Back
[00:53:13] <Roklobsta_> damn my experiment with bakin gpaper in the lase rprinter just went horribly wrong.
[00:53:27] <Roklobsta_> the waxproof paper is wrapped all around the selenium drum!
[00:55:39] <theBear> :(
[00:55:52] <Roklobsta_> it's super jammed
[00:56:38] <Roklobsta_> mrs roklobsta is not impressed
[00:57:03] <alexh> why would you feed baking paper into the printer?
[00:57:21] <Roklobsta_> oh to see if it's any good for transferring an image to a pcb
[00:57:29] <alexh> heh
[00:57:38] <Roklobsta_> no it isn't
[00:57:52] <alexh> I kinda inferred that from your first statement :P
[00:58:13] <theBear> hehe, thanks for the tip
[01:04:10] <Roklobsta_> oh fsck, the whole a4 sheet is somehow scrunched into the toner cartridge!
[01:06:04] <nevyn> Roklobsta_: magazine paper works well apparently
[01:06:12] <Roklobsta_> yeah...
[01:06:16] <nevyn> cause it's shiny so you can iron onto the pcb
[01:06:33] <nevyn> the bigger problem is. the splattery nature of laser printing
[01:06:54] <nevyn> you can't get below 1mm pitch sortof level.
[01:07:06] <nevyn> because the toner is actually quite. blobby.
[01:07:31] <CapnKernel> nevyn: speaking from experience? About toner transfer that is
[01:07:41] <nevyn> yeah.
[01:08:18] <nevyn> you might get an order of magnitude on that. but forget BGA or really fine packages
[01:08:23] <CapnKernel> I haven't had any problem printing 6mil with my laser printer
[01:08:46] <nevyn> hrm 0.1mm rather.
[01:08:50] <nevyn> 1mm is doable.
[01:08:53] <CapnKernel> Which is 0.15mm
[01:10:15] <Roklobsta_> hrm, it's a new catidge. any advice on opening them up without making a mess?
[01:10:56] <mrfrenzy> Roklobsta_: no
[01:10:59] <mrfrenzy> you will make a mess
[01:11:20] * Roklobsta_ grumbles. He should have listened to Mrs Roklobsta
[01:12:34] <mrfrenzy> you could try un-refilling the cartridge, blowing it out with compressed air and then opening
[01:12:40] <mrfrenzy> I would just buy a new cartridge
[01:21:07] <CapnKernel> Take it outside fo'sure
[01:21:23] <CapnKernel> Spilling toner inside is grounds for divorce\
[01:22:28] * CapnKernel has a wedged firefox: While the print dialog was open, GMail popped up a dialog. Neither can be closed :-(
[01:23:52] <theBear> killall -9 firefox
[01:24:21] <CapnKernel> I used -1, more chance that the 70 tabs I have open will be kept
[01:24:55] <theBear> i find the last few years ffox is pretty good at saving tab state sometime before close
[01:27:42] <mrfrenzy> yeah, if you just kill the power most of your tabs will come back
[01:27:48] <mrfrenzy> it seems to save very five minutes or so
[01:42:37] <Roklobsta_> i tried to open it. the paper was stuck so hard in some place i couldn't get it out and i also managed to bleed into the worm gears. scratch one brand new cartridge
[01:43:58] <Roklobsta_> at least new ones are only $25 now
[01:45:20] <wollw> using malloc on the avrs: bad idea?
[02:12:50] <Roklobsta_> wollw: i say don't bother
[02:22:28] <Fleck> wollw i use memcpy with great success! :)
[02:28:27] <Roklobsta_> memcpy doesn't rely on malloc
[02:28:29] <Roklobsta_> anyway, for small embedded systems you are best to not use malloc/free and actually statically allocate your needed resources and use linked lists
[02:31:26] <Roklobsta_> and buffers
[02:37:40] <wollw> Roklobsta_: yeah, i'm ending up doing that
[02:38:02] <wollw> I'm rewriting some code for a project and I used a lot of mallocs before
[02:49:19] <Roklobsta_> see how you go. are you mallocing the same stuff over and over?
[02:49:39] <Roklobsta_> how do you handle failed mallocs?
[02:50:07] <wollw> Roklobsta_: I was just allocating a bunch of stuff at initialization
[02:50:53] <Roklobsta_> can you just use globals instead?
[02:50:56] <wollw> i actually think this static allocation version i'm working on is going to be more elegant anyway
[02:51:17] <wollw> Roklobsta_: pretty much, i was trying to keep it easy to configure at the time
[02:51:27] <wollw> so i was allocating based on a config header
[02:51:30] <Roklobsta_> how much ram do you have?
[02:51:36] <wollw> 1K
[02:51:46] <wollw> i was somewhere around 600 bytes used i think
[02:51:48] <wollw> if i did the math right
[02:52:02] <wollw> target is a t85
[02:52:26] <wollw> i'm just doing the data structure configuration right in the header file now as constant structs
[02:52:45] <wollw> and another typedef struct with members pointing to them
[04:31:27] <pingec> http://www.newscientist.com/video/1482014294001-robot-helicopters-perform-james-bond-tune.html
[05:14:48] <Roklobsta_> wollw: don't forget you can save RAM by explicitly ensuring some constants are read from ROM.
[05:16:16] <Roklobsta_> have a look at avr/pgmspace.h in the avrlibc manual
[05:16:36] <DanFrederiksen> pingec, cute
[05:18:55] <wollw> Roklobsta_: thanks for the reminder
[05:19:24] <Roklobsta_> i have done that for some strings that only need to be send once to some hardware for config at boot.
[05:19:26] * wollw completely forgot about PROGMEM
[05:25:27] <Valen> I'd like make to automagically increment a build number in my code, that way when i serial out stuff at boot i can be sure i'm running the latest code
[05:25:58] <wollw> Valen: you can pass a define to gcc
[05:27:16] <wollw> like for instance, "gcc main.c -DFOO=123"
[05:27:29] <wollw> will replace any instance of FOO in main.c with 123
[05:27:40] <wollw> just like a define
[05:28:13] <wollw> you could pass it epoch seconds - some start time or something
[05:29:48] <Valen> i spose i could do some bash stuff to make it work then
[05:30:18] <wollw> Valen: you could just add "-DTIMESTAMP=`date +%s`" to your CFLAGS
[05:30:24] <wollw> but yeah
[05:30:38] <Valen> too big
[05:30:40] <wollw> ah
[05:30:48] <Valen> 4 digits is all that'll fit nicely on a lcd
[05:31:16] * wollw has his head stuck in the USART sand.
[05:32:06] <Valen> better than usbn
[05:32:09] <Valen> usb
[05:32:22] <wollw> easier sure
[05:32:35] <wollw> nice for debugging as is the current case
[05:33:25] <Valen> that and flashing lights are all i've ever used
[05:34:04] <wollw> I have a TI Launchpad which lets you connect to it with gdb
[05:34:08] <wollw> which is pretty nice
[05:49:56] <wollw> hurrah
[05:50:25] * wollw got the basics of his cellular automaton rewritten
[05:56:09] <Roklobsta_> johnny 5 is alive!
[05:59:03] <wollw> https://github.com/Wollw/Cellular-Polymaton
[05:59:31] <wollw> here's what i ended up doing for the data:
https://github.com/Wollw/Cellular-Polymaton/blob/master/src/config.h
[06:03:43] <serp__> how does the configuration work?
[06:04:12] <serp__> the two rules and 9 cells have bit patterns, what do they represent?
[06:05:20] <wollw> the rules there are conway's game of life
[06:05:32] <wollw> which is 2 and 3 living neighbors == stays alive
[06:05:47] <wollw> and 3 living neighbors == comes to life
[06:05:55] <wollw> bits 2 and 3 are set on r_live
[06:06:01] <wollw> and bits 3 are set on r_dead
[06:06:07] <wollw> so it's basically rule lookup
[06:06:23] <wollw> the bits on the cells are the array ids of neighbors
[06:06:36] <wollw> so bit 0 on the first cell is itself
[06:06:41] <wollw> (which is never a neighbor)
[06:07:01] <wollw> for square grids this is certainly unneeded
[06:07:19] <wollw> but it means i can use other shapes for cells
[06:07:30] <wollw> and define which cells are neighbors to eachother
[06:07:36] <theBear> oh, the old game of life, damn, that was the most boring game i had on the xt
[06:07:40] <wollw> hehe
[06:08:03] <serp__> { 0b110010110, live },
[06:08:07] <serp__> so what does that mean?
[06:08:37] <theBear> erm, got some context ?
[06:08:38] <wollw> it means that the cells at positions 1,2,4,7,8 in the array are neighbors
[06:08:49] <serp__> theBear: context is the url wollw pasted
[06:08:50] <theBear> oh, i just noticed, i haven't been following the conversation :)
[06:08:56] <theBear> yeah sorry
[06:09:03] <wollw> and the cell starts out alive
[06:09:22] <wollw> in this case it's a 3x3 grid
[06:09:39] <theBear> wollw, sweet jesus, did yuo just remember the rules of the game of life without a reference ? damn !
[06:09:42] <wollw> or, rather, could
[06:09:52] <wollw> heh
[06:09:56] <wollw> yeah, at this point
[06:10:01] <serp__> so the entire world is 3x3?
[06:10:09] <wollw> for that simple example yeah
[06:10:16] <wollw> http://www.youtube.com/watch?v=gBdAOIyHySA&feature=related
[06:10:17] <serp__> are there 9 worlds?
