Back
[00:23:50] <tiwake> XXCoder: flying out to texas thursday night
[00:24:07] <tiwake> coming back on sunday
[04:00:00] <Deejay> moin
[04:59:35] <XXCoder> heys
[05:12:01] <jthornton> morning
[05:12:30] <XXCoder> yo
[05:26:59] <XXCoder> bah I can't wait to get my deburr noga tool
[05:27:32] <XXCoder> microsoft has been looking for "cnc model maker" hard for months
[05:28:03] <XXCoder> https://www.linkedin.com/comm/jobs2/view/150847261
[05:29:31] <archivist> linkedin is a blank screen for me, effin useless JS websites
[05:31:20] <XXCoder> they wants very high end skilled guy
[05:31:35] <XXCoder> engineering, few years experence on it, and 10 years on modeling I think
[05:32:23] <XXCoder> accerated 4 year engineering egree, 4 years cad, 10 years building class a apperence models and prototypes
[05:32:30] <archivist> must know outlook IE and mach3 and blue screen of deat
[05:33:25] <XXCoder> expert cam/cnc programming, expert cad solidworks and/or some others, machining skills vertical mills, lathes and other machine tool equipment
[05:34:00] <XXCoder> finish skills including painting paintings and coatings, computer skills, and of course excellent writing and verbal skills
[05:34:11] <XXCoder> they dont need one guy, they need 3.
[05:34:20] <archivist> and most important, a degree does not mean the person has a clue
[05:34:28] <XXCoder> indeed
[05:35:33] <archivist> I need a job.... but I have standards :)
[05:35:48] <XXCoder> can anyone meet that requirements anyway?
[05:38:05] <XXCoder> wow free books for review batch this month sucks
[05:38:53] <archivist> often job requirements are very badly written
[05:39:34] <XXCoder> man scam emails is making a comeback
[05:39:41] <archivist> I am not sure that many multi skilled people have posh degrees
[05:39:44] <XXCoder> I just got lottery AND "similiar surname" scam
[05:41:10] <archivist> I just got a first name one to a random url, but gmail had parked it in the spam folder
[05:42:54] <XXCoder> thats boring one nlol
[05:43:14] <XXCoder> geez
[05:43:33] <XXCoder> I just got last month ebook I won for review
[05:43:38] <XXCoder> its 1,600 pages
[05:43:53] <XXCoder> that will take me quite a while, hope its not one of those shitty ones.
[05:48:04] <XXCoder> archivist: you like to read?
[05:49:10] <archivist> I read on the throne, only real book need apply
[05:49:27] <archivist> I do not read fiction
[05:49:34] <XXCoder> kindle paperwhite 2 in solid waterproof case
[05:55:47] <XXCoder> I will need new case evenually
[05:56:09] <XXCoder> been reading a lot at work and metal chips is slowly damaging the clear protection on front
[05:56:40] <archivist> I just grab some old books and flick through
[05:56:58] <XXCoder> current reading though Shannara series
[05:57:05] <XXCoder> one of best fantasy series of all
[05:57:06] <archivist> adding bookmarks if I find anything
[06:10:43] <XXCoder> cool
[07:23:00] <KreAa> archivist: I am the opposite, only fiction allowed, reality I've had enough of
[07:23:23] <KreAa> I don't need to read real plots and shit from a book, I can just use my eyes and ears.
[07:25:28] <archivist> old books have old techniques
[07:26:00] <archivist> some of which are being re invented by noobs too lazy to read :)
[07:59:33] <CaptHindsight> when was the last period in history where the ignorant, lazy and stupid wore it like a badge of honor?
[08:01:22] <archivist> french revolution iirc
[08:01:57] <jdh> imagine if they had facebook
[08:06:34] <Magnifikus> any suggestions on coolant for minimal cooling system for cutting alumium?
[08:06:53] <Magnifikus> where is some stuff not hazardous and not permanently on the part
[08:06:59] <Magnifikus> some strange alcohol derivates
[08:11:25] <archivist> lubrication to stop built up edge
[08:13:19] <archivist> I tend to use a can of "whatever is nearest"
[08:29:26] <Spida> beer?
[08:38:49] <roycroft> wd-40 works very well for aluminium
[08:39:58] <jthornton> dish soap
[08:41:03] <archivist> on a table saw we used to use a wax on the blade before slicing the extrusions
[08:47:23] <Magnifikus> nah something that gets permanently sprayed on :)
[08:47:34] <Magnifikus> http://pasteboard.co/8dzpYXUTW.png also would you use carbide tools in that mill?
