#emc | Logs for 2005-10-22

[00:05:31] <Jacky^> :)
[00:05:44] <Jacky^> CIA-2: ciaoo
[00:05:47] <Jacky^> :P
[00:06:03] <Jacky^> ! CIA-2
[00:06:15] <Jacky^> CIA-2: thanks
[00:06:29] <Jacky^> .google CIA-2
[00:06:33] <Jacky^> :/
[00:06:38] <Jacky^> nothing to do ..
[01:13:11] <Jacky^> :)
[01:15:20] <Jacky^> thanks Phenny_cnc
[01:15:54] <Jacky^> Phenny_cnc: hey ?
[01:15:58] <Jacky^> are U ?
[01:16:22] <Jacky^> thank you Phenny_cnc
[01:16:45] <Jacky^> ouch
[01:16:49] <Jacky^> :/
[01:16:57] <Jacky^> :)
[02:07:26] <Jymmm> les_w you around?
[02:07:44] <les_w> yeah talking to yuga
[02:07:54] <les_w> on skype
[02:12:58] <Jacky^> :)
[02:13:33] <Jacky^> * Jacky^ is on paltalk
[02:22:09] <Jacky^> * Jacky^ ws in pal
[02:22:20] <Jacky^> * Jacky^ goes to bed
[02:22:23] <Jacky^> :)
[02:22:27] <Jacky^> ciaooo
[02:22:33] <Jacky^> see tomorrow
[02:22:43] <Jacky^> Jacky^ is now known as Jacky^afk
[03:10:07] <Jacky^afk> anyone at home ?
[03:10:20] <Jacky^afk> Jacky^afk is now known as Jacky^
[03:11:08] <Jacky^> les_w: you around ?
[03:20:10] <Jacky^> Jacky^ is now known as Jacky^afk
[03:20:38] <Jacky^afk> night
[03:36:45] <les_w> I am still around
[03:36:59] <les_w> for a little while anyway
[03:54:20] <pfred1> <furrywolf> and, as I said, emc only supports steppers.
[03:54:27] <pfred1> say it isn't so!
[04:04:44] <djb_rh> The motion control can operate true servos (usually analog) with the feedback loop closed by the EMC software at the computer, or open loop with "step-servos" or stepper motors.
[04:05:28] <pfred1> djb_rh like running the servos with some special gecko drive?
[04:05:35] <djb_rh> no idea
[04:05:39] <djb_rh> I just cut and pasted that
[04:05:41] <pfred1> I thik i saw a servo to stepper driver on their site once
[04:05:44] <djb_rh> http://emc.sourceforge.net/cgi-bin/emcinfo.pl?About_EMC
[04:05:46] <djb_rh> :)
[04:06:04] <djb_rh> I've never used EMC or a mill or lathe
[04:06:11] <djb_rh> I'm just reading up and popped in here
[04:06:26] <pfred1> me either yet still a bit hung up with how i am going to power it all over here
[04:06:30] <djb_rh> I was currently reading the tutorial on Sherline's page
[04:06:48] <pfred1> 24V at 10 amps isn't that easy for me to come by it seems
[04:07:12] <djb_rh> eBay?
[04:07:23] <djb_rh> I have a friend that bought a 28V/10A that way
[04:07:29] <djb_rh> dunno what you'd search for exactly
[04:07:34] <pfred1> I've nad less than great luck with ripbay lately
[04:07:43] <djb_rh> since "power supply" is probably gonna be too generic
[04:07:52] <pfred1> it's like nothing but a sea of scammers anymore it seems
[04:08:07] <djb_rh> well, I'm doubting for something like *this* you'd find many scammers
[04:08:08] <pfred1> nothing like it used to be
[04:08:19] <pfred1> pfft if there's a buck to be made
[04:08:33] <djb_rh> I've never found it *that* hard to weed the wheat from the chaff
[04:08:39] <pfred1> well i almost had it with a rewound microwave transformer
[04:08:46] <pfred1> but the primary turned out to be shot on that unit
[04:09:07] <pfred1> it's really really easy to rewind microwave transformers and get some serious power out of them
[04:09:26] <pfred1> but you're not goind to rewind the primary
[04:10:07] <djb_rh> sounds a little too involved for me
[04:10:21] <pfred1> all it takes is a hammer and a chisel
[04:10:26] <pfred1> can you count to maybe 20?
[04:10:44] <pfred1> that's al lthe skill it really takes :)
[04:11:08] <pfred1> bash off the old secondary get as much 12 gauge wire as yo ucan in there
[04:11:09] <djb_rh> well, I understand that
[04:11:47] <pfred1> I used a pneumatic whizzer to great effect on the original secondary
[04:12:07] <pfred1> what that didn't get the hammer and chisel did
[04:12:21] <pfred1> and while it worked it kicked ass i thought i had it!
[04:12:37] <pfred1> cept the unit had a weakened primary when I started i believe
[04:12:57] <djb_rh> http://cgi.ebay.com/ebaymotors/NEW-LAMBDA-INDUSTRIAL-SWITCHING-POWER-SUPPLY-24V-10A_W0QQcmdZViewItemQQcategoryZ6755QQitemZ4583638491QQrdZ1QQsspagenameZWDVW
[04:13:00] <djb_rh> no bids
[04:13:04] <djb_rh> starting at $99
[04:13:09] <djb_rh> if that's a scammer, I'll eat my shorts
[04:13:26] <pfred1> yeah I'm not going to lay out no >$100
[04:13:34] <pfred1> for osmeone else's garbage
[04:13:36] <djb_rh> ah
[04:13:46] <pfred1> any asshole can spend money
[04:14:00] <djb_rh> I suggest a hammer and chisel, then.
[04:14:02] <djb_rh> * djb_rh ducks
[04:14:06] <pfred1> * pfred1 wants to be a special asshole though his mommy told him he was way back when
[04:14:31] <pfred1> yeah another beat microwave and I'm in
[04:14:34] <djb_rh> hey, whatever floats your boat
[04:14:57] <pfred1> I don't get a whole lot of satisfaction from the act of simply spending money
[04:15:06] <djb_rh> understood
[04:15:30] <djb_rh> sometimes I'd rather just have a working tool rather than a tool I built and probably screwed up somewhere. :)
[04:15:32] <pfred1> I mean if that were the case i could blow down 4 grand and get a turnkey CNC system
[04:15:46] <djb_rh> yeah
[04:16:00] <pfred1> but where's the satisfaction in that?
[04:16:20] <djb_rh> the friend of mine that "built" his got an equiv CNC system for about $1000, and that included the eBay route for the power supply and steppers
[04:16:34] <pfred1> somethings i can justify spending money on like motors I'm not about to make my own motors just yet
[04:16:47] <djb_rh> yeah, i was going to ask what parts you were going to make your motors out of
[04:16:48] <djb_rh> :)
[04:17:00] <pfred1> someday i may
[04:17:15] <djb_rh> so, getting back to your original question, what would you use instead of steppers?
