#emc | Logs for 2010-04-08

[00:02:24] <skunkworks> I need more screen realestate... http://imagebin.ca/img/E1TpV5I.png
[00:03:18] <Jymmm> skunkworks: 1024x768?
[00:03:33] <skunkworks> yes
[00:03:52] <Jymmm> skunkworks: And you already have it, see those lil icons in the bototm right corner =)
[00:04:16] <skunkworks> old fugitsu lifebook
[00:04:23] <skunkworks> heh
[00:04:32] <Jymmm> I have one of those... I like it
[00:04:45] <Jymmm> but mine is a portable desktop more than a laptop
[00:04:59] <Jymmm> and it goes up to 1400 iirc
[00:05:13] <skunkworks> tablet?
[00:05:58] <skunkworks> it is a nice 'kick around' machine
[00:07:02] <MattyMatt> my first laptop was one of those. I still keep the carcass like a momma chimp
[00:08:15] <Jymmm> skunkworks: http://support.fujitsupc.com/CS/Portal/supportsearch.do then select: NOTEBOOK > C Series > C2240
[00:08:20] <MattyMatt> sleeping on sofa with boots on. I thought I dreamed i'd kicked the screen in, but when I woke up it was true
[00:09:20] <skunkworks> this is a T4010
[00:09:57] <Jymmm> skunkworks: This has 45 minute battery life =)
[00:10:24] <MattyMatt> NiMH packs can't be hard to make from new cells
[00:11:13] <Jymmm> serial/parallel, 4x usb, fw, SD slot, PCMCIA, DVDRW, FDD 10/100, v90, bg,
[00:11:15] <skunkworks> this one has a 10 minute battery life....
[00:11:21] <skunkworks> nice
[00:11:28] <Valen> MattyMatt can be a bit expensive
[00:11:32] <Jymmm> MattyMatt: No, that's with a BRAND NEW battery
[00:11:39] <Valen> the hard part is they are usually "stick" batteries
[00:11:52] <Valen> stick packs rather where the cells are joined end to end
[00:11:54] <MattyMatt> I've got 3 Li-ion for my T30, none last a whole hour
[00:12:27] <Jymmm> I think I want a new unibody, 7-8 hour battery life
[00:13:43] <Valen> I have an eee pc
[00:13:47] <Valen> origional
[00:14:03] <MattyMatt> fuel cells are taking their time hitting the shelves
[00:14:14] <Jymmm> I have a Lenovo IdeaPad S10 - LUST IT!!!!!
[00:14:30] <Valen> heh fuel cells arent going to happen for a long time
[00:14:31] <MattyMatt> we were supposed to be using methanol by now
[00:14:32] <Valen> too messy
[00:14:43] <MattyMatt> * MattyMatt takes one on a plane
[00:14:43] <Jymmm> Though, it has to be shipped back tomorrow for engineering to look at it =(
[00:15:25] <cstop> any ideas about a fix for a thinkpad R51 that runs purrrfect on line cord, but locks up under battery only power?
[00:15:31] <Jymmm> I found a BIOS/CMOS bug in it =(
[00:15:35] <skunkworks> Valen: what os is on you eeepc
[00:15:38] <Jymmm> cstop: "locks up" ?
[00:15:40] <skunkworks> your
[00:15:44] <Valen> MattyMatt, like you couldn't punch a hole in a plane with a decent LiPoly fire anyway
[00:15:45] <MattyMatt> cstop, new battery?
[00:15:51] <Valen> skunkworks 9.10
[00:16:17] <cstop> yep. screen freeze, no mouse, no keyboard response. The screen content remains though
[00:16:17] <Jymmm> MattyMatt: Sure you could =)
[00:16:31] <skunkworks> I just had to clean up a EEEpc with xp - 8gb ram - <1gb free
[00:16:36] <MattyMatt> I could kung-fu kick a window through, but as long as I don't mention it, they let me on :)
[00:16:44] <cstop> brandy new battery, the $140 solution to nothing
[00:17:11] <Jymmm> cstop: try a different battery and see if the problem continous
[00:17:47] <cstop> did that, I have three good batteries, all proven, as I have two of these lap tops
[00:18:19] <Jymmm> cstop: define "locks up" ?
[00:18:26] <cstop> actually, there are four batteries in the trial
[00:19:00] <cstop> input from key board or mouse is not effective.
[00:19:27] <Jymmm> internal or external mouse?
[00:19:38] <MattyMatt> do you have a smart slot battery? i.e. one that replaces the DVD drive?
[00:19:42] <Valen> skunkworks sucky
[00:19:45] <cstop> usb or tracpoint both
[00:19:56] <Valen> skunkworks with 9.10 my 601 boots in ~15 seconds or so
[00:20:01] <Jymmm> cstop: XP ?
[00:20:03] <cstop> I have no smart slots
[00:20:16] <cstop> yup
[00:20:21] <cstop> XP
[00:20:23] <MattyMatt> are you sure? is your dvd removable?
[00:20:31] <Jymmm> cstop: Have a restore disc?
[00:20:54] <cstop> yes removable DVD, I took the entire thing apart yesterday afternoon
[00:21:19] <cstop> No discs made yet, It has the "recovery partition"
[00:21:54] <MattyMatt> and no extra battery connector behind the dvd?
[00:22:12] <Jymmm> cstop: Ok, make the recovery discs, then copy them so you have two sets - it will save you some grief.
[00:22:17] <cstop> But the thing works perfect as a plug in desk top, just takes forever to boot Windoze. I think I should change that poiece out to Linux, but if it's hardware, why bother?
[00:22:38] <pfred1> linux runs hardware differently
[00:22:46] <skunkworks> boot off of the live cd...
[00:22:59] <MattyMatt> anyway those smart slot batteries are rare, and it would only let you test the physical battery connector really
[00:23:34] <cstop> It may have a battery connector behind the DVD, have not looked, I just meant I don't have such a battery
[00:23:43] <Jymmm> cstop: then... Power off the system completely. unplug all devices and power. Power on the system and get into BIOS. If you have kybd on battery in the BIOS, then restore the system adn it sound slike your ACPI/pwr mgnt drivers/utilities fubared.
[00:24:13] <Jymmm> IBM/Lenovo utilities can be funky that way.
[00:24:24] <MattyMatt> it's the same connector as the main battery, but upside down. I'm too scared to test it with a normal battery :)
[00:25:13] <cstop> At time, It runs for some minutes before the lock up occurs. Like 30 seconds to 3 or 4 minutes. If that means anything
[00:25:31] <Jymmm> cstop: Sounds liek power managemtn utilities/settings
[00:25:48] <Jymmm> cstop: Do what I said above as a test.
[00:26:12] <Jymmm> stay in the BIOS for like 20-60 minutes and see if you still have kybd control while running only on battery.
[00:26:26] <cstop> I looked, but will look again. I believe I turned off all, "shutdown/ hibernate/sleep options etc. Too many though
[00:26:29] <Jymmm> no extrnal devices connected
[00:26:31] <pfred1> why won't eagle autoroute to this library element I made?
[00:26:43] <Jymmm> cstop: No, dont look. just do what I said.
[00:27:18] <Jymmm> cstop: Keep is simple stupid, why overthink things =)
[00:28:05] <MattyMatt> that's what they said to galileo
[00:28:17] <MattyMatt> >:)
[00:28:27] <Jymmm> I must have worked on hundreds of ThinkPads, and close to a thousand laptops over the years.
[00:28:32] <morfic> <skunkworks> boot off of the live cd... <-- 1 out of 3 dells at work booted the livecd, one failed with squashfs bad block, one other had some ata issue that kept it from even getting there, 3rd ran it flawless, don't think this speaks for dell, never had this much trouble with a ubuntu livecd
[00:28:46] <pfred1> Jymmm you're a laptop thief?
[00:29:44] <Jymmm> pfred1: You ever see those things where they setup dominos then they tip one and it starts a chain?
[00:29:45] <cstop> Jymmm test underway IBM rescue and recovery ;-)
[00:30:06] <Jymmm> cstop: Huh? oh, your making the discs. ok
[00:30:12] <pfred1> Jymmm yes
[00:30:29] <Jymmm> pfred1: I've done that with stacks of MacBooks =)
[00:30:30] <skunkworks> Jymmm: how many power plugs have you had to re-solder in?
[00:30:32] <skunkworks> or replace
[00:30:45] <Jymmm> skunkworks: on the laptop or PS ?
[00:30:47] <pfred1> Jymmm you think different thats for sure!
[00:30:52] <skunkworks> laptop
[00:31:13] <pfred1> now who here knows Eagle?
[00:31:24] <Jymmm> skunkworks: Never, I always send it off to the depot
[00:31:31] <cstop> Jymmm I'm glad this topic allowed me to inquire of a PRO!
[00:31:59] <cstop> So far....it remains working....
[00:32:01] <Jymmm> cstop: skunkworks is the professional, I'm just the Village Idiot
[00:32:13] <cstop> Ohh If done a few things with Eagle
[00:32:15] <skunkworks> heh
[00:32:31] <steverob> Just finished an eagle project.
[00:32:42] <Jymmm> cstop: did you say you have two IDENTICAL thinkpads?
[00:32:46] <MattyMatt> * MattyMatt learning gEDA
[00:33:13] <pfred1> MattyMatt I learned to avoid geda
[00:33:29] <skunkworks> Jymmm: I would not wish that on anyone... more times than not the motherboard needs to be totally removed.
[00:33:34] <steverob> Got my new BOB completed with optos instead of direct wiring to the controllers. Getting ready to test it this evenng.
[00:33:43] <cstop> No A T42, and an R51 both with the same battery, after that ???
[00:33:57] <Jymmm> cstop: oh, ok. nm then.
[00:33:59] <pfred1> steverob ever make a library element the autorouter would refuse to route to?
[00:34:19] <Jymmm> cstop: If you need T42 bateies I know where you can get them cheap
[00:34:49] <steverob> No... I've generally copied an existing lib and modified it for my situation. The autorouter always found the part OK.
[00:35:22] <pfred1> I've made dozens of parts lots from scratch but I've never seen this behavior before
[00:35:26] <cstop> I may need one, but not if this R51 keeps up it's troubles. (It's still testing OK!)
[00:35:31] <steverob> Wonder if the part is too far off the grid or something like that?
[00:36:03] <morfic> SWPadnos: http://www.acrosser.com/products/AR-ES0892_detail_id_413.html is that the one you got? did you add gfx or was the integrated one enough? most comments so far were integrated ==
[00:36:03] <Jymmm> cstop: Then fans should be on and loud. Correct?
[00:36:07] <morfic> BAD
[00:36:32] <steverob> I generally route mine by hand anyway. They're fairly simple singlesided boards so, there's no real advantage to the autorouter for me.
[00:37:14] <Jymmm> cstop: On the bottom of the R51 there is a "Machine Type" number in the form of xxx-xxxx, get that for me please
[00:37:53] <pfred1> steverob eagle's autorouter can be awfully brain dead at times but I like to run a few autos just to get going
[00:38:02] <cstop> no Fan
[00:38:13] <Jymmm> cstop: O_o
[00:38:18] <pfred1> steverob I can hand route to the part and I didn't tkeepout it I already checked that
[00:38:56] <steverob> Yeah... Like you said it can be braindead at times.
[00:38:59] <cstop> Machine type 2883GKU from the Current Hardware Information
[00:39:23] <pfred1> well I've never seen it completely refuse ot route to a part at all
[00:39:24] <cstop> Fan works on line power
[00:40:00] <steverob> I think I'd recreate the part and see what happens.
[00:40:10] <pfred1> heh yeah right that'd only take a day
[00:40:23] <cstop> At least the fan was on when I unplugged it a moment ago ;-)
[00:40:35] <MattyMatt> make a copy and chop it in half. see if each half works
[00:40:54] <steverob> Otherwise, post to their forum. It's pretty active with a lot of knowledgabe folks there.
[00:40:54] <MattyMatt> ^advice for pfred1 and cstop
[00:41:02] <steverob> Maybe get some help there.
[00:43:39] <steverob> I'm really happy with how my BOB came out. Just waiting for some parts so, I can build the newest revision of the controllers.
[00:43:55] <cstop> My only experience with the auto router was "it's a start"
[00:44:10] <steverob> This is about the third iterration and each time they get a little better.
[00:44:18] <pfred1> cstop if i fool with it for a while i can usually make it work
[00:44:26] <Jymmm> cstop: FYI... Hardware Mainenance Manual ThinkPad R50/p,R51 http://download.lenovo.com/ibmdl/pub/pc/pccbbs/mobiles_pdf/39t6190.pdf
[00:44:46] <pfred1> steverob I made a BOB of sorts
[00:45:01] <steverob> Of sorts?
[00:45:29] <pfred1> http://img716.imageshack.us/img716/575/ppbb1r.png
[00:46:30] <Jymmm> cstop: NOTE: creating the recovery discs can take a few hours to complete.
[00:47:03] <pfred1> routed http://img690.imageshack.us/img690/9582/ppbbr4brd.png
[00:48:04] <pfred1> built: http://img191.imageshack.us/img191/4185/ppbbpic1.jpg
[00:48:15] <cstop> Thanks for the link Jymmm It's in my docs folder!
[00:48:29] <pfred1> running: http://www.imagechicken.com/uploads/1269814572086495500.jpg
[00:48:46] <steverob> Hmm... Considerably more complex than mine.
[00:49:20] <pfred1> steverob all it really does is boost the current out of the parallel port
[00:49:53] <pfred1> and optically isolates 4 of the inputs too
[00:50:36] <pfred1> and gives you a nice barrier terminal strip to connect to :)
[00:50:50] <steverob> I don't have a power supply on my board nor isolated inputs (yet). Figured I'd make a new one when I figure out EXACTLY what it is going to do ;-)
[00:51:00] <skunkworks> I like eagle http://www.electronicsam.com/images/KandT/servostart/mounting.JPG
[00:51:11] <steverob> what are the 245's for?
[00:51:30] <pfred1> steverob they're buffer ICs
[00:51:56] <pfred1> steverob they take a few microamps for input and give you 24ma output
[00:52:27] <pfred1> steverob keeps you from burning up your port and allows you to drive more loads
[00:53:15] <cstop> Jymmm What is a USB floppy? An external floopy drive?
[00:53:20] <pfred1> steverob some parallel ports are hard pressed to haldle much more than 3 ma
[00:53:29] <pfred1> handle even
[00:54:18] <steverob> I ran a test earlier with my PP and it was giving a steady 15ma across a resistive load.
[00:54:41] <cstop> but it must be settings and software, 'cause it still keeps ticking!
[00:54:42] <pfred1> steverob yeah Pports vary widely
[00:54:58] <steverob> I wanted to make sure it could drive the optos which require ~5ma.
[00:55:35] <pfred1> some optos are optimal at closer to 10 ma
[00:55:43] <pfred1> most in fact
[00:55:49] <steverob> Although.. I see you have schmitt triggers to square up the waveforms. That's probably a goodidea.
[00:56:00] <pfred1> possibly ...
[00:56:38] <pfred1> I've driven an OC 7406 straight off an opto too and gotten fairly square waveform out of it
[00:56:53] <steverob> I tested the optos (MOC8204 from my junk boox) and the wave forms looked good at 5ma. So, I went with that.
[00:57:08] <pfred1> which suprised me being as some optos give more of a ski slop looking waveform
[00:57:15] <pfred1> ski slope even
[00:57:36] <pfred1> but they were nice MCT2Es
[00:58:40] <steverob> Mine is a homemade mill and I really don't expect that it will run very fast anyway. I'll sleep better knowing that I'm not gonna fry my PC by it not being isolated.
[00:59:03] <steverob> Don't have a camera handy, or I'd post a pic.
[00:59:09] <pfred1> I got a bunch of toshiba TLP521-2 coming now haven't messed with them yet
[00:59:29] <pfred1> steverob I isolate right at the motor drivers
[00:59:39] <pfred1> isolating twice doesn't really gain anything
[00:59:55] <pfred1> and all optos are dog slow compared to TTL speeds
[01:00:04] <steverob> I had plenty of real estate on the BOB, so I put the optos there.
[01:00:26] <pfred1> yeah but who has extra us to waste?
[01:00:49] <steverob> I may consider putting schmtt triggers on the driver boards to square up things.