[08:48:31] <archivist> depends
[08:49:09] <Magnifikus> the new spindle is much smoother now (had a kress before)
[09:03:39] <KreAa> Magnifikus: I set up one of those misters with water in it, just pure water. I adjusted the jet so the water is added in low enough quantity to not make the part wet, but high enough to give a freezing blast
[09:03:48] <KreAa> Magnifikus: so far it seems to work very well
[09:04:31] <KreAa> I'm trying to buy some of the cool-mist stuff but no suppliers in Norway :(
[09:04:46] <KreAa> The water stuff seems to work well on acrylics too
[09:04:53] <Magnifikus> KreAa, water is no option, cause corosion
[09:05:05] <KreAa> Magnifikus: there is no "wet"
[09:05:17] <KreAa> I add the water to increase the cooling effect of the airjet
[09:05:21] <KreAa> not as a lube
[09:05:40] <Magnifikus> i know we had spiritus and water last time and even the "mist" was causing the kress to corode
[09:05:49] <KreAa> also, oil based stuff is quite dangerous in misters due to inhalation
[09:06:01] <KreAa> I don't do mist
[09:06:08] <Magnifikus> hmm i found some stuff that is fine as the manufacturer states
[09:06:09] <KreAa> I do air that has absorbed some extra moisture
[09:06:32] <KreAa> yes I found some lubes that are fine as well, and they all mix with water for misters
[09:06:45] <KreAa> not found any that is used plain without respiratory protection
[09:07:03] <Magnifikus> http://www.steidle-gmbh.de/english/produkte/kuehlschmierstoffe/aluquick.html
[09:07:20] <Magnifikus> thought about this now, costs 25€/l but they claim 10-20ml/h is fine
[09:07:25] <KreAa> https://www.youtube.com/watch?v=CoHu0reG47A
[09:07:28] <KreAa> cutting alu
[09:07:47] <KreAa> wow nice and low usage atleast
[09:07:49] <KreAa> :)
[09:08:55] <Magnifikus> the thing is if the 1k spindle is ruined i wont get a new one
[09:09:11] <Magnifikus> so i'm cautioness with water
[09:09:15] <KreAa> sure
[09:09:27] <KreAa> btw, that stuff is hazardus and potentially fatal in airways
[09:09:48] <Magnifikus> http://www.steidle-gmbh.de/english/produkte/kuehlschmierstoffe/alufleckfrei.html eh this one
[09:09:49] <Magnifikus> sry
[09:10:10] <KreAa> also the first one you linked to is explosive in air mixes LOL
[09:10:13] <KreAa> nasty shit!
[09:10:17] <Magnifikus> yeah
[09:10:30] <Magnifikus> the 2nd one not
[09:10:41] <KreAa> hmm, sds not found
[09:10:50] <Magnifikus> yeah you speak german?
[09:10:57] <Magnifikus> http://www.steidle-gmbh.de/produkte/kuehlschmierstoffe/TECH020-Rev19.pdf
[09:11:51] <KreAa> I get a file not found
[09:11:59] <KreAa> searched and found it elsewhere
[09:12:18] <KreAa> http://www.steidle-gmbh.de/english/produkte/kuehlschmierstoffe/alufleckfrei.html
[09:13:10] <KreAa> If room is well ventilated you can forgoe the mask yes
[09:13:13] <KreAa> but only then
[09:13:25] <KreAa> also, low flashpoint means it can ignite easily
[09:14:13] <Magnifikus> the 2nd one has 156°
[09:14:18] <KreAa> yes
[09:14:25] <KreAa> not very high when cutting metals
[09:14:39] <KreAa> still, in low quantitys probably ok
[09:14:41] <KreAa> :)
[09:14:48] <KreAa> Infact I'd love to try some
[09:14:56] <Magnifikus> yeah we used like 50ml isopropanol/h before
[09:15:41] <KreAa> no sellers in norway for this either, dang
[09:16:09] <KreAa> Magnifikus: as you can see in the vid though, the airjet seems to help
[09:16:16] <KreAa> light loads though
[09:22:14] <Magnifikus> airjet we got
[09:22:26] <Magnifikus> but stuff like alumium sticking to the mill etc
[09:23:05] <archivist> sticking to the cutter I care about
[09:33:40] <Magnifikus> esp if you got stock you dont even know the material :=)
[11:16:10] <toast-work> if you're cutting aluminum, get some ZiN endmills and run them super fast
[11:16:15] <toast-work> no need for coolant
[11:16:27] <toast-work> part stays pretty dang cool and no built up edge
[11:20:32] <CaptHindsight> http://www.megaprocessor.com/Images/megaprocessor-tour1-2mbps.mp4
[11:21:42] <archivist> there is a youtube version which is faster to load/view
[11:22:46] <CaptHindsight> https://www.youtube.com/watch?v=G6_XlrQOFAs
[11:22:59] <CaptHindsight> skip the original link
[11:24:24] <archivist> when I first went to his site this morning was expecting prototype boards.... sure went the hard way to do it properly
[11:36:12] <FloppyDisk> Wow- that's pretty crazy.
[11:38:39] <archivist> dedicated
[12:16:49] <andypugh> He’s clearly a lot better at building CPUs than playing Tetris.
[12:25:16] <archivist> see how many boards failed during making and failure to fix
[12:33:50] <MattyMatt> I've been thinking how to use EL wire for that kind of demo circuit
[12:34:18] <MattyMatt> the trouble is, it'll glow the same colour whether it's sinking or sourcing
[12:35:24] <MattyMatt> so everything would have to be open collector logic, or I'd have to fake the EL with hidden wires and buffers etc
[12:35:43] <MattyMatt> or make all the kids wear IR goggles
[12:37:03] <andypugh> MattyMatt: EL wire running alongside signal wires with a diode in the end?