[06:10:23] <wollw> this is an actual implementation
[06:10:24] <serp__> since you have 9 entries
[06:10:26] <wollw> no, 9 cells
[06:10:30] <wollw> each of those is a cell
[06:10:47] <wollw> for things like the above video i have to define manually what is a cell to what
[06:11:31] <theBear> 3x3=9
[06:12:14] <serp__> wtf why is this so hard to understand
[06:12:23] <serp__> 19:56:28 <wollw> bits 2 and 3 are set on r_live
[06:12:24] <serp__> 19:56:34 <wollw> and bits 3 are set on r_dead
[06:12:30] <wollw> sorry :|
[06:12:31] <serp__> what bits?
[06:12:44] <serp__> the bits in cells are static arent they?
[06:12:54] <wollw> yeah
[06:13:12] <wollw> the bits in cfg_rules
[06:13:31] <wollw> the first bit pattern defines the rules for cells that are alive
[06:13:37] <wollw> the second for cells that are dead
[06:14:34] <serp__> so let
[06:14:39] <wollw> when updating the cells it checks the bit number equal to the number of living neighbors
[06:14:46] <serp__> let's say a cell is dead, the rule is 0b00001000
[06:15:03] <serp__> oh I see lol
[06:15:29] <wollw> yeah, sorry i was having trouble explaining it
[06:15:43] <serp__> no, I'm a bit thick, that's all
[06:15:48] <serp__> cool stuff
[06:16:02] <wollw> it's fun
[06:16:06] <theBear> i found it very self explanatory the first time i got a half-explanation and saw the app (this was like, 20 years before youtube existed :)
[06:16:19] * wollw is building a few more like in the video out of PVC
[06:16:39] <serp__> do you have any larger/weirder displays than the one in the video?
[06:16:58] <wollw> serp__: that's the first i've made that wasn't out of like cardboard
[06:17:02] <theBear> oh wow, hw vbersion
[06:17:11] <serp__> ok nice, keep at it :)
[06:17:30] <serp__> be sure to post videos as you make more
[06:17:38] <wollw> serp__: I plan to
[06:18:06] <wollw> there's a quick page i threw up in december too
[06:18:07] <wollw> http://wollw.github.com/automaton-avr/
[06:18:14] <wollw> i'm planning to update it with better information
[06:18:15] <theBear> http://www.youtube.com/watch?v=XcuBvj0pw-E this has too much crap pre-start, but is much more like what i remember
[06:18:18] <wollw> parts lists and such
[06:18:31] <theBear> just like a petri dish
[06:18:49] <wollw> theBear: yeah, that would take a lot of LEDs
[06:18:56] <theBear> ooh, the 'trick' designs that start around 1:30 are kinda cool, conceptually
[06:19:09] <theBear> i just mean as far as concept for serp__ , i wouldn't suggest building it
[06:19:19] <wollw> ah
[06:19:30] <theBear> probly be tricky to program/do hw for random updates on that many leds too :)
[06:19:30] <wollw> night & day is a neat rule set too
[06:19:35] <wollw> yeah
[06:19:53] <wollw> it's easy to do lots of cells with a grid too
[06:20:08] <wollw> the computation that is
[06:21:14] <theBear> indeed, simple computation <grin>
[06:21:52] <serp__> I've never seen other rule sets than game of life
[06:22:04] <theBear> how about "do unto others" etc ?
[06:22:08] <theBear> that's a rule set
[06:22:17] <wollw> hard to program that one :P
[06:22:27] <wollw> Day & Night:
http://en.wikipedia.org/wiki/Day_%26_Night
[06:23:54] <theBear> heh
[06:23:58] <serp__> wow nice
[07:22:27] <norbi> hi
[07:22:27] <tobbor> norbi! like, totally tell us about the project!
[07:22:44] <norbi> guys do you know any example in C for self programming via usart?
[07:23:06] <norbi> doesnt really mater it to be in usart, but example in ci
[07:27:31] <scuzzy> anyone here built a high power brushless controller?
[07:30:18] <amee2k> norbi: hmm doesn't the original arduino boot loader do that?
[07:30:29] <amee2k> not sure what it is written in though
[07:31:04] <amee2k> but for most intents and purposes, C is just a fancy macro processor for assembly
[07:35:12] <norbi> amee2k: the problem is that i dont like and want to programm in asembly
[07:35:32] <norbi> amee2k: i dont use arduino
[07:36:05] <amee2k> well, the idea was that once you see an assembly example, implementing it in C wouldn't be too hard
[07:36:18] <norbi> amee2k: i could do it in assembly but the problem is understanding self programming, there are some details that im not catchin
[07:36:29] <norbi> hoped that freekz would give me some tut
[07:36:32] <norbi> but i cant find
[07:36:47] <amee2k> in that case, studying an assembly example would be just as good IMO
[07:37:56] <amee2k> isn't fappurino all open source?
[07:39:59] <norbi> IMO?
[07:40:01] <norbi> what is that?
[07:40:11] <amee2k> In My Opinion
[07:40:17] <norbi> ah
[07:40:18] <raek> scuzzy: how much power is high power?
[07:40:41] <norbi> yea i did, but i want to understand it clearly now i found something usefull on freaks
[07:40:49] <amee2k> norbi:
http://download.berlios.de/arduino/bootloader.zip << check out atmegaboot.c in that one
[07:41:09] <norbi> amee2k: thank you a lot!
[07:41:45] <scuzzy> raek, like 20kw
[07:41:56] <scuzzy> (peak)
[07:42:00] <amee2k> looks like the part that does the actual flash programming is written in assembly anyway
[07:42:41] <amee2k> but no real surprise there... the C optimizer can do weird things with critical code sometimes
[07:43:01] <raek> scuzzy: then no (I once wrote an ESC for a quadrotor)
[07:43:18] <scuzzy> raek: that's still pretty cool
[07:43:32] <scuzzy> I've seen you can buy the controllers, but they are upwards of $500
[07:43:48] <scuzzy> seems WAY too much for controlling a high power BLDC motor
[07:43:57] <raek> you just need an AVR and some FETs
[07:44:29] <amee2k> at that power level i'd point at IGBTs
[07:44:30] <scuzzy> I figured, but I'm sure there is more to it than that
[07:44:48] <scuzzy> really?
[07:44:55] <raek> I wrote a custom firmware for this design:
http://home.versanet.de/~b-konze/blc_4428/blc_4428.htm
[07:45:44] <raek> for the Artificial Intelligence and Integrated Computer Systems Division at Linköping University
[07:46:30] <raek> 20kW... what are you using it for?
[07:46:38] <scuzzy> not anything yet
[07:46:45] <scuzzy> but, thinking about starting an EV project
[07:46:53] <scuzzy> strip an old motorbike chassis
[07:48:08] <amee2k> sidecar full of batteries?
[07:48:15] <scuzzy> lol
[07:48:16] <scuzzy> no
[07:48:29] <scuzzy> only need to do like 100km max on it
[07:48:33] <scuzzy> maybe even 50
[07:48:44] <scuzzy> I've seen EV bike builds that get that sort of distance
[07:48:48] <scuzzy> on lead acid batteries
[07:48:57] <amee2k> hmm
[07:49:13] <OndraSter> daddy's home!
[07:49:16] <amee2k> i'd rather mod one of these tiny italian 3-wheeled pickup thingies
[07:49:23] <scuzzy> sealed lead acid batteries
[07:49:38] <scuzzy> amee2k: well, as I said, I haven't actually planned anything yet
[07:49:44] <scuzzy> just doing a little bit of research
[07:49:59] <scuzzy> but you can pick up a nice motorbike chassis for quite cheap
[07:50:13] <scuzzy> like, old bikes "written off"
[07:50:49] <scuzzy> I wouldn't have thought a BLDC controller is rocket science
[07:51:14] <scuzzy> you can even buy IC's to do it for you
[07:51:51] <amee2k> http://en.wikipedia.org/wiki/File:1957_Daihatsu_Midget_01.jpg << mod one of these :P
[07:52:07] <scuzzy> yeah... not so much
[07:52:11] <OndraSter> woo
[07:52:14] <scuzzy> too heavey
[07:52:14] <OndraSter> add EV to that..
[07:52:16] <OndraSter> would be cool
[07:52:29] <OndraSter> how much cost they :P
[07:52:46] <amee2k> scuzzy: well, the original had a 7kW air cooled 2 stroke enigne
[07:53:03] <amee2k> so if you drop in a 20kW electric drivetrain it should be plenty of power
[07:54:53] <scuzzy> well, there is also the fact that I just don't wanna ride something like that
[07:55:28] <amee2k> meh :(
[07:55:47] <amee2k> how about welding your own tube frame to make it look like a single seater dune buggy!
[07:56:29] <OndraSter> haha
[07:56:40] <amee2k> http://en.wikipedia.org/wiki/File:Fast-Attack-Vehicle-1.jpg
[07:56:40] <OndraSter> how about getting hammer and smashing atoms so you make your own tube!
[07:57:08] <scuzzy> that would be cool
[07:57:19] <scuzzy> but, to be honest, I don't have the skillz
[07:57:27] <scuzzy> it's easier to start with an already existing frame
[07:58:16] <amee2k> full sized LSV is only just under a ton. i wonder how light it becomes if you remove the top frame on the rear, the diesel engine and rear seats
[07:58:30] <scuzzy> LSV?