[01:01:31] <steverob> Like I said my mill is gonna be slow anyway. Another 3us won't matter tooooooo much.
[01:01:34] <pfred1> if you're driving your opto inputs as low as 5 ma it may be a factor for you
[01:02:02] <pfred1> optos I've looked at seem happier at around 10 ma or so
[01:03:03] <steverob> From my testing, it looks like I'm seeing about a 2-3 us delay turn on. and 4-5 us delay turn off.
[01:03:19] <pfred1> thats pretty good
[01:03:40] <pfred1> is the base of your opto transistors exposed or buried?
[01:03:55] <pfred1> like is it connected to a pin?
[01:04:01] <steverob> The corners are kinda rounded though ;-)
[01:04:05] <steverob> exposed.
[01:04:07] <pfred1> some are most aren't
[01:04:18] <pfred1> you leaving it just hang or you have it driven to ground?
[01:04:41] <steverob> Hanging right now.
[01:05:07] <pfred1> ah yes you can put a 47K to 1meg resistor from it to ground I believe it is to saturate the junction it helps switching
[01:05:16] <pfred1> start at a meg and work your way down
[01:05:34] <pfred1> it reduces the parasitic capacitance of the transistor I believe
[01:05:47] <pfred1> in plain english it makes your opto faster
[01:06:03] <pfred1> that and you don't use the base as an antenna to pick up noise
[01:06:03] <steverob> Yep... Figured it could be biased or whatever to gain performance. I'll just wait and see what happens.
[01:07:08] <pfred1> I just slapped some 1 meggers in my circuit and called it a day
[01:07:14] <Jymmm> cstop: Yep, or a "Drive Bay Floppy" - it replaces the DVD drive
[01:07:15] <steverob> It'll all go in a metal enclosure so, RFI shouldn't be a big concern. Other than what's generated by the mil it's self.
[01:07:42] <pfred1> away from your motor drivers?
[01:08:02] <WalterN> ugh
[01:08:03] <steverob> Hey... I'm just figureing out this stuff as I go.
[01:08:08] <pfred1> I think marconi would have really liked stepper drivers
[01:08:23] <WalterN> this is so much easier to draw up on a piece of paper...
[01:08:34] <WalterN> stupid CAD programs
[01:08:48] <pfred1> WalterN I'm with you there i hate CAD software!
[01:09:02] <steverob> I'll be fine when I finally stop zapping power transistors. I've killed quite a few trying to get the drivers tuned up.
[01:09:04] <pfred1> and i took like 7 years of mechanical drawing in school too
[01:09:32] <pfred1> steverob what kind of motor drivers did you make?
[01:10:11] <steverob> Simple microprocessor driving darlington transistors.
[01:10:42] <pfred1> steverob not doing any sort of current control are you?
[01:11:13] <steverob> Nope. Just put a finger on the resistors and see which one's get hot.
[01:11:33] <pfred1> steverob so what sort of speed are you getting out of your motors with your drivers?
[01:11:55] <steverob> I've been changing the duty cycle on each phase to control the current.
[01:12:30] <pfred1> this is my best motor driver I've built to date: http://img683.imageshack.us/img683/6498/stepproto1.jpg
[01:12:34] <cstop> Jymmm hmm all I would have is a thumb drive or the iPod
[01:12:41] <pfred1> its based on an Allegro SLA7026M
[01:12:50] <pfred1> PWM chopper
[01:13:05] <Jymmm> cstop: Ok, why do you need a floppy for?
[01:13:13] <pfred1> it'll do 2500 RPM at 23VDC
[01:13:52] <pfred1> which ain't too shabby
[01:14:01] <steverob> Oh gosh. I really couldn't say. I been putting out other fires and haven't really had a good test of the system.
[01:14:38] <steverob> This is my first attempt at building a machine so, I wanted to keep it REALLY simple.
[01:14:38] <cstop> Jymm that is what the directions in the recovery app say, Five formatted floppies
[01:14:42] <pfred1> initially I measured the clock pulses and just did the math but I think emc2 will give numbers you can calculate motor speeds with too
[01:15:14] <pfred1> realistically I can only run to around 800 RPM and still have usable torque
[01:15:20] <steverob> I'll be able to test better when my next order of transistors comes in.
[01:15:53] <pfred1> steverob why didn't you just go with an integrated stepper motor module?
[01:16:07] <pfred1> I mean they do make parts for this today
[01:16:24] <steverob> I'm using what I already have on hand.
[01:17:06] <pfred1> I like to junk box it too but some stuff i know i have to get
[01:17:36] <pfred1> I just ordered up a bunch of TB6065AHQs to play with
[01:17:41] <steverob> I find it more rewarding to do things the hard way. Anyone can buy custom chips and slap together a working system.
[01:17:45] <pfred1> look up the datasheet on those
[01:18:05] <pfred1> steverob heh google TB6065AHQ you may find that isn't so true ;)
[01:19:14] <pfred1> after i ordered mine digikey jacked up the prices on them
[01:19:28] <pfred1> they're such a hot item now
[01:19:28] <steverob> Google doesn't want to find TB6065AHQ?
[01:20:08] <pfred1> oh its 6560 sorry
[01:20:20] <steverob> All I'm seeing is boston terrier stuff!
[01:20:28] <pfred1> try 6560
[01:20:46] <pfred1> they look pretty sweet to me but they're only for smaller motors
[01:20:55] <pfred1> like < 2 amps
[01:21:15] <pfred1> my plan is to put more motors on each axis though
[01:22:17] <pfred1> it looks like you can implement them with something like 6 external parts
[01:23:01] <pfred1> couple current sense resistors an osc cap a couple filter caps and go
[01:23:47] <steverob> OK. I was just looking at the datasheet. Look like an interesting chip.
[01:24:01] <pfred1> it sure does at $4.73 a pop!
[01:24:24] <pfred1> you could build a motor driver out of it for about $10
[01:24:58] <pfred1> and that is tough to beat with discrete transistors
[01:26:05] <pfred1> but the word out is don't try to tun this IC over 24V or 1.75 amps per phase
[01:26:22] <pfred1> or it'll probably blow up in your face
[01:26:31] <steverob> It looks like you'd still need some "brains" to make it work. There's a lot of inputs that need to be defined.
[01:26:59] <pfred1> no, you can just tie them all to the logic levels you want they're pretty much either ors
[01:27:12] <pfred1> thing even has built in sinking resistors
[01:27:32] <pfred1> so you can just slap + to what you want to raise high
[01:27:39] <steverob> You plan on using them for your next set of drivers?
[01:27:44] <pfred1> they thought of everything
[01:27:52] <pfred1> yeah i have some on the way now
[01:28:07] <pfred1> last i checked they were floating around in Ohio
[01:28:50] <steverob> My biggest hinderance getting started was the lack of "requirements". Ireally didn't know what it was supposed to do other than move the steppers.
[01:28:51] <pfred1> but i need to fix my library element i made of them being as Eagle doesn't want to route to the damned things!
[01:29:21] <pfred1> move is nice mover faster is better
[01:29:56] <pfred1> what kind of motors do you have?
[01:30:17] <pfred1> like Voltage amps per phase wiring ?
[01:30:35] <steverob> I'm glad mine moves slow. I've had things crash into each other. In this case SLOW is good.
[01:31:20] <pfred1> I'm testing everything out on my bench first
[01:31:28] <pfred1> motors are just sitting there
[01:31:36] <steverob> Once I figure out this stuff, I'll build a faster, bigger, stronger machine. Until, then slow is good.
[01:32:16] <pfred1> my first machine isn't going to set any land speed records I'll be lucky to get 48 ipm out of it
[01:33:57] <steverob> I've got the machine basically finished. Just trying to get things working correctly. I've had some ground loop problems and hopefully the new BOB will fix that. Like I said, I've toasted a butt load of transistors and am waiting for the next order so that I can go back and get things working smoother.
[01:34:21] <pfred1> steverob do you have freewheeling diodes in your motor drivers?
[01:34:58] <pfred1> thats probbaly why you're blowing transistors
[01:35:04] <steverob> Last time I ran the machine, last weekend, it was mostly working but skipping steps. It's mostly work then ski a step, then work, then skip a step.
[01:35:27] <steverob> Yep. I've got diodes on the transistors.
[01:35:52] <pfred1> ok are they slow recovery ? with good on speed
[01:36:24] <steverob> I was blowing them because of thermal runaway. I'd dink with the power supply, the timing in the micro, and before I knew it, I'd smoke a transistor.
[01:36:53] <pfred1> are they configured as an H bridge or just unipolar?
[01:37:19] <steverob> Unipolar. Common nema23 motors.
[01:37:34] <pfred1> ok not deadband then
[01:37:50] <pfred1> still could be back emf whiplash
[01:38:45] <pfred1> if I could get them I'd use SLA7062s for unipolars
[01:39:01] <pfred1> they are the same as i use cept they have built in sequencers
[01:39:20] <steverob> One of problems I was having is the micros weren't comlpetely resetting when powered off. when I powered the machine, it would be in an unknown state. If one of the PORTS happened to be high, it would fire that transistor and cook it before I realized there was a problem.
[01:39:24] <pfred1> dmos pwm etc etc
[01:40:21] <pfred1> yeah i don't know how stepper modules get around that deal but they do
[01:40:42] <steverob> I thin the new BOB will stop that from happening. I also put a resistor across the power supply to bleed any cap when powered off.
[01:40:48] <pfred1> least I've never had it happen to me
[01:41:40] <steverob> I'm pretty sure it was a ground loop problem. The micros were getting enough energy from the PP so that they didn't reset.
[01:41:40] <pfred1> what step sequencing are you using?
[01:41:53] <pfred1> full half wave micro?
[01:42:10] <steverob> Now that the PP isolated, that shouldn't (fingers crossed) happen again.
[01:42:15] <steverob> Half step.
[01:42:22] <pfred1> yeah good choice
[01:42:34] <pfred1> there's two kinds of half stepping too
[01:42:36] <steverob> Didn't want to tackle microstepping yet.
[01:42:53] <pfred1> well maybe 3 in fact
[01:43:02] <pfred1> one's a high Z half step
[01:43:40] <steverob> Only familiar with one type of half step.
[01:44:08] <pfred1> one raises current to compensate for imbalance in the sequence
[01:44:52] <pfred1> every other step you only have 50% torque
[01:45:32] <steverob> As Ted Kennedy said "I'll drive off that bridge when I come to it". For right now, I gotta keep it simple.
[01:45:42] <pfred1> so if you weaken the full step you can balance it out
[01:45:53] <steverob> Otherwise, it'll get done.
[01:46:02] <steverob> ^never
[01:46:24] <pfred1> hey at least now you're aware of it right?
[01:46:35] <steverob> Yeppers!
[01:46:59] <pfred1> but you can see how in half of a half stepping sequence there isn't full power right?
[01:47:47] <pfred1> the steps where only one phase in energized
[01:48:43] <steverob> Yeah. I've looked at the torque curves. I guess if you're looking for optimal performance, it's the kinda thing you gotta do. For me, that is not a consideration at this point.
[01:50:08] <steverob> I just grab the motor shaft. If it seems to have good torque and speed my guestimations for the drive parameters are probably close.
[01:50:41] <pfred1> odly sometimes loading a stepper motor shaft makes it run better it changes the motor resonance
[01:51:14] <pfred1> stepper motors are some odd beasts
[01:51:18] <steverob> If the torque or speed is poor, I tweak things until I hit the right combo. making a mental note of what changed.
[01:52:01] <pfred1> when you mess with steppers you have to throw out everything that seems intuitive as far as running motors goes
[01:52:35] <pfred1> like when i was tweaking my drive i had to lower current to get max perfoemance out of my motor
[01:52:47] <steverob> Yeah. They don't work like one would expect.
[01:53:03] <pfred1> there are a lotta curveballs with stepper motors
[01:53:35] <steverob> I think the coils can get magnetically saturated and the motor won't work as well as you'd expect.
[01:53:52] <pfred1> thats why you run them over volted
[01:53:58] <pfred1> and current controlled
[01:54:07] <pfred1> speeds the flux change
[01:54:43] <steverob> Yep. Seems like it'd toast the motors putting in that much voltage.
[01:54:55] <pfred1> nah you control the current to limits
[01:55:43] <pfred1> but you're doing it old school with dropping resistors right?
[01:56:22] <pfred1> or are you a light bulber?
[01:56:31] <steverob> Yep. Got a handfull of low ohm power resistors. Still looking for the best combo.
[01:56:43] <pfred1> do you know about the light bulb trick?
[01:57:08] <pfred1> light bulbs increase in resistance the more current you put through them
[01:57:24] <pfred1> so they make great current controllers
[01:57:55] <pfred1> just match bulb wattage to your circuit
[01:58:26] <pfred1> and they're cheaper than big honking resistors too!
[01:58:40] <steverob> Yep. Although they have negative resistance at a certain temperature. That's a pretty narrow band so, it doesn't factor in to this situation.
[01:58:53] <pfred1> people do it
[01:59:32] <pfred1> I built a dropping resistor motor driver once and it was just silly
[01:59:44] <pfred1> it was more like building a space heater
[02:00:08] <steverob> I've scraped plenty of old electronics so, things like power resistors are generally easy for me to find.
[02:00:15] <pfred1> I was using like the hugest resistors i had and they were still getting burning hot
[02:00:37] <pfred1> like 38 watt wire wounds
[02:01:08] <steverob> I've got some really nice 100W mil spec resistors in finned aluminum packaging.
[02:01:27] <pfred1> and the performance was still pretty dismal so I said the heck with that 80s tech
[02:01:53] <steverob> Once I find the optimum value, I'll screw them to a heat sink.
[02:01:53] <pfred1> yeah you need them
[02:02:15] <pfred1> and it still can't compete with PWM
[02:02:21] <steverob> As cold as this winter was, we can use the heat!
[02:02:40] <pfred1> yeah i was doing it over the winter too come to think about it
[02:02:57] <pfred1> but when i do electronics burn to the touch is never a good thing to me
[02:03:19] <steverob> That's generally a sign of bad things to come.
[02:03:57] <pfred1> yeah i only had a few really big resistors enough to make one driver and I wasn't going to go looking for any more after I'd fiddled with it some
[02:04:15] <steverob> Well. It's been nice chatting with you but, I gotta go. Good luck with you new drivers.
[02:04:22] <pfred1> thanks
[02:06:36] <Jymmm> cstop: ah, an older system. Yeah, that's what you need... a USB FDD + 5 Floppies.
[02:07:12] <pfred1> I tried to make a floppy boot disk a little while ago I couldn't get 2 floppy drives to work
[02:07:23] <Jymmm> cstop: Though, that seems really small... ~5MB. No CD option?
[02:07:42] <Jymmm> pfred1: one drive, two floppies.
[02:07:55] <pfred1> they got the cutest little ball screws in floppy drives though!
[02:08:19] <Jymmm> no balls, just screw
[02:08:35] <pfred1> Jymmm i wanted to boot a machine with a floppy I'd made in another
[02:08:46] <pfred1> it just wouldn't work
[02:09:11] <pfred1> both drives worked in each machine but the one disk from the one machine wouldn't boot the other
[02:09:45] <Jymmm> There's no reason why that shouldn't work.
[02:09:56] <Jymmm> Maybe bios setting
[02:10:05] <pfred1> but all my floppy drives seemed pretty siezed up and stuff i had to take them apart and actually oil the shafts in them to get them to run right again
[02:10:18] <Jymmm> lol
[02:10:30] <pfred1> yeah it is pretty laughable
[02:10:41] <cstop> Jymmm I'm reading the service book now
[02:10:47] <pfred1> I had a stack of like 10 floppy drives and i kept trying one after the other
[02:11:02] <pfred1> and a few I took apart and lubed
[02:11:12] <pfred1> hey it helped!
[02:11:40] <pfred1> I sprayed some PB blaster into the can cap and used a little brush to brush the lube around
[02:11:59] <Jymmm> whatever works
[02:12:06] <TD-Linux> I plugged a usb fdd into my ps3 :D
[02:12:06] <tom3p> is the 'cgc' stuff lawrence/flyingelectron talks about something he's done or something he'd like to do?
[02:12:07] <tom3p> his text sounds like these gcodes extensions already exist.
