53:07] <andypugh> Maybe. It's hard to say when a new HAL component is useful enough to be part of the distribution.
[21:53:22] <andypugh> Admittedly they don't do any harm.
[21:54:08] <Jymmm> Just seems a "user contrib" seems sei-useless sitting on a wiki somewhere
[21:54:13] <Jymmm> semi
[21:54:29] <Jymmm> "at your own risk" of course =)
[21:54:47] <andypugh> It might be worth someone (who?) going through the contributed comps looking for ones that could be useful. For example the motherboard sensors one.
[21:55:13] <andypugh> I can imagine plugging remote instrumentation into motheboard plugs, even.
[21:56:03] <WillenCMD> yeah i actually thats the first of seen of that page you just sent me andy
[21:56:15] <Jymmm> Maybe add a notation to the wiki... "If you want your contrib included in a future release... check it into the repostory"
[21:56:19] <WillenCMD> i have been staring at halcomps for the past 3 months
[21:56:37] <WillenCMD> if nothing else a simple link to the ueser components would be nice
[21:56:42] <WillenCMD> user*
[22:00:02] -!- gallenat0r has quit [Quit: This computer has gone to sleep]
[22:03:51] -!- CyberGabber [[email protected]] has parted #linuxcnc
[22:07:27] -!- syyl_ [[email protected]] has joined #linuxcnc
[22:08:31] -!- syyl has quit [Ping timeout: 252 seconds]
[22:11:35] -!- davec_ has quit [Ping timeout: 246 seconds]
[22:13:54] -!- GH-12341 [[email protected]] has joined #linuxcnc
[22:15:43] -!- delkin [[email protected]] has joined #linuxcnc
[22:15:48] <delkin> What drill bits do you think are the most important for pcb milling? I am thinking about 0.2, 0.3, 0.8, 1.0 and 2.0 [mm]. Would you suggest something else?
[22:16:18] -!- GH-1234 has quit [Ping timeout: 264 seconds]
[22:19:56] <pcw_home> The spare ones to replace the ones you break?
[22:22:21] <delkin> pcw_home: If you had to make an arduino pcb what drill bits sizes would you use?
[22:22:42] <delkin> in general
[22:24:19] <pfred1> delkin get calipers measure pins or read the datasheets for the parts
[22:24:43] * pfred1 uses calipers
[22:26:10] <Nick001-Shop> Where does one get Anp Mini Mate-N-Lock plugs for the mesa cards
[22:26:18] <Nick001-Shop> Amp
[22:28:53] <DaViruz> farnell
[22:29:10] <andypugh> Nick001-Shop: You don't. Only PCW has a secret stash of that size.
[22:29:49] <DaViruz> are you sure? i believe farnell stock the entire range
[22:30:40] <skunkworks__> WillenCMD: just jump in and start playing with comps.. That is what I did.
[22:31:20] <andypugh> DaViruz: Yiou may be right. I know RS don't have them.
[22:39:46] -!- tron has quit [Client Quit]
[22:40:16] <Nick001-Shop> if i buy 7i30 cards I need them to connect them to my stuff
[22:41:50] <WillenCMD> my halcomp is one hundred percent functional
[22:41:55] <WillenCMD> :)
[22:43:06] -!- FinboySlick has quit [Remote host closed the connection]
[22:46:59] -!- karavanjo has quit [Quit: Leaving]
[22:54:27] -!- netman86 [[email protected]] has joined #linuxcnc
[22:55:10] <netman86> This looks like a group of people that collects hobbies. Does anyone here play around with metal casting? I'm looking for information on refractory.
[22:55:34] -!- dcarter21 [dcarter21!d03626b6@gateway/web/freenode/ip.208.54.38.182] has joined #linuxcnc
[22:55:58] <pfred1> I've made my own blast furnace
[22:56:12] <dcarter21> I'm sure this has been asked before, but what has happened to the buildbot?
[22:56:19] <pfred1> what kind of metal do you plan on melting?
[22:56:35] <netman86> I'd like to cast some iron.
[22:56:55] <pfred1> I've only done aluminum
[22:57:10] <netman86> Endgame, anyways. I'd start with aluminum, but I've got the heat capability for steel- I've set a goal of making a cast iron pan.
[22:57:10] <andypugh> dcarter21: It is dead. It is being worked on
[22:57:15] <pfred1> I hear iron can be a bit trickier
[22:57:40] <netman86> I've done a fair bit of reading, but I can't find too much good information on refractory rated for these temps.
[22:57:43] <pfred1> aluminum is so easy you could dig a pit and do it
[22:58:00] <netman86> Right. I've melted aluminum with torches.
[22:58:03] <dcarter21> Kind of what I figured. I've tried for several days. Was just hoping it wasn't dead for good.
[22:58:07] <pfred1> don't overthink aluminum is what I'm saying
[22:59:36] <pfred1> netman86 you been to Lionel's website?
[23:00:08] <WillenCMD> i have melted aluminum just by rubbing my hands together
[23:00:13] <netman86> Yes, I was just there actually.