[07:58:38] <amee2k> the dune buggy i linked to
[07:59:50] <scuzzy> Ahhh
[08:00:01] <scuzzy> like, a suzuki banti is a nice tube framed bike
[08:00:06] <scuzzy> bandit even
[08:00:17] <scuzzy> and it looks pretty nice
[08:00:23] <amee2k> i can't find anything on the engine but 130kph with that frame means the diesel probably accounts for a significant portion of the weight
[08:00:25] <scuzzy> you can pick up the frames and forks and stuff quite cheap
[08:00:33] <amee2k> nice :)
[08:00:39] <scuzzy> I dunno
[08:00:45] <scuzzy> maybe a long term project
[08:00:58] <scuzzy> with the price of petrol, and gathering how far away my office might be soon
[08:01:12] <theBear> is continental shift messing with your commute ?
[08:01:12] <scuzzy> it seems a nice little EV bike might be the way to go
[08:01:26] <scuzzy> theBear: like you cannot believe!
[08:01:29] <dirty_d> scuzzy, how far do you need to ride?
[08:01:30] <amee2k> also, if you lose the .50 BMG, the M60 and M249 and the bazooka it'll get a lot lighter too
[08:01:34] <theBear> hehe
[08:01:41] <scuzzy> dirty_d: probably around 40 - 50 km a day
[08:01:43] <scuzzy> tops
[08:01:53] <scuzzy> would be good to be able to get 100km out of it tho
[08:01:54] <dirty_d> youd need a lot of batteries
[08:01:56] <scuzzy> on a single charge
[08:02:00] <dirty_d> too many for a bike i think
[08:02:05] <theBear> 100km ? jeez, that's a lot for a bike yeah
[08:02:10] <scuzzy> really?
[08:02:20] <amee2k> scuzzy: then you should aim for 100+km range on good batteries to account for wear. otherwise you'd end up buying new ones every 6-9 months
[08:02:20] <scuzzy> I've seen guys build bikes on lead acid batteries that to like 50
[08:02:25] <theBear> lookup how far ev cars with a boot full of batteries go
[08:02:35] <dirty_d> i had 40lbs of lead acid, i could go 20km
[08:02:42] <scuzzy> theBear: an EV car also weighs about 8 times as much tho
[08:02:44] <amee2k> how about lots of RC car lipo packs?
[08:02:45] <theBear> then notice they don't weigh much more than a steel motorbike chassis
[08:02:53] <dirty_d> amee2k, thats the best way to go
[08:02:53] <theBear> scuzzy, only with all those batteries :)
[08:03:02] <dirty_d> youre never gonna beat the cost/perfomance of RC lipo
[08:03:15] <amee2k> readily available, lots of juice, high power density per unit of weight
[08:03:16] <scuzzy> amee2k: really?
[08:03:28] <theBear> expensive and not the longest life in the world tho
[08:03:35] <amee2k> scuzzy: really what?
[08:03:37] <theBear> as far as charge cycles etc
[08:03:45] <scuzzy> RC lipo packs
[08:03:46] <dirty_d> i just bought four 14.8v 5Ah packs
[08:03:52] <dirty_d> each were $22
[08:04:06] <dirty_d> you gotta get it from HK though
[08:04:10] <amee2k> scuzzy: sure, why not?
[08:04:33] <dirty_d> rated for 100A continuous
[08:04:55] <theBear> yeah, but a 20ah gel cell is under $10... as a single off the shelf item
[08:04:59] <theBear> extrapolate
[08:05:06] <dirty_d> $10 no way
[08:05:16] <amee2k> scuzzy: the tesla roadster is a lotus frame full of laptop batteries that puts 250hp on the rear wheels
[08:05:35] <dirty_d> lead acid gives nowhere near the rated capacity either
[08:05:37] <theBear> yes way, dude
[08:05:38] <scuzzy> amee2k: I'm looking at 15 - 20 kw absolute peak
[08:05:40] <dirty_d> unless youy discharge at C/20
[08:05:46] <theBear> maybe not stupid american dollars, they're worth shit now :)
[08:05:48] <dirty_d> theBear, from where?
[08:05:51] <scuzzy> 90% of the journey is going to be like 3kw
[08:05:54] <scuzzy> maybe even less
[08:05:58] <theBear> i'm not saying to use lead acid, i'm just saying that lipo aren't cheap
[08:06:09] <amee2k> as a side note, you might consider an induction motor over brushless DC too... i'd suspect induction motor is lighter and cheaper
[08:06:12] <theBear> dirty_d, altronics would be the first place i looked
[08:06:19] <theBear> but then again they are local to me
[08:06:24] <scuzzy> I was hoping to go with like 4 smallish lead acid's
[08:06:27] <dirty_d> amee2k, no way
[08:06:30] <scuzzy> sealed lead acids
[08:06:34] <scuzzy> or gell lead acids
[08:06:39] <dirty_d> amee2k, brushless RC motors
[08:06:56] <amee2k> RC motors do 20kW now?
[08:06:59] <theBear> wtf ? brushess rc motors for a pushbike ?
[08:07:04] <dirty_d> amee2k, i think so
[08:07:18] <dirty_d> amee2k, i bought a 6000W 8 horsepower motor for $100
[08:07:22] <scuzzy> theBear: it's not a pushbike if it's got a motor
[08:07:23] <amee2k> i think you want to define what the "rc" there means
[08:07:26] <theBear> that's ridiculous ! what kinda rc needs 20kW
[08:07:44] <dirty_d> theBear, i think you underestimate how large they make them
[08:07:49] <theBear> scuzzy, pfft, moped, pfft, it's a pushbike if it's got pedals and a lightweight frame... oh wait, i forgot what we were talking about
[08:08:15] <theBear> i know rc goes up to 1/5th in cars, and the modern 'brushless jet engines' are pretty outrageous, but NOTHING like 20kW
[08:08:52] <dirty_d> theBear,
http://www.hobbyking.com/hobbyking/store/__14427__Turnigy_CA120_70_Brushless_Outrunner_100cc_eq_.html
[08:09:09] <theBear> err, 100cc ?
[08:09:28] <amee2k> umm.
[08:09:32] <dirty_d> http://www.hobbyking.com/hobbyking/store/__19964__Turnigy_RotoMax_100cc_Size_Brushless_Outrunner_Motor.html
[08:09:33] <dirty_d> etc
[08:09:46] <theBear> err, what makes you think that is 20kW or even close to ?
[08:10:03] <dirty_d> its not
[08:10:05] <dirty_d> im just saying
[08:10:17] <theBear> saying what ? that's TINY compared to 20kW ...
[08:10:19] <dirty_d> they make larger than that i believe
[08:10:23] <dirty_d> but you dont need 20kw for a bike
[08:10:24] <theBear> ever ridden a 100cc motorbike ?
[08:10:25] <amee2k> i want one of these for my 4" RC dune buggy :P
[08:10:30] <dirty_d> theBear, its 8000W
[08:10:40] <theBear> dirty_d, and what makes you say that ?
[08:10:50] <scuzzy> you can buy like 10 - 20 kw brushless DC motors
[08:10:51] <dirty_d> theBear, Watts: 7992w
[08:11:02] <scuzzy> it doesn't HAVE to be Brushless
[08:11:05] <scuzzy> it just lasts longer
[08:11:12] <theBear> where ? that page doesn't list watts
[08:11:15] <amee2k> whats wrong with induction motors?
[08:11:18] <dirty_d> theBear, the second page
[08:11:40] <theBear> plus it uses around err, 11*20 watts just to idle, that's a stupid motor
[08:11:44] <dirty_d> amee2k, finding a suitable one, and driving it
[08:12:23] <theBear> it also says 167kv is the RPM... i'm not sure i trust that page
[08:12:37] <scuzzy> I'm talking about EV's here, not RC
[08:13:01] <dirty_d> theBear, whats not believable?
[08:13:15] <theBear> for a start, 167kv isn't a measure of rpm
[08:13:16] <amee2k> an EV is just a big RC that you can sit on
[08:13:20] <scuzzy> dirty_d: where did you buy that motor?
[08:13:26] <dirty_d> scuzzy, i didnt buy that one
[08:13:43] <theBear> second, it's very hard to believe you can pump 8kw into a 2kg motor without melting the whole thing
[08:13:46] <dirty_d> scuzzy, i have this one
http://www.hobbyking.com/hobbyking/store/__5142__Turnigy_80_100_B_130Kv_Brushless_Outrunner_eq_70_55_.html
[08:13:48] <scuzzy> no, the 6000w one
[08:13:57] <dirty_d> scuzzy, that link
[08:14:09] <scuzzy> that is a big motor
[08:14:16] <dirty_d> theBear, its air cooled by a propeller
[08:14:18] <theBear> without a calculator, i suspect that they are just multiplying max voltage by max current
[08:14:19] <dirty_d> yea
[08:14:43] <amee2k> theBear: i think the RPM rating was related to this:
http://en.wikipedia.org/wiki/Brushless_DC_motor#Kv_rating
[08:14:54] <scuzzy> could just chain up 4 of those!
[08:15:09] <dirty_d> theBear, its rpms per volt
[08:15:10] <theBear> look how big a 2kw 'pushbike kit' motor is
[08:15:11] <amee2k> the editor of that page is an asshat for confusing "kv", "Kv", and "kV" though
[08:15:32] <dirty_d> theBear, its smaller because it makes the power with more rpm rather than torque
[08:15:36] <theBear> so 670*70, yer gonna have one noisy ass gearbox to get that down to usable speeds
[08:15:45] <dirty_d> motors with more torque to rpm have to be larger
[08:15:54] <dirty_d> theres no way around that
[08:16:02] <pingec> Does anyone use putty for serial communication? Is it possible to make it output special chars like \CR\LF
[08:16:22] <theBear> so in exchange for this apparently magical feature of high rpm motors, yer gonna have a huge noisy power sucking gearbox next to it... doesn't sound good to me
[08:16:51] <dirty_d> theBear, people have used belt drives to make efficient reduction drives for them
[08:17:08] <theBear> belt drive at err, around 40krpm ? sounds touchy to me
[08:17:57] <dirty_d> where did you get that number?
