Back
[00:05:09] <djdelorie> question: what's the upper limit on step rate that LinuxCNC can output via the parallel port?
[00:36:52] <mrsun> djdelorie, my guess is that its quite high, but you got latency in the computer so you have to keep it quite low :P
[00:36:58] <mrsun> for accurate stepping
[00:37:24] * djdelorie wonders if latency is an issue because of the cut, or because of how badly stepper behave...
[00:37:53] <djdelorie> my test rig only goes up to 10 uS per step
[00:38:06] <djdelorie> I *think* the hardware will accept far faster than that though
[00:39:06] <djdelorie> 1200 RPM with microstepping is only 80,000 steps/sec, but it could go as high as 4000 RPM with the real power supply...
[00:39:46] <djdelorie> I seem to recall the parallel port itself places a limit on changes/sec too though
[00:47:31] <srsdriver> Anyone used a 7i73 for input? I can see the pins in my hal watch, but referencing them in nets causes errors processing the hal file.
[00:49:01] <srsdriver> net scale1 mux4.0.sel1 <= hm2_7i43.0.7i73.0.1.input-22 causes an error, but I can view this in the watch, and references to the encoder work fine
[01:03:21] <djdelorie> ah, better tester, now tracking at 1,000,000 steps per second!
[01:06:34] <djdelorie> that's 15,000 RPM at the most microsteps, far faster than the motor can go...
[02:59:00] <Loetmichel> mornin'
[03:15:03] <Valen> evenin
[04:16:01] <DJ9DJ> moin
[05:43:49] <Mjolinor> Abstract thought for the day: The HAL layer of LinuxCNC seems to provide most of what you need for an electric vehicle control system.
[05:53:18] <pingufan> Does somebody here know how the shaft of a KRESS1050 FME is fixated on the inner side of the machine? I cannot see a thread on the schematics drawing of this engine on the rear side.
[05:54:05] <pingufan> The shaft is pushed through two ball bearings, but it must be fixed somehow against mowing out again?
[05:56:05] <pingufan> I ask because I think about making a reciption for the bearings on a lathe and to buy the original shaft and bearings from Kress.
[05:56:44] <pingufan> Should result in a very high precision mill.
[06:40:22] <maximilian_h> Hello, are there packages from the buildbot for 64bit lucid or precise beta ?
[06:50:32] <jthornton> the Direct Pitch Activator has been activated!
[06:50:51] <Jymmm> FIRE!
[06:51:29] <jthornton> 100 billion critters are just waking up
[07:00:28] <Jymmm> ☃
[07:01:23] <Jymmm> ☣
[07:03:55] <Jymmm> t⢙⠏⠝
[07:40:07] <pingufan> can somebody from here make me a "turned part" on a lathe for low money? A relatively simple part, a tube, outside diameter 40mm and on the inner side some different diameters for bearings?
[07:40:53] <pingufan> I am from Austria, so somebody from Austria or close surrounding would be best.
[07:41:43] <pingufan> This is a general question, only to know if I can calculate with that.
[07:47:51] <jdhnc> I will do it for e1234 with free shipping.
[08:14:30] <pingufan> I order one dozen of it. :)
[08:14:30] <elmo40> Mjolinor: electric vehicle, eh? sounds like a good project ;) Would you go so far as to add electric steering and brakes?
[08:14:30] <jthornton> I'll do it but you have to come pick it up
[08:14:42] <elmo40> lol
[08:14:57] <elmo40> im in Canada... shipping would be prohibitive ;)
[08:15:08] <elmo40> and what material, steel?
[08:15:18] <pingufan> Yes.
[08:15:35] <archivist> we need a web site like code monkeys spose call it metal monkeys
[08:16:05] <jthornton> what does code monkeys do?
[08:16:39] <archivist> there are a few sites where you bid for work
[08:16:41] <elmo40> what do you think code monkeys do :-P
[08:16:47] <pingufan> Actually I buy a cheap mill for woodworks, disassemble it, and put the shaft into a much slimmer tube.
[08:17:29] <Loetmichel> cl: i think ist alsmost impossibel to meet the law requirements for drive by wire (steering/braking)
[08:17:36] <jthornton> I didn't have a clue as to what code monkeys do :P
[08:17:37] <Loetmichel> for a private man
[08:17:42] <Loetmichel> not a big company
[08:18:11] <Loetmichel> at least in germany
[08:20:47] <elmo40> Loetmichel: what law requirements. if it works you can use it :-P
[08:20:52] <elmo40> just at your own risk...
[08:21:16] <elmo40> you should see the crap they allow on the road in N. America! Just insane. Yet they give them permits
[08:21:38] <Loetmichel> elmo40: obviously you are not familiar with the german "TUEV"
[08:21:40] <Loetmichel> ;-)
[08:21:41] <elmo40> and drive by wire has been around for years. Corvette had it long ago. Mercedes too
[08:22:39] <Loetmichel> elmo40: IIRC the Breaking and steering systems have been carefully desinged to fall back on mechanical systems if damaged
[08:22:55] <Loetmichel> so no, taht qualifies not as "drive by wire"
[08:23:46] <Loetmichel> the acceleratopr pedal electrically isnt a problmen regulation wise 'cause you can always turn the igniotion off if something gooes awry...
[08:23:55] <Loetmichel> (if you are not driving a tojota ;-)
[08:24:29] <elmo40> americans crashed their toyotas... they don't know how to turn the key off to prevent acceleration
[08:25:08] <Loetmichel> tahts what oi meant ;-)
[08:26:37] <Loetmichel> abnd the reason for the mechanical fallbach is the impossibility to prove the failsafe of the nbraking/steering systems if "by wire"
[08:26:56] <Loetmichel> i.E. the "costiness" of the proof
[08:27:25] <Loetmichel> as i said earlier: at least in germany
[08:27:43] <elmo40> then just make a go-cart :-P
[08:34:37] <Loetmichel> elmo40: already on the way: have just ordered two 2kW brushless to motorize the rolling chair of my wife ;-)
[08:34:56] <Loetmichel> (disabled chair?)
[08:35:05] <TekniQue> wheelchaqir
[08:35:09] <TekniQue> wheelchair
[08:35:39] <Loetmichel> thx
[08:35:58] <jdhnc> 5HP wheelchair
[08:36:23] <Loetmichel> ' cause i am tired of shoving her throu malls 4hrs at a time ;-)
[08:37:13] <Loetmichel> jdhnc: model Bruishless motors. the 2kw is peak, if you dont wamt them to overheat one should limit the cont. wattage to about 500W
[08:38:04] <Loetmichel> but yes, IF i deactivate the limiters the chair would accelerate REALLY fast to about 60 MPH ;-)
[08:38:36] <Loetmichel> (If my calculations are correct)
[08:49:04] <DJ9DJ> re
[08:49:47] <frysteev_> mooo
[08:49:56] <skunkworks> http://www.youtube.com/watch?v=QILxRo_E2dQ
[08:49:58] <frysteev_> got the old cnc cpu on the laser booting last night
[08:50:47] <skunkworks> related to one earlier..
http://www.youtube.com/watch?v=b7cjo4kTYMw&feature=channel
[09:02:11] <cradek> skunkworks: that sure looks like a slow way to cut that hole
[09:02:47] <cradek> I'd use a 3/4 2 flute and do it in one helix and then one full circle
[09:03:22] <cradek> at about 1/10th the spindle speed
[09:03:35] <JustinXJS> skunkworks, did you build the spindle?
[09:12:16] <JT-Shop> wow a tool changer slower than my BP 308!
[09:13:08] <JT-Shop> pretty dam cool machine for a homebuilt
[09:13:48] <cradek> oh are those two videos the same machine? I take back what I said then - I didn't realize it was such a light machine.
[09:13:53] <frysteev_> pics?
[09:21:58] <pingufan> skunkworks: Nice mill! Do you have such a sensitive microphone, or is this mill so loud?
[09:28:19] <skunkworks> Not mine.. Just a random video I found
[09:33:19] <pingufan> Oh.
[09:36:54] <FinboySlick> I was about to ask what sort of device was used for the mist coolant.
[09:42:48] <A2Sheds> Loetmichel:
http://www.youtube.com/watch?v=EI8bWlJGyc8&feature=related
[11:10:05] <ssi> I wonder if I can pull the supplies out of this control and make them into bench supplies
[11:10:13] <ssi> one of them is a 50A 5V supply
[11:10:14] <ssi> hahaha
[11:28:43] <archivist> a little one
[11:30:27] <archivist> can I interest you in an Astec 180A 5v 50A 5v 10A 12v 10A5v
[11:31:52] <archivist> found a pair of these in a gigarouter by netstar inc
[11:35:35] <ssi> lul
[11:36:42] <archivist> dunno if I should ebay them or mine the copper out of them, only switchers though
[11:37:12] <mazafaka> what dricer parameters would be in a huge machine? Something like 42 V 100 A ?