[02:12:10] <pfred1> it brought two drives back to life
[02:12:35] <SWPadnos> tom3p, done - check the link in his first email on the subject
[02:12:42] <cstop> I can't get the floppy drive on this linux Desk top think center to go, and I even put in a replacement
[02:12:53] <tom3p> thx ( just got new accounts )
[02:12:58] <SWPadnos> sure
[02:13:00] <pfred1> TD-Linux I was finding files like from 1995 on floppies i have laying around and stuff
[02:13:24] <SWPadnos> damn. this talk reminds me that I need to make a DOS boot floppy so I can do a BIOS update
[02:13:37] <tom3p> whats the link to browse older emails for emc-user list?
[02:13:47] <pfred1> SWPadnos when was the last time you used your FDD?
[02:13:56] <SWPadnos> years ago
[02:14:03] <SWPadnos> probably the last time I did a BIOS update :)
[02:14:08] <pfred1> SWPadnos don't be too shocked if it doesn't work
[02:14:09] <skunkworks> http://news.gmane.org/gmane.linux.distributions.emc.user
[02:14:11] <cstop> I just turned off the "BIOS update by wire" option ;-)
[02:14:15] <SWPadnos> yeah, I know
[02:14:16] <tom3p> skunkworks, thx
[02:14:22] <skunkworks> np
[02:14:34] <pfred1> SWPadnos I found rubbing the read heads with some tissue paper helped out somewhat too
[02:14:42] <SWPadnos> I don't think that option exists on this particular motherboard, but I might be able to use a bootable CD instead
[02:14:48] <TD-Linux> my bios updater works with cds
[02:14:51] <SWPadnos> yeah, like cleaning an old cassette deck
[02:15:06] <SWPadnos> this is a somewhat old SuperMicro server motherboard
[02:15:17] <pfred1> but the drives i was messing with were messed up I mean they sounded sick and stuff
[02:15:40] <skunkworks> * skunkworks loves supermicro motherboards.. (atleast older ones...)
[02:15:45] <pfred1> the white grease in them was like crust
[02:16:31] <SWPadnos> mine seems to spin OK
[02:16:33] <pfred1> but hey I have to give a thumbs up for PB Blaster when it comes to reviving old FDDs
[02:16:56] <pfred1> SWPadnos I found exersizing the drive by having it format a disk a few times helped too
[02:17:46] <SWPadnos> this one seems to work, I just read an old disk with it
[02:17:50] <SWPadnos> at least partway
[02:18:05] <pfred1> yeah mine seemed to have some troubles at extremes
[02:18:19] <SWPadnos> heh - the disk has a Win95 Y2k patch, and an old BIOS for the motherboard I want to update :)
[02:18:42] <pfred1> try to like run a scan on it that should make it read it all
[02:18:47] <SWPadnos> oh no, that's the old motherboard - the dual Pentium Pro
[02:18:58] <SWPadnos> wow, that's an old disk :)
[02:19:14] <pfred1> I found formatting and scanning helpful in getting FDDs back running again
[02:19:22] <SWPadnos> yep
[02:19:31] <tom3p> wow is flyingelectron's pick&place done with emc? (love the vacuum and centering 'nudger')
[02:19:36] <SWPadnos> the only problem there is that a misaligned drive may make a disk that only it can read
[02:20:35] <pfred1> I was sort of pissed when floppies went by the wayside but after my last experience with them good riddance I say now
[02:22:15] <pfred1> I bet every floppy i ever wrote wouldn't even fill up one CD-ROM disc
[02:22:47] <skunkworks> tom3p: link?
[02:24:30] <tom3p> skunkworks, about 1/2 way down this page is a link to the video http://tsemsb.blogspot.com/2010/04/cgcc-gcode-with-c-constructs.html
[02:26:10] <grommit> The last floppies (the 2" hard plastic ones) were 1.44MB, so 520 or so would fit on a 750MB CD. Earlier floppies were even less capacity....
[02:26:19] <pfred1> gee I love whatever google did to their maps that makes it so I can no longer zoom
[02:27:03] <morfic> SWPadnos: your mobo does not do bios updates via cdr image?
[02:27:31] <SWPadnos> well, it probably could if I would make a DOS bootable CD
[02:27:46] <pfred1> grommit depends what disks you had the 1.44 ones were the high density disks
[02:27:56] <morfic> SWPadnos: oh, a PPro? the real i686? :) i skipped over that before
[02:27:58] <SWPadnos> I don't think it has a flashing routine in the ROM, it needs a DOS executable to do the actual burning
[02:27:58] <pfred1> grommit some were as low as 720K though
[02:28:06] <SWPadnos> yeah, I had a dual PPro board
[02:28:09] <cstop> USB drives are so handy
[02:28:13] <SWPadnos> was really hot stuff in 1996 when I bought it
[02:28:37] <SWPadnos> and then upgraded it with the Overdrive CPUs - 333MHz you know
[02:28:39] <morfic> oh man, don't get me started talking old cool hardware, or i tell you about my 88MB syquest
[02:28:44] <SWPadnos> heh
[02:28:59] <SWPadnos> I'd have to tell you the story of the BBS we made in high school :)
[02:29:13] <SWPadnos> (the one that ran on 4 floppies until we finally raised enough money to buy a hard drive)
[02:29:15] <morfic> SWPadnos: oh yeah, bbs is before my time
[02:29:20] <cstop> I still can't type faster than an 8088 can post to screen
[02:29:54] <Valen> morfic how old are you?
[02:30:02] <Valen> I used to bbs ;->
[02:30:03] <morfic> SWPadnos: compuserve was first ISP in my town (no really i ran ppp through their system, rather than use the compuserve software)
[02:30:11] <morfic> Valen: 37.5
[02:30:22] <Valen> i just turned 27
[02:30:34] <morfic> * morfic points out the .5 were much more important when you were 14.5 :P
[02:30:34] <skunkworks> tom3p: holy crap!!
[02:30:49] <pfred1> morfic I wasn't going to say anything ...
[02:31:05] <pfred1> half the time i can't even remember how old i am anymore
[02:31:14] <cstop> morific is obviously not a woman, for there are no admitted women of that age until they no longer care to remember ;-)
[02:31:14] <pfred1> one year is much like the next
[02:31:59] <morfic> skunkworks: thanks for screaming, i missed the link otherwise
[02:32:20] <morfic> cstop: why would you even consider the fact i may have been? :P
[02:32:36] <tom3p> is it run with emc2? (who is flyingelectron when he's on irc?)
[02:33:05] <cstop> have ben female? As in used to be?
[02:33:47] <morfic> cstop: ok, we kind of took a wrong turn in my grammar.....
[02:33:51] <cstop> You are among friends
[02:34:36] <cstop> Sorry play with words seems so innocent, but without the facial expression and hand movements....It's just not the same
[02:34:48] <morfic> cgcc sounds interesting
[02:35:20] <cstop> If you have such application cgc
[02:35:20] <SWPadnos> tom3p, I think he's flyingelectron
[02:35:41] <skunkworks> tom3p: You should ask him. (I don't know)
[02:35:52] <cstop> I stumble over multi level symbolisms
[02:36:30] <morfic> someone should tell the author of the video that PCB Board is as good as a LCD Display
[02:38:03] <morfic> i would be so bummed if boss wouldn't pick emc, the more i see of emc, the more i think this could actually be awesome
[02:38:03] <skunkworks> I did my first 'surfing' through my dads account thru the univeristy. (I really didn't know what I was doing) sometime after that - I got on using the brother-in-laws compuserve. :)
[02:38:21] <skunkworks> I think it is awesome every day :)
[02:38:49] <pfred1> morfic hey no one is stopping you from using EMC even if you boss doesn't want to
[02:39:05] <skunkworks> you can do so much.. ;P http://imagebin.ca/img/E1TpV5I.png
[02:39:05] <morfic> SWPadnos: the acrosser model you gave me has integrated gfx, and you give it your emc blessing?
[02:39:15] <SWPadnos> well, yes and no
[02:39:29] <skunkworks> goal3? ;)
[02:39:33] <morfic> pfred1: my finances beg to differ, i am into photography already..
[02:39:34] <SWPadnos> the system I put those in didn't use X, or networking
[02:39:50] <SWPadnos> skunkworks, no, some embedded machines I made a couple of years ago
[02:39:54] <pfred1> morfic people build CNC machines out of dumpster parts
[02:39:55] <skunkworks> ah
[02:40:15] <SWPadnos> in fact, those machines didn't run EMC2, they only used an SMP RTAI kernel that seemed to work at the time, and HAL on top of that
[02:40:17] <morfic> pfred1: i have pretty snobby standards? else i wouldn't be here ;)
[02:40:40] <pfred1> morfic people build radical nice CNC machines out of dumpster parts!
[02:40:58] <morfic> SWPadnos: i'll look at acrosser for something similar with a separate gfx card, based on the rest of the specs
[02:41:20] <SWPadnos> morfic, one thing to remember (for your boss) - when you get something added to EMC2, you get to keep using it, on as many machines as you like. If you decide to buy something, you have to keep paying for it over and over as you convert more machines
[02:41:24] <SWPadnos> not to mention support contracts
[02:41:48] <pfred1> morfic thing is I've seen machines slapped together using drawer slides as linears onto MDF and they're just as impressive as THK rail systems in their own way
[02:41:53] <SWPadnos> well, it was OK - I think the latency was no higher than 15000 or so even before I disabled all the crap we didn't need
[02:42:17] <SWPadnos> but I got it down under 2000, and usually around 200, by disabling all the things we didn't need
[02:42:33] <pfred1> morfic more when you think its basically a computerized drawer dong the work
[02:42:46] <SWPadnos> I actually didn't really like the motherboard, to be honest. the PCI chipset has no read combining, so it was dog slow, even reading from a PC Mesa card
[02:42:48] <SWPadnos> PCI
[02:42:58] <cstop> plotters run on "string drives"
[02:43:30] <morfic> SWPadnos: he understands that already, what is important to him is to have someone who says "yes, i know how to integrate it, this is what it will cost to get me on the phone, this is what it will cost to show up on your doorstep should you get dead stuck"
[02:43:42] <SWPadnos> sure
[02:43:43] <pfred1> cstop sort of gives show string budget a whole new meaning don't it?
[02:43:47] <pfred1> shoe string even
[02:43:57] <cstop> yup
[02:44:25] <skunkworks> some use cables to keep their large gantrys square..
[02:44:37] <morfic> SWPadnos: his idea is once we fully understand the first, we should only require the occasional programming service
[02:44:50] <pfred1> skunkworks yeah that looks awefully hokey to me
[02:44:58] <cstop> From the youtube clips, you would think 99% of EMC is 2D plotting with laminate trimmers
[02:45:37] <morfic> SWPadnos: he understands opensource better than any other obss i ever had, which is why he would like to make sure he can pay for coding, he inderstands he pays for the service of coding, not for the product of code, since the code would go right back into the emc code base
[02:45:37] <cstop> but then you look at that cinci ball test....whoa!
[02:46:03] <morfic> "goes right back" is more "always part of emc"
[02:46:38] <morfic> SWPadnos: it was great seeing cradek's enthusiasm based on the code sample i put on the wiki, i hope that's good news.
[02:47:19] <SWPadnos> it probably is :)
[02:47:31] <morfic> cstop: i am going to send myself an email to work with 4 youtube videos, and cgcc link, to think about that some more, the hardinge link i think is most impressive next to the ball video
[02:47:58] <SWPadnos> all right. 2 working floppies
[02:48:04] <SWPadnos> drives even
[02:48:07] <pfred1> morfic failing that you can always crash your next local LUG meeting http://en.wikipedia.org/wiki/Linux_User_Group
[02:48:17] <morfic> hearing a dev say "i would like to see this in our project" is a good start
[02:48:37] <morfic> pfred1: crash it with a dumpster part cnc you mean?
[02:48:51] <SWPadnos> and boy is it funny to see text mode DOS on a 9 megapixel monitor :)
[02:48:59] <pfred1> morfic no crash it looking for a high school kid to write code for you
[02:49:40] <morfic> pfred1: well, yes that too, boss likes idea someone who knows the project inside out already doing the work, rather than a "let me check this out, i think i know what they did"
[02:50:02] <morfic> this way you avoid a solution that seems to work on paper but causes a "this changes too much we didn;t want to change"
[02:50:05] <pfred1> does anyone know all of EMC2?
[02:50:12] <skunkworks> no
[02:50:18] <pfred1> I didn't think anyone would
[02:50:20] <SWPadnos> heh, no way :)
[02:50:34] <morfic> thanks all
[02:50:40] <morfic> 'night again
[02:50:44] <SWPadnos> but there are a couple of people who know it very well, and cradek is right there at the top
[02:50:46] <SWPadnos> see you morfic
[02:51:10] <pfred1> well yeah they got a leg up on everyone else but someone good at coding they have their methods
[02:51:49] <pfred1> problem resolution is formulatic in nature
[02:52:01] <pfred1> some get it most don't
[03:12:37] <Jymmm> you can't find your incense burner, what do you do???
[03:12:58] <Valen> get eaten by a rat?
[03:13:06] <Valen> wait this *is* nethack right?
[03:13:25] <Jymmm> You grab a 1x4" aluminum angle, drill a hole, and instant burner! woo hoo
[03:14:37] <Valen> ahh
[03:15:32] <Jymmm> nethack?
[03:17:23] <pfred1> Jymmm text game
[03:17:30] <Jymmm> ah
[03:17:42] <pfred1> sort of like a single user mud
[03:18:09] <pfred1> text myst?
[03:18:54] <pfred1> wth is wrong with this library part i made?
[03:34:04] <jimbo> 'when using PYVCP can you lay a led color over a button?
[03:34:40] <jimbo> IE. Have the button color be dynamic.
[03:35:20] <pfred1> oh I'm an idiot i know why it won't route to the part
[03:39:37] <cstop> why?
[03:40:06] <pfred1> when i laid out the pads i used some crazy grid now they're not on the grid
[03:40:22] <pfred1> I'm getting gridlocked ;)
[03:40:31] <cstop> Pray tell
[03:40:44] <cstop> Locked out by a wire
[03:41:02] <pfred1> its a really stupid package eagle can't hit things unless they're on grid crosshairs
[03:41:46] <cstop> every book has it's story
[03:41:46] <pfred1> it happens to me all the time in the schematic section its usually not an issue in the board section
[03:42:03] <cstop> details details...
[03:43:26] <pfred1> hmm it still won't route to this part
[03:45:01] <pfred1> yes it will!
[03:45:21] <cstop> Joy and pleasure!
[03:46:49] <cstop> Off now Ta Ta
[04:05:39] <pfred1> oh this is driving me bananas
[04:23:05] <Jymmm> (short drive)
[04:28:21] <tom3p> has anyone tried the CGCC ? i loaded the sample, and get 'near line 3... bad char 'c' used'.
[04:28:22] <tom3p> CGCC dir is in the rip dir in use, and I added '$HOME_DIR/CGCC: into the path line in scripts/emc-environment
[04:28:49] <tom3p> (it reads fine in the AXIS display
[04:30:40] <elmo40> cgcc? why use that one?
[04:32:20] <tom3p> CGCC (not some gcc ) is a new extension to the gcode processor for emc2, it allows writing c-style syntax to create gcode
[04:32:51] <tom3p> http://tsemsb.blogspot.com/2010/04/some-cgcc-features-and-examples.html
[04:34:49] <elmo40> ahh, I see
[04:35:03] <elmo40> g-code c-compiler thingy
[05:11:55] <tom3p> to get CGCC to work, add a directory ~/CGcode2 and put the unzipped dir 'CGCC' into that. when unzippiung make sure the directories are rw and that the files are not executable
[05:13:00] <tom3p> hmmm looks like it misses the last row last column in the example
[05:14:59] <tom3p> hell it misses the 1st row 1st column!
[05:26:52] <tom3p> it misses a LOT of positions, very odd pattern too ( 2nd col in 2nd row, 3rd col 3rd row...)
[05:27:07] <tom3p> this explains 1st col 1st row & last col last row, at least its consistant
[05:29:00] <tom3p> oh he does have a 'if x == y clause, nice example, obfuscated enuf to be really helpful to a user
[05:30:01] <tom3p> ok, remove the tricky index skipper, and its a nice grid of 10x10 now
[05:58:06] <awallin_> anyone using Metrol tool-length sensors? http://www.9xx-parts.com/temp/elekpak.jpg
[05:58:30] <awallin_> the one lower right is for a lathe
[06:10:29] <tom3p> the cross like bit? ( not the small prox sensor )
[06:17:49] <Jymmm> Hey damnit! I've been looking for that armored cable!