[08:18:13] <dirty_d> at 70V the max speed would be 11,000rpm, and thatts with no load
[08:18:15] <theBear> i multipled 670 * 70 in my head
[08:18:33] <theBear> wow, 167 isn't 670
[08:18:36] <dirty_d> youd probably not want 70V and you would more likely be artounf 5000rpm
[08:18:49] <theBear> evem at 11krpm a belt is err, tenuous at best
[08:18:57] <dirty_d> its not gonna be that high
[08:19:02] <dirty_d> but motorcycle engines rev that high
[08:19:06] <dirty_d> and they manage
[08:19:16] <scuzzy> they go even higher
[08:19:24] <scuzzy> sometimes up to 20
[08:19:58] <theBear> but they got incredibly finely machined gearboxes and not a lot, if any belts
[08:20:01] <dirty_d> http://www.youtube.com/watch?v=kVN8AERoUOA
[08:20:19] <theBear> nice big gearboxes with metal gears in them
[08:20:50] <scuzzy> don't BLDC motors have a pretty linear relationship of torque to speed?
[08:21:02] <scuzzy> ie: higher speed -> less torgue
[08:21:04] <scuzzy> torque
[08:21:05] <dirty_d> scuzzy, yes if current is constant
[08:21:14] <scuzzy> fair enough
[08:21:16] <dirty_d> torque is proportional to current
[08:21:20] <dirty_d> neglecting saturation
[08:21:31] <scuzzy> I have no idea what that means
[08:21:55] <dirty_d> it means that if you have 20Nm of torque at 10A you will have 40Nm of torque at 20A
[08:22:00] <dirty_d> regardless of anythign else
[08:22:15] <dirty_d> as long as you are below the current where the iron saturates
[08:22:21] <theBear> but with 'traditional' drive circuits, your current will go down as the speed goes up
[08:22:32] <dirty_d> theBear, yes
[08:22:42] <theBear> unless you anticipate, but hmmm...
[08:22:52] <dirty_d> im just saying if you compensated for that with mroe voltage to maintain current, torque would be constant
[08:22:54] <theBear> i don't know any rc stuff that does
[08:23:04] <dirty_d> it doestn
[08:24:08] <inflex> hrm?
[08:24:49] <dirty_d> scuzzy, rc stuff is definitely the cheapest way to go to build something like this, the main challenge is a gearbox
[08:25:01] <scuzzy> dirty_d: I was hoping to not have one
[08:25:08] <scuzzy> just a direct chain/belt drive
[08:25:21] <dirty_d> well not so much a gearbox as 2 belts
[08:25:30] <dirty_d> well, one belt, one chain
[08:25:33] <scuzzy> I don't know what the RPM's of these bigger motors are
[08:25:41] <scuzzy> you think I should just but the biggest outrunner I can?
[08:25:46] <dirty_d> youd need to use something like an e-tek motor to have one chain
[08:25:54] <dirty_d> very big and torquey
[08:26:02] <scuzzy> dirty_d: I was planning on using something like an etek
[08:26:06] <dirty_d> scuzzy, i dont think you need one that big
[08:26:09] <scuzzy> you can buy brushless etek's
[08:26:15] <scuzzy> dirty_d: maybe not
[08:26:15] <dirty_d> yea those are like $800
[08:26:22] <dirty_d> i believe
[08:26:25] <dirty_d> or is it $400
[08:26:28] <amee2k> hmm can you gear two smaller ones together for a weight and/or efficiency advantage?
[08:26:29] <scuzzy> 20kw was like, the super high end
[08:26:41] <scuzzy> amee2k: possibly
[08:26:42] <dirty_d> yea thats rediculously high
[08:26:48] <scuzzy> maybe 10kw?
[08:26:49] <amee2k> well, weight/cost/efficiency
[08:26:51] <dirty_d> that motor i have will get you to at least 50mph
[08:27:06] <dirty_d> scuzzy, how fast do you need to go?
[08:27:08] <scuzzy> dirty_d: really? I don't wanna excede 70km/h
[08:27:11] <dirty_d> my 1200W ebike went 45mph
[08:27:14] <amee2k> also, how tollerant are these motors about short-term overload?
[08:27:32] <scuzzy> amee2k: apparently quite tollerant
[08:27:45] <scuzzy> they are usually rated like 8kw nominal, 20 kw peak
[08:27:51] <amee2k> from what i understand the main limitation is thermal, so couldn't you boost a 10kW motor up to 20 for like 30 seconds or so to climb a gradient
[08:27:57] <scuzzy> and the can sustain this peak for between 60 - 120 seconds
[08:28:07] <dirty_d> scuzzy, ok, 45mph is 72km/h, and i did that with 1200W
[08:28:08] <scuzzy> amee2k: indeed
[08:28:19] <scuzzy> dirty_d: I live in hill country
[08:28:22] <dirty_d> and crappy lead acid batteries
[08:28:22] <amee2k> just make sure you have thermal overload protection so it'll just throttle back when it runs out of juice
[08:28:26] <amee2k> instead of blowing the motor
[08:28:30] <scuzzy> dirty_d: I was planning to go lead acit
[08:28:32] <scuzzy> adic
[08:28:34] <scuzzy> acid *DSF*SDF)SFD
[08:28:50] <scuzzy> amee2k: that's quite an easy thing to build in I guess
[08:29:09] <dirty_d> amee2k, usually theres some kinda current limiting
[08:29:10] <amee2k> well, find the hottest part in it and superglue a 1wire temperature sensor to it
[08:29:19] <scuzzy> dirty_d: you think I could get 8kw out of an outrunner?
[08:29:27] <scuzzy> sustained?
[08:29:34] <dirty_d> scuzzy, if you hadd a fan to it
[08:29:40] <amee2k> when temp exceeds so-many degrees, override throttle control and progressively lower power
[08:29:50] <scuzzy> dirty_d: don't want the bike pulling sideways! (;
[08:29:57] <dirty_d> scuzzy, isnt 8kw overkill though?
[08:30:02] <scuzzy> dirty_d: I don't know
[08:30:18] <scuzzy> dirty_d: I mean, it's hard to gauge this stuff without riding it
[08:30:33] <dirty_d> if you have hills multiple gears is a good idea
[08:30:37] <amee2k> scuzzy: if you live in a hilly country, considering regenerative braking would come to mind too
[08:30:38] <scuzzy> 8kw Emotor vis 8kw petrol motor, and it's a completely different game
[08:30:45] <dirty_d> some people power the crank so you cna still use the bikes gears
[08:30:51] <scuzzy> amee2k: oh, definitely
[08:31:10] <scuzzy> dirty_d: how much power loss would you get?
[08:31:13] <dirty_d> an 8kw electric motor will feel a lot more powerful than a 8kw petrol
[08:31:16] <amee2k> well, an 8kW combustion engine won't do 20kW for 40 seconds
[08:31:19] <scuzzy> don't gear boxes chow energy?
[08:31:28] <dirty_d> scuzzy, hardly any, a bikes drive train is like 95% efficient
[08:31:28] <scuzzy> dirty_d: exactly! that's my point
[08:31:37] <scuzzy> it's hard to gauge how powerful 8kw is
[08:31:40] <amee2k> so that 8kW electric motor would yield much better handling than an 8kW gasser
[08:32:19] <dirty_d> id use that 6000W motor and gear it to only use 4000W
[08:32:30] <dirty_d> or less
[08:32:37] <dirty_d> gear low so you have power on hills
[08:32:38] <scuzzy> dirty_d: it might be a good start
[08:32:44] <scuzzy> I mean, I guess I could change things in future
[08:32:47] <dirty_d> yea
[08:33:19] <scuzzy> maybe start with like a 6kw motor
[08:33:24] <scuzzy> dirty_d: did you make the controller yourself?
[08:34:20] <dirty_d> i did but never got to use it
[08:34:26] <scuzzy> can I have yours?
[08:34:27] <scuzzy> (;
[08:34:29] <dirty_d> i havent used the 6000W motor yet
[08:34:37] <dirty_d> i fired the controller, lol
[08:34:39] <dirty_d> fried
[08:35:11] <scuzzy> oh nice
[08:35:12] <scuzzy> lol
[08:35:13] <dirty_d> http://endless-sphere.com/forums/download/file.php?id=11592&t=1
[08:35:16] <dirty_d> thats the bike i made
[08:35:19] <dirty_d> with the 1200W motor on it
[08:35:27] <scuzzy> have you got the schematics for your controller?
[08:35:34] <dirty_d> somewhere
[08:35:49] <dirty_d> i installed hall sensors in the motor to make driving it easier
[08:36:07] <dirty_d> the 6000W brushless one i mean
[08:36:11] <dirty_d> the 1200W one is brushed
[08:36:27] <scuzzy> yeah, I head some controllers can cope without the hall sensors, some can't
[08:36:47] <scuzzy> does the sensor make it smoother?