[11:37:26] <mazafaka> *driver parameters
[11:57:59] <pcw> big machines use line operated servos (~360VDC for 220)
[11:58:53] <jdhnc> do servo drives need regulated power supplies?
[11:59:20] <mazafaka> probably not...
[12:00:03] <mazafaka> so it's cheaper to add a computer and CNC to servo motors already having ones
[12:00:14] <pcw> no not generally (and fancier drives have a current control loop which makes torque independent of PS voltage))
[12:00:50] <mazafaka> we have electric boards on ISA or PCI for each axis, this is not reliable
[12:01:50] <mazafaka> I think what if our plant convert this to modern-type CNC, not Pentium 166 MHz with Win 98
[12:02:15] <mazafaka> the machine turns off with vibration in the shop.
[12:02:49] <mazafaka> there's an electric closet ...with electrics, and schemas in it are probably 20 years old
[12:09:43] <mazafaka> At our plant, we're dreaming of better times... :)
[12:14:26] <pcw> its not that bad Pentium 166 is only 18 years old :-)
[12:14:52] <ssi> yeah seriously... the machine I'm ripping up doesn't have a CPU... just a couple hundred pounds of TTL logic :P
[12:16:27] <jdhnc> real machines have tubes, relays, maybe a drum programmer.
[12:17:19] <pcw> a couple hundred lbs of TTL might be a CPU...
[12:17:36] <ssi> pcw: on the 7i49, if you're not using all six resolver/pwm ports, is it possible to use them for anything else?m
[12:17:49] <ssi> pcw: haha well it's not exactly "central"... more of a "DPU"
[12:18:15] <pcw> Heck I had a PDP8S that had only transistors and fit on a table top
[12:19:52] <pcw> ssi: I suppose the some PWM pins could have alternate uses but I doubt its worth messing with the cabling to that extent
[12:20:02] <pcw> that some
[12:20:11] <ssi> fair enough
[12:21:14] <pcw> And I guess you could use unused resolver inputs as Analog inputs
[12:21:34] <ssi> sounds like it might take some firmware hacking to make that work
[12:21:43] <pcw> (with some special firmware)
[12:22:25] <flyback> http://1saleaday.com/
[12:22:25] <flyback> <flyback> <flyback>
https://github.com/wikireader/wikireader/wiki/Hardware-Specifications
[12:22:44] <flyback> mabye you can use the sdk and gpio on it ti make a touoch interface for a cnc machine
[12:23:03] <pcw> yeah it would just be a custom firmware with say 3 resolver inputs and 6 analog inputs
[12:23:45] <ssi> who's hosting linuxcnc.org?
[12:23:50] <ssi> it seems like it's balls slow all the time
[12:24:06] <pcw> Yes it got really slow a day or two ago
[12:24:33] <ssi> should get my 7i42s in the mail today
[12:24:35] <awallin> dreamhost ?
[12:24:36] <ssi> and my 10gal oil
[12:24:47] <ssi> I should be making chips in no time! :D
[12:25:10] <jdhnc> anyone converted an X2?
[12:25:40] <pcw> note that my desire to hack the resolver code is not terribly high
[12:26:34] <ssi> understandable
[12:26:37] <ssi> does it work as is?
[12:26:42] <ssi> that's all I likely care about :)
[12:26:44] <pcw> yes
[12:26:54] <pcw> has for a long time
[12:26:57] <ssi> ok good
[12:27:10] <ssi> someone in here was expressing concern about its finishedness
[12:27:14] <ssi> (was it you flyback?)
[12:27:33] <flyback> no
[12:27:39] <ssi> was someone effy
[12:27:42] <flyback> I just tinker with hw
[12:27:43] <ssi> might have been frysteev
[12:27:56] <ssi> yea it was frysteev_
[12:27:57] <flyback> but I got a canuckload of steppers, servo's etc I saved from printers and copy machines over the yrs :)
[12:28:07] <flyback> I might eventually build some kind of cnc
[12:28:10] <ssi> get on it!
[12:28:15] <pcw> There were some bad bitfiles around (turns out Webpack 9.2 which I have used for a long time mucks up small counters in spartan3)
[12:28:18] <ssi> I scratch built my first two
[12:28:29] <flyback> heh I got this brushed dc motor servo drom IBM mainframe hivh speed dot matrix
[12:28:33] <flyback> it looks like a car starter
[12:28:44] <flyback> it's like $120
[12:30:01] <ssi> good lawd this site is slow
[12:35:54] <frysteev_> what do you guys recomend for cam software for use with emc?
[12:36:08] <Tom_itx> mastercam
[12:36:14] <Tom_itx> i use smartcam
[12:36:44] <Tom_itx> surfcam is another good one
[12:36:47] <Tom_itx> none are free
[12:36:55] <Tom_itx> none of those anyway
[12:37:00] <Tom_itx> but they are good
[12:37:41] <syyl> proE/NC
[12:37:42] <syyl> ;)
[12:37:50] <syyl> but at home its sheetcam
[12:38:30] <Tom_itx> there's addons for solidworks and acad too
[12:38:37] <Tom_itx> the acad addon is ncplolaris
[12:39:25] <syyl> there is also cambam
[12:39:25] <Tom_itx> i've seen it used but it's kindof a kludge
[12:39:39] <syyl> its pretty interesting for its price..
[12:41:53] <frysteev_> im looking for mainlly 2d stuff outputted form autocad
[12:42:04] <frysteev_> mastercam kind of overwhelmed me last time i looked at it
[12:42:28] <frysteev_> i also imagine you have to take some time and setup a profile for the machine you want to output gcode for..
[12:42:33] <pcw> someone on the forum was playing with gsimple (which is free)
[12:43:11] <frysteev_> ive basically used gcode converters in the past, for laser cutting where we didnt need acuracy
[12:46:46] <frysteev_> do all the good ones have about the same learning curve?
[12:46:46] <Tom_itx> frysteev_ it's easier to set up a profile but you don't have to if it's a simple part
[12:46:53] <Tom_itx> yeah
[12:47:07] <Tom_itx> but they're all quite similar
[12:47:41] <Tom_itx> i personally liked smartcam because of the way you could make your own post for different machines instead of buying them from the vendor
[12:48:33] <frysteev_> is that typical?
[12:48:35] <Tom_itx> at the time, it and mastercam were the ones to have and at the time, smartcam was easier to use and learn
[12:48:39] <Tom_itx> it varies
[12:48:59] <Tom_itx> if you get catia, you don't even get a post with it
[12:49:11] <frysteev_> for my laser, i will be using its original controlled, the rest of my machines have emc
[12:49:29] <Tom_itx> and at that time the owner paid $75k for a seat of it
[12:50:22] <Tom_itx> also you may want to look at ncgui
[12:51:58] <frysteev_> eep
[12:53:03] <ssi> https://p.twimg.com/AoC48fSCIAAZ-rY.jpg
[12:53:04] <ssi> oil!
[12:53:22] <FinboySlick> ssi Surface grinder!
[12:53:28] <ssi> yea that too :)
[12:53:41] <FinboySlick> Looks old but in good shape.
[12:53:46] <ssi> yeah, it's a Reid 2C
[12:53:52] <ssi> pretty good shape, spindle's tight
[12:54:00] <ssi> I bought it on ebay for $400
[12:54:07] <ssi> and it cost me $1500 to bring it home and unload it :)
[12:54:30] <Tom_itx> why can't i find a link to ncgui now?
[12:54:35] <ssi> ngcgui
[12:54:46] <FinboySlick> Hehe, where I live, 400$ machines would still cost $2k shipping.
[12:55:12] <FinboySlick> Irony, there's a train station right in front of my home :P
[12:55:15] <ssi> heheh
[12:55:33] <ssi> well this one was about 300 miles from me, and I dragged a trailer down to fetch it
[12:55:43] <ssi> but then I had to rent a forklift when I got home, and that cost me like $650
[12:55:44] <Tom_itx> http://www.linuxcnc.org/docs/2.5/html/gui/ngcgui.html
[12:55:52] <ssi> which is why this time when I brought the hardinge home, I BOUGHT a forklift
[12:55:57] <ssi> not going through that nightmare again :)
[12:56:13] <FinboySlick> ssi: You mean the bycicle repair platform?
[12:56:19] <FinboySlick> bicycle
[12:56:19] <ssi> that's the one!
[12:56:23] <jdhnc> should have gone down to home depot and hired 10 guys to carry it for you!
[12:56:38] <ssi> actually until yesterday, that surface grinder was still on the pallet that it was on when I got it
[12:57:04] <ssi> but I managed to sneak the lift into the garage and pick up the grinder so I could get the pallet out from under it
[13:00:22] <djdelorie> when I bought my metal lathe, I also had to buy the engine hoist next to it so I could move it around...