[04:17:26] <pfred1> they are fairly close tolerance precision devices though
[04:17:43] <pfred1> well seems like anyone that does CNC for a while moves on to servos
[04:17:50] <pfred1> steppers got squat for real speed
[04:17:58] <pfred1> seems like in the end it's all about speed
[04:18:11] <djb_rh> for production, I guess
[04:18:22] <djb_rh> but for the hobbyist, it may not be as important
[04:18:26] <djb_rh> unless I'm missing something
[04:18:29] <pfred1> no sitting there for 10 hours wiaiting for something I think would get to anyone after a bit
[04:18:41] <pfred1> watching your spindle burn out
[04:18:51] <pfred1> IPS
[04:19:50] <pfred1> with the bit I've messed with steppers here it seems like i can get them to run so so but to get them to reliably reverse i have to really slow the pulse train down on them
[04:20:11] <pfred1> it's something I will have to work on some more i suppose
[04:22:41] <djb_rh> so I take it you are doing your own controller?
[04:22:50] <pfred1> did
[04:22:52] <pfred1> driver
[04:22:52] <djb_rh> well, of course you are
[04:22:58] <djb_rh> what am I saying?
[04:23:06] <pfred1> yeah that's been done for quite some time now wans't too hard
[04:23:13] <pfred1> I did PWM
[04:23:17] <djb_rh> anyone who would be making their own power supply would be... :-)
[04:23:46] <pfred1> yes I've been doing electronics for about 30 years now so wasn't too difficult for me
[04:23:56] <djb_rh> oh, they aren't bad
[04:24:08] <pfred1> i used some allegro module
[04:24:13] <djb_rh> I am barely 30 years old, but I'm sure I could build one, too
[04:24:17] <pfred1> was like $7
[04:25:18] <djb_rh> I'm a little different in that I can afford to buy an off the shelf package and would prefer to spend that time learning about machining, though. Not that there's anything wrong with either approach...
[04:25:30] <djb_rh> mine will just cost a lot more. :)
[04:25:45] <pfred1> I still got my IAM book laying around someplace here
[04:25:48] <icee> I'm building my own power supply/drive ;)
[04:26:00] <djb_rh> I'm just trying to learn a little bit about the machining side so I can make some intelligent purchase decisions.
[04:26:12] <icee> I have a nice working xylotex setup running steppers now.
[04:26:17] <pfred1> djb_rh mo bigger mo betterer
[04:26:17] <djb_rh> that and learn enough to even be able to evaluate what I want in a mill.
[04:26:18] <icee> Servos, imo, are the only way to go for a big mill.
[04:26:30] <pfred1> that's the golden rule in machining
[04:26:39] <djb_rh> yeah, but I've got no need for a big mill
[04:26:44] <pfred1> because even a huge machine can do small jobs
[04:26:57] <icee> djb: well, what kind of work are you looking to do?
[04:26:57] <djb_rh> I don't want to devote the space nor deal with the power requirements, though.
[04:27:04] <pfred1> I ran this hting called a milwaukee no 3
[04:27:10] <pfred1> looked like the prow of a battleship
[04:27:15] <icee> a lot of the smaller machines are rigidity limited
[04:27:19] <djb_rh> oh, mostly plastic and aluminum parts that are pretty small
[04:27:25] <pfred1> still could do stuff to like .002
[04:27:34] <djb_rh> yeah, I do want a good machine, even if small
[04:27:39] <pfred1> man don't get a sherline toy
[04:27:52] <djb_rh> technically I do own a mill already
[04:27:52] <pfred1> you'll learn nothing!
[04:28:06] <djb_rh> pfred1: oh? why not?
[04:28:07] <pfred1> cept that little things are underpowered
[04:28:13] <djb_rh> heh
[04:28:30] <pfred1> and have severely limited range
[04:28:33] <icee> the sherlines are fine, for small parts, occasional duty, by a very patient machinist
[04:28:40] <pfred1> yup
[04:28:45] <pfred1> they take cuts like a geinder can
[04:28:49] <pfred1> grinder even
[04:28:56] <pfred1> woo lets shave off .002
[04:29:07] <pfred1> whoah hold it down from the chatter!
[04:29:08] <djb_rh> I own a Bridgeport CNC mill, but the CNC stuff is hella-old and basically needs to be replaced. I'm not interested in learning *that* much about building mills.
[04:29:20] <pfred1> djb_rh yeah they're rebuildable
[04:29:27] <icee> djb: yah, convert it.
[04:29:33] <pfred1> I'd sooner rebuild that than throw away money on a sherline
[04:29:49] <pfred1> in the end you'd actually have a real machine tool not a toy
[04:30:07] <djb_rh> I was hoping to find someone who could do the upfit for me, but that seems hard to find.
[04:30:16] <pfred1> people do it
[04:30:33] <pfred1> I've seen pages where people have redone the CNC 1 i believe they are
[04:30:35] <djb_rh> there are some pro outfits out there, but you're talking about paying some insane travel costs and shipping to get the person and parts to your place yada yada yada
[04:30:45] <icee> djb: yah.. do it yourself.
[04:31:00] <djb_rh> still, I don't want to be stuck with such a big machine taking up space
[04:31:02] <pfred1> man the hard stuff is done o nthem gibs and ways and whatnot
[04:31:02] <icee> right now there's a heck of a deal on servo motors, and several of us here are collaboratively designing a drive for them
[04:31:11] <pfred1> the rest is strip off and bolt on
[04:31:20] <pfred1> get greasy and just do it!
[04:31:30] <djb_rh> stupid thing needs three phase power!
[04:31:40] <pfred1> so?
[04:31:42] <djb_rh> I have a phase converter from another project, though
[04:31:51] <pfred1> they make inverters for that today
[04:32:04] <icee> djb: yah, three phase is easy, whether you use another motor to do it or an inverter
[04:32:05] <pfred1> don't evne gotta play with the old moter crap anymore
[04:32:17] <pfred1> nah the other motor deal is a thing of the past now
[04:32:18] <djb_rh> wow, I didn't know that
[04:32:25] <djb_rh> already have the motor, though. :)
[04:32:29] <pfred1> today it's all inverters for it
[04:32:43] <icee> pf: well, the 'old motor crap' spins up three phase motors faster than all of the reasonably priced inverters
[04:32:54] <pfred1> pfft
[04:33:01] <pfred1> once it spins up run it all day
[04:33:06] <djb_rh> I wonder if the mill I have has servos or steppers now
[04:33:09] <pfred1> that's what it's built to do
[04:33:21] <pfred1> it'll run day in day out for 2 lifetimes
[04:33:40] <pfred1> that's why bridgeport went out of business
[04:34:10] <pfred1> we'd fire them puppies up i nthe morning stop them when it was time to go home
[04:34:19] <icee> http://www.surpluscenter.com/item.asp?UID=2005102121491658&catname=&item=10-2207
[04:34:40] <icee> djb: get 3-4 of those and build the drive we're building
[04:34:50] <pfred1> icee nice!