[08:36:50] <dirty_d> you might be able to get away with just using an RC ESC though
[08:37:07] <dirty_d> scuzzy, im not sure, starting from 0rpm it definitely would
[08:37:16] <dirty_d> starting is the downfall of sensorless
[08:37:26] <scuzzy> yeah
[08:37:28] <dirty_d> its ok for a propeller, but not so great for a wheeled vehicle
[08:37:31] <scuzzy> because it's guesswork isn't it?
[08:37:35] <dirty_d> sorta
[08:37:38] <scuzzy> you're making assumptions
[08:37:50] <dirty_d> its kind of just hoping that the motor will be able to follow
[08:37:55] <scuzzy> yeah
[08:37:57] <dirty_d> if theres too much torque it wont
[08:38:13] <scuzzy> well, I'd put in a hall sensor I think
[08:38:18] <scuzzy> it's not expensive anyway
[08:38:29] <dirty_d> not, less than a dollar each i think
[08:38:30] <dirty_d> just need 3
[08:38:35] <scuzzy> yeah
[08:38:35] <amee2k> so you just drive it open loop at low speed until the motor catches on and starts turning?
[08:38:40] <dirty_d> you need a lot of mosfets though
[08:38:52] <scuzzy> dirty_d: your controller use fets?
[08:38:52] <dirty_d> amee2k, i guess so
[08:39:08] <dirty_d> scuzzy, yea its not high enough voltage to make igbts a better option
[08:39:21] <scuzzy> I'm thinking like 48volts
[08:39:23] <scuzzy> something like that
[08:39:40] <scuzzy> would you mind if I have a look at your controller?
[08:39:50] <dirty_d> scuzzy,
http://www.youtube.com/watch?v=nQtVXJVdxR0&feature=channel
[08:39:56] <dirty_d> thats it shortly before i blew it
[08:40:13] <dirty_d> i didnt set the hard current limit pot, and hadnt perfected the software current limiting yet
[08:40:14] <dirty_d> lol
[08:40:25] <scuzzy> nice
[08:40:58] <scuzzy> this is basically the problem, I'm not good enough to do current limiting
[08:40:59] <dirty_d> i think i have the schematic on my computer at home, maybe its on that forum, ill check
[08:41:28] <scuzzy> is that the 6.5kw motor?
[08:41:33] <dirty_d> you should be able to do it with a pid algorithm
[08:41:36] <dirty_d> yea it is
[08:42:31] <scuzzy> I guess I need to know the theory first
[08:44:43] <dirty_d> i found it
[08:44:55] <scuzzy> you're a legend
[08:45:00] <dirty_d> http://endless-sphere.com/forums/download/file.php?id=14355
[08:45:01] <dirty_d> lol
[08:45:17] <dirty_d> it probably needs tweaking, it did blow up after all
[08:45:27] <dirty_d> but i think it was just because i didnt set the current limit first
[08:46:57] <scuzzy> you have a pid algorithm in there?
[08:47:03] <dirty_d> i dont think so
[08:47:16] <dirty_d> oh wait, theres no code for the atmega
[08:47:31] <dirty_d> thats just the schematic
[08:47:40] <scuzzy> I've used pid algorithms before, just not in any relation to hardware
[08:48:40] <scuzzy> I don't know the electronic theory for liming current
[08:50:15] <dirty_d> well the only thing you can do in this situation is lower the voltage
[08:50:21] <dirty_d> by lowering the pwm duty cycle
[08:50:33] <scuzzy> Hmmm
[08:50:58] <dirty_d> what i did was instead of make throttle increase voltage, i made the throttle control current
[08:51:00] <scuzzy> easier than I thought then I guess
[08:51:11] <scuzzy> how?
[08:51:25] <dirty_d> just make the throttle the setpoint of the pid algorithm
[08:51:38] <dirty_d> if motor current is higher than the throttle is telling to it be decrease pwm
[08:51:47] <dirty_d> other way around increase pwm
[08:51:58] <dirty_d> thats the idea basically
[08:52:17] <dirty_d> just gotta tune it right
[08:52:22] <dirty_d> so it doesnt oscillate etc
[08:52:23] <scuzzy> yeah, I guess
[08:52:25] <dirty_d> i never got that far
[08:52:56] <dirty_d> i didnt know pid well enough then
[08:53:02] <dirty_d> im using it now for a quadcopter
[08:53:44] <dirty_d> scuzzy, it may make things easier if you write some kind of simulation involving a motor and a throttle and try to write an algorithm that can control it well before you do it for real
[08:54:17] <dirty_d> thats what i did for the quadcopter, i made a physics simulation with the quad, motors, esc gyro, accelerometer etc and figured out how pid works
[08:54:26] <scuzzy> considering how much these things cost, that's a very good idea
[08:54:31] <dirty_d> yea
[08:54:48] <dirty_d> i would have desttroyed 20 real quadcopters if i hadnt done that, lol
[08:55:06] <dirty_d> you can make that pretty simply with a time step simulation
[08:55:21] <scuzzy> yeah, it wouldn't be too difficult
[08:56:00] <dirty_d> i think what youd have to model would be the motors electrical charactaristics, resistance kv, inductance etc, the moment of inertia, the throttle and thats it i think
[08:56:29] <dirty_d> maybe some kinda opengl graph of current, voltage and pwm drive
[08:58:51] <scuzzy> Hmmm
[08:59:11] <scuzzy> thanks for the help, I gotta run
[08:59:40] <dirty_d> no problem, later
[08:59:46] <scuzzy> well, thinking about it, you can only use the variables you can monitor
[08:59:59] <scuzzy> ie: current, voltage, rpm
[09:00:03] <scuzzy> throttle
[09:00:14] <dirty_d> yea, in the pid part
[09:00:23] <dirty_d> but everything has to be in the physics part
[09:00:24] <scuzzy> yup
[09:00:41] <scuzzy> assuming I've moddled it correctly. (;
[09:00:45] <scuzzy> right, gotta go
[09:00:45] <dirty_d> yea
[09:00:47] <dirty_d> later
[09:53:04] <dirty_d> hmm, wacky behavior
[10:29:03] <dirty_d> if i have two separate signals going to two pins that are next to each other that trigger interrupts, what might cause one of the signals to trigger the other pins interrupt?
[10:29:35] <dirty_d> im not sure if i have a capacitor across GND-VCC on that side of the chip
[10:29:39] <dirty_d> would that might do it?
[10:30:31] <amee2k> CapnKernel: \o/
[10:32:19] <CapnKernel> amee2k: Why hello there!
[10:32:50] <amee2k> CapnKernel: do you sell (official) eval/dev boards too?
[10:33:12] <CapnKernel> I can probably find them, but I wouldn't bet on it
[10:33:34] <CapnKernel> I do see shops with eval/dev boards, but I don't know how reliable they are.
[10:37:01] <CapnKernel> Re XRC vs XRE, are you able to do some looking and find out a newer model for the red emitters?
[10:37:13] <CapnKernel> amee2k: yes you :-)
[10:37:25] <amee2k> yeah, i'll take a look
[10:37:34] <amee2k> (PM, btw)
[10:38:21] <amee2k> CapnKernel: well, i've been scratching my head over the STM32 discovery boards and was wondering if i could get one with my parts order from you too
[10:40:02] <CapnKernel> If you can give me a web page that shows me precisely what you want, I might be able to do something. My standard M.O. is to print it out and walk around waving it in the air.
[10:40:16] <amee2k> http://www.st.com/internet/evalboard/product/252419.jsp
[10:40:20] <amee2k> STM32F4DISCOVERY
[10:41:48] <theBear> my standard mo is neatly trimmed straight above the lip and maybe a little bit at the top so as to suggest a little curve down from the nostrils...
[10:42:07] <theBear> wish i had a big fat one, then i could do a line in it like Cheech Marin
[10:42:09] <CapnKernel> theBear: LOL
[10:42:40] <CapnKernel> amee2k: It may well be more expensive than if you bought it locally, as the import duty is pretty stiff.
[10:42:47] <CapnKernel> But I'll put it on the list of stuff to look for.
[10:43:57] <amee2k> cheapest i've found so far is 16EUR + 6EUR shipping
[10:44:27] <amee2k> considering that the ST website lists them at under 8$, how bad can it come?
[10:44:33] <CapnKernel> I'd be surprised if I could get it for that.
[10:44:43] <amee2k> lol wat?
[10:44:49] <CapnKernel> Price advantage here comes from massive competition
[10:45:05] <CapnKernel> I bought cyanide some nice Panasonic caps today
[10:45:14] <CapnKernel> If there's no competition, they can charge what they like
[10:45:28] <amee2k> okay
[10:45:44] <CapnKernel> Can you look into the red emitters please?
[10:46:01] <amee2k> looking right now
[10:46:01] <CapnKernel> Also, can you sent me your short list of driver chips, and I'll get prices
[10:46:09] <CapnKernel> How many units are you ultimately looking to buy?
[10:46:21] <amee2k> CapnKernel: check your PM
[10:46:28] <CapnKernel> As in, some time down the track, how many driver boards and emitter boards?
[10:47:22] <amee2k> first run is only 5 boards to see if they work and for the immediate deployment (which is somewhat overdue by now anyway :/ )
[10:48:08] <CapnKernel> Re your assorted resister and cap packs, might be better to look on eBay, they are quite cheap there
[10:48:27] <CapnKernel> What I want to do that you can't do on eBay is get a BoM's worth of components
[10:49:18] <amee2k> yeah, on ebay i can pay 500$ and have a huge closet with a resistor set in my room
[10:50:08] <CapnKernel> I'd look for you, except I'm currently connected via my phone and counting bytes.