[06:17:57] <tom3p> i think the CGCC cant handle index-- in c style for loop eg for ( x = lastcolumn; x >=0; x-- ) { seems to borq the interp
[06:18:04] <Dave911> Did a Git clone the other day and compiled the master..2.5_pre. looks like the TKEmc display is broken for some reason.. Not much on the dmesg and an error window doesn't even pop up. Axis and Touchy are fine.
[06:18:15] <tom3p> the kind that makes the 'zip'sound as it gets dragged around
[06:18:31] <Jymmm> tom3p: Yep =)
[06:19:11] <Jymmm> or phone booth type. hard to find, and EXPENSIVE!
[06:19:31] <Jymmm> I need to make sure it's not zinc coated
[06:20:22] <Jymmm> ...and MSDS's dont apply to armored cable.
[06:24:16] <tom3p> looked plastic coated
[06:24:43] <Jymmm> I think it's actually paint.
[06:25:21] <tom3p> the touch off device looked interesting, like 4 rollers on a maltese cross, i image it pivots or has some plate over it
[06:26:14] <tom3p> the prox swx is just a Baumer or Festo 3mm 3 wire (blue 0 Brown + Black signal)
[06:26:37] <Jymmm> http://sfbay.craigslist.org/sby/tls/1681779626.html
[07:06:50] <tom3p> hacked CGCC examples and notes on the -- and -= operators http://imagebin.ca/view/k-XN6b2C.html http://pastebin.ca/1858633
[09:09:39] <piasdom> g'mornin all
[10:36:12] <awallin> who decides what happens with the ubuntu updates? I've been running 9.10 and now suddenly my firefox is called "Namoroka"... WTF?
[10:37:09] <Valen> thats interesting
[10:37:53] <Valen> ahh, thats just a version name
[10:38:02] <Valen> https://wiki.mozilla.org/Firefox/Namoroka
[11:23:00] <frallzor> dust!
[11:24:32] <JT-Dev> weee the flags showed up on the user map today
[11:26:48] <alex_joni> JT-Dev: nothing changed ;)
[11:27:01] <JT-Dev> did on this end :P
[11:27:04] <alex_joni> ah
[11:27:38] <frallzor> My mill is up and running now! Its working like a charm except I dont get the "sharp" corners I want, aka bad acceleration =)
[11:27:49] <frallzor> trial and error
[11:28:05] <Valen> you could drop the feed rate to compensate
[11:28:18] <frallzor> what is the G-code to make it follow it more by itself?
[11:28:24] <frallzor> so it slows down at corners and such
[11:28:47] <Valen> I was reading that recently
[11:28:53] <frallzor> G-something
[11:28:59] <frallzor> =P
[11:29:00] <JT-Dev> http://www.linuxcnc.org/docview/html//common_User_Concepts.html#r1_1
[11:29:02] <Valen> yeah with some numbers after it
[11:30:11] <JT-Dev> and http://www.linuxcnc.org/docview/html//gcode_main.html#sub:G61,-G61.1,-G64:
[11:30:46] <frallzor> G64 P seems what I want
[11:31:08] <frallzor> *to be
[11:31:18] <frallzor> speed and a tolerance
[11:31:41] <frallzor> wonder if one could tell artcam to implement those lines
[11:32:53] <JT-Dev> does it have a post that you can edit?
[11:32:58] <frallzor> hmm
[11:33:04] <frallzor> good question
[11:33:34] <JT-Dev> I changed my sheetcam post to give me more or less what I wanted for my plasma cutter
[11:35:39] <frallzor> I honestly cant say if artcam has that
[11:35:50] <frallzor> I know I can edit my solidcam PP to perfection
[11:36:07] <JT-Dev> * JT-Dev wonders why the email notice from the forum from line is "EMC Support Forum Forum"
[11:36:38] <SWPadnos> because the name of that forum is "EMC Support Forum", and the software adds "Forum" to the end
[11:36:50] <JT-Dev> ah ok
[11:37:39] <alex_joni> obviously
[11:37:57] <alex_joni> at least it's not EMC Support Forum Forum Forum
[12:40:14] <JT-Hardinge> so after you do a T1 M6 if you don't do a G43 before touching off G54 you get a suprise
[12:40:24] <JT-Hardinge> surprise
[12:41:18] <frallzor> what? =)
[12:41:35] <JT-Hardinge> you don't end up where you thought you would
[12:41:54] <JT-Hardinge> tool turret on the lathe
[12:42:19] <skunkworks> I had issues with tool length going from program to program because I didn't turn off tool lenght comp. (took a bit to figure out what was happining)
[12:44:39] <JT-Hardinge> with g49?
[12:45:37] <JT-Hardinge> that was scary moving the parting tool to within 0.125" of the collet face to part off my part
[12:47:39] <JT-Hardinge> hmm should have drilled the part a bit deeper but it will be ok
[12:48:42] <JT-Hardinge> * JT-Hardinge heads off to the other shop to mill the rest of the features on this tiny little part
[12:51:04] <skunkworks> yes
[12:51:13] <skunkworks> well when he is back on
[13:08:05] <skunkworks> this is interesting... http://cncloisirs.com/Construction/Entra%EEnementVictorSendas
[13:08:25] <skunkworks> JT-Work: yes - g49
[13:08:43] <JT-Work> ok thanks
[13:10:29] <skunkworks> the issue was - I would set the tool length in mdi between programs but the last tool used was still in effect. Then when the new program started - it used the first tool of that program - effectivly changing my offset.
[13:10:51] <mikegg> huh. I wonder if he turns the inner race to mate with the threads?
[13:11:13] <skunkworks> mikegg: doubt it.
[13:11:30] <JT-Work> ok, so you had two offsets added together when you did the mdi offset
[13:11:59] <mikegg> looks neat, but I bet with a long lead screw it would jump
[13:12:32] <JT-Work> how do you do the wayback machine I forget emc logger or something like that
[13:12:56] <skunkworks> logger_emc: bookmark
[13:12:56] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2010-04-08.txt
[13:13:17] <JT-Work> thanks
[14:09:28] <cradek> * cradek wonders how clip9's tests are going
[14:11:07] <skunkworks> :)
[14:11:28] <skunkworks> it is a little unnerving for something like that to show up - isn't it ;)
[14:12:40] <awallin> what?
[14:13:26] <alex_joni> awallin: some part gouging
[14:13:47] <awallin> due to CAM or EMC2?
[14:13:59] <skunkworks> so far - the hope is that he had hardware issues (memory check showed errors)
[14:14:00] <alex_joni> cradek: seen http://svarteper.com/images/ram_errors.jpg ?
[14:15:56] <skunkworks> awallin: the machine path didn't match programed path for a bit in a large program
[14:16:04] <elmo40> RAM errors suck. Plus, DDR RAM is getting more expensive these days
[14:17:02] <skunkworks> because it is old ;)
[14:17:50] <skunkworks> we still have machines with pc100/pc133 (I think that is the oldest we have now)
[14:20:17] <skunkworks> no - I think we still have one with simm in it
[14:22:10] <cradek> skunkworks: yes I saw that - wondering if replacing the ram fixed it.
[14:24:12] <skunkworks> cradek: if you do a g64p0 - will it exact stop?
[14:24:26] <skunkworks> or is it the same as g64nothing
[14:24:54] <cradek> same as nothing I think
[14:25:20] <skunkworks> * skunkworks doesn't have a machine booted
[14:40:36] <JT-Work> skunkworks: http://www.linuxcnc.org/docview/html//gcode_main.html#sub:G61,-G61.1,-G64:
[14:42:24] <elmo40> skunkworks: simm? DAMN, that is old.
[15:01:48] <cradek> skunkworks: I have some 256k simms in the drawer in the bathroom. I don't know why they're there - they've been there a really long time.
[15:07:03] <skunkworks> heh - that is funny.. I have a lot of things like that. a p4 proccessor in the kitchen junk drawer...
[15:07:27] <skunkworks> (doesn't everyone?)
[15:07:29] <cradek> heh, that's brand new junk.
[15:07:35] <cradek> oh probably
[15:08:14] <skunkworks> I think it was a litter faster then the one in the main computer at home.. but I never got around to switching it out.
[15:21:15] <skunkworks> cradek: my portable config writing machine. (I need a bigger screen) http://imagebin.ca/img/E1TpV5I.png
[15:21:22] <skunkworks> ;)
[15:25:38] <tom3p> skunkworks, that cockeyed bearing idea is interesting. but he's really just playing with the connection of 'nit' to screw and his way is iffy for a big torque reversal (read: jump a thread)
[15:25:53] <tom3p> nut
[15:26:02] <skunkworks> right
[15:27:06] <skunkworks> I would want 2 - torquing opositely (not that I would ever use that in a real machine)
[15:27:15] <skunkworks> opposite
[15:27:39] <frallzor> http://www.vimeo.com/10774348 besides it taking freedom on the curves, pretty nice? =)
[15:28:24] <skunkworks> freedom on the curves?
[15:28:42] <frallzor> yeap, not going G64 PX =)
[15:29:11] <cradek> did you know that increasing your accel setting will help too
[15:29:16] <frallzor> yup
[15:29:21] <cradek> looks fast - maybe you have more accel available
[15:29:27] <frallzor> but I cant go higher than 2200
[15:29:35] <frallzor> then it stalls a bit
[15:29:38] <cradek> ah
[15:29:48] <cradek> yep then G64 P is your solution
[15:29:57] <frallzor> but that was when I was doing back and forth on a dime manually
[15:30:08] <frallzor> might do better when running code
[15:30:21] <cradek> yeah but you don't want to risk that
[15:30:40] <skunkworks> oh look - it is a video. duh
[15:30:41] <frallzor> woow I shot most of the mdf except the milling
[15:31:49] <frallzor> Im pretty happy with it
[15:32:00] <skunkworks> that is quick! nive work
[15:32:02] <skunkworks> nice
[15:32:14] <frallzor> and ppl say steppers suck =(
[15:32:17] <frallzor> shame on them
[15:32:59] <frallzor> im surprised I got the alignment of the spindle perfect, perfect flat milled bottom too =)
[15:35:42] <frallzor> if you look at the 2nd letter in the video
[15:36:05] <frallzor> why does it do a big curve like that there and not in the lower left?
[15:36:21] <frallzor> upper left = bigger but lower is a good as it can get with 3mm tool
[15:36:28] <frallzor> *is as
[15:38:08] <skunkworks> because that is where it starts and stops the path
[15:38:20] <frallzor> ah
[15:38:43] <frallzor> but G64 Px will make all look like that then? if set at a reasonable value?
[15:38:50] <skunkworks> yes
[15:39:05] <frallzor> will try that later then =)
[15:42:41] <JT-Work> nice a Samsung TFT LCD touch screen just walked in the door
[15:43:31] <tom3p> dman it has hands AND legs
[15:44:15] <JT-Work> and a complete computer hanging off the back side
[15:45:37] <Jymmm> QUICK! KILL IT!!!!!
[15:47:23] <Jymmm> Anyone seen one of these before? Is it lower one quill, the others lower at the same time? http://sfbay.craigslist.org/sby/tls/1682139057.html
[15:47:56] <SWPadnos> it doesn't look like they're linked
[15:48:37] <clip9> cradek: I've done 4 dry runs and one real milling. No errors.
[15:49:00] <SWPadnos> they may have power feed, which would improve throughput by allowing a spot drill/drill/ream cycle, even for a single hole in a workpiece
[15:49:01] <clip9> It's looking good.
[15:49:01] <tom3p> its a manual station take part in fixture pull handle 1, move to next dowel pin pon table, pull handle 2, ... (i'm an engineer :)
[15:49:42] <Jymmm> SWPadnos: ok
[15:50:39] <SWPadnos> even without power feed, it's like tom3p said - just drilling a small hole at high speed before making the full size hole improves speed and tool life
[15:50:52] <JT-Work> hmm, it is missing a file digitouch.exe
[15:53:33] <JT-Work> * JT-Work wonders where to find drivers for the touch screen
[15:54:00] <SWPadnos> evtouch
[15:54:00] <Jymmm> http://TouchScreenDrivers.com/
[15:54:12] <Jymmm> * Jymmm ducks
[15:54:29] <DaViruz> that looks like a highly reputable site
[15:54:48] <JT-Work> Jymmm: thanks
[15:55:00] <Jymmm> JT-Work: It was a joke
[15:55:11] <JT-Work> thanks
[15:55:26] <atmega> I have touchscreens from lots of different mfgrs, they are all ELO
[15:55:40] <JT-Work> what is ELO?
[15:55:44] <Jymmm> But this might help... http://www.google.com/search?q=evtouch
[15:55:52] <atmega> http://www.elotouch.com/Support/Downloads/default.asp
[15:57:21] <Jymmm> JT-Work: where did you get the LCD from?
[15:58:01] <cradek> JT-Work: many brands are supported by xorg packages that you install from synaptic
[15:58:03] <JT-Work> it walked in the door this morning
[15:58:19] <JT-Work> ok
[15:59:38] <JT-Work> I'll plug in a cd drive when I get it home and load EMC
[16:00:39] <SWPadnos> if you have an Ubuntu machine nearby, you could try just plugging it in (the USB or serial connection)
[16:00:59] <SWPadnos> oh right, it's got a computer wart on it - nevermind
[16:06:45] <Jymmm> pc wart?
[16:14:38] <frallzor> T1M6
[16:14:38] <frallzor> G64 P0.1
[16:14:38] <frallzor> G0Z5.000
[16:14:38] <frallzor> G0X0.000Y0.000S17000M3
[16:14:38] <frallzor> G0X201.387Y159.561Z5.000
[16:14:39] <frallzor> G1Z-6.000F3000.0
[16:14:39] <frallzor> G1X203.154Y165.324F10000.0
[16:14:42] <frallzor> this look correct?
[16:14:50] <frallzor> if I want the G64
[16:15:15] <awallin> yes I think so
[16:15:36] <frallzor> before or after tool shouldnt make a difference,
[16:15:45] <frallzor> (according to me) =P
[16:15:49] <awallin> no...
[16:16:46] <frallzor> time for Ze test
[16:19:07] <kanzure> does anyone know what HPGL is?
[16:19:34] <bill2or3> yes.
[16:19:36] <Jymmm> yes
[16:19:49] <kanzure> righto, so i'm sick and tired of trying to get coreldraw installed on an xp box, and i prefer linux
[16:20:01] <bill2or3> it's Hewlett Packard Graphics Language, what you send to pen plotters.
[16:20:02] <kanzure> i have a laser cutter attached via lpt/parallel/centronics
[16:20:18] <bill2or3> kind of like simpler g-code.
[16:20:21] <kanzure> i can do rastering with the machine, but not vector anything- even when printing from inkscape and so on
[16:20:28] <kanzure> so, will it do cutting with HPGL input?
[16:20:43] <bill2or3> dunno. does it accept hpgl input?
[16:20:48] <kanzure> apparently, yes
[16:20:52] <Jymmm> What version of CorelDraw?
[16:20:59] <kanzure> 15 (X5) or something
[16:21:05] <bill2or3> you can send hpgl with vector-lines..
[16:21:17] <kanzure> i spent four days rebooting and unbooting and debooting windows a few times just to get microsoft .NET installed (ugh)
[16:21:38] <kanzure> i'm tempted to just rip out the boards from the cutter and replace it with something that emc will play with
[16:21:48] <kanzure> but since it's working at the moment, that's probably not a good idea
[16:21:52] <bill2or3> it probably has specific pen numbers for different levels of laser power.
[16:22:04] <bill2or3> examining some of the hpgl that works would probably help you figure it out.
[16:22:08] <Jymmm> HPGL is a well documented standard
[16:22:12] <kanzure> so, do i just cat HPGL-commands to /dev/rs232-or-whatever ?
[16:22:13] <bill2or3> it's just text, humans can read it.
[16:22:53] <kanzure> yes but how do i get it to the machine
[16:22:53] <bill2or3> yeah, but you still need to know the settings that are specific to that 'printer'
[16:23:00] <kanzure> or how do i convert an svg file to hpgl
[16:23:07] <Jymmm> export
[16:23:15] <kanzure> export from where to what?