[12:37:14] <andypugh> (or wrapped round)
[12:37:16] <MattyMatt> yep that kinda thing
[12:37:41] <MattyMatt> or those yellow filaments of LEDS if I'm gonna do that
[12:38:14] <MattyMatt> 70V logic :)
[12:38:33] <MattyMatt> TTL = triode triode logic
[12:40:12] <MattyMatt> plan B, do it all in CGI. youtube is my classroom
[12:41:06] <MattyMatt> or a game. Kerbal Electronic Warfare
[12:41:59] <MattyMatt> does he say what cpu he's done? I've seen that wire wrapped 6502 already
[12:44:17] <archivist> 6502 would have been sensible being a static design
[12:45:47] <MattyMatt> and easy to find coders
[12:46:11] <MattyMatt> the 8080+ in a gameboy also
[12:46:33] <archivist> and should we cry foul for using the LEDs in the ram as a display!
[12:51:05] <MattyMatt> he coulda used a CRT as ram, but that needs a special tube iirc
[12:51:32] <archivist> mercury delay line
[12:52:35] <MattyMatt> they had to fake that when they rebuilt one of the early ones, because the comp sci students aren't allowed to play with mercury anymore
[12:52:53] <MattyMatt> baby or eniac, one of those
[12:54:35] <MattyMatt> edsac
https://www.youtube.com/watch?v=xGEAPVCuwvY
[13:45:07] <JT-Shop> yea I sold a cable
[13:46:43] <saldot> hi. trying to get the "config" of a machine using the config files. in what file is steps/rev, leadscrew pitch et.c stored and what are they called?
[13:47:06] <JT-Shop> usually in the ini file
[13:47:22] <saldot> ok, thnx
[13:48:14] <saldot> for steps/rev, is it "scale"?
[13:48:28] <JT-Shop> yes usually
[13:48:40] <saldot> right, thnx
[13:48:48] <saldot> and leadscrew pitch?
[13:49:06] <JT-Shop> scale covers it all, how many per user unit
[13:49:22] <JT-Shop> pitch is only used to calculate scale
[13:50:18] <JT-Shop> http://gnipsel.com/images/blast-cabinet/Blast%20Cabinet%2041.JPG
[13:50:28] <saldot> Ok. In the stepconf wizzard i'm asked what my motor step/rev is and also leadscrew pitch.
[13:51:02] <saldot> and i have an old config that i know is correct which i'm trying to get the values from
[13:51:25] <JT-Shop> right, and stepconf uses the entries to calculate scale
[13:51:54] <saldot> ok. so i'll have to reverse engineer it than
[13:54:47] <JT-Shop> you need to know the leadscrew pitch for something?
[13:55:03] <saldot> for the stepconf wizard. =)
[13:55:06] <JT-Shop> do you still have the .stepconf file in configs?
[13:55:17] <saldot> yes
[13:55:32] <JT-Shop> open the old correct file with stepconf
[13:55:32] <saldot> an, no, sorry. only old hal and ini files
[13:55:55] <saldot> the old config was created in a really old version of linuxcnc.. like 3 year old one
[13:56:05] <JT-Shop> just enter any numbers till the scale is correct
[13:56:57] <JT-Shop> or put 1 for everything and the scale for steps per rev and scale will be correct
[13:58:40] <saldot> yeah, i guess that would work
[15:09:18] <tiwake> _methods: poke
[15:13:30] <Sync> JT-Shop: neat
[15:16:31] <saldot> i've tried this guide to connect to a huanyang using a rs485 adapter:
https://www.youtube.com/watch?v=1NF8EFoPlrk
[15:17:00] <saldot> i am able to control the vfd using spindle talker 2 on my windows laptop, but not using LinuxCNC. Any idea how i can troubleshoot that?
[15:17:30] <saldot> i can see the device if i do ls /dev/ttyUSB0
[15:22:51] <saldot> i guess i could do debug logging.. where are the logs stored? /var/log ?
[15:27:06] <saldot> JT-Shop: thnx for the help, setting those to 1 helpded and i got the scale correct
[15:28:14] <MarkusBec> if you start linuxcnc in a shell ther will some deebug infos
[15:31:29] <saldot> thnx.
[15:31:51] <saldot> Lots of errors there. "error reading setup from VFD"
[15:32:44] <XXCoder> CaptHindsight: cool link
[15:33:30] <gregcnc> will it run linuxcnc and control a 3d printer?
[15:59:41] <saldot> anyine running a huanyang VFD with a rs485 adapter and got it working? Not sure what i'm doing wrong but linuxcnc is not able to talk to the VFD.
[16:00:14] <saldot> i've set baudrate et.c. with values matching what's set in the VFD.
[16:00:37] <XXCoder> yet another amazing application of 3d printing, guy gets new jaw
[16:00:49] <XXCoder> I wonder if someday we can print entire human :P
[16:01:10] <SpeedEvil> There are only ~20000 protiens in the human
[16:01:17] <SpeedEvil> So, you'd need a lot of inks
[16:01:40] <XXCoder> or assembler of proteins
[16:02:07] <XXCoder> honestly "whole human" wont happen for long time. theres few organs and bones prints so far though
[16:03:32] <SpeedEvil> You can buy human printers for $1K in africa.