[10:50:26] <CapnKernel> Seriously, you can buy a book of common values for (guesses) USD10
[10:51:17] <dirty_d> just buy 1k, 10k, 100k etc
[10:51:24] <dirty_d> and series parallel to your needs
[10:51:40] <amee2k> well, the small sets don't have values <10R
[10:51:40] <dirty_d> im unorganized
[10:51:46] <dirty_d> so i cant deal with more than 5 kinda of anything
[10:51:50] <theBear> heh, and end up with every project you make having big stacks of weird parallel/series resistors sticking out the top
[10:52:01] <amee2k> dirty_d: i want to see you solder a huge stack of 1k resistors to get a 0.47 shunt >_>
[10:52:33] <dirty_d> amee2k, fine 0.1 ohm, 1 ohm, 100, ohm, 1k, 10k, 100k, 1M
[10:52:34] <dirty_d> lol
[10:53:27] <dirty_d> my heads killing me
[10:54:50] <CapnKernel> Seriously, you can buy a strip of every E12 value for a few dollars.
[10:56:59] <dirty_d> amee2k, what are you making
[10:58:54] <amee2k> dirty_d: trying to get my hands on some LED driver ICs without having to pay 5$ a pop from a distributor
[10:59:10] <amee2k> i'm working on some small form factor LED driver board to retrofit fluo fixtures
[10:59:41] <amee2k> swap the ballast board for the driver, put 24V on the wires instead of mains, and drop a LED board into the tube socket
[11:00:43] <dirty_d> cool
[11:00:53] <dirty_d> how much power goes through it?
[11:00:55] <CapnKernel> $5 for a prototype, or $5 for a product that you're going to make a few of?
[11:02:32] <dirty_d> amee2k, what would happen if you rectify 120vac to 170vdc and use enough leds in series to take that voltage with some kinda current limiting
[11:03:27] <amee2k> got any thoughts on how to implement the current limiting? ;)
[11:03:58] <dirty_d> a current sense resistor and a to-220 300V transistor or something
[11:03:59] <amee2k> also, that would be between 40 and 60 LEDs
[11:04:13] <dirty_d> drop whatever voltage across the transistor to get the current you want
[11:04:32] <amee2k> the 6 LEDs i'm using are already overkill. the fixtures we want to upgrade are between 8 and 20W
[11:05:11] <dirty_d> ok
[11:05:52] <amee2k> 2 would probably be bright enough, i'm only using 6 so i can use normal FR4 boards and increase service life
[11:05:57] <dirty_d> so not those big shop flourescent fixtures?
[11:06:01] <dirty_d> the long tube one
[11:06:02] <dirty_d> s
[11:06:31] <amee2k> no, thats for the second led board project that is on hold right now due to lack of budget
[11:06:34] <dirty_d> is your board going to do the 120vac to 24V conversion?
[11:07:05] <amee2k> no, we'll get some 24V SMP module for running them
[11:07:24] <amee2k> off line drivers are nasty because they almost always need transformers
[11:07:35] <dirty_d> can you just make a buck converter?
[11:07:41] <dirty_d> with current sensing
[11:07:54] <dirty_d> that way you can series as many leds as you want and your design wont have to change
[11:08:24] <dirty_d> 120Vac to whatever V dc to meet the current requirement of however many leds you have
[11:08:32] <dirty_d> plug and play
[11:09:00] <amee2k> 20V led string driven off 320Vdc would need a duty cycle of 6%
[11:09:11] <dirty_d> 320?
[11:09:18] <amee2k> rectified 230V
[11:09:27] <dirty_d> its not 120v?
[11:09:35] <dirty_d> europe or something?
[11:09:38] <amee2k> no, we have real electricity here :P
[11:09:59] <dirty_d> your real electricity is making your job harder
[11:10:00] <dirty_d> lol
[11:10:08] <amee2k> high voltage ultra low impedance transistors, large inductors, huge current pulses, and i'd need an auxiliary supply for the regulator IC too
[11:10:43] <amee2k> http://i.ebayimg.com/00/$%28KGrHqJ,!hIE458p9CevBOTQb!WrD!~~48_35.JPG << these things are 20-30EUR for a 150W PSU
[11:11:09] <amee2k> doesn't look stylish but they seem to work fine and the build quality is no worse than your average ATX supply
[11:12:09] <amee2k> one of these can run at least a dozen led fixtures at once
[11:14:00] <dirty_d> sweet
[11:14:07] <dirty_d> how many amps per fixture?
[11:14:38] <CapnKernel> I'm fairly sure you can buy LED replacements for long fluoros...
[11:14:45] <amee2k> actually 24V PSUs seem to be easier to find than 12V ones. the cheap "electronic transformers" for halogen lighting are useless... they give some 20kHz AC with a huge amount of mains hum riding on it
[11:15:03] <amee2k> CapnKernel: for 3EUR + cost of the LEDs?
[11:15:52] <CapnKernel> You have to consider that the cost of the off-the-shelf ones include manufacturing? What is it costing you per hour to mess around with this, and if and when you settle on a design, what's your manufacturing costs?
[11:17:05] <amee2k> dirty_d: the small driver boards are intended for 10W worth of LEDs tops. the larger board will start with 10-15W too and the second version will have a 30W driver. or at least thats the plan
[11:17:22] <amee2k> CapnKernel: well, i don't get any experience from buying an off the shelf module ;)
[11:17:45] <amee2k> i want the LM3404 on my samples list because thats the driver i'm using on the large board too
[11:18:18] <amee2k> and the thought makes my skin crawl because i haven't even seen one of these irl, much less actually used them
[11:22:01] <dirty_d> amee2k, i think you can make a better cuck converter yourself and cheaper
[11:22:04] <dirty_d> buck*
[11:22:41] <dirty_d> a schmidt trigger sensing the output current and driving a mosfet would work i think
[11:23:02] <dirty_d> the hysteresis and inducotr size would determine frequency
[11:23:10] <CapnKernel> op amp?
[11:23:15] <dirty_d> yea
[11:23:20] <CapnKernel> I take that back
[11:23:36] <dirty_d> what
[11:24:22] <CapnKernel> op amp. now don't mention it again, as I've already realised it's a dumb idea
[11:24:31] <dirty_d> why?
[11:25:19] <CapnKernel> Piss off troll!
[11:25:30] <dirty_d> what? lol
[11:26:13] <CapnKernel> laffs all round
[11:30:22] <dirty_d> nah it will work
[11:33:01] <theBear> "THIS WILL WORK !" "huh, wha, what are you doing ?" "reviving you ... i've seen it like a billion times on tv" <throws homemade tesla coil in the swimming pool>
[11:34:28] <dirty_d> lol
[11:50:12] <_abc_> Is there some popular little board based on low end avrs with an ascii 44770 display and keyboard matrix on it?
[11:54:38] <CapnKernel> There are some Arduino shields that have such a thing
[11:54:49] <CapnKernel> OMG he said the forbidden word!
[11:58:21] * specing enters rage mode
[11:58:37] <specing> Hey CapnKernel
[11:59:07] <specing> Besides making custom PCBs, do those chinese also make custom chips?
[12:03:09] <CapnKernel> specing: Hey there
[12:03:09] <CapnKernel> Yes
[12:03:27] <CapnKernel> But you better have a wheelbarrow of money and be prepared to buy in the millions
[12:03:59] <CapnKernel> For smaller designs, folks often use CPLDs and FPGAs
[12:13:00] <specing> CapnKernel: ah
[12:13:32] <CapnKernel> For example, that's what Atmel do. They don't have their own fabs any more. Someone else makes the chips for them.
[12:13:52] <specing> The local university has a fab
[12:14:44] <CapnKernel> Cool! I wish I was there, I'd be trying to develop a hobbyist friendly FPGA
[12:15:45] <RikusW> that would be nice
[12:16:04] <RikusW> with open specs probably too :)
[12:16:52] <CapnKernel> Yes, and open tools
[12:17:18] <CapnKernel> You could do some cool things, like make a programmable logic analyser. Set up your match conditions on the PC, compile it to something, and squirt it down to the FPGA capture engine
[12:18:46] <RikusW> configuring an fpga incorrectly can result in drive contention ;)
[12:18:51] <RikusW> sw break hw....
[12:19:37] <CapnKernel> Don't Do That[tm]
[12:20:06] <RikusW> hehe
[12:21:19] <mapee> hi
[12:53:20] <specing> CapnKernel: they go down to 2 um
[12:54:46] <CapnKernel> So it's about 25 years old :-)
[12:54:53] <CapnKernel> Or 20
[12:55:14] <CapnKernel> Because when I did VLSI design in 1993, 1um was The Thing
[12:55:40] <CapnKernel> It would still be useful.
[12:56:03] <theBear> heck, if i could make my own chips i wouldn't care if they were double the size of '93 grade chips
[12:57:25] <specing> CapnKernel: Yes, its from the communist times, thats when some industry was still here
[12:57:45] <CapnKernel> Interesting
[12:57:59] <CapnKernel> Could have been used to copy chips from the west :-)
[12:59:32] <scuzzy> hey RikusW
[12:59:36] <scuzzy> how you doing man?
[13:00:03] <RikusW> Hi scuzzy
[13:00:15] <RikusW> fine, just a little tired
[13:00:18] <RikusW> how about you ?
[13:00:20] <scuzzy> yeah?
[13:00:24] <scuzzy> urg, I'm ok
[13:00:31] <scuzzy> managed to tear the graft in my knee
[13:00:38] <scuzzy> so, back under the knife for me
[13:01:03] <scuzzy> why are you tired?