[04:34:56] <icee> it'll be cheep cheep and get much better performance than any little cnc machine you get
[04:35:10] <icee> pf: they had 1kW ones, too, which I bought 6 of; now they're sold out :P
[04:35:16] <pfred1> yeah sherlines are like HO model trains not like a real locomotive by any stretch
[04:35:45] <pfred1> the bridgeport will jold close tolerances
[04:35:47] <icee> djb: the only real question is.. is the mill you have in good enough shape to make the parts you need for the conversion
[04:36:01] <pfred1> heck in the long run the bridgeport will be cheaper to run because yo ucan use real tooling on it
[04:36:09] <pfred1> not dinky little bits
[04:36:11] <icee> though depending on what you have you may be able to find bolt on parts
[04:36:31] <icee> pf: well, plus you can bury cuts deep
[04:36:36] <icee> so you're not just always wearing out the end of the mill
[04:36:39] <pfred1> well yeah
[04:36:56] <pfred1> it'll run like an inch and a half hogger
[04:37:12] <pfred1> eat steel like it's farina
[04:37:49] <pfred1> I'd like to see how a sherline would run with a 3/4 endmill in it
[04:38:36] <djb_rh> hmm, can't find the model of the CNC stuff on my Bridgeport right now
[04:38:53] <djb_rh> I just remember it has a monochrome computer of some type with 8" floppy disks
[04:39:05] <icee> djb: so it's a bridgeport series 1 someone else converted?
[04:39:09] <pfred1> I htink they were called CNC 1s
[04:39:15] <djb_rh> I think so
[04:39:17] <pfred1> the ones that no one knows how they work today
[04:39:29] <djb_rh> it supposedly worked when it was removed from service
[04:39:30] <icee> djb: well, that is what you want to use. I strongly recommend it :P
[04:39:31] <pfred1> and their control boxes are just rats nests anymore
[04:39:36] <djb_rh> I paid $1500 for it
[04:39:40] <pfred1> yeah cept no one quite knew how it worked
[04:39:43] <pfred1> I know all about them
[04:39:56] <pfred1> every machine shop has one collecting dust
[04:40:04] <djb_rh> hmm
[04:40:17] <pfred1> usually yo ucan get them for just hauling them away
[04:40:25] <djb_rh> *shrug*
[04:40:27] <pfred1> they take up valuable floorspace and all
[04:40:28] <djb_rh> should have negotiated harder
[04:40:29] <djb_rh> :)
[04:40:45] <pfred1> yeah the clincher is yes show me how it works
[04:40:51] <pfred1> they usually get pretty embarassed then
[04:40:58] <pfred1> being as they don't evne know how to make it come on
[04:41:14] <djb_rh> so there isn't much chance that the servo/steppers on what I have are usable?
[04:41:16] <pfred1> then out comes the binder
[04:41:24] <icee> djb: hey, they might be.
[04:41:30] <pfred1> I believe they are steppers
[04:41:39] <pfred1> them big motors with the fins on the axises
[04:41:42] <djb_rh> even if so, that would be all that I'd want to keep of the existing CNC stuff, right?
[04:41:51] <icee> djb: yah.
[04:41:52] <djb_rh> just start ripping off the controller and wiring
[04:41:54] <pfred1> nah i hear they even suck
[04:41:56] <djb_rh> and start over
[04:42:10] <pfred1> but the basic frame of the machien is awesome
[04:42:16] <icee> djb: seriously, you should join our project. I don't expect it's going to take all that much to build a working drive
[04:42:19] <pfred1> like none better
[04:42:34] <djb_rh> I'm not sure how much help I'd be
[04:42:47] <djb_rh> seeing as the most complex thing I've ever done machining wise is use a drill press
[04:43:04] <icee> well, this is just electronics.
[04:43:04] <pfred1> drilling is the most important machining task
[04:43:06] <djb_rh> ooh, and a cold-cut saw
[04:43:20] <djb_rh> oh, electronics I can do
[04:43:27] <djb_rh> and these computer things
[04:43:31] <icee> i'm designing a board around a dsPIC to decode quadrature from encoders and to drive PWM and control current in a closed loop
[04:43:48] <djb_rh> ircee: so I assume this driver would be run by EMC?
[04:43:51] <icee> that speaks the IEEE 1284 protocol and pretends to be a pico systems board (so EMC will already work with it) over the parallel port
[04:43:52] <pfred1> I'm pretty good at designing PCBs
[04:43:58] <djb_rh> ah
[04:44:01] <djb_rh> okay
[04:44:25] <icee> And then it's just taking that PWM and amplifying it; so rectifying AC power, and an appropriate fet driver and fets.
[04:44:30] <djb_rh> not sure that I'd be much help with the driver
[04:44:34] <icee> (in my case, IGBTs because i'm 1kW)
[04:44:35] <djb_rh> I'm a digital guy
[04:44:41] <pfred1> here's the last one i just made ismple but that's because i wanted it simple
[04:45:05] <icee> well, lawrenceg is designing a driver for the 300W steppers. i'm going simpler
[04:45:08] <icee> er
[04:45:12] <icee> he's going simpler, that is.
[04:45:40] <icee> anyways, we've already bought the motors and are pretty serious about things
[04:45:46] <djb_rh> well, I gotta find out if my buddy sold my mill or not
[04:45:53] <djb_rh> he was trying to get rid of it for me
[04:46:04] <icee> once I have them in hand monday, and I confirm that the encoders work the way I expect, I'll be doing all the design entry
[04:46:08] <pfred1> your bridgeport?
[04:46:16] <djb_rh> pf: yeah
[04:46:23] <pfred1> damn
[04:46:29] <pfred1> oh well
[04:46:32] <icee> djb: well, if you're interested in the big servos, you'll need to hurry about it. 300 of the 1kW ones sold out in 4-5 days, and there's only 70 of the 300W ones left
[04:46:36] <djb_rh> I don't think he's sold it yet
[04:46:44] <djb_rh> icee: okay
[04:46:58] <pfred1> yeah sherlines I mean they're OK if you're in an apartment but they're not real tools
[04:47:32] <djb_rh> icee: anything else I need from that same place? I'm ordering them now.
[04:47:45] <icee> djb: nope, that's it, just the motors
[04:47:46] <icee> well
[04:47:51] <icee> you should decide whether you want the ones with a brake or not.
[04:48:12] <djb_rh> oh, like I have a clue
[04:48:13] <icee> the brake can be a hassle (you'll need a 24V DC supply to keep it 'open')
[04:48:14] <pfred1> can't you make an electric brake?
[04:48:23] <icee> but it's a safety thing
[04:48:35] <djb_rh> oh, so you can have a big "all stop" switch?
[04:48:37] <pfred1> like a regenerative brake?
[04:48:40] <icee> http://www.surpluscenter.com/item.asp?UID=2005102121491658&catname=&item=10-2208 is the one without the brake
[04:48:56] <icee> pf: it's a 24VDC brake; you need to feed 24V to feed the coil that disengages it
[04:49:02] <icee> pf: an actual mechanical brake
[04:49:10] <icee> djb: yah.. that cuts power and thus engages the brakes
[04:49:15] <djb_rh> how much current does the brake need, though?
[04:49:18] <djb_rh> can't imagine it's too much
[04:49:19] <icee> djb: if you do engage the brake, it stops hard and is bad for the machine
[04:49:28] <pfred1> heh
[04:49:36] <icee> djb: i'm not sure on your motor.. one second, let me find the data sheet
[04:49:42] <pfred1> yeah might not want the brake
[04:50:22] <pfred1> djb_rh well for what a sherline costs I'm sure yo ucould go a long way towards gettign the bridgeport retrofitted
[04:50:32] <pfred1> if not all the way if you're careful
[04:50:37] <icee> djb: 70mA to keep it disengaged; about 100-150mA instantaneously to disengage it
[04:50:49] <djb_rh> that's not bad
[04:50:56] <djb_rh> which do you recommend?