[16:23:27] <Jymmm> what are you viewing the SVG in?
[16:23:32] <bill2or3> there's a cli tool called "hp2xx", if you're on linux.
[16:23:40] <kanzure> lots of different tools- let's say, inkscape
[16:23:52] <Jymmm> ok, export from inkscape
[16:23:56] <kanzure> bill2or3: cool, thanks
[16:23:57] <skunkworks> JT-Work: are you going to use touchy?
[16:24:04] <kanzure> Jymmm: once i export it to the hpgl file format, then what?
[16:24:09] <bill2or3> but you need to know the specifics of that laser cutter, like what "pen" (hpgl has the concept of different pens), does it use for a cut, etc.
[16:24:10] <kanzure> how do i transmit it to the machine?
[16:24:15] <Jymmm> print it
[16:24:32] <kanzure> Jymmm: how does inkscape know not to send a raster or vector version of the hpgl file?
[16:25:04] <Jymmm> hpgl is vector, look athte file in a text editor like bill2or3 said
[16:25:33] <kanzure> oh, it's the only vector format that my machine will support?
[16:25:44] <kanzure> then why hasn't inkscape been working when i try to print a .svg in vector mode?
[16:25:51] <bill2or3> stop.
[16:25:53] <bill2or3> stop.
[16:26:00] <kanzure> sorry.
[16:26:10] <Jymmm> No, HPGL is a plotter language.
[16:26:27] <bill2or3> the machine accepts HPGL, which is basicially instructiosn like "draw a black line from 23, 45 to 37,44"
[16:26:32] <kanzure> bill2or3: right
[16:26:46] <kanzure> bill2or3: so the question is how to get software (maybe inkscape?) to transmit that to the laser cutter
[16:26:49] <bill2or3> but you have to know what your cutter expects for "draw black"
[16:27:03] <Jymmm> kanzure: I told you, print the hpgl file.
[16:27:03] <kanzure> well, let's separate two issues going on here
[16:27:05] <bill2or3> the same way anything transmits anything to any printer.
[16:27:24] <kanzure> Jymmm: but i asked a follow-up question because that didn't make sense :/
[16:27:38] <bill2or3> "cat cutstuff.hpgl > /dev/ttyUSB1" may very well work, if the hpgl file is correct.
[16:27:54] <bill2or3> what software does the cutter use now?
[16:28:00] <kanzure> custom windows printer driver
[16:28:13] <kanzure> btw i'd be happy to use linux. i would prefer it.
[16:28:14] <bill2or3> can you print to a file?
[16:28:19] <kanzure> yes
[16:28:31] <kanzure> so, if i print a hpgl-file i could examine it and see the differences i suppose
[16:28:38] <bill2or3> so make a graphic of a simple square, print it to a file, and then look at that file.
[16:28:45] <Jymmm> kanzure: Just drop the hpgl file on to the printer (cutter) icon, that easy.
[16:28:53] <bill2or3> it will make a lot more sense.
[16:29:09] <kanzure> have you successfully used cat stuff.dat > /dev/something for commercial machines without hacking it?
[16:29:10] <bill2or3> hpgl is dead simple.
[16:29:34] <sealive> yes hpgl is stadard
[16:29:37] <Jymmm> kanzure: Look, it's REAL EASY, you're making complicated. just do what I said
[16:29:45] <Jymmm> kanzure: Just drop the hpgl file on to the printer (cutter) icon, that easy.
[16:29:56] <kanzure> Jymmm: i will, i will. calm down. i just don't like the windows bullshit, is all, so i try to ask questions
[16:30:11] <Jymmm> kanzure: fine, good luck!
[16:30:31] <sealive> all paint programms does have a hpgl output
[16:30:35] <sealive> as blender also
[16:30:54] <kanzure> sealive: "all paint programs"? i took a half-decade hiatus from windows, ms paint doesn't seem to have hpgl output
[16:31:06] <bill2or3> there's a plugin to inkscape called 'graphtecprint', that's for driving a vinyl cutter, you could probably hack it into working with your cutter.
[16:31:34] <sealive> ms paint does use all the windows HP modes
[16:31:44] <bill2or3> vector != bitmap
[16:31:49] <kanzure> * kanzure agrees
[16:32:32] <kanzure> cool ok, inkscape has the ability to save in .hpgl
[16:32:34] <bill2or3> so you want to be using inkscape, or coreldraw, or illustrator, or something.
[16:32:56] <sealive> corel is the best way to do HPGL
[16:32:58] <bill2or3> but that doesn't mean that it will 'print' without some changes.
[16:33:10] <sealive> you only have to be on 1layer
[16:35:23] <bill2or3> this may help too: http://pldaniels.com/hpgl-distiller/
[16:35:24] <sealive> i used that when i wars on Windows with NC-FRS free CNC
[16:35:46] <sealive> since i am on linux emc is first choice
[16:39:43] <bill2or3> send me the laser cutter, I'll make it work.
[16:40:47] <sealive> how got actually problems with hpgl i misset the first posts
[16:41:13] <bill2or3> kanzure has a laser cutter, and too much coffee.
[16:41:27] <sealive> on witch OS is he
[16:41:28] <Jymmm> bill2or3: what you have?
[16:41:54] <kanzure> when i "print to file" via inkscape (for a ".hpgl" file) and look at the resulting file, it's totally bonkers. when i print it through the laser cutter, it's in raster mode
[16:41:56] <bill2or3> just a diy cnc mill'enstien
[16:42:13] <bill2or3> http://protovision.com/2010/01/31/cnc-mill/
[16:42:20] <kanzure> sealive: i'm on linux most of the time, but the laser cutter is attached to a Windows-box (because of driver issues)
[16:42:24] <bill2or3> bonkers how?
[16:42:26] <sealive> kanzure: witch OS ?
[16:42:30] <bill2or3> and what are you using to look at it?
[16:42:33] <kanzure> bill2or3: i can upload the file if you'd like to look
[16:42:37] <kanzure> bill2or3: notepad.exe
[16:42:37] <Jymmm> bill2or3: that's not a laser!
[16:42:38] <bill2or3> yes, do.
[16:42:52] <bill2or3> jymmm, I know right, and it doesn't even take hpgl.
[16:42:58] <kanzure> bill2or3: also, there's no coffee in my system :) i don't drink the stuff
[16:43:05] <Jymmm> bill2or3: Well, why the hell not?! lol
[16:43:31] <bill2or3> but I learned up on hpgl when making one of those 'Hektor' style vertical plotters.
[16:43:39] <sealive> kanzure: are you on linux?
[16:44:00] <bill2or3> what make/model cutter is it?
[16:44:21] <kanzure> sealive: i just told you?
[16:44:23] <kanzure> bill2or3: uls25ps
[16:45:09] <sealive> oh yea so kanzure why dont you use qcad
[16:45:20] <sealive> google qcad hpgl
[16:45:27] <sealive> and there are all the answers
[16:45:40] <sealive> nice and easy 2d cat
[16:45:40] <kanzure> bill2or3: http://designfiles.org/~bryan/uls25ps/hpgl_to_file.txt
[16:46:12] <sealive> best open CAD 2d soft with the best docu
[16:46:20] <kanzure> sealive: i am having trouble understanding you
[16:46:39] <kanzure> i'll put qcad on the todo list
[16:46:43] <kanzure> right after hp2xx
[16:46:58] <bill2or3> that's not hpgl, it looks like some sort of print-queue entry.
[16:47:09] <kanzure> bill2or3: right
[16:47:15] <kanzure> so, Jymmm's advice didn't seem to work?
[16:47:40] <kanzure> save a .hpgl file, open it up in inkscape, go file->print, set up the printer to print to a file, upload the file and cry :(
[16:47:50] <bill2or3> add in my advice.
[16:48:02] <kanzure> send it to you? :P
[16:48:09] <kanzure> graphtecprint?
[16:48:13] <bill2or3> make a square in the windows software, save it, look at it.
[16:48:26] <kanzure> yeah that's what i did, except i uploaded what the printer is sent instead of the .hpgl file
[16:48:33] <Jymmm> kanzure: I didn't say open the hpgl file in inkscape.
[16:48:41] <bill2or3> *export*, not 'print to file'
[16:48:43] <kanzure> you said to drop it to the printer icon
[16:48:49] <kanzure> export to what?
[16:48:55] <bill2or3> kanzure, listen.
[16:49:02] <Jymmm> kanzure: Yes, the icon in WINDOWS, not Inkscape
[16:49:17] <kanzure> but that makes no sense. why would that be the only way to make this work? O_o
[16:49:31] <bill2or3> it makes no sense *to you*.
[16:50:08] <kanzure> so.. how does it work then?
[16:50:26] <bill2or3> you create your image, or whatever.
[16:50:32] <Jymmm> kanzure: you need to chill out if you want any help in here.
[16:50:33] <bill2or3> save it in hpgl
[16:50:48] <bill2or3> edit it, so that it's got all the proper settings that your cutter expects.
[16:50:51] <bill2or3> send it to the cutter.
[16:50:52] <bill2or3> done.
[16:50:54] <kanzure> btw inkscape only has an "export bitmap" option
[16:51:07] <kanzure> Jymmm: calm down..
[16:51:20] <sealive> kanzure: sorry i'am german
[16:51:28] <kanzure> bill2or3: and the "send it to the cutter" part is a drag-and-drop operation?
[16:51:35] <kanzure> sealive: german isn't that bad :)
[16:52:00] <bill2or3> no, it's a abstract explanation of the next step in the overall process.
[16:52:21] <mozmck_work> ah, sprechen sie deutch?
[16:52:25] <sealive> if you construct with qcad you have a exelent and acurate hpgl file
[16:53:12] <sealive> moz ja
[16:53:17] <bill2or3> in inkscape, file->save_as, file-type "hewlitt packard graphics language file"
[16:53:35] <kanzure> bill2or3: when i open up the resulting .hpgl file it only has the contents: "IN;SP1;"
[16:53:49] <kanzure> (yes i drew a square)
[16:54:07] <bill2or3> * bill2or3 trys the same thing.
[16:54:18] <bill2or3> you drew a square in inkscape, or in the windows program?
[16:54:27] <kanzure> my version of inkscape is 0.47
[16:54:40] <kanzure> i drew a square in inkscape running under windows
[16:55:02] <kanzure> * kanzure tries it on linux
[16:56:15] <kanzure> hm 0.46 can't save to .hpgl?
[16:56:16] <bill2or3> it has to be paths
[16:56:22] <kanzure> not strokes?
[16:56:24] <bill2or3> 'object to path' on everything.
[16:56:46] <bill2or3> paths.
[16:57:14] <kanzure> that did it
[16:57:20] <kanzure> now the file has more content
[16:57:27] <bill2or3> it's simple.
[16:57:51] <bill2or3> PU1741,3268 is "pen up" and coordinates, PD is pen-down, and so on.
[16:58:06] <bill2or3> "SP1" Select Pen #1
[16:59:31] <cradek> RPRPRP
[17:01:41] <skunkworks> well r and p are too far apart to be cat induced... unless they are kneeding..
[17:02:15] <cradek> my mill needs an implementation of RP
[17:02:36] <cradek> (that's hpgl2 for "replot" - make another one just the same as the last)
[17:06:02] <kanzure> bill2or3: dragging and dropping the .hpgl file to the printer icon on windows gives an error: "This file cannot be printed. Please try it in .." (the native application)
[17:06:24] <bill2or3> sorry,I cant help you with windows.
[17:06:55] <kanzure> well. let's move on to linux then
[17:07:15] <kanzure> should i try to use cupsd to work with hpgl output?
[17:08:10] <bill2or3> no.
[17:08:15] <bill2or3> Hell No.
[17:08:31] <kanzure> the device seems to want to be treated as a printer though
[17:08:54] <bill2or3> do you need queueing, and network printing support?
[17:08:57] <kanzure> no
[17:09:04] <bill2or3> cups is overkill.
[17:09:08] <bill2or3> by a lot.
[17:12:56] <bill2or3> there's some stuff you'll need to read on about page 46 of this -> www.hltlasers.com/reference/25-50PSRevisionB1.pdf
[17:13:14] <kanzure> * kanzure has the document :)
[17:13:35] <JT-Work> skunkworks: if I can get the touch screen to work I will
[17:14:01] <kanzure> however, the manual was written for the old 2.74 and 2.76 versions of the driver which worked on windows 3.1
[17:15:52] <bill2or3> the old driver doesn't matter, unless they changed the machine since then.
[17:16:09] <bill2or3> but there''s some details in there about setting the pen, and line-width, etc.
[17:17:18] <kanzure> * kanzure sees it.
[17:20:41] <steverob> Sorry, I got in on the conversation late but, I do have SOME experience using HPGL from windows. Anything I might be able to help with.
[17:22:00] <kanzure> hi steverob. i'd be happy to hear your advice, but i'm seriously contemplating throwing windows out and just going emc
[17:23:00] <steverob> Older versions (up to W2000) had drivers for penplotters. If you set up the computer to use a driver like HP7550a, and redirect the output to a file, you wind up with a text (human readable) file with all the HPGL commands.
[17:24:08] <kanzure> steverob: how do i "properly" send it to the machine after i generate the .hpgl file?
[17:24:15] <kanzure> this looks useful: http://poespoes.nl/inkscape_hpgl/Linux%20inkscape%20save%20as%20HPGL%20file%20for%20pen%20cutting%20plotter%20extension.html
[17:24:23] <kanzure> in particular: cat /myvondutch.plt > /dev/ttyS0
[17:24:32] <kanzure> so, i guess someone else was able to do it via just dumping the file to the serial port
[17:24:49] <steverob> I was contimplating using a script to convert the HPGL to g-code. then run that in EMC.
[17:25:47] <steverob> Sure... I used to just copy the file to the serial port that the plotter was attached to and viola. It'd print/plot the image.
[17:25:55] <bill2or3> that's a lot of overkill.
[17:26:13] <steverob> If your device speaks HPGL, tat should do it.
[17:26:14] <kanzure> huh, i was expecting a communication protocol on top of the serial port
[17:26:24] <kanzure> other than hpgl i mean, like an inbetween layer
[17:26:26] <kanzure> but that's good
[17:26:52] <steverob> Not if the computer is connected directly to the device.
[17:27:46] <steverob> To see how it works, you can use a dumb terminal (or terminal emulator) and watch the commands from the computer to the device.
[17:28:37] <kanzure> steverob: i think i'll opt to install linux (er, run a livecd) on the machine (my laptop here doesn't have serial or parallel (wtf)) and try out dumping the file straight down the pipe
[17:28:57] <sealive> at 9600baud
[17:29:17] <sealive> on my cutter is a dip field
[17:30:58] <tom4p> i tried the smi fix from the wiki, emc2 failed to load saying it cant find rtap_smi.ko my cpu is a P3 coppermine but i think emc doesnt believe its intel.
[17:30:59] <tom4p> the rtapi.conf http://pastebin.com/LPMAicH9, the error from emc2 http://pastebin.com/JbpDsjgn
[17:31:28] <tom4p> the mobo has 2 such cpus
[17:32:23] <steverob> Before building my own machine, I was contimplating using an old plotter and a dremel to make PCB's. I could use eagle and print to the plotter and it would outline all the traces. If I had a way to mount a cutting tool, It wold make the PCB's for me.
[17:32:47] <bill2or3> it might, but only big pcb's.
[17:32:54] <bill2or3> p.s. dremels suck for that.
[17:33:42] <sealive> steverob: http://www.youtube.com/watch?v=PJIA3-U23Cs
[17:34:00] <sealive> steverob: my one
[17:35:47] <steverob> Or you can use virtually any CAD program and generate the hpgl just by "printing" the file. If you capture the hpgl to a file, convert it to G-code, you have everything you need to mill a 2D thngy.
[17:36:13] <sealive> yea :D
[17:39:00] <steverob> I did a little milling on a PCB with my machine and I think it will (eventually) work OK. I am still loosing some steps and the machine isn't quite tight enough to give good results. Once I get those issues resolved, I think it'll work. It may not do SMD or other close routing but for the simple boards that I've been making, it will probably be OK.
[17:40:48] <tom4p> why would rtapi fail to recognize my PIII coppermine as intel ?
[17:40:49] <tom4p> ( why does emc2 fail to load after failing to find rtapi_smi.ko )?