[13:01:14] <RikusW> oh no... sounds bad
[13:01:33] <scuzzy> it's a bit crap
[13:02:50] <RikusW> seems like a dairy allergy or something
[13:03:06] <RikusW> whenever I eat any dairy it makes me feel bad
[13:05:45] <scuzzy> give it a try
[13:05:47] <scuzzy> stop eating it
[13:05:52] <scuzzy> I had to give up cheese
[13:06:01] <scuzzy> you get used to it after a while
[13:07:00] <RikusW> I gave it up more or less
[13:07:10] <RikusW> just ate some yesterday, seems like a mistake
[13:07:22] <RikusW> so why did you give up cheese ?
[13:08:00] <theBear> i'd rather die than give up cheese
[13:10:04] <RikusW> theBear: love it too much ?
[13:10:30] <theBear> well lets just say it's not cos i don't like it :)
[13:11:43] <scuzzy> I have high cholesterol
[13:11:51] <scuzzy> so, I will die if I keep eating it
[13:12:08] <theBear> i'm a chronic alcaholic and will probably die from it, but i aint slowin down
[13:12:11] <CapnKernel> Out of consideration for me, can you guys please stop talking about cheese???
http://capn.blogspot.com/2012/01/cravings.html
[13:12:16] <theBear> hehe ok
[13:13:27] <CapnKernel> Yay, I'm into the second month of my tour of duty, I guess I can open my second sacred block of cheese!
[13:16:13] <Steffanx> All Aussies are cheese-addicts?
[13:16:44] <theBear> sssh, stop mentioning it
[13:16:49] <theBear> and probly not all
[13:16:57] <Steffanx> I thought we were… the dutch
[13:17:19] <theBear> you are the dutch
[13:17:25] <theBear> i know i'm not
[13:17:49] <scuzzy> The only time I eat cheese now, is on home made pizza
[13:17:53] <Steffanx> I mean, I thought we were the cheese people :P
[13:18:04] <dirty_d> amee2k, CapnKernel, i messed around a little with ltspice, something like t his would work for an LED driver right?
http://i1186.photobucket.com/albums/z372/dirty_d2285/led_driver.png
[13:18:08] <scuzzy> it's my... luxury
[13:18:08] <scuzzy> hehe
[13:19:57] <theBear> dirty_d, it's so small
[13:20:05] <dirty_d> yea
[13:20:16] <dirty_d> you cna see the general idea
[13:20:20] <theBear> woah, and the fet is all sideways, took me till now to find it
[13:20:43] <theBear> and what is v2 sposed to be ? a spare battery ?
[13:20:44] <dirty_d> low side mosfet switch, differential amp for shunt voltage, schmidt trigger for current control
[13:21:03] <amee2k> dirty_d: i can't see anything on that schematic ;)
[13:21:06] <dirty_d> oh that was for the lower comparator, 6V max
[13:21:31] <dirty_d> amee2k, you could make that a lot cheaper than buying those led drivers couldnt you?
[13:21:40] <dirty_d> and this could handle however much current you want
[13:22:21] <specing> CapnKernel: Better start designing it then
[13:22:22] <amee2k> what are the voltages on these two voltage sources?
[13:22:57] <theBear> i find simulations with more than one dc source are err, hard to reproduce in real life
[13:23:00] <amee2k> did you scale that screenshot down or something??
[13:23:05] <dirty_d> http://i1186.photobucket.com/albums/z372/dirty_d2285/led_driver_big.png
[13:23:07] <dirty_d> thats bigger
[13:23:10] <dirty_d> no the website did
[13:23:53] <dirty_d> youd need a driver for the mosfet etc, and choose the appropriate parts
[13:23:53] <theBear> hmm buck eh
[13:24:04] <amee2k> theBear: not neccessarily, but simulation has a way to make every problem look easy by popping in a lot of auxiliary supplies
[13:24:06] <dirty_d> i just threw it together with whatever ltspice had that was in the ballpark
[13:24:13] <theBear> amee2k, it does :)
[13:24:19] <dirty_d> you dont neccissarily need the 5v supply
[13:24:25] <dirty_d> just use a comparator that can handle 24V
[13:24:38] <CapnKernel> dirty_d: Sorry, I'm on a bandwidth diet tonight
[13:24:48] <theBear> oh that is supply.. damn, you draw everything sideways and upside down
[13:24:51] <dirty_d> i upped to 30V because whatever leds those are i couldnt get over 6A with on 24V
[13:24:56] <theBear> not to mention backwards :)
[13:24:58] <dirty_d> theBear, i know im horrible
[13:25:13] <theBear> you are, and i may never forgive you
[13:25:15] * theBear smiles
[13:25:32] <dirty_d> lol
[13:26:24] <amee2k> o.O
[13:26:42] <amee2k> well, all you have to do now to convince me is make a prototype of your "simple" circuit
[13:26:57] <amee2k> and a video that shows it popping 6 amps into a dummy load
[13:26:59] <dirty_d> ok, make a prototype
[13:27:00] <dirty_d> lol
[13:27:08] <dirty_d> oh i have to make it?
[13:27:08] <dirty_d> lol
[13:27:13] <theBear> hehe
[13:27:27] <amee2k> (doesn't have to be LEDs, take a pile of 4007s if you're cheap)
[13:27:32] <theBear> don't forget you gotta find a 6amp dummy load
[13:27:54] <amee2k> 10ohm/meter heating wire is cheap on ebay
[13:27:56] <theBear> what is d6 rated voltage wise ?
[13:28:10] <dirty_d> i think it was 60V
[13:28:19] <theBear> errr, why ?
[13:28:27] <amee2k> but for this you'll want the 4001 dummy load for the diode-like IV curve
[13:28:35] <dirty_d> i dunno, im not actually makin this i just wanted to see if the concept was valid, lol
[13:28:48] <amee2k> dirty_d: yes. i mean, that circuit is really dead simple, no?
[13:28:54] <dirty_d> yea pretty much
[13:29:01] <amee2k> i expect to see pics of your running prototype in 20 minutes then :P
[13:29:08] <dirty_d> im at work, lol
[13:29:31] <theBear> so it should be easier than in your home workshop :)
[13:30:02] <dirty_d> all i have is a computer
[13:30:44] <dirty_d> amee2k, ive pondered the same idea except in boost configuration to charge a cap to like 1000V from low voltage
[13:30:54] <amee2k> ...
[13:31:05] <dirty_d> (i wanted to make a railgun)
[13:31:37] <dirty_d> coilgun i mean
[13:31:44] <dirty_d> damn my head hurts
[13:31:47] <amee2k> i bet that thing is going to cook at 6 amps
[13:31:58] <dirty_d> what will
[13:32:04] <amee2k> the FET
[13:32:23] <amee2k> not sure about the opamp that does the driving
[13:32:36] <dirty_d> you would have that drive a totel-pole etc
[13:32:40] <dirty_d> totem*
[13:32:47] <amee2k> o.O
[13:32:51] <dirty_d> they make mosfets that will handle whatever current you need
[13:32:55] <amee2k> all i can see in that schem is an opamp
[13:33:09] <dirty_d> amee2k, are you looking at the bigger picture?
[13:33:26] <amee2k> http://i1186.photobucket.com/albums/z372/dirty_d2285/led_driver_big.png << this one?
[13:33:36] <dirty_d> yea
[13:33:51] <amee2k> the opamp output goes straight into the gate
[13:33:58] <dirty_d> i know
[13:34:05] <dirty_d> it was just for simulation purposes
[13:34:09] <dirty_d> it worked, so i didnt change it
[13:34:22] <dirty_d> you wouldnt do that in real life
[13:34:24] <amee2k> whats ltspice saying about dissipation on that FET?
[13:35:02] <amee2k> so err, you're complaining that i'm overcomplicating things based on a simplified circuit that you're admitting would never be practical outside a simulation?
[13:35:17] <dirty_d> im not complaining
[13:35:29] <dirty_d> but thats 2 transistors that would need to be added
[13:35:36] <dirty_d> nothing serious
[13:36:52] <amee2k> whats the switching frequency? looks like some 150kHz from the pic
[13:36:53] <dirty_d> isnt there a way to see an average to get average power dissipation?
[13:37:00] <dirty_d> about 100kHz
[13:37:12] <amee2k> alt+left click i think
[13:37:15] <dirty_d> with a 100u inductor it drops to 10Khz
[13:37:19] <theBear> average on that graph just draw a line near the middle
[13:37:36] <theBear> bout 6.4a
[13:37:52] <dirty_d> i mean for power dissipation of the fet
[13:39:20] <dirty_d> amee2k, ahh ok, 11W
[13:39:24] <dirty_d> peaks at 200W
[13:39:52] <amee2k> 11W is enough to warrant a heatsink
[13:39:58] <dirty_d> yup
[13:40:00] <amee2k> whats the peak gate current?
[13:40:36] <amee2k> the 1720 is rated for an absolute maximum of 20mA. with only 11W dissipation i have a hunch ltspice isn't modelling that right
[13:40:55] <dirty_d> hold on im gonna add totem-pole
[13:41:29] <amee2k> i'd rather replace the compartor with an inverting gate driver IC
[13:42:15] <amee2k> gate drivers are like 50 cent these days
[13:42:16] <dirty_d> you could do that
[13:43:07] <amee2k> or replace everything with a 555
[13:43:31] <CapnKernel> Or an ATtiny
[13:44:00] <dirty_d> amee2k, how would the current limiting work?