[13:00:57] <djdelorie> but a friend helped me get it into the basement:
http://www.delorie.com/photos/southbend-lathe/img_2418.html
[13:01:31] <ssi> any HNC folks lurking around?
[13:03:40] <frysteev_> ok cool. thats what i was thinking too
[13:04:14] <cradek> ssi: I think so but check your schematics to be sure
[13:04:24] <frysteev_> on my drives, the tach circuit is part of the overcurrent circuitry
[13:04:26] <ssi> the schemes for the control itself are a big mess
[13:04:34] <cradek> keep the special twisted/shielded wire if you can
[13:04:55] <cradek> frysteev_: do you mean overspeed?
[13:05:21] <cradek> frysteev_: I think a speed limit is common (can detect runaway, sort of)
[13:06:28] <frysteev_> to the best i have figured it out,, if its trying to drive the motor, and thinks its moving, and the tacho voltage isnt matching, it will trip the overcurrent on mine, evener if it has overcurrented actually
[13:07:29] <ssi> cradek: it calls out 25V for the limit switches everywhere... how are you running your limits?
[13:07:53] <archivist> djdelorie, I have a big old Southbend here too
[13:08:30] <cradek> I don't remember, maybe they're on 12v
[13:08:42] <djdelorie> that one was my father-in-laws. He had it since the 70's. His friend had it since WWII.
[13:08:45] <cradek> they're all proxes, even the double home switches
[13:08:47] <ssi> seems ike it wouldn't make a difference
[13:08:58] <djdelorie> they kept it in good shape; I even have all the gears!
[13:09:01] <ssi> hrm not sure how the prox switches work actually
[13:09:08] <ssi> hall effect?
[13:11:12] <ssi> djdelorie: that's a pretty nice old lathe
[13:11:17] <djdelorie> thanks!
[13:11:19] <ssi> you have it running?
[13:11:43] <djdelorie> it ran when I brought it home, but I took the motor off to get it into the basement and I've been cleaning it up before I actually use it.
[13:11:53] <djdelorie> need to replace the dirty grease with clean grease, remove rust, etc.
[13:11:57] <ssi> yea
[13:12:03] <ssi> I saw your rust electrolysis pics
[13:12:06] <ssi> looks like it's working pretty well
[13:12:08] <djdelorie> the previous owners *used* it
[13:12:18] <djdelorie> it does, it just takes time. About one part per day
[13:12:50] <FinboySlick> djdelorie: Can I peek? Rust removal tricks could come in pretty handy for me in the near future.
[13:13:24] <ssi> it's funny seeing a modern QCTP on that ancient lathe :)
[13:13:41] <djdelorie> http://www.delorie.com/photos/southbend-lathe/ - the last few photos. Just google for "rust electrolysis"
[13:13:56] <djdelorie> ssi: I have the original toolposts too
[13:14:39] <FinboySlick> Wow...
[13:14:40] <ssi> I have a 12x36 clausing from the early 60s
[13:14:44] <ssi> but it's worn out as hell :/
[13:14:51] <ssi> needs to be scraped in, and I dunno that I have the patience to do it
[13:16:02] <frysteev_> ssi: get an intern
[13:16:07] <ssi> hheheh
[13:16:12] <ssi> anyone wanna be my intern? :D
[13:17:36] <FinboySlick> Heh, if I had time I'd volunteer, I'd probably learn a lot even on the menial stuff.
[13:18:12] <djdelorie> my lathe was originally made in 1922, according to SouthBend
[13:18:42] <IchGuckLive> 1922 does it have E-power
[13:18:56] <IchGuckLive> or powerd by Hp
[13:19:00] <djdelorie> it ran off overhead belts. I have the pulleys for that too
[13:19:30] <IchGuckLive> if got the warer wheel running in about 50m Distance
[13:19:42] <IchGuckLive> water
[13:20:33] <IchGuckLive> Transmission belt from outsourced power source that is the futher keep it in stock for refoundation of the technology
[13:22:22] <djdelorie> I have to keep them; that's what the motor connects to :-)
[13:24:21] <FinboySlick> djdelorie: What's your solution for electrolysis, just some acid and water?
[13:24:40] <djdelorie> no, sodium carbonate (washing soda). It's electrolysis, not electroplating
[13:25:36] <djdelorie> the process changes the orange (hard) rust to black (soft) rust, which you can clean off with a kitchen scrubbie or brass wheel.
[13:26:06] <FinboySlick> OK.
[13:29:04] <djdelorie> the sodium carbonate is only there to make the water conduct electricity, the current breaks the water into hydrogen and oxygen, and the oxygen changes the rust
[13:29:18] <djdelorie> (at least, that's what I've read, it works so I'm not complaining :)
[13:30:00] <ssi> sweet, got the rest of my mesa gear
[13:30:03] <flyback> wonder how well sodium carbonate + sodium percarbonate would work
[13:30:06] <flyback> aka oxyclean
[13:30:28] <ssi> whoa what are these little orange stars that came with them?
[13:30:40] <djdelorie> from what I read, you get better results by more current, not more chemicals. I.e. get out taht 48V 10A power supply :-)
[13:30:52] <flyback> and don't you have that backwards
[13:30:52] <ssi> just current?
[13:30:57] <flyback> normal orange rust is soft
[13:31:06] <flyback> hard rust is a protective film like al or copper
[13:31:06] <ssi> what about this 50A 5V supply I have?
[13:31:34] <FinboySlick> djdelorie: I see. My guess would have been that the carbon steals oxygen from the iron oxide and the electrolysis causes oxygen to leave the solution.
[13:31:36] <djdelorie> ssi: just current. The water has resistance, higher voltage leads to higher current, current = electron flow = oxygen ions
[13:31:57] <ssi> need lower resistance water then!
[13:32:20] <djdelorie> FinboySlick: all this is "from what I read". They say the gas produced is hydrogen, not oxygen. I don't know the chemistry. I know the gears go in orange and come out black, and the black cleans right off
[13:33:07] <djdelorie> the "sacrificial anodes" get a layer of gunk on them you have to scrape off occasionally, too.
[13:33:15] <FinboySlick> djdelorie: I'm not very good at this stuff either.
[13:34:07] <djdelorie> the black stuff is like a fine powder, I use a brass wheel on my dremel between the teeth and it takes very little effort to clean them.
[13:36:03] <ssi> ok lets see what pncconf has to say
[13:38:45] <pcw> ssi: note that you should not connect daughtercards unless you know the
[13:38:47] <pcw> firmware matches the daughtercard(s) and the daughtercard(s)
[13:38:49] <pcw> are on the correct connectors
[13:39:44] <ssi> good to know
[13:39:49] <ssi> I haven't attached anything yet
[13:40:22] <ssi> hrm pncconf doesn't have a comprehensive list of firmwares
[13:41:30] <pcw> Ive tried to convince Chris M to just use the installed firmware as the data source (and not have 4 billion xml files)
[13:42:00] <pcw> at least as an option
[13:44:30] <ssi> hrm so perhaps the resmod bitfiles don't come preinstalled
[13:44:42] <ssi> I'm looking for something like SVRM6.bit yeah?
[13:45:19] <pcw> I dont think so, LinuxCNC firmware sources are more than a year old
[13:45:46] <ssi> are they up to date in git?
[13:45:50] <ssi> or is there some other source for them
[13:45:54] <pcw> no
[13:49:08] <cradek> ssi: we're all hoping one day pcw will use git
[13:49:13] <ssi> haha
[13:49:13] <cradek> hoping hoping hoping
[13:50:26] <pcw> Seb was going to update the buildbot but Hes probably too busy
[13:50:57] <cradek> pcw: we hope to work on it Mar 28-30
[13:51:23] <cradek> pcw: ideally we'd have updated firmware packages built to release with 2.5 near Mar 31
[13:52:46] <ssi> so... I'm confused
[13:53:02] <ssi> there are firmwares built that work for resolvers
[13:53:10] <ssi> but I can't have them, because noone has put them on the internet?
[13:53:11] <pcw> Yes
[13:53:47] <cradek> pcw: is there also somehow hal support for resolvers?
[13:53:53] <frysteev_> ssi
[13:54:00] <cradek> (as encoders?)
[13:54:07] <frysteev_> it also depends on what typ of resolvers,
[13:54:18] <frysteev_> transmitter resolvers, or reciever resolvers
[13:54:34] <ssi> I don't know the answer to that... cradek might
[13:54:35] <pcw> yes Andy wrote a driver
[13:54:45] <pcw> its in 2.5
[13:54:48] <cradek> ah
[13:54:53] <cradek> how did I miss that.
[13:55:07] <ssi> so I'll need to move up to 2.5 then
[13:55:08] <pcw> seems to work OK, havent tried the simulated index however
[13:55:31] <frysteev_> ssi: wht type of resolvers do you have?