[04:51:07] <icee> well, I bought ones with brakes because they're big scary 1kW motors
[04:51:37] <icee> it's up to you. it's just one more thing to integrate against (a 24VDC power supply/connectors)
[04:51:39] <pfred1> icee what are you using your motors on?
[04:51:51] <icee> pf: I still actually need to get a mill for them.
[04:52:07] <icee> pf: though one of them will become an 'addressable spindle' for my mini mill
[04:52:11] <pfred1> icee what kind are you thinking about picking up?
[04:52:20] <icee> pf: something big.
[04:52:21] <djb_rh> * djb_rh buys motors
[04:52:38] <icee> Still divided between whether I want a worn out bridgeport or a brand new piece of chinese crap
[04:52:39] <djb_rh> fuggit, if I got no mill I can always eBay them
[04:52:54] <pfred1> icee heh yeah them chinese ones are pretty hurting
[04:53:17] <pfred1> icee and even if a bridgeport looks worn out well if you shop around I'm sure yo ucan find a fairly tight one
[04:53:41] <icee> pf: Yah, I'm keeping my eyes open when machine shops etc are shutting down
[04:53:46] <pfred1> yup
[04:53:51] <pfred1> or downsizing
[04:53:59] <djb_rh> or upsizing
[04:54:10] <pfred1> guy i know got some nice machines for next to nothing I htink he did get his CNC 1 for nothing
[04:54:25] <icee> i'm in the sf bay area, though.. and there's still a lot of machine shop demand from tech.. so the market for tools isn't as good here as it is elsewhere in the country
[04:54:30] <pfred1> got a lathe with a 12 foot bed for $200
[04:54:35] <djb_rh> I thought my buddy who hooked me up with the mill would want to use it
[04:54:40] <djb_rh> he's got a CNC plasma table
[04:54:44] <djb_rh> and loves running it
[04:54:54] <djb_rh> but he just doesn't seem to care to have a mill for some reason
[04:54:56] <djb_rh> I find that odd
[04:55:05] <pfred1> djb_rh make him jealous and CNC a waterjet
[04:55:08] <djb_rh> but then again, I love tools
[04:55:17] <djb_rh> ooh
[04:55:18] <pfred1> plasma cutters blow next to water jets
[04:55:20] <djb_rh> I'd love a waterjet
[04:55:35] <djb_rh> pfred1: for $20k that plasma cutter is pretty sweet
[04:55:42] <pfred1> pfft
[04:55:48] <pfred1> slob jobs
[04:55:55] <pfred1> waterjets are like magic
[04:55:57] <djb_rh> 1/10 the cheapest waterjets, I believe
[04:56:03] <djb_rh> well sure
[04:56:13] <pfred1> unreal the cuts they make
[04:56:17] <pfred1> I think they're nicer han lasers
[04:56:18] <djb_rh> no doubt
[04:56:40] <pfred1> stuff looks like yo uhand sanded it all day long
[04:56:49] <djb_rh> I've seen references to making your own with hypodermic needles and washing machine pumps. I can't think that would last long, though.
[04:56:58] <djb_rh> the needle, I mean
[04:57:03] <pfred1> ah, no
[04:57:26] <pfred1> cheezewhiz try to avoid it
[04:57:33] <djb_rh> icee: I ordered four of those
[04:58:51] <pfred1> djb_rh so you've never really machined?
[04:59:09] <djb_rh> I thought I made that clear. :)
[04:59:12] <djb_rh> Nada.
[04:59:26] <pfred1> djb_rh well do yo uhave any tooling at all for your mill?
[04:59:29] <pfred1> collets
[04:59:34] <pfred1> a wobbler
[04:59:39] <pfred1> you'll need one of those
[05:00:37] <pfred1> wobblers are cool such a simple idea
[05:00:51] <pfred1> djb_rh_ what'd you miss?
[05:00:52] <djb_rh_> damned dialup
[05:01:07] <pfred1> djb_rh well do you have any tooling at all for your mill?
[05:01:15] <djb_rh_> <pfred1> djb_rh well do yo uhave any tooling at all for your mill?
[05:01:15] <djb_rh_> <djb_rh> nope.
[05:01:15] <djb_rh_> <djb_rh> wouldn't even know where to get it
[05:01:15] <djb_rh_> <djb_rh> it didn't come with anything useful
[05:01:15] <djb_rh_> <djb_rh> ie. everything it did come with was obviously beat to hell
[05:01:19] <pfred1> collets or a wobbler
[05:01:29] <djb_rh_> no idea
[05:01:36] <pfred1> do you know what taper it is?
[05:01:44] <pfred1> because offhand i don't evne know
[05:01:59] <pfred1> you need to figure out it's spindle taper
[05:02:09] <djb_rh_> no idea
[05:02:12] <pfred1> so you can figure out what tools go into it
[05:02:23] <djb_rh_> I know nothing about it other than it's a bridgeport and it was cnc with big 8" floppy disks. :)
[05:02:23] <pfred1> well I'm telling you how to figure it out
[05:02:53] <pfred1> once yo uknow what taper the machine is then you cna buy some tooling for it
[05:03:01] <pfred1> I mean you're goingto need collets
[05:03:10] <pfred1> you have ot be able to hold tools in it
[05:03:37] <pfred1> you're goingto need tee slot setup stuff too
[05:03:53] <pfred1> stuff so you can fix work to the table
[05:04:18] <pfred1> * pfred1 used to make J&S hold down clamps
[05:04:25] <djb_rh_> yeah, but it doesn't do much good to tell me this stuff until I have the unit at my shop and can at least look at it
[05:04:28] <djb_rh_> though I appreciate the info
[05:04:32] <pfred1> I was a tool and die maker not just a plain old machinist :)
[05:05:09] <pfred1> heck I made this one gadget that cost like 6 grand and fit into the palm of your hand
[05:05:33] <pfred1> but you won't need one of those
[05:06:48] <pfred1> djb_rh_ want to se a pic of one?
[05:18:58] <djb_rh_> djb_rh_ is now known as djb_rh
[06:13:34] <CIA-2> 03paul_c * 10emc2-auto/wiki/ (12 files in 7 dirs): "Auto update wiki from a cron job. Sat Oct 22 05:30:01 BST 2005 "
[07:06:40] <CIA-2> 03jmkasunich * 10emc2/configs/ (mazak_rf.hal mazak_rf.ini): more mazak work - Y axis added, spindle drive signals added
[11:35:11] <alex_joni> anonimasu_ morning
[11:35:18] <alex_joni> anonimasu: morning
[11:37:14] <anonimasu> mornin
[11:37:15] <anonimasu> g
[11:39:28] <alex_joni> yeahh.. too early for both of us :)
[11:40:20] <anonimasu> yeahah
[11:41:11] <alex_joni> so.. what should I do today... wondering...
[11:42:25] <anonimasu> * anonimasu wonders too
[11:42:31] <anonimasu> I woke up 4 minutes ago
[12:04:59] <alex_joni> * alex_joni drops a needle
[12:44:32] <alex_joni> cradek: don't suppose you're around..
[12:58:17] <etla> hi
[12:58:23] <alex_joni_> hello
[12:58:44] <alex_joni_> alex_joni_ is now known as alex_joni
[12:59:10] <fenn> kill.. killl!