[16:03:42] <XXCoder> :P
[16:11:27] <XXCoder> http://interestingengineering.com/costa-rica-makes-history-announcing-discharge-high-temp-plasma/
[16:14:46] <XXCoder> centaurs. lol
https://i.imgur.com/XekOPPb.jpg
[16:16:15] <andypugh> saldot: Yes, I have usd the hy_vfd component. It worked really easily :-)
[16:17:22] <XXCoder> alternate version
https://i.imgur.com/UfBUIDU.png
[16:17:57] <saldot> andypugh: i must be doing something wrong than. although i can get it working connecting it to my laptop
[16:18:48] <andypugh> saldot: Are you runnign from LiveCD? You will have a permissions problem if not running from an installed system
[16:19:05] <saldot> nope, i have it installed
[16:19:33] <andypugh> There might be something useful in this thread:
https://forum.linuxcnc.org/forum/24-hal-components/31081-rs485-huanyang-speed-control#75830
[16:21:32] <saldot> andypugh: thanks' i'll read through it
[16:26:49] <saldot> andypugh: not the same issue i'm having i'm affraid
[16:27:22] <saldot> i get "error reading setup from VFD, retrying".
[16:28:00] <saldot> so it's able to talk to the vfd i'd say
[16:29:48] <saldot> i'll continue troubleshooting some other day.
[16:36:51] <Deejay> gn8
[16:52:43] <saldot> andypugh: solved it. replaced the USB->rs485 adapter. even thought it was working in windows it must have been faulty. Thought i'd tell you if someone else comes in with similar issues as me. =)
[16:53:09] <andypugh> Great
[16:53:27] <andypugh> I took apart my USBRS485 and put a motherboard header on it.
[16:54:38] <andypugh> saldot: If you need a neater installation:
https://forum.linuxcnc.org/forum/18-computer/30675-on-motherboard-modbus-rs485-connection#72351
[16:57:38] <saldot> looks nice, .Byt the cnc belongs to the local makerspace so i dont dare to modify it to much =)
[17:06:18] <saldot> on a different subject. How much backlash is acceptable on a cnc mill?
[17:06:29] <saldot> I'm measuring about 0.1mm in the X-axis.
[17:09:03] <saldot> no idea if that's good or bad
[17:09:40] <Sync> ideally, 0
[17:10:53] <saldot> yeah, ideally. But is that even possible on a ~20 yr old ~€4000 machine?
[17:22:35] <gregcnc> what you have is what you have. The problem with backlash and old machines is that it will be variable due to wear.
[17:25:02] <Sync> he already left, but well, ballscrews are wear items
[19:06:52] <Duc> what is a the range for RPM speeds for a NEMA 23 or 34 motor
[19:08:07] <malcom2073_> 0-X
[19:08:21] <Duc> whats the maxium
[19:08:26] <malcom2073_> Depends on voltage
[19:08:41] <malcom2073_> I've run my 23's up to 7000rpm with 60volts, but that's past the torque dropoff
[19:08:58] <malcom2073_> And they're really low impedance motors
[19:09:09] <malcom2073_> higher voltage, lower impedance, both allow for faster speeds
[19:09:18] <malcom2073_> Oh wait, motor
[19:09:19] <malcom2073_> nm
[19:09:21] <malcom2073_> I'm thinking stepper
[19:09:21] <malcom2073_> heh
[19:09:35] <Duc> yea stepper motor
[19:09:38] <malcom2073_> Ah then yeah
[19:09:56] <malcom2073_> Depends on a *lot*, but usually the datasheet will give you a torque dropoff vs rpm at a particular voltage
[19:12:29] <enleth> Duc: "typical" stepper motors will usually start dropping off significantly at 3000rpm or so
[19:13:41] <enleth> many shittier ones will start dropping off at 1500rpm, staying reasonable but increasingly weaker up to that 3000 or so
[19:13:44] <malcom2073_> typical motors are a typical voltage and typical current in ideal settings :P
[19:14:03] <Duc> I dont need above 500 RPM with the machine Im designing
[19:17:04] <enleth> malcom2073_: pretty much yeah, but what I said will usually hold for what you can get second-hand on ebay from dismantled plotters and such
[19:17:05] <TurBoss> anyone here using blendercam?
[19:17:44] <enleth> Duc: then you're probably going to be fine with whatever hybrid stepper of this size you find
[19:18:15] <Duc> enleth: based on my math I need about 500RPM
[19:18:20] <Duc> or less
[19:21:19] <enleth> Duc: and that's below the 1000rpm "typical" motors reach within the flat part of the speed/torque curve
[19:21:49] <Duc> Enleth: where can I find a good chart about info
[19:22:03] <enleth> motor datasheets
[19:23:17] <enleth> take the first motor you can google within your price range, download the datasheet, find the torque over speed chart
[19:23:40] <Duc> ok
[19:23:51] <Duc> lucky I dont need a bunch of torque
[19:24:07] <enleth> if you don't happen to find some special-purpose motor by accident, most other motors of similar size and price will give you similar performance
[19:24:08] <Duc> sending it all thru a 81:1 reducer
[19:24:46] <enleth> but 500rpm sounds like something you should be able to comfortably maintain even with a random second-hand stepper from ebay
[19:27:33] <enleth> https://www.youtube.com/watch?v=MzbXJN2_LVo well it works, I just had to bump the brightness and contrast up quite a bit
[19:27:55] <enleth> somehow the built-in screen driver gets a much brighter text
[19:28:51] <enleth> also, I knew the built-in TV was made in the UK even before I found the "made in UK" label on the PCB - because the brightness potentiometer was labeled "BRILLIANCE"
[19:29:23] <Duc> enleth: Slowly building a gun mount like on "The Jackal"
[19:32:41] <enleth> Duc: tired of mowing grass the easy way?