[17:47:49] <tom4p> in /list i dont see any rtai rt or rtapi groups, anyone know where to find one?
[18:00:37] <tom4p> no luck with smi , so removed that edit to rtapi.conf
[18:02:25] <tom4p> how do i edit xorg.conf to try the 'vesa' driver instead of the 'nv' driver?
[18:02:25] <tom4p> the text of xorg.conf says Section "Device" Identfier 'Configured Video Device'
[18:02:39] <SWPadnos> driver "nv"
[18:02:51] <SWPadnos> or driver "vesa"
[18:03:17] <SWPadnos> these days, anything left out is auto-detected, so the file is essentially blank
[18:03:35] <tom4p> supplant "Identifer 'Configured Video Device'" with 'driver 'vesa'' ?
[18:04:04] <SWPadnos> add the driver line to the "Configured Video Device" section
[18:04:14] <SWPadnos> or replace any existing driver line
[18:10:01] <frallzor> anyone into artcam?
[18:10:16] <frallzor> want to fix the "issue" that they use short lines for arcs instead of real arcs
[18:10:30] <tom4p> SWPadnos: thx, now in vesa ( and thank god it wasnt 640x800 AXIS hidden button game )
[18:10:38] <SWPadnos> heh
[18:18:12] <frallzor> when using G64, is it possible to make it a bit more "fluid" somehow?
[18:18:22] <frallzor> feels a bit jerky sometimes
[18:19:11] <SWPadnos> use G64 Pxx
[18:19:42] <SWPadnos> the larger the "xx", the smoother the path will be (but it will have higher deviation from the programmed path)
[18:20:50] <frallzor> I guess slowe speed would help too
[18:20:55] <frallzor> *r
[18:27:58] <skunkworks> frallzor: the higher accelleration your machine has - the more it can keep the speed up..
[18:28:54] <frallzor> I know, but if fast accel = still jerky as i call it =)
[18:29:08] <frallzor> but It seems its an artcam issue
[18:29:19] <frallzor> arcs = small lines
[18:29:25] <frallzor> = lots of movementpoints
[18:30:17] <frallzor> seems to be a proper arc output, will try that
[18:30:20] <cradek> G64 Px is called the naive cam detector for a reason - it was done just to help with this kind of shit gcode
[18:30:52] <frallzor> Ill try both arcs + G64PX
[18:33:24] <tobs> what types of homing switches did you use for an exact homr
[18:39:04] <cradek> do you have a servo machine with encoders? if so, the switches don't matter much, because the encoder's index pulse is what gives you the exact home position.
[18:39:20] <cradek> * cradek just wasted seconds of his life
[18:47:20] <Jymmm> cradek: just set ALL your clocks backwards and regain those precious moments
[18:47:40] <cradek> Jymmm: suddenly I'm young again
[18:54:11] <Dave911> cradek: This might be a stupid question.... but is Touchy only grayscale or am I missing something? Jymmm... a man with a solution.. ;-)
[18:55:39] <cradek> it uses your system theme for colors, so it matches other applications
[18:55:56] <cradek> it's true there are no rainbows and stuff - I keep meaning to add them, but I haven't yet. :-)
[18:56:59] <Dave911> Ahhh... see I told you a stupid question.. :-) So I need to alter my system theme.. probably a linux setting I need to find... rainbows ... I've chased a few of those.. :-)
[18:58:17] <Dave911> When you run Glade 3.0 does it bog down your PC?? I'm trying to run 3.6.7 on Karmic and my PC almost goes to sleep waiting for screen updates... have you seen anything like that?? Not sure what to do..
[18:59:04] <Dave911> The CPU goes to 100% and stays there for 10+ seconds sometimes.... I can't believe that everyone has this problem with Glade..
[18:59:08] <cradek> I used glade2, I haven't seen glade3
[18:59:52] <Dave911> OK, well I guess I'll visit the Glade IRC and see what they say... way too slow.. Thanks! :-)
[19:00:35] <micges> glade 3.6 is mess
[19:00:52] <Dave911> Uh oh....
[19:01:02] <micges> don't know if it will be usabe without deeply rewrite
[19:01:16] <Dave911> Is that why they call 3.6.7 the horizontal version .. :-(
[19:01:26] <micges> hi cradek, Dave911
[19:01:55] <Dave911> Hi micges :-)
[19:02:46] <Dave911> micges: Have you tried the 3.7 version they have out now.... suppose to be a lot of 3.6 issues fixed.... I thought I could get by with 3.6.7
[19:03:15] <micges> nope
[19:03:56] <micges> If I have time I'll import few of my projects to 3.7
[19:04:26] <motioncontrol> micges, excuse for quastion . the 2.4 version the emc2 have the y axis tool lengh compensation ?
[19:04:33] <Dave911> OK, well I guess I will "git" it and see what happens before I go back to an older version..
[19:04:38] <micges> motioncontrol: yes
[19:05:19] <motioncontrol> micges, use a new format for table file ?
[19:05:28] <micges> yes
[19:05:48] <micges> but it will import old one and convert it automatically
[19:07:12] <motioncontrol> micges have some pin for select the tool compensation active example machine with 2 spindle, one orizzontal and one vetical '
[19:08:07] <micges> no
[19:08:25] <micges> you have one compensation for one axis only
[19:09:50] <motioncontrol> how setect the compensation axis ?
[19:11:24] <motioncontrol> micges, excuse my english . how selection the axis on i want tool lengh compensation ?
[19:16:18] <micges> motioncontrol: if you change it dynamically you must use 'G43.1 Xn Yn'
[19:16:20] <tom4p> do i throw these tyan S1832D's in the trash?
[19:16:20] <tom4p> unexpected realtime delays : afik tried all the tricks on wiki...removed usb, killed all APM/ACPI in BIOS, using vesa, unsuccessful try of smi 'fix' ( cant find rtapi_smi.ka thinks the P3 coppermines arent Intel).
[19:22:24] <skunkworks> tom4p: onboard video?
[19:22:43] <SWPadnos> one CPU or two?
[19:22:55] <tom4p> no agp , geforce 5000 was using 'nv' now using 'vesa'
[19:23:02] <tom4p> two cpu
[19:23:10] <andypugh> That's a relief! My eBay "AC servo motors" actually are the permanent magnet BLDC motors I hoped they were. Now I just need to work out how to drive them :-)
[19:23:27] <skunkworks> AMC drives!
[19:23:47] <SWPadnos> tom4p, well, you could try using only one CPU
[19:24:28] <SWPadnos> I think Intel motherboards have a shared memory bus for all the CPUs, so when the second CPU accesses memory, the first one has to wait
[19:24:34] <andypugh> The slight fly-in-the-ointment is that they are resolver-only feedback. But I can think of several solutions, only some of which cost >$500
[19:24:40] <tom4p> ok, thx, anything else while i tear down ?
[19:25:12] <tom4p> andypugh: jon elsons' i/f for resolvers?
[19:25:24] <skunkworks> andypugh: jonE has resolver > encoder converters..
[19:25:27] <skunkworks> heh
[19:25:42] <andypugh> Yes, but that doesn't immediately help with commutation.
[19:25:43] <tom4p> bbl
[19:25:52] <tom4p> //part
[19:26:00] <skunkworks> the resolvers did the comutation?
[19:26:21] <andypugh> Well, the resolvers go into the drive, and that understands where the poles are.
[19:26:36] <andypugh> But I don't have the drive.
[19:26:38] <PCW> Resolver are almost always used for commutation
[19:26:50] <andypugh> Hi Pete.
[19:27:16] <PCW> Hi, finally dragged myself into work..
[19:27:17] <andypugh> I should point out that I have been chatting to Pete in email much of the afternoon about thism
[19:27:25] <skunkworks> heh
[19:27:45] <skunkworks> i suppose - they are absolute for one revolution.
[19:28:05] <PCW> Yes sometimes 1/2 rev
[19:28:12] <andypugh> Resolvers are in many ways nice things.
[19:28:40] <PCW> tough as the motor they are attached to. maybe tougher (no magnets)
[19:28:47] <andypugh> And I think I could quite easily add hall sensors to the motors to use a normal dumb drive.
[19:29:00] <Dave911> no glass disks :-)
[19:30:04] <Dave911> Aren't those motors like 4KW each ??
[19:30:14] <PCW> You could also synthesize the Hall signals from the absolute angle and have EMC output them
[19:30:18] <andypugh> Although another idea might be to use the AD2P1200 chip which converts resolver to quadrature (for EMC2) and also 10-bit parallel and find some simple way to use that parallel data to simulate hall sensors via a very simple look-up table.
[19:30:26] <Dave911> I remember your Ebay accident ...
[19:31:01] <andypugh> The motors are 540W (7.5Nm peak, they feel good and strong on 30VDC on one phase)
[19:31:22] <Dave911> Oh.. I thought they were much larger...
[19:31:30] <Dave911> Nice size
[19:31:46] <andypugh> Yes, about NEMA 34 size. Any bigger would bend my lathe bed..
[19:32:23] <PCW> Ya, that's a good servo size for a home machine, 220v?
[19:32:47] <andypugh> 345V, but I doubt I will go that high.
[19:33:15] <andypugh> As I say, they hit rated steady-state current at 30V
[19:33:22] <PCW> 345 is probably DC bus rating, so 240V
[19:34:02] <PCW> which means 30V is fine as long as they dont move :-)
[19:34:38] <andypugh> I was just making absolutely sure they were permanent magnet machines.
[19:35:33] <PCW> put the wires in your hand and give it a good spin, that will tell you...
[19:35:53] <andypugh> The manufacturer datasheet I was pointed at by the eBay ad described them as asyncronous and synchronous on the same page.
[19:36:41] <andypugh> I paid £50 for three, I see the same motor second-hand on eBay for $500 BiN (not that anyone seems to be doing so)
[19:38:08] <PCW> I dont think they call induction motors "servo motors" very often
[19:38:10] <PCW> good deal! (if shipping was not too bad)
[19:38:19] <andypugh> Shipping was £7
[19:39:54] <PCW> Maybe you can be a guinea pig for the 7I49. making it work well with a range of resolvers promises to be fun
[19:40:07] <andypugh> Does adding Hall sensors and a little magnet sound plausible? I saw a web page of electic bicycle enthusiasts adding them to sensorless outrunners to use more robust drives.
[19:41:10] <andypugh> 7i49 + the Hostmot2 3-phase PWM certainly sounds like a nice elegant solution.
[19:41:36] <JT-Hardinge> pcw: Plasma is back up and running... thanks for the quick turn around on the 5i20
[19:42:40] <andypugh> I find myself wondering if the sin-cos resolver output could go directly into the invclarke comp module as the cartesian inputs to output PWM magnitudes for the phases.
[19:42:52] <PCW> JT-Hardinge: NP sorry for the bad one, looks like a lemon lot of Xilinx
[19:43:20] <skunkworks> PCW: sure - blame the supplier... ;)
[19:43:36] <PCW> Of course...
[19:44:04] <skunkworks> I still have to purchase another 7i33
[19:44:27] <skunkworks> deciding if I should by 2 with terminals and save the one I have for another project.
[19:45:49] <skunkworks> *screw terminal
[19:46:06] <PCW> andypugh should work but multiplied by the PID output (haven't looked at the invClark comp)
[19:47:32] <PCW> (note that the sin/cos resolver output have the carrier component that needs to be removed)
[19:47:34] <andypugh> I think you would need to threshold (possibly at zero) to control the normal drivers.
[19:48:51] <andypugh> Do you need to remove the carrier? I have been thinking about this, and I think that the _ratio_ of the two signals gives the position regardless of the instantaneous absolute magnitude?
[19:49:21] <andypugh> (as long as there is enough magnitude to actually divide)
[19:49:50] <PCW> Well the magnitude is 0 twice a cycle...
[19:50:06] <andypugh> Details!
[19:50:14] <andypugh> :-)
[19:50:33] <PCW> Yes atan(2) of the magnitudes gives you the pos
[19:51:08] <andypugh> I am pondering buying a PIC programmer. I used to be able to write machine code..
[19:51:51] <andypugh> And I am _so_ short of projects :-/
[19:52:48] <PCW> A DSPIC should be able to read 1 resolver fairly well...
[19:52:50] <PCW> "And I am _so_ short of projects :-/" ---->>> Hows your clock coming?
[19:53:44] <andypugh> This is a key enabling step to the clock project. Actually it isn't, I have no excuse not to start making the clock tonight, I have a perfectly functional CNC mill, lathe and hobber.
[20:09:37] <tom4p> SWPadnos: so far, removing cpu2 seems to have fixed it.
[20:09:38] <tom4p> i found out writing a while [1 EQ 1] ... wrapper around 3DChips wasnt a good idea ( emc preloads the whole file )
[20:10:08] <tom4p> thx much
[20:10:57] <andypugh> A friend just suggested an approach: Use the 12-bit output of the Analog Devices resolver interface as the address lines of an EEPROM which is programmed with the required commutation pattern. Sledgehammer to crack a nut, but pretty cheap.
[20:11:40] <Jymmm> Dave911: Wouldn't have a clue about touchy
[20:11:50] <Jymmm> Dave911: Wouldn't have a clue about touchy, sorry.
[20:11:59] <SWPadnos> tom4p, great! glad it worked
[20:12:28] <tom4p> gonna let it cook for a few days
[20:12:44] <SWPadnos> andypugh, plus additional logic to actually strobe the read line
[20:12:47] <Dave911> Hi Jymmm.... Cradek answered my question... I think I am set.
[20:13:04] <Jymmm> Dave911: cool beans =)
[20:13:49] <Dave911> Thanks Jymmm... very cool indeed :-)
[20:15:37] <tom4p> haha new smi fix... rip the 2nd core outta the die
[20:15:52] <SWPadnos> heh
[20:16:39] <cradek> andypugh: I tried a similar thing with a traditional eprom and it was way (way) too slow. Check before you try...
[20:17:28] <SWPadnos> an AVR is less expensive than an EEPROM (I bet), and can also be used as a lookup table :)
[20:18:20] <SWPadnos> but you'll only get ~1MHz updates, unless you use an external crystal/resonator (I think 8MHz are the fastest internal clocks they have)
[20:20:03] <andypugh> cradek: 150nS read time sounds OK?
[20:20:38] <motioncontrol> good day at all i close
[20:21:26] <tom4p> 3x (adc settling + lookup + ??) sounds like a lot to do at high rpm
[20:22:50] <andypugh> tom4p: I wasn't planning on any ADC, this is just 12 bits of data in, 3 digital bits out.
[20:23:33] <tom4p> whered the 12 bits come from? wasnt there a conversion from sinus to number
[20:24:10] <andypugh> And I guess you could throw away the lower-order 8 bits quite happily (or use them to strobe the clock on a ROM)
[20:25:00] <tom4p> gotta eat, sat at this box since sunrise,
[20:25:04] <tom4p> thx all bbl
[20:25:09] <andypugh> I was looking for quick and easy, and the AD2S1200 chip does all the hard stuff. It connects to a resolver and outputs quadrature and 12 bits of binary
[20:25:20] <tom4p> cool
[20:25:27] <tom4p> nice hdwr
[20:26:44] <andypugh> You know, I bet you could do something with simple TTL logic on the top 4 bits..
[20:28:20] <SWPadnos> an AVR is probably less expensive than a couple of TTL chips ;)
[20:34:23] <andypugh> Yes, but I don't have any way to program an AVR. I wouldn't even know where to start.
[20:35:18] <atmega> you start at http://www.avrfreaks.net/
[20:35:46] <andypugh> atmega: I suspect you are a fan?
[20:36:08] <atmega> how did you guess?
[20:36:24] <JT-Hardinge> hmm, this sony touch screen mount fits the hole in the Hardine almost exactly
[20:36:38] <andypugh> I am reading the Wikipedia page in order to find out what the heck they are
[20:38:09] <skunkworks> micges: do you know of this guy? :) http://cnczone.com/forums/showthread.php?t=98554&highlight=axis
[20:38:41] <mozmck_work> heh, a tip for avrfreaks: if you want to search it forget their search function and do a google search with site:avrfreaks.net
[20:40:10] <andypugh> http://u2.mattair.net/ looks useful.