[12:59:39] <etla> guess who bought some sanyo denki P5s this morinng ? ;)
[12:59:56] <alex_joni> alex_joni has changed the topic to: "Welcome to the Enhanced Machine Control forum - Support and development of a linux based CNC control. | Homepage www.linuxcnc.org | Regular Developer's meetings every Sunday between 14:00 & 18:00 GMT | wiki up @ http://wiki.linuxcnc.org | Six people have ops capabilities - Lucky dip !"
[13:00:07] <alex_joni> etla: I did
[13:00:26] <etla> mee too !
[13:00:27] <fenn> damn i better keep up with the joneses and buy some motors
[13:00:36] <fenn> er, keep up with the joni's :)
[13:00:39] <etla> you know what drive you will use yet ?
[13:00:48] <alex_joni> etla: was kidding
[13:00:54] <etla> oh...
[13:00:54] <alex_joni> get geckos :D
[13:01:06] <etla> yeah, brushless dc ??
[13:01:26] <alex_joni> hmm.. don't think they have for brushless :/
[13:01:44] <etla> and I'm not sure step/dir from parport is really optimal
[13:03:29] <alex_joni> if you want a nice machine.. get a servo card
[13:03:33] <alex_joni> motenc or the like
[13:03:41] <alex_joni> but that's pricey
[13:03:51] <alex_joni> * alex_joni goes away for a while..
[13:03:53] <alex_joni> later guys
[13:04:12] <etla> what about the way jon elson uses the parallell port ?
[13:12:56] <etla> hmm... everyone is gone....
[13:15:04] <fenn> * fenn is seriously considering buying those motors...
[13:15:13] <fenn> is there a datasheet anywhere?
[13:15:58] <etla> for what ?
[13:16:14] <fenn> the sanyo denki p5...
[13:16:53] <etla> there are links in the cnczone thread
[13:16:58] <etla> rutex forum
[13:17:20] <fenn> thanks
[13:17:49] <etla> http://www.cnczone.com/forums/showthread.php?t=13909
[13:18:43] <etla> do you know how to save g92 etc offsets into the var file ?
[13:20:25] <fenn> there was a big hoohah about this on the mailing list and the general consensus is to use g54 et al, instead of g92
[13:20:53] <fenn> it also depends on whether you end your program with % or M05 (?)
[13:22:34] <fenn> etla: have you read this? http://www.linuxcnc.org/Dropbox/g92test1.pdf
[13:23:32] <fenn> might not be that useful though
[13:23:57] <etla> thanks, I found my problem...
[13:24:14] <etla> I'm using axis from bdi 4.30
[13:24:30] <etla> when I quit axis using the cross in the upper right corner it does not save to the var file
[13:24:44] <etla> when I quit using the menu file/quit it works fine
[13:24:52] <fenn> hmm..
[13:28:49] <fenn> i hate cnczone
[13:29:37] <etla> no attachment without registering ? ;)
[13:32:51] <fenn> that, and the over-commercialization of public space
[13:33:15] <fenn> and i hate forums in general
[13:35:42] <etla> ok
[13:37:33] <etla> you convinced about the sanyos yet ?
[13:39:15] <fenn> yeah
[13:39:40] <fenn> i'm kinda glad i missed out on the 1kw's
[13:39:46] <fenn> i think i woulda broke something :)
[13:42:06] <etla> I got the 400Ws
[13:42:16] <etla> should be plenty when geared 1:2 or 1:3
[13:42:54] <fenn> yeah, i'm gonna have 6 of them going at once
[13:42:59] <fenn> should be overkill
[13:43:26] <etla> 6-axis ?
[13:43:30] <fenn> hexapod
[13:43:36] <etla> wow !
[13:43:44] <fenn> well, it's not built yet
[13:43:56] <etla> whats your plan for the actuators ?
[13:43:56] <fenn> er, not even designed yet actually :)
[13:44:06] <fenn> acme threaded rod from bridgeport retrofits
[13:44:30] <fenn> i've already got a pile of those
[13:44:43] <fenn> i'm going to make some desktop-sized versions first
[13:44:54] <fenn> see what works and what doesnt
[13:45:01] <etla> ok
[13:45:26] <etla> I think we'll go traditional 3-axis...
[13:46:17] <anonimasu> damn
[13:46:29] <anonimasu> I with I could have shipped 4 1kw ones
[13:47:15] <etla> I think 400w will be plenty...
[13:47:23] <etla> just have to get/build a drive first
[13:48:31] <etla> there might be plenty of the 1kWs on ebay real soon ??
[13:48:47] <anonimasu> no clue
[13:48:54] <anonimasu> it dosent matter I cant ship them
[13:49:34] <anonimasu> I paid more for the shipping of 3 400w's then for the servos
[13:49:49] <etla> :)
[13:50:36] <anonimasu> they'll be a good upgrade to my machine later..
[13:53:20] <etla> so what about drives ?
[14:00:28] <fenn> three-phase power and a magnetic contactor :)
[14:04:12] <etla> might not be high-tech enough :)
[14:04:15] <anonimasu> heh
[14:04:22] <anonimasu> fenn: can you do positioning with that?
[14:05:44] <fenn> well, you can get 1/3 of a rotation, that's probably good enuf
[14:05:51] <etla> icee was planning on building something based on a dsPIC
[14:06:06] <etla> which would be compatible with the pico-systems parport protocol
[14:06:07] <fenn> yes he certainly sounds like he knows what he's talking about
[14:06:40] <etla> I'd like to help also
[14:07:09] <anonimasu> fenn: I dont know what world you live in but 1/3 of a rotation is crap.
[14:07:12] <anonimasu> :)
[14:08:11] <anonimasu> *grins*
[14:08:24] <fenn> anonimasu: it was a joke..
[14:08:38] <anonimasu> I know ;)
[14:08:49] <anonimasu> * anonimasu is insanely slow today
[14:09:51] <etla> more like sinewave commutation based on both hall sensors and encoder
[14:16:38] <Jacky^afk> Jacky^afk is now known as Jacky^
[14:16:55] <fenn> i've actually got a 550W servo sitting here that i've never touched..
[14:18:15] <etla> BLDC ?
[14:18:27] <fenn> yeah, with an "fa-coder" on it
[14:18:30] <fenn> whatever that is
[14:18:50] <fenn> its a 2000 cpr encoder with a bunch of other crap
[14:19:11] <fenn> no way to test out the motor though
[14:19:45] <fenn> actually i'm not sure it's a bldc, it's got U, V, W wires
[14:22:29] <etla> hey if anyone is interested I've got three Gecko G340's for sale
[14:28:57] <anonimasu> *yawns*
[14:45:11] <fenn> ok i got 7 400 watters
[14:45:25] <fenn> we'll see what happens
[14:45:37] <fenn> fenn is now known as fenn_afk
[14:47:14] <etla> ok
[14:47:29] <etla> any ideas for drives ?
[14:48:53] <fenn_afk> i liked some of icee's ideas
[14:49:19] <fenn_afk> i will probably try to do some kind of home=built thing based on ebayed stuff, and it will blow up :)
[14:51:05] <etla> I'll probably wait for the motors to arrive
[14:51:17] <etla> in the mean time could order some of the dsPICs that icee was proposing
[14:52:35] <etla> do you know where the logs from this channel are ?