[13:55:40] <ssi> frysteev_: whatever's in the HNC :)
[13:55:44] <ssi> frysteev_: I don't know how to tell you the answer to that
[13:55:50] <pcw> (the resolver firmware makes it pretty easy, it just has velocity and position regs)
[13:56:09] <cradek> oh look, resolver.c
[13:56:46] <frysteev_> im NOt an expert on this, one way, energizes 2 of the coils, one with sin, and the other with cosin, the third coil is the analyized against that
[13:57:01] <pcw> other way
[13:57:13] <cradek> frysteev_: one driving coil and sin/cos receivers
[13:57:40] <ssi> pretty sure my resolvers take a driving signal and return sin/cos
[13:57:50] <pcw> Yep
[13:58:14] <ssi> haha this is my favorite part
[13:58:15] <ssi> int hm2_resolver_get_param(int param){ // This function will eventually return parameters from the 7i49 // but for now, it just returns the number 6 return 6;
[13:58:18] <ssi> }
[13:58:36] <frysteev_> thats handy
[13:58:50] <ssi> andy's patented six generator
[13:59:04] <pcw> Actually thats readable from the firmware (but Andy punted)
[14:01:31] * frysteev_ wonders why there isnt a band called 'the resolvers'
[14:04:24] <pcw> I'll make a simple 5I23 resolver bitfile up today
[14:04:26] <pcw> theres freeby.mesanet.com/rmsvss6_8.bit which should work
[14:04:28] <pcw> but has a 8 channel sserial interface as well
[14:04:43] <ssi> thanks, that'll get me started
[14:04:57] <ssi> I think I just saw something in a google summary about resmod not fitting in a 5i20
[14:05:01] <ssi> is that accurate?
[14:05:04] <ssi> (good thing I got a 5i23 :P)
[14:05:09] <pcw> Yes too big
[14:05:34] <ssi> yikes
[14:05:37] <ssi> dodged that bullet :D
[14:06:17] <pcw> and too much BlockRAM (5i20 has only ~7Kbytes total)
[14:06:55] <ssi> I need to brush up on my vhdl
[14:07:02] <ssi> it's been almost ten years since I messed with them :/
[14:07:03] <pcw> Pretty sure I could cram it in a Spartan2, pretty sure I dont want to...
[14:07:56] <pcw> the resolver interface is all brute force, If I did it again I would do it smaller/smarter
[14:12:06] <pcw> I realized later I could get rid of about 30% of the 600 or so lines of assy language
[14:12:07] <pcw> with a smarter A-D front end, but thats deep into IIABDFI territory
[14:14:09] <ssi> yea I understand that :P
[14:14:23] <cradek> if you're looking for redemption you're in the wrong channel
[14:14:39] <cradek> nobody likes code they wrote "a while back" :-)
[14:15:03] <pcw> Gets uglier and dumber
[14:15:16] <jdhnc> "wtf did this idiot write this crappy code like this.... oh, it was me"
[14:17:18] <pcw> or just coming back after a few months and having it look like Greek
[14:19:54] <mrsun> how fun that my pcb etchant didnt come ... or it did, 1 bag and it said 5 bags ... so have to call the company and complain :/
[14:20:24] <cradek> yes, the feeling of "I used to be smart enough to understand how this works" can be a real drag.
[14:20:50] <ssi> I get that a lot these days :(
[14:20:52] <ssi> getting old sucks
[14:22:52] <pcw> I'm happy I just got ARP working on a Ethernet FPGA card (hostmot2!)
[14:22:53] <pcw> Spent 3 days on a desperately stupid bug (a macro smashing the accumulator)
[14:22:55] <pcw> (I'm probably older than quite a few here)
[14:23:16] <cradek> pcw: has andy showed up in CA yet?
[14:23:35] <pcw> Not due till April
[14:27:43] <mrsun> pcw, how old are you ? :)
[14:28:25] <pcw> three score
[14:28:32] <mrsun> huh ? :P
[14:28:35] <ssi> hahaha
[14:36:22] <ssi> this machine has a pile of limit switches
[14:37:24] <cradek> do you have the silly cutoff slide?
[14:37:28] <ssi> no
[14:37:45] <ssi> I kinda wish I did, but everyone says it doesn't work well
[14:37:48] <ssi> still looks awesome :)
[14:38:35] <ssi> so why is there a spindle resolver AND a spindle encoder?!
[14:41:31] <jdhnc> one for the control, one for the drive?
[14:41:37] <archivist> pcw, never mind gene77 is older than us
[14:42:14] <cradek> ssi: good question - also notice the encoder only has one channel
[14:42:15] <pcw> Yep
[14:42:42] <ssi> cradek: I can't tell anything about the encoder from this drawing other than it's got six lines coming off it
[14:46:12] <archivist> how old code looks
http://www.bonkersworld.net/building-software/
[14:48:08] <frysteev_> so the encoder is just a tach then?
[14:48:54] <ssi> really not sure yet :P
[14:49:01] <ssi> also trying to dig up some info on the prox switches
[14:49:16] <frysteev_> ssi: you better be taking pics! :P
[14:49:22] <ssi> haha
[14:49:25] <ssi> tell me what you want pics of!
[14:54:19] <frysteev_> your conversion,
[14:54:36] <frysteev_> cnc geekery,
[14:54:51] <frysteev_> not those pics you post over int hat OTHER irc channel :P
[14:57:01] <jdhnc> you promised to never mention those.
[15:05:10] <ssi> frysteev_: I have some pics I'm putting up just for you :)
[15:07:16] <ssi> http://sphotos.xx.fbcdn.net/hphotos-snc7/416980_811175596152_71107655_35471859_1798297854_n.jpg
[15:07:26] <ssi> http://a3.sphotos.ak.fbcdn.net/hphotos-ak-ash4/421789_811175646052_71107655_35471860_1184776647_n.jpg
[15:07:31] <ssi> http://sphotos.xx.fbcdn.net/hphotos-ash4/421833_811175720902_71107655_35471862_290089005_n.jpg
[15:07:38] <ssi> http://a4.sphotos.ak.fbcdn.net/hphotos-ak-snc7/427591_811175785772_71107655_35471864_1274608115_n.jpg
[15:07:45] <ssi> http://a2.sphotos.ak.fbcdn.net/hphotos-ak-snc7/422069_811175920502_71107655_35471867_379063615_n.jpg
[15:10:41] <archivist> too much floor space...needs more toys
[15:11:24] <Loetmichel> archivist: indeed
[15:11:41] <Loetmichel> ocer here i have som ways of about 60-80cm
[15:11:59] <cradek> I love the whiteboard where you're supposed to write down your tool offsets because it'll probably forget them when the bank of D-cells runs down
[15:12:04] <Loetmichel> the test of the 12m^2 worspace is occupied by desks and machines ;-)
[15:12:33] <Loetmichel> rest
[15:17:26] <Thetawaves> ssi, second to last photo, looks like an old school computer back there
[15:18:20] <ssi> it's the GE550 control for this hardinge
[15:18:39] <skunkworks> ssi: still operational?
[15:18:46] <ssi> yep
[15:19:31] <frysteev_> ssi: megacyclelabs.ca my mess
[15:21:29] <ssi> i want a laser! D:
[15:21:33] <archivist> not much spare space in that garage
[15:22:16] <frysteev_> ive tossed out like 2 dumpsters worth in the last week
[15:22:31] <ssi> mine is quite a bit cleaner now than it was 2 weeks ago
[15:22:35] <ssi> I've had a friend over here helping me with that
[15:24:19] <ssi> there's three lathes, two mills, three bandsaws, a surface grinder, a 4x4' CNC plasma table, two blast cabinets, a TIG welder with 16' of bench space, three grinders, a buffer, and a drill press in 1200sqft of space :)
[15:24:33] <ssi> the forklift is parked outside sadly
[15:25:46] <frysteev_> i only have the laser, wich take up like half my shop, a small cnc router, a small 40watt laser cutter and a robot arm
[15:26:03] <ssi> what's the power on the big laser?
[15:31:55] <frysteev_> 1.5kw
[15:32:04] <frysteev_> still working on getting that machine going
[15:32:16] <frysteev_> working on the motion contorl stuff now,
[15:32:24] <ssi> I want a machine like that!
[15:32:25] <alex4nder> joe9: what steppers did you go with?
[15:33:00] <frysteev_> ssi: me too :P
[15:33:03] <joe9> the keling ones, not sure of the specs. would have to look it up.
[15:33:42] <alex4nder> ah
[15:35:51] <ssi> frysteev_: is it stationary gantry, sliding table?
[15:37:07] <frysteev_> sliding table.
[15:37:28] <frysteev_> table is rated for 800lbs of workpiece
[15:37:35] <ssi> nice
[15:37:45] <ssi> what's the max steel pierce?