[20:40:11] <mozmck_work> you can get an AVRISP MKII to program the chips for about $35, and there's a port of gcc for avr to write programs in C
[20:40:31] <atmega> andypugh: get an arduino to play with, you can also use it to program non-arduino'ed AVR's
[20:41:28] <micges> skunkworks: I've heard about his machine on polish forum
[20:41:32] <micges> nothing more
[20:50:18] <andypugh> atmega: 6 lines of A/D sounds a lot like 3 resolver inputs. Hmm. Pity there are not 18 digital outputs, that would be ideal.
[20:53:04] <mozmck_work> andypugh: these days you can get an avr32 as cheap as an 8 bit avr and they pack a lot of power.
[20:53:27] <SWPadnos> andypugh, very easy to program. I use assembly pretty much exclusively, it's so simple
[20:54:03] <andypugh> I am looking at the Arduino Mega which has 54 IO pins. That would work.
[20:54:16] <SWPadnos> eek
[20:54:51] <bill2or3> fwiw, seeedstudio.com has their "Seeeduino Mega", which is a few bucks cheaper, and has some extra features.
[20:55:18] <mozmck_work> SWPadnos: did you see that they have an fpu on some of the avr32 uc chips now?
[20:55:24] <SWPadnos> I did see that
[20:55:37] <SWPadnos> but I didn't notice any available chips with it
[20:55:48] <mozmck_work> that could be interesting at some point.
[20:56:02] <tom3p> skunkworks, the 4 rail Z axis is good idea, hes got maybe a foot cross section on Z and at most 18inches on the traveling beam at top. thats not much for how far he's extended the spindle
[20:56:05] <SWPadnos> it was more like "we're developing an FPU peripheral for the 32-bit micros"
[20:56:19] <mozmck_work> I don't know if you can get them yet or not, but I think it will be on the new uc3c series.
[20:56:30] <SWPadnos> could be
[20:57:05] <SWPadnos> they've got an interesting architecture for the AVR32 and ARM chips - they more or less use the same bus internally, so they can re-use peripheral designs for either chip line
[20:58:00] <mozmck_work> interesting. I've heard that ARM can't toggle IO as fast as AVR32 - is that right? AVR32 can do up to 33mhz
[20:58:29] <alex_joni> I think it's multi-cycle for an ARM
[20:58:39] <SWPadnos> I think the ARM has a half-speed I/O bus, but the internal clock is faster (like 66 MHz), so I think the net is about the same
[20:58:43] <alex_joni> but otoh an 400MHz ARM might do it as fast
[20:58:44] <SWPadnos> but I don't know for sure
[20:59:15] <alex_joni> depends on the ARM core though
[20:59:15] <SWPadnos> well, I'm thinking of the SAM7X series or similar
[20:59:23] <SWPadnos> not the 9xx or other fast ones :)
[20:59:26] <alex_joni> * alex_joni was thinking ARM9
[20:59:31] <mozmck_work> The AVR32 runs at up to 66mhz and the IO can toggle at 33mhz
[20:59:48] <SWPadnos> hmm
[21:00:09] <alex_joni> driven by a program or pwmgen?
[21:00:11] <andypugh> That 5-axis machine is not messing about, is he?
[21:00:57] <mozmck_work> alex_joni: I don't know about that. I think the program can do it that fast if it does nothing else :)
[21:00:59] <mozmck_work> bbl
[21:02:49] <andypugh> Darn it to heck! Looks like I will be messing about with an Arduino this weekend.
[21:03:54] <alex_joni> andypugh: no worries, anything is better than a PIC
[21:04:00] <tom3p> cool, i think the poisons professor from harry potter is a dr.who incarnation http://www.bbc.co.uk/iplayer/episode/b00rw1t1/Doctor_Who_Doctor_Who_and_the_Giant_Robot_Episode_1/
[21:04:47] <andypugh> It looks like fun. I am thinking it might be good for running my parent's central heating, for example. (5 zones, 4 pumps, 6 valves)
[21:05:17] <SWPadnos> huh. it looks like the I/Os on the AT91SAM7X are a bit slower
[21:05:31] <andypugh> Ah! Gyros! Self-levelling headlights for the R1 suddenly look plausible.
[21:05:46] <SWPadnos> there's a couple clock delay before the write occurs, and the peripheral bus access takes 2 cycles, as does the actual peripheral write
[21:06:20] <SWPadnos> tom3p, that's "potions" :)
[21:06:26] <SWPadnos> and no, it's not Alan Rickman
[21:07:14] <tom3p> no? anyways it ends up being read by Tom Baker ( the real DR Who fer me )
[21:07:23] <SWPadnos> yep
[21:08:26] <kbarry> Question about IRC
[21:08:35] <kbarry> I am trying to speak in another channel, and am getting
[21:08:46] <kbarry> "Cannot send to channel"
[21:08:52] <kbarry> anyone knwo whats happening?
[21:09:14] <alex_joni> SWPadnos: yeah, but for stepping you'd probably just use the pwmgen
[21:09:24] <SWPadnos> maybe the channel requires you to be registered and signed in or something
[21:09:35] <SWPadnos> you don't have "voice" in that channel
[21:09:42] <skunkworks> tom3p: I agree totally. (I have most of tom baker on vhs)
[21:09:43] <SWPadnos> alex_joni, yes, I probably would :)
[21:09:48] <SWPadnos> or for PWM ;)
[21:09:57] <SWPadnos> PAL VHS!
[21:10:24] <alex_joni> reminds me of a fanta 4 song
[21:11:00] <kbarry> how do i get voice?
[21:11:02] <alex_joni> http://german.about.com/library/blmus_fanta4mfg.htm
[21:11:06] <alex_joni> kbarry: you ask an op
[21:11:10] <skunkworks> tom3p: I don't think that is the posions teacher.. (I like him - he would make a good dr who)
[21:11:19] <SWPadnos> potions! ;)
[21:11:27] <kbarry> how do i identify an OP?
[21:11:44] <mozmck> alex_joni: heh, if you program a PIC in C they are really not bad.
[21:11:46] <skunkworks> heh
[21:11:48] <skunkworks> sorry
[21:12:10] <andypugh> PCW: I think that the 7i49 still looks the easiest way. The Arduino just looks like too nice a toy not to have.
[21:13:25] <alex_joni> mozmck: says you
[21:13:55] <alex_joni> mozmck: I did about 3 or 4 projects with PIC's and I had troubles with all of them
[21:14:05] <alex_joni> and not software problems.. spooky things
[21:14:19] <alex_joni> one of them would only work when I was holding the PCB
[21:15:09] <alex_joni> another one was a compiler problem.. it seems it truncates too long lines without any warning
[21:15:22] <alex_joni> I had a bigger bidir array defined
[21:15:41] <alex_joni> and it truncated the longer lines, which caused the array to overflow
[21:16:11] <alex_joni> * alex_joni could go on.. but I'd rather not
[21:16:33] <Jymmm> SWPadnos: http://sfbay.craigslist.org/sby/tls/1682776271.html
[21:17:14] <skunkworks> tom3p: http://en.wikipedia.org/wiki/Matt_Smith_(actor)
[21:18:26] <micges> kbarry: http://freenode.net/faq.shtml
[21:21:04] <SWPadnos> Jymmm, http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=200321439924
[21:31:10] <andypugh> MattyMatt: You are playing Dr Who now as well?
[21:31:26] <frallzor> lo UK andypugh and MattyMatt
[21:56:39] <frallzor> hmm how common is it for 3phase vfds to be able to be run from 1phase?
[21:56:56] <Jymmm> SWPadnos: If it'll run off 120VAC@15A I'll take it, with free delivery and setup of course.
[21:57:45] <SWPadnos> heh. only slightly harder to move than the one you pointed out
[21:57:45] <DaViruz> frallzor: very common
[21:57:53] <DaViruz> although they can't handle full power
[21:58:13] <frallzor> 11kw>2kw
[21:58:21] <frallzor> 1phase, plausible?
[21:58:47] <frallzor> you know what Ive got, want to guess about it? :)
[21:58:57] <DaViruz> you won't get full voltage either, so the power and revs you can get out of the motor will be limited
[21:59:07] <frallzor> bah
[21:59:09] <frallzor> nm then
[21:59:12] <frallzor> f00k it
[21:59:28] <frallzor> G64 worked nicely btw =)
[21:59:33] <DaViruz> hmm, did you have two phases?
[21:59:41] <frallzor> but artcam is suckier than suckiest
[21:59:42] <DaViruz> that'd probably give you 2kW at least
[21:59:46] <frallzor> 1
[22:00:01] <frallzor> 2 where I cant hook it up any easier than 3phare
[22:00:02] <frallzor> *s
[22:00:08] <DaViruz> oh.
[22:00:45] <DaViruz> a wild guess is that you'll get about half the power from the motor
[22:01:02] <frallzor> well next week im getting a permanent installation anyways, I can live with cables everywhere until then )
[22:01:06] <frallzor> =)
[22:05:17] <andypugh> You are probably only limited by the rectifier, as the first stage is to convert to DC.
[22:05:38] <andypugh> Though there may be monitoring that turns it off if there is a phase missing.
[22:20:05] <PCW> Right, only difference is DC bus supply 1 phase vs 3 phase VFD is really a question of how much filter capacitor you wish to buy...
[22:20:08] <PCW> (a 3 phase only VFD may have minimal filters, but a 1 phase VFD depends on the filter capacitor to supply power when the line can't)
[22:20:43] <frallzor> any idea on how different brands generally are?
[22:20:57] <PCW> No idea
[22:20:59] <tom3p> if you had a shop full of cnc's with servo's , would a centralized dc supply be valid or is on the spot dc generation from ac superior?
[22:22:51] <tom3p> we've played with similar ideas where there were a hundred edm's, each with their own trafos & bridges
[22:23:56] <PCW> Good question, with three phase its not much of an issue since you have minimal ripple to begin with and at 360 or 300 Hz to boot
[22:25:31] <Valen> andypugh, decent VFD's will do power factor correction on the convert to DC stage
[22:26:16] <Valen> tom3p: I like the idea of DC power distribution
[22:26:42] <Valen> but until its the "standard" your doing it on your own with not much of a benifit
[22:26:43] <PCW> Like big Arcs?
[22:27:04] <tom3p> i was trying to remember the arguments between Edison & Tesla ( dc vs ac distribution )
[22:27:05] <andypugh> I have used a VFD with external DC Bus connectors for commoning.
[22:27:10] <Valen> most of the servo amps out there (in the larger sizes) seem to take AC mains and servo inputs
[22:27:43] <Valen> at the time AC was the only way to do it, because you can use iron transformers to boost voltages for transmission and then step it down for use
[22:28:01] <Valen> thing is with silicon you can do the same much more efficently with DC these days
[22:28:28] <Valen> and anything you plug in to mains other than your fridge and vaccume cleaner converts it to DC anyway
[22:28:30] <tom3p> i worked in theatres in NYC who still had Edison DC panels behnd all the ropes and crap
[22:29:02] <Valen> edison electric turned off their last DC supply a few years ago I believe
[22:29:30] <tom3p> these werent in use, but were scary to uncover
[22:29:36] <Valen> heh
[22:30:14] <Dave911_> Dave911_ is now known as Dave911
[22:31:05] <tom3p> any conversion is a loss, i was wondering how big a loss, is it 85% like trafo's?
[22:31:52] <tom3p> (15% loss 85% eff)
[22:32:18] <Valen> DC -DC step down can be up around 99%
[22:33:10] <Valen> 90+ is common even if your not really trying
[22:34:37] <andypugh> Is the 7i43 BIT file the actual binary executable file that runs on the FPGA? (if an FPGA actually "runs")
[22:37:51] <PCW> Yes, its the actual file of bits that fill up the LUTS and RAM and connection pips that determine the FPGA operation
[22:39:54] <PCW> And yes the FPGA can "run" for example the 7I49 interface has a 32 bit processor that runs code out of one of the BlockRAMS
[22:43:33] <andypugh> I am finding it all a bit baffling :-).
[22:45:12] <PCW> All proprietary unfortunately, Xilinx wont tell you what the bits mean (other than BLockRam contents and metadata)
[22:45:16] <andypugh> Where does the Hostmot2 code run? The hm2_7i43 code runs on the Linux machine and interfaces with EMC2 HAL, I think I have that part clear.
[22:46:12] <Valen> hostmot 2 runs in the FPGA
[22:46:22] <andypugh> (Guess who is looking through source code and finding a maze of little files all alike)
[22:46:54] <andypugh> So Hostmot2 is compile-time configured to be part of each BIT file?
[22:48:05] <PCW> So the hostmot2 driver (both pieces) connects between HAL and the hardware
[22:48:07] <PCW> which is just a bunch of addressable hardware like PWM gens and encoder counters
[22:48:08] <PCW> Each bit file is compile tile configured for a specific pinout and set of functionality
[22:48:24] <PCW> (compile time)
[22:49:19] <PCW> the set of functions and pinout is determined by the pinout file (its all or mostly all pinout driven)
[22:49:47] <andypugh> OK. So the BIT file contains a bunch of (say) PWM modules as hardware, and also a CPU (as hardware) on which Hostmot2 runs?
[22:50:21] <andypugh> (for suitably loose defnitions of "hardware" in this case)
[22:50:24] <PCW> No normally there's no CPU
[22:50:37] <PCW> Just addressable registers
[22:51:21] <andypugh> I think I need to weld something. Software is all too clever for me.
[22:51:38] <Valen> hostmot2 is the "cpu" as it were
[22:51:57] <PCW> When I get the same feeling I go out and talk to my sheep...
[22:53:41] <andypugh> Hostmot2 is written in vhd?
[22:53:54] <Valen> I believe so
[22:53:59] <SWPadnos> yes
[22:54:14] <SWPadnos> there's no CPU, by any definition I'm aware of
[22:54:32] <PCW> The USB version of HostMot2 (not for EMC) has a CPU in the FPGA for running the interface protocol
[22:54:34] <PCW> and the 7I49 interface uses a CPU for data acquisition and calculations at the resolver carrier rate
[22:54:35] <PCW> An the twiddler interface will use a CPU for general bit twiddling
[22:54:39] <SWPadnos> (Mesa has something called SoftDMC, which does include a CPU, but that's not supported by the EMC2 driver)
[22:55:59] <andypugh> So 7i49 is likely to be very much unsupported by EMC2?
[22:57:04] <Valen> I meant "cpu" as in a thing that does stuff, IE the counters for the PWM's etc, pretty much overkill for a FPGA but still
[22:57:40] <Valen> you would think, given the number of empty gates there must be, that a crossbar wouldn't be out of the question to allow you to shift the input and output pins around
[22:57:43] <PCW> We will probably also include a CPU for supporting a HS serial interface for things like the 8I20
[22:57:45] <PCW> 7I49 should be easily supported as it presents the angle and velocity in simple read only registers
[22:58:57] <PCW> Cross bars are too large for most of the FPGAs (the 5I22 would be OK the 5I20 no way)
[23:01:45] <andypugh> I think as a first step I might tape a magnet to the motor shaft, arrange a few Hall sensors on the mounting flange with Blu-Tak and see what my low-voltage BLDC driver does with these motors.
[23:01:49] <morfic> what's a reliable source for say 10HP servos, and say 30-40HP spindle servos? (drive/servo combos more like it)
[23:02:14] <SWPadnos> 10HP servos are BIG
[23:02:26] <Valen> mag motor might go up around that size in servo but that is a really really big servo
[23:02:33] <Valen> like 100kg range
[23:03:03] <Dave911> 10 hp Servos .... The big motor companies.. Siemens... GE (I would guess.. )
[23:03:27] <Dave911> Siemens can deliver Servos that are hundreds of hp ... if you gotta the money .. ;-)
[23:04:33] <Valen> and a semi ;->
[23:04:42] <Valen> postage would be a bitch lol
[23:04:52] <Dave911> Just need the $$$$
[23:05:04] <morfic> SWPadnos: both numbers thrown out by my boss, guess what, designer is still looking for data sheet of that one machine we would retrofit first, found all the other ones
[23:05:14] <SWPadnos> hah
[23:05:31] <Valen> 400W servo's would more than likley run that machine you have
[23:05:40] <morfic> he promised me "something" tomorrow morning
[23:05:41] <Valen> suitably scarily
[23:05:51] <Dave911> I have seen Siemens motors that have side access doors on. You can walk into them without bending over ... crazy big.. Steel mill motors
[23:06:57] <morfic> Valen: how'd you figure 400W?