[14:52:44] <etla> icee talked about his plans a day or two ago
[14:57:03] <anonimasu> logger_aj: bookmark
[14:57:03] <anonimasu> See http://solaris.cs.utt.ro/irc/irc.freenode.net:6667/emc/2005-10-22#T14-57-03
[15:10:03] <k4ts> hello
[15:10:10] <k4ts> hello Jacky^
[16:07:38] <djb_rh_> djb_rh_ is now known as djb_rh
[16:23:40] <fenn_afk> if you run a servo at a higher voltage, does that decrease the current of back emf?
[16:23:43] <fenn_afk> fenn_afk is now known as fenn
[16:27:27] <rayh> Hi Ben
[16:28:23] <jmkasunich> not sure I understand the question
[16:28:36] <jmkasunich> back EMF is approx. equal to armature voltage
[16:28:50] <jmkasunich> do you mean running it at a higher bus voltage, or armature voltage
[16:30:30] <fenn> um, not quite sure i understand you jmk
[16:30:39] <steves_logging> steves_logging is now known as steve_stallings
[16:30:53] <fenn> how do you run a motor at a higher armature voltage?
[16:31:11] <alex_joni> fast :)
[16:31:20] <fenn> ah
[16:31:36] <dmess> faster... yu mena
[16:31:40] <dmess> mean
[16:31:47] <alex_joni> dmess: oui
[16:32:02] <fenn> so for example if i run these 12V rated servos at ~160VDC
[16:32:15] <dmess> but it could jeopordize service life
[16:32:46] <alex_joni> fenn: with proper cooling
[16:33:05] <dmess> or you'll just let the smoke out
[16:33:53] <jmkasunich> if you put 24V into a 12V motor, and it is unloaded, it will just run twice as fast
[16:34:01] <alex_joni> * alex_joni is mad at trackers
[16:34:30] <fenn> what happens if you're decelerating a motor that's running not very fast?
[16:34:35] <jmkasunich> if twice as fast is too much for bearings and such, then you get smoke
[16:36:30] <jmkasunich> depends on what you are doing to decel it... fixed armature current? fixed armature voltage? shorted armature? (the last is a special case of fixed arm volts)
[16:37:36] <fenn> if i connect the reverse DC voltage to the motor, that's fixed armature voltage right?
[16:37:43] <jmkasunich> yep
[16:38:10] <jmkasunich> the problem with fixed voltage (any voltage, forward or reverse) is that current is totally determined by load and motor characteristics
[16:38:16] <jmkasunich> and current is what burns up motora
[16:38:44] <jmkasunich> stall it, and you get current = Varm / Rarm
[16:39:16] <jmkasunich> for any moderately powerfull motor and Varm near the max rating, Rarm is small enough that the stall current will cook it
[16:39:31] <jmkasunich> servo amps directly control the current, so even at stall you only get a limited amount
[16:39:54] <alex_joni> right, even the "simpler" ones do current limiting
[16:40:28] <etla> I guess you need to limit the time that max current is applied also ?
[16:41:13] <jmkasunich> etla: yeah, most motors have peak and continuous ratings, peak can be several times higher than cont
[16:41:33] <fenn> so if i make a current-limiting driver, is there anything really bad about running 10-15x the rated voltage?
[16:41:52] <etla> you get real good acc.
[16:42:16] <etla> components will be more expensive
[16:42:32] <jmkasunich> you only get real good accel because the high voltage lets you push a lot of current thru the motor
[16:42:47] <fenn> well i'm looking at igbts on ebay, they seem real cheap and i want to use them for what they're worth
[16:42:48] <alex_joni> usually on the servos I worked with you have 2 tunings (I2t and instant current)
[16:43:19] <etla> fenn: what kind of driver are you building ?
[16:43:19] <jmkasunich> igbt's are cheap, but only a small part of a drive
[16:43:43] <jmkasunich> igbts are also mostly for higher voltages (over 240VAC/300VDC)
[16:43:57] <alex_joni> * alex_joni is gone
[16:43:59] <alex_joni> later guys
[16:44:05] <rayh> see you.
[16:44:09] <fenn> etla: a simple one, probably pwm, to drive 6 100w brushed 12v dc servos
[16:44:20] <jmkasunich> MOSFETs are much better under 100V or even 200V
[16:44:37] <jmkasunich> trust me, you don't need (or want) IGBTs for a 12V drive
[16:45:29] <etla> microchip had an application note on IGBTs vs mosfets
[16:45:43] <etla> their app note on brushed dc motor control is probably also worth a read
[16:47:32] <jmkasunich> yeah - if you are gonna build a motor drive, you really need to grok motor behavior, current loops, etc, in more depth than can be explained over IRC
[16:47:45] <fenn> any suggestions on what to read?
[16:48:10] <fenn> btw i bought some of those surplus center motors :P
[16:48:14] <jmkasunich> start with the microchip app notes, google for info on DC motor control, PWM choppers, etc
[16:48:29] <etla> fenn: good boy !
[16:48:44] <jmkasunich> brushelss DC, aren't they? That is a whole 'nother ball of worms - makes brush DC drives look simple
[16:49:01] <etla> I'd like to talk to icee, he was talking about a dsPIC based BLDC for the surpluscenter sanyos
[16:49:34] <etla> someone was saying on cnczone that you absolutely need a sinusoidal drive for these...
[16:49:53] <etla> so take both hall sensor inputs and encoder value to determine three phases
[16:51:18] <steve_stallings> As I understand it the penalty for non-sine drive of a BLDC motor is extra motor heating and torque ripple.
[16:51:34] <jmkasunich> that's about what I'd expect
[16:52:06] <etla> the non-sine trapezoidal drive is described in some detail as a microchip app note
[16:52:14] <etla> should not be that hard to build
[16:52:36] <etla> icee was thinking of using the jon elson/pico interface to emc since there is already code for that
[16:52:54] <jmkasunich> to me, the trap vs. sine difference is the distinction between brushless DC (trap) and permanent magnet AC or AC servo (sine) motors
[16:53:06] <jmkasunich> dunno if that matches industry's definition or not
[16:53:42] <fenn> when they say a sinusoidal drive, does it mean sinusoidal current or voltage?
[16:54:06] <etla> the motor winding is linear after all so both will be sinusoidal
[16:54:28] <jmkasunich> it means the counteremf is sinusoidal
[16:55:03] <fenn> * fenn goes back to school
[16:55:08] <jmkasunich> the current comes from the difference between the CEMF and the applied voltage
[16:55:31] <jmkasunich> so if you apply trap voltage to a sine motor (or vice versa), the current waveform will be all fscked up
[16:55:43] <fenn> "back emf" means back electromotive force, right?
[16:55:47] <jmkasunich> yes
[16:56:24] <fenn> which is basically saying voltage, right?
[16:56:30] <jmkasunich> with a permanent magnet motor (all of the ones we are discussing), the CEMF is the voltage that the motor would generate if you spun it (like a generator)
[16:56:50] <fenn> ok
[16:57:04] <jmkasunich> that "generator" voltage is always there whenever it spins, whether it is actually functioning as a motor or a generator
[16:57:27] <steve_stallings> .... and it there were zero torque load, the motor would spin at an RPM such that the counter EMF matched the applied voltage
[16:57:34] <steve_stallings> it/if
[16:57:45] <jmkasunich> the difference between CEMF and applied voltage causes current to flow
[16:58:17] <jmkasunich> if applied voltage is more than CEMF, current flows in, and it is a motor - it consumes power and delivers torque
[16:58:35] <etla> so how do we know if the sanyo denki P5s want trapezoidal or sine drive ?