[15:58:40] <PCW__> something is really broken with linuxcnc.org ~25% packet loss ~30 second page refresh
[16:04:51] <cradek> the lossage is at the last hop, so inside dreamhost's compound
[16:21:11] <ej_> evening! (from my point of view anyway :)
[16:22:08] <ej_> I just learned that moschip parallel pci card from DX doesn't work and I could use suggestions for alternatives
[16:22:22] <ej_> I have pluto board, so EPP would be needed
[16:24:04] <PCW__> Yep MOSChip (at least the PCI ones) are nogo fpr EPP
[16:24:06] <PCW__> usually motherboard ports are the most likely to work
[16:24:49] <ej_> I have one that doesn't accept pure EPP. it always puts EPP/ECP and doesn't work :/
[16:24:58] <ej_> or then I messed something up
[16:25:35] <djdelorie> my motherboard doesn't have a parallel port at all...
[16:25:41] <PCW__> I would kind of expect most motherboard ports to work (but I have no Pluto experience)
[16:26:29] <PCW__> I know the oxsemi (now PLXTech I think) based cards work with our EPP interface
[16:27:19] <PCW__> probably many PCI parallel ports should work just not MOSChip/NetMOS
[16:27:39] <ej_> in theory they should
[16:29:09] <PCW__> MOSChip/NetMOS are known to not work (chip error in EPP inplementation)
[16:29:12] <ej_> I just wouldn't want to spend the time for iterating DX offerings
[16:29:15] <Thetawaves> i've got a startech pci parallel port card that claims to support EEP/ECP
[16:29:40] <ssi> I'm not having much luck with anything this evening :(
[16:29:48] <PCW__> MOSChip ones claim this also
[16:30:11] <PCW__> but have a broken EPP imlpementation
[16:31:17] <ej_> I found that out after waiting a month for a pci card and setting it up :)
[16:35:30] <ej_> is there some way I can check the lpt setting from the linux side? I'm not that linux-literate
[16:35:30] <DJ9DJ> gn8
[16:41:17] <PCW__> The pluto driver should set the port into EPP mode (not that it will help with MOSChip)
[16:42:11] <ej_> ok, thanks!
[16:43:46] <ej_> now I get: parport0 does not support mode 4
[16:44:39] <ej_> so I have to try to get another BIOS to enable epp only :/
[17:23:14] <PCW__> The BIOS will not deal with non motherboard port modes
[17:27:42] <Nick001-Shop> I have a line Linksp FloodOn <+parport.0.pin-02-out - this pin-02-out needs to be inverted. How do I accomplish this as the manual only covers in pins as far as I can find. Right now M8 and M9 are doing the reverse of what they're supposed to be doing.
[17:27:59] <Nick001-Shop> <=
[17:30:03] <skunkworks__> parport.<p>.pin-<n>-out-invert (bit) Inverts an output pin.
[17:30:42] <Nick001-Shop> I'll go try that and be back
[17:31:03] <skunkworks__> so you would do a parport.0.pin-02-out true
[17:31:22] <skunkworks__> so you would do a parport.0.pin-02-out-invert true
[17:31:26] <skunkworks__> sorry
[17:32:11] <Nick001-Shop> ok
[17:35:45] <Nick001-Shop> comes back with an error Link SP requires two arguments, three given
[17:40:20] <Tom_itx> where's the link to the stepper voltage calculator?
[17:40:41] <Nick001-Shop> The first one errors pin parport.0.pin-02-out-invert not found
[17:42:06] <SadMan> try setp parport.0.pin-02-out-invert 1
[17:44:34] <Nick001-Shop> ok be back
[17:47:37] <Nick001-Shop> error linksp requires2 arguments 4 given Been going round and round with this one
[17:48:07] <SadMan> you need 2 separate lines
[17:48:20] <SadMan> your original and that invert one
[17:49:00] <SadMan> setp ...
[17:49:02] <SadMan> linksp ...
[17:49:52] <Nick001-Shop> Ok back again
[17:56:15] <ssi> so I think my coolant motor might be boogered
[17:56:21] <ssi> VFD won't run it up to full frequency
[17:56:26] <ssi> and it gets hot and trips overcurrent after awhile
[17:57:28] <SolarNRG> Hi guys
[17:57:33] <SolarNRG> I;m making a cnc machine
[17:57:42] <SolarNRG> i got these 3 3.1nm stepper motors today
[17:57:52] <SolarNRG> and i got an arduino
[17:57:58] <Tom_itx> http://cnc4you.co.uk/resources/60BYGH301B.PDF
[17:57:59] <Tom_itx> those
[17:58:16] <SolarNRG> yeah them
[17:58:18] <SolarNRG> thanks tom
[17:58:39] <SolarNRG> http://i.imgur.com/rh3SU.jpg
[17:58:42] <SolarNRG> heres a pic
[17:59:01] <Tom_itx> i'm trying to steer him toward linuxcnc but i need a cheering section here
[17:59:01] <Nick001-Shop> Linksp FloodOn <= parport.0.pin-02-out
[17:59:03] <Nick001-Shop> setp parport.0.pin-02-out-invert 1
[17:59:05] <Nick001-Shop> These 2 lines work and now M*8 and M9 are working correctly Thanks
[17:59:27] <SolarNRG> help
[17:59:39] <SolarNRG> i'm stukc with a load of motors and i dont know how to make them turn
[17:59:53] <jdhnc> SOlar: I'll take 3 :)
[18:00:22] <jdhnc> that's like... 400+ oz
[18:00:51] <Tom_itx> yup
[18:01:01] <SolarNRG> 439 oz inch yep
[18:01:11] <SolarNRG> They seemed good for what I want to make
[18:01:16] <SolarNRG> an aluminium router
[18:01:19] <SolarNRG> I got the router head
[18:02:37] <jdhnc> what drivers?
[18:02:55] <Tom_itx> he doesn't have any yet
[18:03:03] <Tom_itx> i think he wants to build them
[18:03:03] <djdelorie> SolarNRG: "Go! Go! Rah rah rah!"
[18:03:05] <SolarNRG> http://i.imgur.com/RVKBw.jpg
[18:03:12] <Tom_itx> but i suggested he get some
[18:03:17] <djdelorie> that's all the cheering section I do...
[18:03:25] <Tom_itx> thanks
[18:03:26] <Tom_itx> :)
[18:03:33] <SolarNRG> This is going to be my router head
[18:03:46] <jdhnc> that's a lot of plastic
[18:03:47] <SolarNRG> Thank you for your time and support people
[18:04:22] <jdhnc> drivers are good :)
[18:04:24] <SolarNRG> Also I got thread I can use
[18:04:25] <Tom_itx> jdhnc, there's a router in there somewhere
[18:04:40] <jdhnc> Tom: you mean, there's a dremel in there somewhere.
[18:04:45] <Tom_itx> :)
[18:04:54] <jdhnc> it will work, for a while.
[18:05:03] <Tom_itx> yeah that may not be heavy enough for what you plan to do
[18:05:07] <djdelorie> SolarNRG: what part are you needing help on? Theory? Electronics? Software?
[18:05:08] <jdhnc> probably long enough to make a mount for a better one.
[18:05:36] <SolarNRG> I'm building a CNC machien from scratch
[18:05:47] <Tom_itx> i think he want's to use an arduino to control it but i'm trying to talk him out of it
[18:05:52] <SolarNRG> I have experience in controlling stepper motors with an arduino, uln2003 and a power adaptor
[18:05:55] <frysteev_> like as in hand winding the motors by hand?
[18:06:01] <JustinXJS> I started with a dremel, they dont last long cutting aluminum
[18:06:06] <djdelorie> I suspect my idea of "scratch" is different than yours... but if it's not, I'll help :-)
[18:06:11] <SolarNRG> But these steppers are like loads beefier than my little project ones
[18:06:26] <SolarNRG> I've used a PC power supply to get a windscreen wiper motor to turn around
[18:06:28] <jdhnc> they are 4 amp motors, buy drivers
[18:06:42] <SolarNRG> But don't the drivers need a power supply?
[18:06:44] <JustinXJS> A uln2003 will not cut it
[18:06:54] <djdelorie> jdhnc: unless you're my kind of crazy, then make your own drivers from scratch :-)
[18:06:56] <Tom_itx> jdhnc, will the G540 drive those?
[18:06:58] <SolarNRG> I thought about using some 8am N type mosfets to drive them
[18:07:06] <SolarNRG> Gecko drive?
[18:07:10] <Tom_itx> yes
[18:07:24] <djdelorie> SolarNRG: see
http://www.delorie.com/electronics/bldc/ for a "DIY controller"
[18:07:37] <djdelorie> (although mine is for bldc, not stepper, the power rules are similar)
[18:08:27] <jdhnc> you could run them off a TB6560, just not full current
[18:08:53] <SolarNRG> Why can't I just trigger 4 power mosfets in sequence to drive the motors?