[19:33:19] <Duc> enleth: Yes sir. figure a AR15 loaded with a beta mag will start the grass on fire
[19:44:04] <Frank_12> I finnaly have in my hands the 5i25 7i76 combo!!
[19:44:47] <Frank_12> is there an easy way if everything works good? it has been through a lot of travels, and i dont want to be dissapointed when i finaly plug it in..
[19:45:37] <Duc> Plugging it in will need to happen
[19:48:13] <andypugh> Frank_12: Plug in the 5i25 by itself first. The halrun / loadrt hostmot2 / loadrt hm2_pci / show pin
[19:48:31] <TurBoss> Frank_12 congrats i'm dreaming in having that combo
[19:48:36] <Frank_12> okey, dont be mad but i dont even have linuxcnc installed :D
[19:48:39] <TurBoss> and 7i77
[19:48:59] <Frank_12> or linux for that matter... hehe
[19:49:12] <andypugh> TurBoss: Don’t ignore the 50-pin cards. They have a wider range of daughter cards, and more total IO
[19:49:27] <TurBoss> ohhhh
[19:49:40] <TurBoss> any model in concrete¿?
[19:50:36] <andypugh> I am using a 5i24 at the moment. (Actually, a 6i24, which is the PCIe version of the same card)
http://store.mesanet.com/index.php?route=product/product&path=83_85&product_id=298
[19:51:10] <TurBoss> thx!
[19:51:40] <andypugh> But I mainly used that because I wanted to use the resolver card, 7i49:
http://store.mesanet.com/index.php?route=product/product&product_id=101&search=7i49
[19:52:35] <TurBoss> I have 2 projects in mind
[19:52:55] <TurBoss> build a gantry with servos and retrofit a kuka arm
[19:52:58] <andypugh> It depends on what you need. The 7i76 is enough for most purposes, and the on-board expansion port for a smart-serial card can run any one of these:
http://store.mesanet.com/index.php?route=product/category&path=83_88
[19:53:32] <Frank_12> whats the card for andy, im still an ingnorant on that topic
[19:54:01] <andypugh> Frank_12: The 7i49? It’s for servo motors that have resolver feedback rather than encoders.
[19:54:31] <Frank_12> i see
[19:54:36] <andypugh> Resolvers give an absolute analogue feedback, the 7i49 converts that into a 16-bit angle.
[19:54:52] <Frank_12> up to 6 axes?
[19:55:04] <TurBoss> the kuka arm has resolver IIRC
[19:55:12] <andypugh> Yes. But you could connect 3 of them to one 6i24
[19:56:01] <Frank_12> so 9.. lol what would be the application? automation line?
[19:56:18] <TurBoss> joints no?
[19:56:28] <enleth> andypugh: good to know, a friend has a pair of Siemens 1FK7022-5AK71 servos with resolvers, I'll let him know there's a mesa card for this
[19:56:29] <TurBoss> multiple motors per axis
[19:57:18] * TurBoss is taking notes
[19:57:51] <enleth> TBH I think what would be super useful for newcomers to Mesa products is a feature matrix/filter page
[19:58:16] <Frank_12> guys when handling the cards for inspection, is it just fine by having no contact to ground? i mean feet up plus a wooden chair?
[19:59:28] <andypugh> To give an idea of the setup I am using, it is a 6i24 FPGA card with 3 x 50-pin connectors (72 IO pins and lots of GND pins). One header carries a 7i49 for the resolvers. Another header has a 7i44 smart-serial adaptor into which I could connect 8 smart-serial devices, but I only actualy have a 7i84 IO card, 2 x 8i20 2.2kW servo amps and a 7i73 panel-interface that looks after the IO in the lathe apron.
[19:59:46] <enleth> with so many products, most of which overlap with some of their features and differ in others, it's quite difficult to navigate their product list
[20:00:02] <andypugh> Frank_12: I have never taken any precautions and haven’t killed a Mesa card yet.
[20:00:09] <skunkworks> Just figure out what you need - and buy the right cards.. ;)
[20:00:37] <enleth> skunkworks: yeah, but I don't know what the right cards are because I may not know a card even exists for what I need
[20:00:46] <andypugh> You need to start at what you need by way of IO and interfaces then work back towards the PC really.
[20:00:47] <skunkworks> error on the 'more i/o' side.
[20:00:49] <enleth> how do I know I do not know something?
[20:00:51] <TurBoss> at simple view is posibe to diference encoders or resolvers?
[20:01:38] <cradek> are you asking how to identify which kind you have?
[20:01:55] <andypugh> You might have noticed that my 6i24 stlll has one complete IO port free, so might have been overkill.
[20:02:06] <TurBoss> yes
[20:02:12] <enleth> TurBoss: resolvers are magnetic, usually heavier, bigger and sturdier, but an encoder in a beefy enclosure will not be externally distinguishable
[20:02:25] <cradek> the wires are different too
[20:02:26] <TurBoss> :\
[20:02:29] <TurBoss> oh
[20:02:30] <cradek> do you have a photo?