[23:07:10] <Valen> gut feel, and seeing similar ones
[23:07:14] <Valen> its a common size
[23:07:15] <andypugh> Lenze seem to top out at 5.6kW (but were very helpful with my query about my eBay motors). http://www.lenze.co.uk/lenze.uk_en_active/images/Produktinformationen/Servoantriebe/Servomotoren/Asynchron-Servomotoren_MCA/Katalog/englisch/Servo_motors1_Catalog_Lenze_en.pdf;jsessionid=2E4A487A1E1BCF09D1853B3BBDDF37C1
[23:07:29] <SWPadnos> a 400W AC servo is pretty big, but then again, we don't know what kinds of machine you have
[23:07:35] <SWPadnos> reasonably big anyway
[23:07:39] <Valen> he posted pictures before
[23:07:45] <SWPadnos> oh, I don't remember that
[23:07:46] <Valen> a decent sized lathe
[23:08:11] <SWPadnos> morfic, http://www.automationdirect.com/adc/Shopping/Catalog/Motion_Control/Servo_Systems/Medium_Inertia_(1KW_-_3KW)_Servo_Systems
[23:08:12] <morfic> http://www.zerorealm.org/12LC/3-12-10_002.jpg 15" chuck for reference
[23:08:29] <andypugh> Consider that a human is a few hundred watts and can work a lathe without breaking a sweat.
[23:08:49] <Valen> andypugh, might break a sweat on that one ;->
[23:09:06] <Valen> mainly trying to keep up with the bastard
[23:09:10] <SWPadnos> the 3KW ones are a respectable size, and <$3k for the motor and drive
[23:09:24] <Valen> thats a pretty decent servo
[23:09:32] <SWPadnos> I'd get tired spinning the tool turret :)
[23:09:32] <Valen> I'd run that as a spindle ;->
[23:10:03] <morfic> SWPadnos: which is why it would be SO nice to have that damn data sheet
[23:10:11] <SWPadnos> as for spindle motors, a servo shouldn't be necessary, an induction motor will usually do fine
[23:10:42] <andypugh> Unless you want to index and use live tooling.
[23:10:47] <SWPadnos> a couple of other brands, though I don't have a distributor/dealer: Baldor, Yaskawa, Danaher, Parker (probably too small)
[23:11:13] <morfic> i will now start this in reverse, goolging machine specs daewoo published
[23:11:20] <SWPadnos> yes, any application where you want a stopped motor to be kept in position might not work as well with an induction motor
[23:11:38] <SWPadnos> Kollmorgen
[23:11:48] <andypugh> Panasonic
[23:11:48] <SWPadnos> (which I think is part of Danaher motion now)
[23:11:49] <Valen> thats why I'd want one as a spindle, on the mill that is
[23:11:56] <SWPadnos> yep, Panasonic, Omron
[23:12:04] <morfic> SWPadnos: while we do not have live tooling, we do want motor to be held in any angle, it's one thing boss made sure camsoft supports
[23:13:46] <morfic> http://www.usedmachinerysales.com/equipment/2747.html so far they all only list spindle, not servo info
[23:13:48] <andypugh> I think a 10hp servo on the axis would be pointless expense. It might be right for the spindle, but 10hp is still quite a lot for a lathe spindle.
[23:14:36] <andypugh> OK, 35HP spindle? That's a seriously powerful lathe then.
[23:14:56] <pfred1> andypugh you think?
[23:15:30] <pfred1> I'm so excited I got my TB6560AHQs today!
[23:15:34] <morfic> we machine parts larger than the 15" chuck, we take some nice depth of cut at a nice SFM, the 12LC is one of 2 machines we don't stall too often
[23:15:53] <andypugh> I think conventional lathes rarely went above 5HP. You run out of tool and workpiece rigidity at some point.
[23:16:23] <Dave911> Another place to get cheap servos is MachMotion.com.. (no affiliation with them at all... ) They sell Teco drives and motors.. I have three here I need to test tonight still. I suspect that they make Automation Directs Servos... way similar. About half the price of Automation Direct...
[23:16:29] <SWPadnos> production environment != "conventional" equipment :)
[23:18:04] <andypugh> 10Hp servo for you :-) http://cgi.ebay.co.uk/New-Fanuc-Servo-Spindle-Motor-A06B-0854-B190_W0QQitemZ300412739040QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item45f1fe9de0#ht_500wt_1182
[23:19:25] <morfic> SWPadnos: we make shit loads of money off of our manual machines, just need more manual machinists
[23:19:51] <andypugh> Out of interest, what is wrong with the servos already on the machine?
[23:20:18] <SWPadnos> the cost to repair them (or their drives), I imagine
[23:20:43] <Dave911> Hey AndyPugh.. that servo is not to far from me....
[23:21:46] <morfic> andypugh: it is more a question of what if they fail, we get away from the rest of the fanuc stuff, so keeping them makes the initial investment lower, but what do we do once we need to replace them
[23:21:55] <pfred1> wasteful Americans with money to burn I imagine
[23:22:26] <morfic> and seeing how much info we have about our drives/servos, it might be more efficient to start with a known drive/servo combo, unless the missing documentation shows up
[23:22:30] <Dave911> >>just need more manual machinists.. interesting that you say that.. We have a major shortage of CNC machinists in the Midwest USA... even with the economy being slow.
[23:23:33] <andypugh> morfic: There are lots of Fanuc manuals on eBay.
[23:23:34] <morfic> Dave911: well, we are really short on manual machinists, but still would hire more CNC machinists, but not operators who think they are cnc machinists, they never last
[23:23:35] <Dave911> pfred1: You must be reading our magazines .... ;-)
[23:24:28] <Dave911> Big difference between an operator and a machinist... is that what you mean? I would agree with that..
[23:26:23] <morfic> Dave911: we tell them, operators don't make it, can you do your setups, can you edit your canned cycles, can you read code, and most seem to be able to do it, gets hard when we get "i know it, i'm good, i just didn't touch a fanuc in a while"
[23:26:40] <morfic> then they last between 1day and 1 month before they crash and burn
[23:26:46] <morfic> some walk out on their own
[23:28:44] <morfic> my boss pays good wages for this area, so he expects a lot, expects you stay at or close to the avg times we have on parts, not from day one, it's still too much pressure for some
[23:29:04] <morfic> Dave911: so send us your cnc machinists, SWPadnos said he would sign off on that :P
[23:29:52] <SWPadnos> sure. nobody can read my signature, but I can sign for you if you like ;)
[23:30:43] <Dave911> Ha ha.... not many around here.. I don't know if you are familar with Craigslist in the US. Usually used to sell stuff on the web and it is free.. If you do a search in the Midwest USA on CNC on Craigslist.. about all you find are companies looking for CNC operators, Machinist, and programmers...
[23:30:45] <Jymmm> SWPadnos: I can, but I can read Rx's too =)
[23:30:50] <morfic> oh, seems designer asked about this Lennox system i am looking at, asked the camsoft guy, interesting answer
[23:31:02] <SWPadnos> Jymmm, I'll have to take your glasses then
[23:31:50] <Jymmm> SWPadnos: won't help, I gots me one of these... http://www.thinkgeek.com/gadgets/electronic/c6d1/
[23:31:58] <morfic> the best thing ever i read though is from the people who make the windows real time extension soft servo uses
[23:32:01] <Dave911> I had one operator I was working with who wanted me to train him on Autocad, CAM, and the CNC lathe control in less than one day... he simply did not understand that it was not possible.
[23:32:27] <morfic> http://www.intervalzero.com/pdfs/RTXWhitePaper-6-09.pdf read about the development cost
[23:32:37] <SWPadnos> yeah? well I have one of these: http://www.thinkgeek.com/homeoffice/kitchen/8ace/
[23:33:03] <morfic> SWPadnos: and just so you know, linux is on its way out and you? or linux? have 0% market share
[23:33:11] <SWPadnos> clearly
[23:34:02] <pfred1> morfic depends what market
[23:34:36] <SWPadnos> I believe he was paraphrasing (sarcastically) the paper he linked
[23:35:48] <morfic> SWPadnos: i was paraphrasing the email i read from camsoft, he sounded kind of rattled
[23:35:55] <SWPadnos> oh, ok
[23:36:23] <SWPadnos> there are about 1200 registered users on linuxcnc.org who might disagree
[23:36:37] <SWPadnos> and who knows how many more unregistered masses :)
[23:36:39] <morfic> he gets a fair shot, same as emc
[23:36:45] <SWPadnos> sure
[23:37:14] <morfic> emc is a much more alive beast than i imagined, based on other projects where there might be channels, but only users around most of the time
[23:37:38] <morfic> and i thought a plug computer is a much more common freenode item than a linux based cnc control :)
[23:37:39] <pfred1> its alive!
[23:37:41] <SWPadnos> yeah, it's a good community
[23:37:55] <SWPadnos> morfic, did you ever sign up for the mailing list(s)?
[23:38:12] <pfred1> SWPadnos I registered at linuxcnc
[23:38:37] <SWPadnos> I guess I did too :)
[23:38:58] <SWPadnos> but I had forgotten about it. good thing computers remember passwords well
[23:39:03] <pfred1> whoot my pads align with the part
[23:40:29] <morfic> SWPadnos: i have not, actually
[23:40:36] <SWPadnos> ah
[23:41:03] <morfic> so busy looking at everything, i did not even think about that once o_O
[23:41:07] <SWPadnos> it was interesting. a couple of days ago a user posted a reply to some question someone had, saying that he had implemented C in G-code
[23:41:34] <morfic> i wish i could suck it all in, in one evening and know it all and get it all done in one day on the machine
[23:41:35] <SWPadnos> you can basically do C programming within the G-code now, with his addition
[23:41:39] <SWPadnos> heh
[23:41:52] <morfic> SWPadnos: cgcc link y'all posted? sounds interesting
[23:41:53] <pfred1> SWPadnos it's too bad i don't see Dennis Ritchie around all the time anymore I'm sure he'd get a good laugh out of it too
[23:42:00] <SWPadnos> yes
[23:42:05] <SWPadnos> heh
[23:42:12] <morfic> SWPadnos: i thought it was a compiler that spat out gcode emc loves
[23:42:17] <SWPadnos> nope
[23:42:23] <pfred1> SWPadnos I used to live a town away from him we shopped in the same supermarket
[23:42:29] <SWPadnos> it runs the "C" inside EMC2/AXIS
[23:42:34] <SWPadnos> cool
[23:42:49] <pfred1> he was even listed in the phone book
[23:43:15] <pfred1> he lived in berkley Heights
[23:43:24] <SWPadnos> what's really funny about the CGCC thing is that the announcement was the first time I had heard of it, and I think that's true for most of us
[23:43:35] <SWPadnos> just out of the blue "hi guys, here's what I've done"
[23:43:43] <morfic> tinkered away in his basement, and then "here, done"
[23:43:49] <SWPadnos> yep
[23:44:16] <SWPadnos> there's a reconfigurable user interface someone is working on too, not quite all there yet, but looks promising
[23:44:41] <SWPadnos> and who knows what else is brewing, outside the "developers" minds
[23:45:03] <morfic> which is what's great about opensource, user contributions
[23:45:09] <SWPadnos> yep
[23:45:24] <pfred1> morfic use the source Luke!
[23:45:38] <SWPadnos> there are a few people who email patches in from time to time, with bugfixes or minor improvemenrts
[23:45:46] <SWPadnos> it's a good community
[23:45:46] <morfic> the project does not even have to accept it, if i find something and i like it and i have instruction on how to build it, i run it, if my linux distribution accepts their offer or not
[23:45:55] <SWPadnos> yes
[23:46:09] <SWPadnos> well, that's an interesting thing that has to be considered for your controls
[23:46:22] <SWPadnos> Windows XP is about at EOL, even for corporate contracts
[23:46:50] <SWPadnos> Vista and Win7 are relative unknowns, but you *WILL* use whatever the controls company tells you to use
[23:46:56] <andypugh> Ford are due to transfer to Win7 this year. That will be a big project!
[23:47:17] <SWPadnos> and if there's a new feature (or even a bugfix) 5 years from now, you'll have to do whatever they want to get it
[23:48:11] <pfred1> if it wasn't for linux Microsoft would still be releasing OSes as buggy as Windows 95 was
[23:48:17] <SWPadnos> we're only providing packages for Ubuntu 6.06 and 8.04 at the moment, and will add 10.04 to the list (I think we may be dropping 6.06 though)
[23:49:21] <morfic> as long as ubuntu has LTS every so often, it's all good
[23:49:24] <SWPadnos> but that's only the packages we build. If you want to patch your system (which might include compiling a kernel), you can build your own packages and install them on the same OS you install today, until you get tired of doing it and decide to move to a "supported" platform
[23:49:32] <SWPadnos> they do LTS every 2 years
[23:49:47] <SWPadnos> and support them for 3 years (desktop) and 5 years (server)
[23:50:07] <pfred1> my LTS is tar up my home dir format and install a new distro every now and again
[23:50:08] <SWPadnos> so there's always at least 1 year of overlap for the changeover (assuming you want to change over)
[23:50:14] <SWPadnos> heh
[23:50:42] <morfic> yeah, ubuntu is a'ight, they just really need to keep an eye on their distupgrades, they are not as smooth as they could be (new installs are still sexy as hell)
[23:51:08] <SWPadnos> yeah, upgrading has never been all that fun, but then again it's usually not all that necessary either
[23:51:38] <pfred1> morfic initially I was impressed with Ubuntu but now I realize i have to clear out half the cruft it has installed
[23:51:55] <SWPadnos> pfred1, you could start with the server install and add what you like instead
[23:52:00] <SWPadnos> that might work better for you
[23:52:15] <morfic> i switched away from ubuntu only because of what they decided to do with arm cpu support (feels like i mentioned this in here before?)
[23:52:39] <SWPadnos> I think there's a neybook remix that's supposed to have ARM (or something like that)
[23:53:00] <pfred1> SWPadnos I should probably just fight with Debian until i get it like I want it
[23:53:06] <SWPadnos> yeah
[23:53:35] <morfic> i run gentoo again, gentoo has always been good for embedded work, and i like running same distro on the arm and on the laptop, otherwise i forget where stuff goes :P
[23:53:49] <SWPadnos> and then write a program that will automagically create an install list for the next version, so you can just say "dpkg --get-all-these-things" to get all the software you like :)
[23:54:06] <pfred1> morfic is gentoo that distro where you have to compile everything?
[23:54:23] <morfic> pfred1: one of them, yes
[23:54:30] <SWPadnos> heh. I ran Gentoo in the kernel 2.5-2.6 development cycle
[23:54:35] <SWPadnos> had a lot of kernels on that one
[23:54:40] <pfred1> morfic no thanks slack broke me of the compiler merry go round
[23:55:14] <pfred1> I must have spent the better part of 7 years compiling everything I wanted to run
[23:55:35] <morfic> i ran sourcemage for a while after i left gentoo, but after making them a temporary amd64 install cd, i ran out of time to work on that distro too and installed ubuntu, so i can just use it, not work on it :/
[23:55:37] <pfred1> now all I wanna do is apt get and go
[23:56:00] <SWPadnos> that's the thing, isn't it
[23:56:11] <SWPadnos> you buy a car that "just works", when you want to just use it
[23:56:13] <pfred1> yeah I can't be bothered adminning my system
[23:57:09] <pfred1> SWPadnos cars are different you can see when parts are falling off them with computer software it isn't always as obvious where attention is needed
[23:57:34] <SWPadnos> well, it's more a question of whether you buy a house because it's near a "performance parts" store
[23:57:38] <morfic> SWPadnos: gentoo was fun when i had time to be part of gentoo, i could commit things that would otherwise just have collected in overlays, i started helping with gcc3.4 patches, back when sloppy coding made people go "gcc3.4 broke my code", tons of ignored user contributed fixes that laid in bugzilla, untouched by the unwilling devs
[23:58:09] <SWPadnos> I wonder if I still have that hard disk around somewhere
[23:58:41] <SWPadnos> the PC was an Athlon 2600-ish, on an Abit AT7 (legacy free!) motherboard, with maybe 1G of RAM
[23:59:02] <SWPadnos> compiling the system would be a bit faster these days, with the quad-core Phenom and 8GB :)