[18:08:58] <jdhnc> you can
[18:09:01] <SolarNRG> Surely all I need is a 48v power source
[18:09:10] <SolarNRG> Surely that would be loads cheaper
[18:09:25] <SolarNRG> And I can use the arduino to turn usb g code into pulses
[18:09:34] <SolarNRG> 12 outputs in total
[18:09:37] <SolarNRG> 4 per motor
[18:09:37] <JustinXJS> With this kind of load you will at least need a good driver chip to drive the mosfets
[18:09:40] <djdelorie> SolarNRG: the tricky part is turning the high-side FET on
[18:10:00] <SolarNRG> Can I use my ULN2003 to turn on the power mosfets?
[18:10:08] <SolarNRG> arduino trggers the uln
[18:10:11] <jdhnc> couldn't you just drive teh mosfet with a uln?
[18:10:12] <djdelorie> if you can find a P-MOSFET that works, it will be easier, but you still need to switch the full motor voltage to drive the gate
[18:10:43] <SolarNRG> ulns are only rated to half an amp
[18:10:52] <SolarNRG> but as a trigger they won't draw too much current, right?
[18:11:04] <djdelorie> right, current rating there just gives you faster switching times
[18:11:17] <JustinXJS> which you want
[18:11:25] <SolarNRG> arduino can only do 1ms pulses
[18:11:43] <jdhnc> you don't want the arduino interpreting gcode
[18:11:48] <SolarNRG> in my original stepper motor experiments I discovered any pulses on off with delays of less than 10ms woudn't turn
[18:12:40] <djdelorie> you can manually commutate a stepper just by tapping the wires by hand
[18:13:45] <djdelorie> with that stepper, you could wire it in "parallel" configuration and use a 5V pc power supply, that might simplify the electronics
[18:13:54] <jdhnc> it woudl be rather slow
[18:14:08] <djdelorie> but it wouldn't be optimal - the power mosftets really want 15v or more on the gate to turn them on *hard*
[18:15:08] <JustinXJS> yes, a driver should be used
[18:16:56] <jdhnc> I have about the same motors (425oz/in) on a xylotex controller at 30v
[18:17:36] <djdelorie> with my controller tests, I'm able to max out the motor current with only 24v. But with 60v I can max it out *faster*
[18:18:17] <SolarNRG> what do you mean by max out?
[18:18:22] <jdhnc> you get current at speed though
[18:18:31] <djdelorie> the motor specs say "maximum continuous current: 2.55 amps"
[18:19:03] <djdelorie> the coil is 5.6 ohms, so 15 volts or so gives you that, but it takes "forever" to build up the magnetic field.
[18:19:04] <SolarNRG> But I can make an arduino based controller for peanuts
[18:19:16] <SolarNRG> why will this not be optimal?
[18:19:18] <jdhnc> Solar: you can, but what woudl it do?
[18:19:19] <SolarNRG> Will this be slow?
[18:19:24] <djdelorie> SolarNRG: the problem isn't the controller, the problem is the *power*
[18:19:37] <SolarNRG> So controller sorted
[18:19:45] <djdelorie> you need to switch lots of amps very quickly, meaning "fast edges" not "fast CPUs"
[18:19:46] <SolarNRG> I need a DC to DC converter
[18:20:05] <jdhnc> more loss, more lag
[18:20:34] <djdelorie> a driver chip will have all the power electronics you need in it, just need to find and choose the right one
[18:20:37] <SolarNRG> I've heard to steer clear of chinese drivers
[18:20:49] <jdhnc> chinese drivers are like any other chinese tool
[18:20:54] <SolarNRG> :))
[18:21:00] <jdhnc> fine for lots of things
[18:21:12] <djdelorie> if you look at the bldc link I posted, the big chip with the white rectangle on it does nothing but convert logic signals to heavy current power signals
[18:21:21] <jdhnc> can't beat the cost of a tb6560 board
[18:21:27] <Tom_itx> what about a HIP4081?
[18:21:34] <Tom_itx> for a driver
[18:21:58] <SolarNRG> Is it parallel port based or usb?
[18:22:00] <JustinXJS> I love the IR2110
[18:22:24] <JustinXJS> It is only high/low side, so you need 2 per bridge though
[18:23:02] <SolarNRG> Its a chip!
[18:23:21] <JustinXJS> yes, to switch the fets
[18:23:30] <SolarNRG> Can I solder this onto prototype board?
[18:23:46] <SolarNRG> So its better than a ULN
[18:24:21] <JustinXJS> yes, you should read the data sheet to see how they work
[18:25:08] <JustinXJS> and buy spares, bridge design is for these power levels is not easy and you are bound to let the smoke out before you get it right
[18:25:10] <JustinXJS> :)
[18:25:44] <SolarNRG> They are not like ULN's at all, they seem very confusing
[18:26:05] <SolarNRG> Why do you like this chip
[18:26:07] <SolarNRG> ?
[18:26:26] <djdelorie> JustinXJS: I've let the smoke out of my test chip twice, it still works.
[18:26:50] <PCW__> Thanks to however fixed the LinuxCNC forum/dreamhost, its back to normal!
[18:28:44] <JustinXJS> SolarNRG, that chip is fairly easy to use, as far as bridge drivers go, that is why I like it.
[18:28:55] <JustinXJS> haha, lucky djdelorie
[18:29:49] <djdelorie> yup, the IRS2110 looks like the right chip for a DIY controller. Need a bunch of them though...
[18:30:08] <SolarNRG> 1 chip controlls 2 inputs
[18:30:11] <SolarNRG> 12 in total
[18:30:13] <SolarNRG> so 6 chips?
[18:30:20] <SolarNRG> minimum for 3 axis?
[18:30:27] <djdelorie> one chip controls one WIRE
[18:30:31] <SolarNRG> ??
[18:30:39] <SolarNRG> What's the difference between Ho and Lo then?
[18:30:49] <djdelorie> for each wire going to the stepper, you either connect it to GND or VCC - that's two driver transistors, so two drive signals
[18:31:18] <djdelorie> so for a stepper, you need four drivers minimum (two wires per phase)
[18:32:00] <djdelorie> in a driver like the IRS2110, the two outputs are both designed to drive an N-MOSFET, but one connects to the "high" voltage, and the other to the "low" voltage.
[18:32:08] <JustinXJS> Yes, steppers need two bridges, so 4 IR2110s and 8 fets
[18:32:21] <djdelorie> the "high" gate drive provides either VCC or VCC+Vgate, the "low" driver is GND or GND+Vgate
[18:32:26] <SolarNRG> With my original unipolar stepper system that I made, the red and green went to 5 volts in, the yellow orange brown black in sequence went to the right side of my ulns 1-4 and when the pins on the 1-4 of the left of the uln were triggered by the arduino it completed the circuit
[18:32:36] <SolarNRG> and I know the half step algorithm
[18:32:49] <PCW__> We did not have so much luck with the IGBT modules when we made mistakes (at 360VDC and 5000 uF things go bad pretty fast)
[18:32:58] <SolarNRG> 1000 1100 0100 0110 0010 0011 0001 1001 etc
[18:33:04] <flyback> mmmmm
[18:33:06] <flyback> IBGT
[18:33:10] <flyback> IGBT
[18:33:21] <flyback> the "CATAPILLER" of transistors :)
[18:33:28] <SolarNRG> 8 fets?
[18:33:31] <djdelorie> if you want to do unipolar, connect common to VCC and you only need the "low" sides, but you still need two drivers per phase
[18:33:32] <flyback> as in CATAPILLER HEAVY DIESELS
[18:33:35] <SolarNRG> 4 really big ones and 4 really small ones?
[18:33:44] <djdelorie> no, 8 all the same size
[18:33:53] <JustinXJS> they should be matched
[18:34:06] <PCW__> IGBTs get huge
[18:34:38] <PCW__> (too slow for step motor drives however)
[18:34:40] <djdelorie> but n-mosfets are easier to get than p-mosfets, hence all the "high side driver" stuff.
[18:34:46] <SolarNRG> Isolated gate bipolar transistors? They make my BJTs look puny
[18:34:54] <djdelorie> PCW__: my driver uses IGBTs at 20 KHz drive
[18:34:55] <JustinXJS> yeah they do, I have 4 matched 1MW IGBTs on my desk, thinking of what to use them in
[18:35:12] <PCW__> Yea but 20KHz is slow for steppers
[18:35:16] <SolarNRG> megawatt?