[20:02:35] <TurBoss> not yet
[20:02:46] <enleth> well yes, if you have the pinout, you can tell
[20:02:48] <andypugh> You can tell a resolver because it will have 6 wires. An encoder will typically need 5 or 8.
[20:02:51] <Frank_12> well andy you certainly are a linuxcnc/mesa entusiasth as you mentioned a few days ago hehe,
[20:03:15] <TurBoss> ok 6 vs 5/8 wires
[20:03:45] <cradek> resolver has three carefully twisted pairs
[20:04:01] <TurBoss> ok thx!
[20:04:28] <andypugh> This is my installation, you can even see that the 7i84 IO card isn’t fully populated yet.
https://picasaweb.google.com/108164504656404380542/Holbrook#6289136202698936898
[20:05:20] <andypugh> The PC and Mesa cards fit in the space vacated by the Selenium rectifiers of the original electronics. I think that’s amazing.
[20:05:36] <skunkworks> my setup for one of my machines
[20:05:38] <skunkworks> http://forum.linuxcnc.org/forum/38-general-linuxcnc-questions/30441-matsuura-mc-500v2-retrofit
[20:05:48] <skunkworks> I did error on the more i/o...
[20:06:42] <andypugh> skunkworks: Not enough pink wire for my tastes :-)
[20:09:23] <skunkworks> heh
[20:09:39] <TurBoss> what about this guy
https://granitedevices.com/store/motioncontrol/argon/argondrives.html ??
[20:09:57] <TurBoss> the argon
[20:10:35] <TurBoss> ++
https://granitedevices.com/store/motioncontrol/argon/argonmotors/950w-low-inertia-ac-servo-motor.html
[20:10:45] <TurBoss> for a gantry
[20:10:50] <TurBoss> 4 motors in total
[20:10:57] <TurBoss> X YY Z
[20:11:23] <TurBoss> i need it for aluminium
[20:11:34] <TurBoss> precission + speed
[20:11:54] <andypugh> granite drives have a good reputation
[20:12:09] <TurBoss> are they expensive ?
[20:12:13] <andypugh> Yes
[20:12:17] <TurBoss> phew
[20:12:21] <andypugh> The price is in the link
[20:12:27] <TurBoss> i know
[20:12:36] <TurBoss> but don't know others
[20:12:37] <TurBoss> :D
[20:12:52] <TurBoss> I mean compared to similar drivers
[20:12:52] <andypugh> Actually, in the context of the market Granite are rather reasonably priced.
[20:13:17] <TurBoss> ++ they have opensource firmware
[20:14:11] <Frank_12> is it normal that the spartan procesor has a drop of paint right in the middle? :s lol
[20:14:34] <andypugh> Yes.
[20:14:35] <TurBoss> fpga?
[20:14:47] <Frank_12> yea
[20:14:48] <Frank_12> blue
[20:15:00] <andypugh> Mine is pink
[20:15:07] <andypugh> Inspection mark, I think
[20:15:56] <TurBoss> so I can go with 6I25 -> 7I49 -> Argon -> 950W servo ?
[20:16:03] <Frank_12> thought so, heh
[20:16:11] <TurBoss> npo
[20:16:14] <TurBoss> no*
[20:16:57] <TurBoss> *6I24-25
[20:16:58] <andypugh> No, 6i25 and 7i49 are not a match. You need 6i24 for 7i49
[20:17:12] <andypugh> Yes, that :-)
[20:17:16] <TurBoss> :)
[20:17:30] <Frank_12> i cant belive they are in my hands, and i cant belive either that i get excited about it, haha, the dark force is calling me
[20:17:31] <TurBoss> and for the dual motor per axis i need Joints
[20:17:33] <andypugh> AC or DC servo motor?
[20:17:47] <TurBoss> AC
[20:18:14] <andypugh> Mesa have this: linuxcnc-uspace-dev:
[20:18:23] <andypugh> No they don't
[20:18:29] <TurBoss> :O
[20:18:32] <andypugh> This:
http://store.mesanet.com/index.php?route=product/product&path=65&product_id=149
[20:19:20] <andypugh> (note that you get 5 for the price of 4 under bulk-pricing)
[20:19:34] <TurBoss> wow
[20:19:43] <TurBoss> but they are for AC?
[20:20:02] <TurBoss> I need it for 50Hz
[20:20:03] <andypugh> It’s a _much_ dumber drive than the Granites. You need to put all the cleverness in the PC.
[20:20:18] <andypugh> But then, with LinuxCNC, you can ;-)
[20:20:47] <andypugh> I think there is here a confusion about what an AC servo is
[20:20:54] <TurBoss> yes maybe
[20:21:13] <Frank_12> just noticed that the cable didnt ship with the combo.. (between 5i25 and 7i76
[20:21:20] <andypugh> AC servos and brushless DC motors are exactly the same thing.
[20:21:36] <TurBoss> ahhhmmm
[20:21:40] <andypugh> Frank_12: It should have done with a plug-n-go kit
[20:21:48] <TurBoss> 3 phase
[20:21:55] <andypugh> TurBoss: I know that that doesn’t make any sense.
[20:22:53] <PCW_> The 8I20 is not for induction motors but rather permanent magnet rotor 3 phase motors
[20:23:18] <andypugh> The 8i20 (and most other AC Servo / Brushless DC) drives take a DC supply and convert it to a commutated AC. But the frquency of the AC is governed by the motor rotation.