[16:58:52] <jmkasunich> if applied voltage is less than CEMF, current flows out, and it is a generator - it consumes torque and produces electricity
[16:59:21] <steve_stallings> The motors for trap and sine should be the same inside, just the ratings and performance vary based on how you drive them.
[16:59:40] <jmkasunich> two ways: read the data sheet (if you can find it), or spin the motor (use a drillpress or lathe, chuck onto the shaft, or just spin by hand) and put an o'scope on the terminals
[17:00:09] <fenn> there are motors that actually generate a trapezoidal waveform?
[17:00:29] <etla> OK, will have to wait until they arrive... there are datasheets out there but I don't recall reading about this
[17:00:42] <jmkasunich> steve: not quite true - trap and sine motors have either different distributions of windings in the slots, or different arrangements of rotor magnets, so the flux in the airgap and thus the CEMF is either trap or sine
[17:01:10] <steve_stallings> OK, guess I am stuck in the past.... 8-)
[17:01:15] <fenn> that's nifty
[17:01:32] <etla> sinewave generation is done by looking at both the halls and the encoder, right ?
[17:01:44] <jmkasunich> I could go on and on about motors and drives (that's my day job after all), but I really should get to work
[17:01:59] <fenn> yeah i'll still be here :)
[17:02:10] <steve_stallings> and you are doing great work on the Mazak, congratulations!!
[17:02:11] <jmkasunich> etla: that's one way to do it
[17:02:29] <fenn> * fenn shoots an ak-47 in the air to celebrate
[17:02:46] <jmkasunich> cheap drives might just attempt to interpplate between the HALLs, especially if the end application doesn't need an encoder
[17:02:59] <jmkasunich> steve: did you see the log page on the wiki?
[17:03:11] <steve_stallings> yes, and thus my comment
[17:03:33] <etla> ok, I need to go also, hope to catch icee later and talk about dsPIC -> surpluscenter sanyo drive
[17:04:12] <steve_stallings> etla: there are some drivers that do a 6 step drive that is slightly better than trap using only hall sensors
[17:05:05] <etla> steve: yes, I read about that. from the post on cnczone it looks like the sanyos might need sine driving
[17:05:15] <etla> hope to be able to do that with a dsPIC
[17:05:55] <steve_stallings> lots of folks in the same boat, time for an open-source project??
[17:07:08] <jmkasunich> might be fun
[17:07:27] <jmkasunich> there is an open source project for brush motor control, might want to work with those folks
[17:07:35] <jmkasunich> yahoo group "osmc"
[17:11:23] <fenn> maybe we should make a new yahoo group just to talk about p5's
[17:11:45] <fenn> or better yet, not a yahoo group
[17:11:50] <fenn> anything but a yahoo group :)
[17:12:00] <etla> google is even worse ;)
[17:12:04] <etla> the user interface
[17:12:15] <fenn> google is just a glossy web interface for usenet
[17:12:21] <etla> a wiki page with info would be nice
[17:12:35] <fenn> that could be arranged pretty easily
[17:12:37] <etla> google groups are not usenet groups afaik
[17:12:54] <etla> ok I need to go, back in a few hours
[17:13:06] <etla> if you catch icee here he is very much thinking along the same lines!
[17:27:56] <rayh> steve_stallings: I tried the upload item on the wiki menu bar. It said it uploaded and returned me a strange link which I tried to use on a wiki page. No success.
[17:28:33] <rayh> I suspect that there are some sort of routing issues to be worked out in order to make that part work.
[17:38:06] <steve_stallings> rayh: I have not tried uploads. Will have a look.
[17:50:51] <steve_stallings> rayh: Yep, its broke. Got to get on ssh and try to see where the uploads actually go and if we can fix the wiki configuration to handle it.
[17:57:49] <rayh> and "she" was so good looking!
[18:00:29] <rayh> Thanks again steve_stallings for all your webmaster work.
[18:05:58] <Jymmm> Mornin Folks!
[18:06:34] <fenn> one of the surpluscenter motors is on ebay for $150
[18:06:49] <fenn> but they're not even sold out yet
[18:06:49] <steve_stallings> suspect you will see more 8-)
[18:07:36] <jepler> think they'll sell at $150? If so, I should take a trip to supluscenter .. it's just down the road.
[18:08:01] <fenn> price new is like $1200
[18:08:15] <fenn> market will probably be saturated though
[18:08:32] <Jymmm> jepler: even if you solde them for $80, would you be willing to buy them for yourself?
[18:09:08] <jepler> Jymmm: I don't need anything remotely that powerful, considering that my milling interests don't go much beyond circuit boards.
[18:10:28] <Jymmm> jepler: Well, put up an auction with all the bells and whistles describing them and they usage as well as how they can be used. somebody is sure to buy
[18:11:12] <Jymmm> You ever see those one-liner auctions?
[18:11:25] <Jymmm> jepler ^^^
[18:12:19] <jepler> I've seen a lot of auctions that underdescribe the item being sold
[18:12:22] <jepler> is that what you mean?
[18:12:28] <Jymmm> exactly
[18:13:17] <Jymmm> describe it's specs, say it works with geckos (or whatever) great on BP, etc. ppl perk up =)
[18:13:37] <Jymmm> "A fool and his money are soon parted"
[18:13:59] <Jymmm> Just like me in about 90 minutes
[18:14:08] <jepler> is it true that they work with geckos?
[18:14:27] <Jymmm> Tapmatic X30 #0-1/4" tapping head for $175
[18:15:10] <Jymmm> jepler: Oh, I don't know, just an example. Maybe find and say what it does work with.
[18:15:19] <fenn> supposedly they work with rutex
[18:15:24] <Jymmm> there ya go
[18:16:13] <Jymmm> jepler you can search ebay for closed/completed auctions too
[18:22:52] <Jymmm> Heh, this is a cool concept.... http://experiencethis.us/et/index.php
[18:25:14] <dmess> where' the hang & para gliding??
[18:25:24] <dmess> kite-surfin
[18:25:49] <dmess> base jumpin' ??
[18:29:14] <steve_stallings> steve_stallings is now known as steves_logging
[18:30:21] <jepler> the gnome desktop has gotten better since the last time I seriously tried using it.. circa redhat 9, I think
[18:42:42] <Jacky^_> Jacky^_ is now known as Jacky^
[18:43:51] <Jacky^> storm :/
[18:44:26] <etla> where ?
[18:47:17] <Jacky^> It
[19:42:01] <etla> very quiet...
[20:07:23] <anonimasu> hello
[20:14:13] <etla> hi :)
[20:17:54] <djb_rh> yo
[20:18:39] <etla> hi
[20:20:01] <djb_rh> http://forums.donniebarnes.com has a forum for discussion of icee and lawrenceg's servo driver for EMC1
[20:20:12] <etla> cool
[20:20:32] <djb_rh> you gotta register to post, and I have approve any registrations
[20:20:46] <djb_rh> but anyone here is welcome to register
[20:21:00] <etla> let's do that.