[18:35:33] <flyback> IGBT OWN YOUR "CANUCK"
[18:35:43] <SolarNRG> Sounds like the Mr. Burns switch
[18:35:46] <Tom_itx> you may wanna rethink the arduino as the driver for your steppers
[18:35:56] <SolarNRG> This street hasn't been paying me enough money, click, power goes off
[18:35:59] <flyback> I think the swede company ABB
[18:36:08] <flyback> has patents on like 100 variants of IGBT
[18:36:37] <JustinXJS> Yeah, 20kHz is a bit slow and yes, 1 megawatt
[18:36:42] <JustinXJS> SolarNRG, Commercial stepper drivers have so many nice features, like low current holdi mode. I would seriously consider buying stepper drivers, designing your own stepper driver (specifically the bridge section) is no trivial task for these kind of loads.
[18:37:15] <PCW__> We use 12 KHz to keep loses within reason (though big IGBTs are often slower = 5 or even 2 KHz)
[18:37:15] <SolarNRG> How is this any different from my old uln setup except scaling everything up?
[18:37:44] <PCW__> fancy step drives are current sources
[18:37:49] <flyback> or do both
[18:38:02] <flyback> buy a fancy driver chip and follow the appnotes for building into a circuit
[18:38:05] <flyback> and add your own spin on it
[18:39:14] <SolarNRG> bridge? i.e turning arduino voltage current into gradually higher and higher voltages and currents until you can switch on and off the largest transistors in the circuit that tell the stepper motors steps to go on and off?
[18:39:58] <Tom_itx> added latency each step
[18:40:46] <SolarNRG> so if I do it my way, it WILL work but will be slow and chuggy as hell, right?
[18:41:02] <JustinXJS> it will have to be, yes
[18:41:28] <SolarNRG> But how much are the good commercial driver/controllers?
[18:41:48] <Tom_itx> is the G540 big enough for those steppers?
[18:42:00] <Tom_itx> that would be a good place to start
[18:42:19] <Tom_itx> or 3 302v
[18:43:03] <Tom_itx> err 203v
[18:43:12] <Jymmm> g540 has the breakout board, and relay controls chargepump, VFD, etc
[18:43:19] <SolarNRG> Come on, 200 dollars
[18:43:21] <SolarNRG> 300
[18:43:39] <SolarNRG> Does that even include a power supply?
[18:43:41] <Tom_itx> all you have now are 3 paperweights
[18:43:45] <SolarNRG> true
[18:43:45] <Tom_itx> nope
[18:43:48] <Jymmm> mid-band compensation
[18:43:48] <SolarNRG> very true
[18:44:48] <SolarNRG> Too much, I'll take my chances with the arduino, s-stage activation mosfet route. Much cheaper, same goal achieved, even if it is sub-optimal
[18:44:48] <PCW__> I think a Arduino would be prettly slow for this
[18:44:50] <PCW__> (we built a step controller with a DSPIC and 40 MIPs was just barely enough at 50 KHz)
[18:45:11] <SolarNRG> This is my 1st project
[18:45:27] <Tom_itx> we're just trying to keep it from being your last
[18:45:56] <SolarNRG> sounds like I'm trying to teach a prize winning marrow grower how to water here
[18:46:22] <SolarNRG> does the gecko drive even have usb complience?
[18:46:34] <PCW__> USB?
[18:46:43] <djdelorie> if you read the linuxcnc wiki, you'll see that USB is not suitable for cnc stepper control
[18:47:01] <PCW__> Universal Serial Botch
[18:47:32] <SolarNRG> Funny, I can use my USB arduino to control my mclennan 1NM stepper motors as is perfectly
[18:47:44] <djdelorie> there's no usb between the mcu and the stepper
[18:48:21] <djdelorie> what are you going to use the *run* the cnc machine? I.e. coordinate the cutting tasks etc?
[18:48:33] <SolarNRG> What you mean is you're grasping onto 20-30 year old tech and you aren't being enlightened to 2st century tech
[18:48:55] <PCW__> People have done it (Arais robotics for example) but its now a funny hybrid drip fed buffered system instead of having all real time control in one place
[18:49:41] <PCW__> EMC has all control in one place, a non real time buffered system (that USB requires) breaks this
[18:49:43] <SolarNRG> In my experience I anticipate with an arduino based system I can realistically hope to achieve about 100 "steps" in all axes per second
[18:49:44] <djdelorie> SolarNRG: sure, an arduino can control a stepper just fine. What it can't do it control three steppers at high speed with precision and determinism while simultenously interpreting a gcode stream
[18:50:06] <djdelorie> my board can do 1,000,000 steps per second
[18:50:09] <SolarNRG> :o
[18:50:38] <SolarNRG> No way
[18:50:38] <Tom_itx> 100 steps per second on a 200 step per rev stepper isn't much
[18:50:45] <PCW__> and especially dealing with real time inputs (without having to port all of EMC to your Arduino)
[18:50:46] <djdelorie> it's not that impressive when you consider microstepping.
[18:51:00] <SolarNRG> half step?
[18:51:29] <djdelorie> works out to 15,000 RPM on my motor, but the motor peaks at 3,000 RPM or 200,000 steps/sec
[18:51:55] <djdelorie> the chip has hardware support for it
[18:52:46] <SolarNRG> So is that why a gecko drive is important?
[18:53:07] <JustinXJS> SolarNRG, if you are set on building your own, run it the steppers in unipolar mode, copy this guy;
http://www.piclist.com/techref/io/stepper/linistep/index.htm put your arduino in place of the pic. The problem is the transistors get smoking hot with these kinds of loads
[18:53:45] <SolarNRG> Even at a mere 100 steps per second?
[18:53:54] <djdelorie> it's not the steps, it's the amps
[18:54:08] <SolarNRG> heatsinks
[18:54:10] <SolarNRG> thermo gel
[18:54:14] <SolarNRG> they are 80 p each
[18:54:19] <Tom_itx> liquid nitrogen
[18:54:31] <SolarNRG> forget it, I'll never afford that
[18:55:13] <JustinXJS> yeah, my first stepper drivers were those linisteppers, the heatsink I needed was 12"x10"
[18:55:17] <Tom_itx> give it a shot but keep in mind the options
[18:56:01] <SolarNRG> linisteppers?
[18:56:53] <JustinXJS> I think he is saying give building your own a shot
[18:57:20] <SolarNRG> But an arduino based dremel based one won't overheat because it'll be going soooooo slowly
[18:58:16] <SolarNRG> Then there's the construction
[18:58:22] <SolarNRG> I was advised to use C-slot
[18:58:33] <SolarNRG> I thought about using 25mm steel bar for rails
[18:58:42] <SolarNRG> And I've got some meter lengths of thread
[18:58:56] <SolarNRG> I thought if I got some c slot welded onto a couple of nuts
[18:58:56] <JustinXJS> I am guessing you will be moving more mass than you were moving when you were using the ULN, what was the load on your motor with the ULN?
[18:59:18] <SolarNRG> Well I stepped the 5v mclennan down even mroe with double gears
[18:59:21] <SolarNRG> I've got a vid
[18:59:45] <SolarNRG> http://www.youtube.com/watch?v=pp8vGB8JwtE
[18:59:52] <SolarNRG> It took 20 minutes to do 1 rotation
[19:00:08] <SolarNRG> Not bad for 5 volts and 50mA
[19:01:20] <SolarNRG> And here's another vid of it with load attached
http://www.youtube.com/watch?v=88zYI2hgFuw
[19:02:57] <SolarNRG> But to machine anything in any realistic timeframe, these tiny geared down steppers I'm guessing are too slow
[19:03:07] <JustinXJS> yes
[19:03:31] <Tom_itx> you gotta consider what the material feedrates dictate too
[19:03:53] <SolarNRG> well steel needs a slow feedrate
[19:03:58] <SolarNRG> Wood and plastic can be higher
[19:04:25] <Tom_itx> plastic will burn if it's not just right
[19:04:36] <JustinXJS> yes, you really need to design with a minimum speed in mind
[19:04:37] <SolarNRG> What about ceramics?
[19:04:57] <SolarNRG> Can I use a diamond coated router head to machine that?
[19:05:24] <JustinXJS> I build my own machine and I designed with 300 inches per minute for rapids in mind
[19:05:55] <SolarNRG> Can you build me one, save me the trouble and how much?
[19:06:14] <JustinXJS> haha
[19:06:34] <SolarNRG> As in "bollocks, you haven't the cash" right
[19:08:07] <JustinXJS> no, but I ended up dropping quite a bit of cash into mine. The problem is all the time, had I been paying myself it would have been far more economical to just buy a machine a do a CNC conversion on it
[19:08:18] <Jymmm> My suggestion would be to build the CHEAPEST thing that's "functional", that will give you the reality check that this isn't a cheap hobby, then you can plan/budget for what you really want.
[19:09:07] <Jymmm> cheapest can include duct tape and bailing wire.
[19:09:25] <JustinXJS> I agree
[19:09:26] <SolarNRG> What do you advise, C-slot? Acme thread? What sort of rails?