[23:00:40] <pfred1> he seems to have cracked the cast iron ceiling with casting
[23:00:41] <netman86> he has some really neat skill, but he says his home made refractory isn't good much past aluminum temps
[23:01:11] <pfred1> well with higher heats you have to expect refractory material to fail I think
[23:01:29] <pfred1> when they fire pottery they usually rebuild the things
[23:01:34] <pfred1> reline them
[23:01:44] <netman86> yes, but not after every run :)
[23:01:55] <pfred1> back in the day with iron they did
[23:02:11] <netman86> well, that's harsh.
[23:02:39] <pfred1> today it is all electrode discharge they don't even use flames
[23:02:44] <pfred1> so it is all different
[23:03:04] <pfred1> they have these things that look like huge welding rods
[23:03:42] <pfred1> bascally I guess that is what they are
[23:03:46] -!- Como| [Como|[email protected]] has joined #linuxcnc
[23:03:53] <Como|> Err. Was I booted, or did my client drop?
[23:04:11] -!- netman86 has quit [Disconnected by services]
[23:04:19] Como| is now known as netman86
[23:04:34] <pfred1> you saw as much as I did
[23:04:34] <netman86> Looks like I dropped. Odd.
[23:04:35] -!- dcarter21 has quit [Ping timeout: 245 seconds]
[23:06:32] -!- Guthur has quit [Ping timeout: 246 seconds]
[23:08:20] <netman86> Looks like lionel labs buys his refractory for iron
[23:08:28] <netman86> guess I should start saving
[23:11:55] <pfred1> netman86 are you going to make a waste oil burner?
[23:13:15] <netman86> pretty much, yeah.
[23:13:50] <netman86> I've got 130 gallons of b100 that I've been burning in my truck- I figure I can use that until its up to temp, and then burn used motor oil/WVO
[23:21:27] <netman86> there's always #2 if I run out of fuel.
[23:22:37] -!- Thetawaves [[email protected]] has joined #linuxcnc
[23:28:03] -!- syyl_ has quit [Ping timeout: 255 seconds]
[23:28:03] -!- FinboySlick [[email protected]] has joined #linuxcnc
[23:29:08] <netman86> I'm reading that a sand/fireclay combo should work fine for these temps, too.
[23:29:28] <pfred1> one thing I have a lot of is sand
[23:29:47] -!- r00t-Shed [[email protected]] has joined #linuxcnc
[23:30:07] -!- factor has quit [Read error: Connection reset by peer]
[23:31:06] <netman86> I'm sure I could find some, I have a shovel.
[23:31:18] <pfred1> that is how I'd get sand
[23:31:33] <r00t-Shed> i got tons of sand around me
[23:31:40] <r00t-Shed> natural sand
[23:31:54] <pfred1> I'm about 6 miles from the beach so
[23:31:58] <Jymmm> rob the neighbors kids sandbox =)
[23:32:09] <pfred1> I dig down about a foot and it is sand to China
[23:32:21] <r00t-Shed> goto lowes, the sell play sand in 80lbs bags
[23:32:43] <pfred1> then I'm a multi gazillionare
[23:32:47] <netman86> I don't need sand so much
[23:33:01] <r00t-Shed> what are you making, glass?
[23:33:05] <pfred1> there's sand pits all around me
[23:33:15] <pfred1> where they pump out sand
[23:33:32] <netman86> No, I'd like to melt iron. I'm reading that a fireclay/sand mixture works pretty well as a refractory.
[23:34:06] <r00t-Shed> just get an old fuel oil furnace
[23:34:18] <pfred1> I got 2 old kilns at the town dumps and i took all the bricks out of them
[23:34:19] <r00t-Shed> they have a refractory chamber in them
[23:34:41] <pfred1> I think the bricks i have are the 2,500F variety the soft bricks
[23:34:55] <pfred1> they're really light and soft
[23:35:02] <pfred1> I can cut them with a hand saw
[23:36:10] <netman86> yeah, I've got some of those
[23:36:24] <netman86> 2500F isnt enough, unfortunately
[23:36:40] <pfred1> nope good for aluminum though
[23:36:54] <netman86> and the refractory in a typical fuel oil furnace won't be rated for much more than 500
[23:37:01] -!- WiillenCMdesign [[email protected]] has joined #linuxcnc
[23:37:28] <netman86> according to the wiki, high grade fireclay is typically good to 3200 or so
[23:37:39] <netman86> IIRC that's about where you want iron for a good pour
[23:38:09] <pfred1> yeah it seems you have to superheat melts a bit
[23:39:22] <netman86> the idea is to have extra heat, not only to remove impurity but to make sure its still all liquid at the end of the pour
[23:40:12] -!- WillenCMD has quit [Ping timeout: 248 seconds]
[23:47:42] <netman86> hm, this place says cast iron pouring temp is closer to 2600F or so
[23:48:11] <netman86> so if I plop down the cash for 3000F refractory it might just last far more than 10 runs
[23:48:35] -!- factor [[email protected]] has joined #linuxcnc
[23:54:15] <r00t-Shed> My binocular fence guy is almost complete:
[23:54:18] <r00t-Shed> http://i.imgur.com/9stCT.jpg
[23:55:05] <r00t-Shed> im gonna carve come eyes