[20:23:56] <Frank_12> wired, i will check again but i havent seen any cable so far, and i ordered the plug and go kit
[20:24:16] <PCW_> (well I guess it would work with a wound rotor synchronous motor like an alternator run as a motor)
[20:26:20] <andypugh> TurBoss: The Granite drives do have the advantage of taking mains AC current input. (Or DC at up to 380V)
[20:26:53] <TurBoss> the i dont need a power supply for the motor no?
[20:26:58] <andypugh> With the 8i20 you need to create or buy a power supply. But that can be as simple as a big capacitor and a rectifier.
[20:26:58] <TurBoss> then*
[20:27:18] <TurBoss> so unregulater supply will do the job?
[20:27:48] <TurBoss> I have bunch of 380 diode bridges and big cops
[20:27:53] <TurBoss> caps*
[20:28:03] <andypugh> Yes, unregulated is better, really
[20:28:57] <andypugh> You will probably need to actually have a soft-start setup of some sort, and a discharge resistor for safety.
[20:29:43] <andypugh> ie, a rectifier and cap will work, but you will tend to trip the breakers with inrush current, and 380V DC hanging about on caps is likely to bite.
[20:29:44] <TurBoss> what do you mean?
[20:29:58] <TurBoss> ahhh
[20:30:12] <TurBoss> well i think that can be dangerous
[20:30:14] <TurBoss> :P
[20:30:39] <andypugh> Granite have (presumably) already done all that.
[20:30:49] <TurBoss> yes the share a resistor
[20:30:55] <TurBoss> ah big one
[20:32:17] <TurBoss> I will go with 6i24-25 + 7i49 for the Kuka retrofit
[20:32:26] <TurBoss> and argon for the gantry
[20:32:40] <TurBoss> thx!
[20:33:06] <TurBoss> i was having doubts
[20:34:34] <andypugh> What I have done (and this is just an example of what can be done, nothing more) is to use the fact that the 8i20 drives report bus voltage back on the serial comms to know when I can safely take the soft-start resistor out of circuit and enable the drives. I also use this to know when the discharge resitor relay (A NC relay that discharges the caps when the power goes off) can safely be energised to stop disharging.
[20:34:34] <andypugh> It’s quite a small relay, absolutely not cabably of breaking 300V DC So I make sure it is never asked to through HAL interlock.
[20:35:17] <andypugh> Yeah, I think that Argons are probably an easier solution than the 8i20 for a first retrofit.
[20:35:40] <andypugh> You could look here too:
http://www.dmm-tech.com
[20:35:42] <TurBoss> the gantry is not a retrofit but yes i prefer easy `safety
[20:37:05] <andypugh> When looking at DMM don’t even consider the Dyn2 drive, it has a bad behaviour in analogue control mode (full speed if the analogue control is disconnected). The Dyn4 does not have this problem.
[20:37:45] <TurBoss> wow 1,8kw
[20:38:27] <andypugh> They only wirk with the matching motors, I think
[20:40:19] <TurBoss> i see
[20:40:33] <TurBoss> they are cheap
[20:40:52] <TurBoss> compare with argon + gmotor
[20:40:54] <TurBoss> *motor
[20:41:24] <andypugh> Look at Teco too.
http://machmotion.com/cnc-products/drives-motors/teco-servo-drives-and-motors.html
[20:41:42] <andypugh> And, possibly Alibaba and direct purchse from China.
[20:42:04] <andypugh> Granite do have a very good reputation for quality.
[20:42:27] <andypugh> But then they don’t actually seem to have any Argon drives or motors in stock…
[20:42:32] <TurBoss> I don't trust chiness
[20:42:35] <TurBoss> e
[20:43:09] <TurBoss> this are more expensive lol
[20:43:19] <TurBoss> Dmm is fine
[20:43:37] <andypugh> DMM is fine if you trust Canadians :-)
[20:43:47] <tiwake> canada makes good stuff
[20:44:07] <TurBoss> :D
[20:44:07] <gregcnc> what's Zeeshan up to?
[20:44:58] <TurBoss> and what is the direrence betwen 6i24-16 / 25???
[20:45:20] <andypugh> TurBoss:
http://www.dmm-tech.com/multiaxissolutions2.html states “Plug and play compatibilty with Mach3 / LinuxCNC” for a mutli-axis set
[20:45:31] <andypugh> TurBoss: FPGA size
[20:45:37] <TurBoss> ahhhh thx
[20:46:12] <andypugh> I _think_ that all current LinuxCNC firmwares fit on the small one. PCW can probably confirm.
[20:46:17] <TurBoss> but this uses parport?
[20:46:38] <andypugh> Yes, parport and step/dir mode.
[20:47:06] <andypugh> But could be 5i25 and step-dir mode, if you wanted.
[20:47:20] <TurBoss> ohhhh
[20:48:07] <andypugh> Personally I don’t like step-dir control of servos. But I am also aware that that is probably irrational.
[20:48:40] <TurBoss> :D
[20:48:49] <andypugh> (And my objection to the Dyn2 drive does not apply in step-dir mode)
[20:49:58] <TurBoss> look good
[20:50:03] <TurBoss> looks *
[20:50:33] <TurBoss> i need big and cheap
[20:50:39] <TurBoss> 1000W >
[20:50:51] <TurBoss> :D
[20:51:13] <TurBoss> 120-DST-A6h51 is 1.8kW
[20:51:46] <gregcnc> what machine is that going on?