[19:09:42] <Jymmm> SolarNRG: Cheapest you can get away from
[19:09:46] <Jymmm> s/from/with/
[19:10:10] <Jymmm> allthread and 2x4's
[19:10:41] <SolarNRG> I've got some allthread
[19:10:58] <SolarNRG> But I was told if there are too many turns per inch it'll be slow as hell
[19:11:03] <SolarNRG> Hence why acme thread is used
[19:11:19] <JustinXJS> Yes, start there, that is what I started with, but ended up with ballscrews and thk linear rails, it was a good learning experience
[19:11:31] <ssi> allthread is awful... acme screw is cheap enough
[19:11:44] <Jymmm> ssi: but the nuts aren't
[19:11:51] <ssi> make your own nuts out of delrin
[19:11:55] <ssi> that's how I built my first machine
[19:11:59] <Jymmm> delrin isn't cheap either
[19:12:02] <SolarNRG> what is delrin?
[19:12:06] <Jymmm> plastic
[19:12:08] <JustinXJS> or you can buy the delrin nuts on ebay, for fairly cheap
[19:12:10] <ssi> high-weight plastic
[19:12:17] <ssi> and delrin is plenty cheap enough
[19:12:39] <SolarNRG> acme thread and delrin nuts
[19:12:55] <ssi> what sort of machine do you intend to build?
[19:13:01] <SolarNRG> how do I attach the delrin nuts to my rail bearings?
[19:13:09] <SolarNRG> One I can mill ceramic casts with
[19:13:18] <SolarNRG> Then pour molten bean cans into
[19:13:21] <Jymmm> Build the ghettoest one you can for as cheap as you can. It'll be an eye opening experience, then you can plan/build the rEAL ting.
[19:13:26] <ssi> molten bean cans, eh
[19:13:41] <SolarNRG> that i melt in a microwave
[19:13:48] <ssi> hahahaha
[19:13:53] <ssi> I like your style
[19:14:03] <SolarNRG> Here's a video of my attempt
[19:14:04] <SolarNRG> http://www.youtube.com/watch?v=VSKl1konZ0U
[19:14:48] * CareBear\ holds breath for rickroll
[19:14:57] <SolarNRG> No it is actuallly my video
[19:15:00] <ssi> heheh
[19:15:04] <ssi> it is indeed
[19:15:09] <ssi> but it looks like a TERRIBLE idea :)
[19:16:11] <ssi> "run like fuck"
[19:16:20] <SolarNRG> Yeah safety first
[19:16:49] <Jymmm> You coulda just plugged n the ext cord as the plug =)
[19:16:56] <Jymmm> s/as/at/
[19:18:01] <Thetawaves> hahaha
[19:18:03] <JustinXJS> that is funny, a friend and I did that a few years ago
[19:18:32] <SolarNRG> Thank you
[19:18:33] * Jymmm waits for a kid to walk thru the back yard
[19:20:05] <Jymmm> I like how the microwaves is aimed at the door you're behind "protecting you"!
[19:20:10] <Jymmm> -s
[19:20:20] <CareBear\> can't see the show otherwise
[19:20:33] <SolarNRG> its for safety
[19:20:40] <CareBear\> entertainment first, safety second
[19:20:42] * flyback smacks CareBear\ around
[19:20:45] <ssi> there's a whole lot of nothing happening
[19:20:52] <SolarNRG> that;s the point
[19:20:55] <SolarNRG> no sparks!
[19:21:00] <flyback> why don't you just make a induction furnace
[19:21:02] <flyback> stupid canuck
[19:21:02] <ssi> I expected more from this 15 minutes of my life!
[19:21:04] <flyback> stupid canuck
[19:21:11] <Jymmm> SolarNRG: aiming the oven at the door is for safety?
[19:21:15] <SolarNRG> yeah
[19:21:19] <SolarNRG> so i can see
[19:21:33] <SolarNRG> but the glass of the door prevents a microwave blowing up in my face
[19:21:40] <flyback> Solar
[19:21:42] <Jymmm> Is that like aiming the cannon at the door for safety too?
[19:21:43] <CareBear\> SolarNRG : another option would be to mount a cheap camera very close to the oven
[19:21:47] <flyback> you can make a gas furnace cheap too
[19:21:50] <flyback> DO NOT
[19:21:56] <flyback> use a hair dryer for the turbocharger
[19:22:08] <flyback> they either have a gay little toy dc motor with metal strips for brushes
[19:22:20] <flyback> or if you buy professional hair dryers like I do which I still manage to burn out in a year $50
[19:22:29] <flyback> they have larger brushes you could replace if you can find them
[19:22:34] <ssi> get a bbq fan
[19:22:37] <ssi> like a pit viper or something
[19:22:37] <flyback> they use a universal motor that was designed with many uses
[19:22:41] <Jymmm> leaf blower
[19:22:44] <SolarNRG> its sorted
[19:22:44] <flyback> so some of those uses would want a brush change
[19:22:48] <SolarNRG> ive bought a crucible
[19:22:57] <flyback> but a normal hair dryer is not going to last long
[19:23:17] <ssi> anyone a wizard at diagnosing 3phase motor problems?
[19:23:19] <SolarNRG> I'm trying again next weekend
[19:23:26] <flyback> i'm on my 5th here in 6 yrs
[19:23:37] <flyback> sadly I could have fixed almost all of them but ran out of time
[19:23:46] <SolarNRG> Anyway, thanks for your help guys, I'm off 2 bed
[19:23:46] <flyback> this was before I was on ebay though and could look for brushes
[19:23:51] <SolarNRG> Thank you all and good night
[19:24:09] <flyback> I kill them drying out pc parts and circuit boards
[19:24:14] <flyback> not drying my hair :P
[19:24:41] * Jymmm wants his 14 minutes back damnit!
[19:25:20] <ssi> Jymmm: SRSLY
[19:27:02] <Thetawaves> flyback, what about a 4" duct fan
[19:27:18] <Thetawaves> i've got a 12v 4" duct fan sitting right here
[19:27:29] <Thetawaves> i was thinking it would be too powerful
[19:27:31] * Tom_itx gives Jymmm a time machine
[19:27:41] <JustinXJS> I am with Jymmm, leaf blower works really well
[19:27:48] <Jymmm> Tom_itx: Y Flux capacitor included?
[19:27:49] <flyback> I bet my furnace draft blower I saved from the old furnace would be awesome for this
[19:27:52] <flyback> but I use it as a sleep machine
[19:28:00] <flyback> and the bearings are close to eol it's almost 20 yrs old
[19:28:06] <flyback> I dunno what I am going to do :/
[19:28:07] <Tom_itx> no, it was removed for safety in shipping
[19:28:13] <flyback> I think I can get the top off and change that one
[19:28:24] <flyback> but they sealed the plastic housing around the centrif blower for life
[19:29:08] <Jymmm> ssi: I'm still working on the lamp shade formula =)
[19:29:18] <ssi> Jymmm: heheh get it done
[19:29:24] <ssi> my coolant pump is being an ass
[19:29:35] <Thetawaves> i just need to get an inline gasoline pump to get a gasoline fired burner made :D
[19:29:37] <ssi> it won't run unless you give it a nudge as you're bringing the frequency up
[19:30:09] <Jymmm> ssi: bring it up to 90%, then back it off
[19:30:19] <ssi> ?
[19:30:23] <flyback> ssi sounds like a bad cap
[19:30:30] <flyback> or your vfd profile is CANUCKED
[19:30:51] <ssi> don't think it has caps!
[19:30:55] <ssi> it's 3ph
[19:30:58] <flyback> ah yeah
[19:31:05] <flyback> your VFD profile sucks ass then
[19:31:08] <flyback> see if you can change it
[19:31:16] <ssi> I've changed it every way I can think of
[19:31:21] <ssi> I think it probably has an open phase
[19:31:22] <flyback> mabye a high freq blitz for 1/10 of a second just to get it turning
[19:31:32] <flyback> then ick
[19:31:38] <flyback> i'd get the dmm now and start checking
[19:31:43] <flyback> before it turns into more problems
[19:31:45] <ssi> yea that's my next step
[19:32:06] <ssi> I'm just not sure if I can just measure from line to line
[19:32:09] <flyback> it might be savable if it doesn't do a statue of liberty torch impersination and warp the rotor etc
[19:32:13] <ssi> or if I need to break the coils apart and measure them independently
[19:32:20] <flyback> feel free to ask in ##electronics
[19:34:38] <JustinXJS> If it is a 3 wire motor you can test for a bad phase, just measure between all 3 combinations of wires (AB, AC, BC)
[21:34:14] * flyback waits for sms messages from his phone to tell him that the work ups's have reported a power failure from the storm (sometimes I am sorry I set this up)