[20:21:03] <icee> hey ;)
[20:21:10] <icee> it's just a placeholder to put up data sheets, etc
[20:21:12] <djb_rh> an email to djb@donniebarnes.com letting me know (unless you tell me here) is a good idea
[20:21:34] <djb_rh> yeah
[20:21:49] <icee> we found a nice part for driving the servos BTW.. irams10up60b for the small ones and iramx16up60b for the big ones
[20:22:17] <etla> I've registered as AndyW
[20:22:32] <etla> mosfet drivers ?
[20:22:35] <icee> all the black magic of designing and amplifier goes away, because it's a nice integrated module
[20:22:45] <icee> combination driver and IGBT and shunt resistor
[20:23:03] <etla> what about price ?
[20:23:13] <djb_rh> etla: you're in
[20:23:24] <icee> etla: $27 for the little one
[20:23:33] <djb_rh> icee: per servo?
[20:23:35] <icee> $42 for the big one
[20:23:35] <icee> yah
[20:23:39] <etla> will it do 400W
[20:23:59] <icee> etla: the little one is rated 10A, (so ~750W); the big one 16A
[20:24:06] <djb_rh> wow
[20:24:26] <djb_rh> icee: so the driver will be an all-in-one unit? just hook up motor leads and go?
[20:24:33] <djb_rh> well, other than the brake, anyway?
[20:24:51] <icee> djb: well, that's the amplifier half.. i know there's ambiguity since i used the word 'driver'
[20:24:56] <etla> hmmm... still need the dsPIC to generate pwm, count encoder signals and interface with pc
[20:25:04] <icee> (there's a need for level conversion to drive the transistors, so that also is a 'driver')
[20:27:44] <etla> icee: what do you think about sinusoidal vs trapezoidal drive ?
[20:27:44] <icee> we're going sinusodial
[20:27:44] <etla> so hall+encoder is used to calc voltage
[20:27:44] <icee> we are a little limited in PWM resolution, so it won't be perfect sinusodial, but it'll be better than trapezoidal
[20:27:44] <icee> well.. the wiring diagrams I have don't show halls, though apparently some of the people who have gotten motors have them
[20:27:44] <djb_rh> stupid question time
[20:27:44] <icee> I'm prepared to drive motors using the shaft encoder alone for commutation (and in fact, the angle information is better, and that's what the really high end drives do, anyways, so..)
[20:27:44] <icee> djb: shoot
[20:27:44] <djb_rh> if you're going to this level of design, why not hack EMC to support a "new" protocol and get rid of the RT requirement by offloading the wave generation to the driver side?
[20:27:44] <djb_rh> which would also let you move to something other than a parallel port for comm
[20:27:44] <icee> the wave generation is on the driver side.. there's still realtime requirements though
[20:27:44] <etla> only encoder feedback: but you need to somehow bootstrap then to know where you are
[20:27:58] <icee> etla: Yah, it would require up to one turn at powerup.
[20:29:48] <etla> do you know if the dsPIC3011 is programmable with a picstart that I normally use for flash pics ?
[20:29:48] <icee> djb: the real heavy lifting here is the trajectory planning-- figuring out how to accelerate and decelerate the axes and to change that into effort levels on each axis
[20:29:48] <icee> etla: still investigating programmer
[20:29:48] <icee> one second
[20:29:48] <icee> hm, the site with the homebrew dspic programmer seems to be down
[20:29:48] <icee> crap.. it seems like everything i try and do with this project,.. the internet is broken
[20:29:57] <djb_rh> hmm, I guess I still don't have a clear picture of how this stuff *really* works
[20:29:59] <icee> microchip was doing maintenance when i needed their data sheets; digikey was broken when i needed to find amplifier parts
[20:30:03] <etla> :) do you have dsPICs and/or motors already ?
[20:30:34] <icee> djb: the computer plans the trajectory for the part, and compares actual position to desired
[20:30:46] <icee> that error runs a PID control loop, which tells the driver board how much effort to exert on each axis
[20:30:51] <djb_rh> ah, I didn't realize there was that much feedback
[20:31:04] <icee> the driver board changes the effort level + current feedback into a PWM waveform for 3 phases
[20:31:14] <icee> and the amplifier changes that PWM waveform into voltage on the motor
[20:32:39] <icee> etla: nope. motors come monday, and then I start working on design entry/bom
[20:33:51] <etla> icee: if emc+pc+epp protocol can do up to 10kHz loop closure rates, how far will that get us ?
[20:33:51] <djb_rh> my introductory tutorials talk about this as fairly simple "robotic control"
[20:33:51] <Jacky^_> Jacky^_ is now known as Jacky^
[20:33:51] <icee> etla: 10KHz is very high end.
[20:33:51] <djb_rh> icee: gotcha
[20:33:51] <icee> that's at max rotor RPM an update cycle per every 20th of a revolution of the rotor
[20:33:51] <etla> ok, and it's doable using the pico systems protocol
[20:33:51] <etla> for how many axes ?
[20:33:51] <icee> that's what the pico guys say they're getting. even if we get 1/5th that i will be happy
[20:33:51] <icee> etla: for 4
[20:33:51] <etla> ok, do you have an idea of how the pc interface will be coded on the dsPIC ?
[20:33:51] <icee> yah.
[20:33:51] <icee> It's very simple.
[20:33:51] <etla> from reading the microchip app-notes I have a fairly good idea of the motor control
[20:33:51] <icee> interrupt lines for the strobes
[20:33:51] <icee> then read stuff, set the ack lines appropriately
[20:33:51] <djb_rh> icee: oh, my mill didn't get sold
[20:33:51] <djb_rh> icee: so it'll come home this coming week
[20:33:51] <icee> wait for the strobe to go away, deassert data, deassert ack
[20:33:51] <djb_rh> gotta get the electricians to wire the 3 phase converter back up
[20:34:00] <djb_rh> hopefully I can get it working as a basic mill fairly quick
[20:34:02] <icee> djb: awesome. You'll also need a circuit for power for the CNC drive
[20:34:13] <djb_rh> yeah
[20:34:17] <etla> so there are two numbers per axis: encoder count -> pc and commanded torque -> dsPIC ??
[20:34:18] <djb_rh> just a 20A 110?
[20:34:22] <icee> etla: yup.
[20:34:25] <djb_rh> or more?
[20:34:36] <icee> djb: still being determined. I think you may be better off with 20A 220
[20:34:48] <djb_rh> okay, I think I can do that easily enough
[20:35:03] <djb_rh> at worst I won't be able to weld and mill at the same time. ;)
[20:35:26] <etla> icee: is the dsPIC3011 the highest end chip available right now ?
[20:35:33] <icee> djb: with a bridgeport, and big servos, you won't have to be waiting long for things to machine ;)
[20:35:39] <djb_rh> lol
[20:35:46] <icee> etla: there's ones with more memory
[20:35:55] <icee> that doesn't really do anything for us though
[20:36:14] <icee> we're really fine on CPU cycles. it would be nice to run the timers a little faster, but..
[20:36:31] <icee> I am thinking we're going to end up with something a lot nicer than most commercial drives.
[20:36:33] <etla> did you decide on pwm frequency
[20:36:39] <icee> etla: looks like 20KHz now;