[20:52:05] <TurBoss> a gantry for alu
[20:52:29] <gregcnc> how big is the spindle?
[20:52:47] <TurBoss> 30000RMP
[20:52:48] <andypugh> I find 750W more than enough on my conventional knee-mill.
[20:53:20] <andypugh> Lots of people run gantry routers on NEMA 34 steppers.
[20:53:31] <gregcnc> how big a tool can you run at 30k?
[20:53:55] <TurBoss> i don't have it yet
[20:54:05] <gregcnc> kw?
[20:54:08] <TurBoss> fast?
[20:54:10] <andypugh> gregcnc: Any tool bigger than about 12” diameter will tend to explode. I guess that’s the limit :-)
[20:54:44] <TurBoss> i don't know
[20:56:10] <andypugh> TurBoss: Might be fun to put numbers in here
http://zero-divide.net/?page=fswizard
[20:56:35] <andypugh> Note especially the tool forces, that might play a part in your calcs.
[20:57:24] <gregcnc> I'm just wondering if you need(can use) 1.8kW drives 3.6kW to move a gantry is quite a bit of power.
[20:57:25] <TurBoss> I don't know that values
[20:57:30] <TurBoss> the machinist does
[20:58:40] <gregcnc> Brother mills with 2000m/min rapids only have 1kW servos.
[20:59:00] <andypugh> TurBoss: fswizard does the calcs, you just really choose material and cutter
[20:59:00] <gregcnc> 2kW on Z
[21:00:36] <andypugh> I just tried a 12mm carbide cutter in 7075 with 12mm DOC. That suggested 10000 rpm and 100 in/min for 3,3HP of spindle power. Cutting force is 100lb in that example.
[21:01:16] <andypugh> (Or, 400N in sensible units)
[21:02:19] <andypugh> At 3mm dia it starts to suggest 30000 rpm. You don’t need a 1kW servo to break a 3mm tool.
[21:03:09] <andypugh> In fact at that point it is 75W spindle power and 26N cutting force
[21:03:24] <TurBoss> 0.39N ?
[21:03:32] <TurBoss> i'm doing wrong :D
[21:03:59] <andypugh> You probbaly chose different flute counts or depth of cut
[21:04:14] <andypugh> But it gives you order of magnitude numbers.
[21:05:08] <TurBoss> I'm goin to use aluminium with a 2mm ball and max 1 - 2 mm prof
[21:05:31] <andypugh> Just make sure you don’t end up with a gantry that has the power to push a very big cutter and a spindle that can’t spin slow enough to use it. What’s the rated power of the 30,000 rpm spindle at 10,000 rpm?
[21:06:16] <TurBoss> don0't know
[21:06:22] <gregcnc> I found the same when I built my spindle. Motor is capable of 30k, but I have yet to run it over 15k and can't run it slow enough in steel with 3/8" cutters.
[21:06:25] <TurBoss> i havent looked at the spindle
[21:07:00] <andypugh> Cutting force is not everything, of course. You need motor power to get acceleration and speed. But very big motors tend to get a bit ponderous themselves. You actually need to “impedance match” the mehanics and the motors.
[21:08:05] <TurBoss> so 750 W are fine?
[21:08:07] <TurBoss> xD
[21:08:13] <TurBoss> even cheaper
[21:08:38] <andypugh> No idea :-) You need to calculate masses and accels. And don’t forget the motor inertias.
[21:08:52] <TurBoss> thx!
[21:09:20] <andypugh> This is design engineering. Clever folk get paid for it. (but not paid a lot, few are clever enough to notice this)
[21:09:46] <andypugh> (I used to be a design engineer :-)
[21:10:02] <andypugh> And if I want to keep the current job, I should go to sleep.
[21:10:04] <Frank_12> yep. checked again, no cable, they forgot to send the 5i25-7i76 cable..
[21:10:05] <TurBoss> so low inertia motors are better
[21:10:21] <TurBoss> gn8 Andy
[21:10:34] <TurBoss> and thank you
[21:11:00] <andypugh> They sell motors in a range of inertias. This is for a reason. Sometims high-inertia motors are what you need
[21:11:43] <andypugh> Frank_12: PCW can probably tell you the type of cable you need, though Mesa will probably send you the right cable if you complain.
[21:14:06] <gregcnc> IEEE-1284 parallel cable, I got one on ebay
[21:14:46] <Frank_12> yeah, but im not in us nor my family member anymore,,
[21:14:53] <Frank_12> most probably i will have to buy one
[21:15:29] <Frank_12> 20usd to get buy it here thou..
[21:51:05] <tiwake> XXCoder: poke?
[22:36:03] <zeeshan> hi greg
[22:36:05] <zeeshan> im alive :P
[22:36:08] <zeeshan> just busy as hell
[23:14:02] <enleth> got some very nice surface finish on the ORAC after fitting the new VFD and adjusting gibs
[23:14:41] <enleth> it might be a small lathe but it seems quite capable
[23:15:02] <enleth> now off to play with it some more, then gut the old control
[23:16:16] <enleth> I have to use reduced speed jogging to control it manually, which obviously won't do for any actual machining