[13:44:27] <amee2k> just amplify the shunt drop and wire the 555 as ghetto schmitt trigger
[13:44:28] <dirty_d> a comparator is super simple and you get cycle by cycle current limiting
[13:45:01] <amee2k> 555/556/558 and 74AS gates are neat choices for ghetto gate drivers
[13:54:13] <norbi> guys, from the api example code:
http://www.nongnu.org/avr-libc/user-manual/group__avr__boot.html
[13:54:56] <norbi> w=UDR in case we are using USART?
[16:05:23] <Rok> tom_itx you there?
[16:35:15] <Tom_itx> yup
[16:35:38] <Tom_itx> you left a min too soon though
[16:39:53] <jadew> 10 actually :)
[16:40:22] * Tom_itx shrugs
[16:42:57] <Hackbat> Would anyone be up to explaining how to set up the makefile and fuses for my project? I understand somethings but I could use some help.
[16:43:51] <jadew> do it by hand first, then make a makefile
[16:44:05] <Hackbat> I have an ATtiny 2313 which needs to accept input from a external 20 mhz clock
[16:44:07] <Tom_itx> this guy wrote somewhat of a tutorial with code examples:
http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/mega168_howto_main_index.php
[16:44:37] <Tom_itx> just change the names to protect the innocent
[16:45:04] <Tom_itx> although i don't have an input capture routine there
[16:45:29] <Tom_itx> you need to capture a 20Mhz signal on an input pin?
[16:45:33] <Tom_itx> good luck with that
[16:46:02] <Hackbat> No it needs to use 20 mhz on Xtal
[16:46:05] <Tom_itx> oh
[16:46:28] <Tom_itx> http://www.engbedded.com/fusecalc/
[16:46:36] <Tom_itx> set it up for a full swing crystal
[16:47:52] <Hackbat> Full swing?
[16:49:13] <Tom_itx> let's just say cksel = 1111 sut = 11
[16:49:48] <Tom_itx> (the last one in the list)
[16:50:18] <Tom_itx> then look at the fuse settings at the bottom of the page
[16:50:20] <Hackbat> kay, the 8 mhz was what confused me
[16:50:35] <Tom_itx> is the 2313 rated for 20Mhz?
[16:50:54] <Tom_itx> better check the data sheet on that one
[16:51:49] <Hackbat> It runs at 20mhz
[16:52:08] <Tom_itx> then use that setting
[16:52:40] <Tom_itx> and decide if you want ckdiv8 enabled or not too
[16:52:48] <Tom_itx> it will divide the clock down if it's set
[16:54:17] <Hackbat> IWhat would be the plus side to that?
[16:54:20] <Hackbat> *What
[17:01:55] <Tom_itx> the timers would be fed with 2.5Mhz instead of 20Mhz
[17:09:42] <Hackbat> Thanks Tom_itx :3
[17:10:10] <Hackbat> Now I just need to set up the make file and I'll be golden
[17:10:10] <Tom_itx> don't screw up your fuse settings
[17:10:17] <Tom_itx> you may end up with a brick
[17:10:32] <Tom_itx> borrow a makefile from my tutorials and mod it
[17:17:10] <xorm> there's a handy fuse calculator in a webapp somewhere
[17:17:31] <xorm> google 'avr fuse calculatpr,' makes it p deck simple
[17:18:51] <Tom_itx> i posted engbedded's
[17:19:03] <xorm> ah that's the one
[18:15:14] <Hackbat> oh...
[18:15:18] <Hackbat> I'm stupid
[18:16:11] <Tom_itx> you were warned
[18:16:12] <Hackbat> The chip won't work because I don't have an external clock hooked up
[18:16:15] <Hackbat> lol
[18:16:30] <Hackbat> Let me pop it into it's circut then >_<
[18:17:38] <OndraSter> lol
[18:17:40] <OndraSter> epic fail
[18:20:00] <Tom_itx> bricking an avr is a reight of passage
[18:35:39] <jadew> hey, what kind of wave do I need as a clock?
[18:36:33] <jadew> can I get away with an almost triangle wave?
[18:36:47] <Tom_itx> why would you want to?
[18:37:01] <Tom_itx> sine would be best i'd say
[18:37:06] <Tom_itx> or square
[18:37:30] <jadew> I was hoping for sine
[18:37:48] <jadew> but the sine I get from this crystal osc is more like a triangle
[18:38:14] <xorm> it might work depending on the jitter
[18:38:32] <jadew> what exactly is jitter?
[18:38:33] <Tom_itx> is your test equipment skewing the result?
[18:38:38] <jadew> I keep hearing about it, I imagine some noise
[18:38:51] <xorm> jitter is frequency instability
[18:39:19] <jadew> Tom_itx, it is possible, but I'm inclined to blame it on the transistor I used
[18:40:44] <jadew> thing is I want to feed this into that wave generator IC
[18:41:05] <jadew> I hope it won't screw up too bad
[18:43:00] <jadew> with a 20Mhz crystal the shape of the wave doesn't seem so distorted
[19:28:25] <jadew> well, the 20Mhz sinewave I'm getting from my oscillator seems a lot cleaner and stable than the 12Mhz square one I'm getting from the clock output pin of a 2313
[19:28:38] <Hackbat> Wooo!
[19:28:42] <Hackbat> Got it working :3
[19:28:50] <jadew> the makefile?
[19:28:57] <Hackbat> The chip @_@
[19:28:58] <Tom_itx> finally dig yourself outta the hole you dug?
[19:29:08] <jadew> congrats :)
[19:29:12] <Hackbat> Lesson learned
[19:29:18] <jadew> what did you do?
[19:29:25] <Hackbat> Don't enable the external clock until you have it in circut
[19:29:33] <jadew> ah
[19:29:33] <Tom_itx> screwed the pooch on the clock source
[19:29:42] <jadew> I think we all did that at one point :P
[19:29:46] <Hackbat> Best way to learn
[19:30:07] <Tom_itx> it came with all the warning stickers attached
[19:31:59] <Tom_itx> Hackbat, when i was using a bunch of smt 2313's i made a little board to program them and attached a crystal to the bottom for that very reason
[19:32:18] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/attiny2313/attiny8.jpg
[19:32:20] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/attiny2313/attiny9.jpg
[19:32:29] <Hackbat> I just made a little break out board for my usbtiny
[19:32:37] <Hackbat> most fancy part is the power led
[19:33:04] <Tom_itx> that was back when i made up a bunch of these:
http://tom-itx.dyndns.org:81/~webpage/attiny2313/mini_usb_batch1.jpg
[19:34:43] <Hackbat> oh god I just nearly freaked
[19:34:59] <jadew> lol
[19:35:04] <Hackbat> my can of pepsi was fizzing
[19:35:16] <Tom_itx> thought it was the chip?
[19:35:31] <Hackbat> thought it was my programmer
[19:36:25] <jadew> I get that feeling when it starts smelling like something's burning
[19:38:08] <jadew> I wonder if my power source is measuring this properly, but this oscillator seems to be eating less than 100uA
[19:38:29] <jadew> because it doesn't detect any power consumption
[19:47:39] <jadew> this is weird, if I connect my ammeter in series with the circuit, it stops oscillating
[19:58:04] <Tom_itx> load
[20:14:47] <jadew> I needed a decoupling cap, checked the voltage between vcc and gnd and it was oscillating with the crystal :P
[20:50:36] <DanFrederiksen> any of you considered hacking a better AI into one of those aimless robotic vacuum cleaners?
[21:06:23] <rue_bed> yea
[21:19:17] <jadew> put an avr in it and it'll cut the grass as well
[21:22:58] <rue_bed> I keep thinking of modding the one I have, but it always leads to me thinking to make a new one from scratch
[21:23:16] <jadew> heh
[21:23:29] <jadew> aren't they slow?
[21:31:44] <jadew> are you guys aware of any project that does usb slave/master for i2c and spi?
[21:37:53] <LoRez> you can use a buspirate if it's temporaryish
[21:38:14] <jadew> what's a buspirate?
[21:38:44] <LoRez> an awesome tool
[21:39:11] <jadew> ah neat
[21:39:16] <jadew> seems to do exactly what I want
[21:39:44] <jadew> I was in the process of building something similar, but only with i2c and spi support and was feeling kinda lazy to write support software for it
[21:40:41] <Roklobsta> tom_itx: where did you say you got your laser printer transfer papepr from?
[21:40:52] <Roklobsta> i tried baking paper yesterday and wrote off a toner cartridge
[21:41:16] <jadew> Roklobsta, for tonner or photo transfer?
[21:42:40] <Roklobsta> to print to the paper then melt onto a PCB
[21:43:03] <Tom_itx> a dude in Au sent it to me
[21:43:21] <jadew> regular photo paper should do it
[21:43:30] <Tom_itx> yeah
[21:43:33] <Tom_itx> non glossy
[21:43:42] <Tom_itx> some has a plastic coating that screws it all up
[21:43:46] <jadew> actually that's the thing
[21:43:54] <jadew> I had great success with that type
[21:44:04] <Roklobsta> ok
[21:44:13] <jadew> and it peals off really nice
[21:44:33] <Roklobsta> at least laser printer cartidges have dropped by over 50% since I bought my last one
[21:44:45] <Roklobsta> non glossy pto paper ok
[21:45:10] <Tom_itx> some i used was inkjet too
[21:45:19] <jadew> that's what I'm using
[21:45:20] <Tom_itx> i've tried quite a few different ones
[21:46:19] <jadew> but I kinda hate the process
[21:46:59] <jadew> especially the fact that you can't be too precise when you put it on the pcb
[21:47:29] <Tom_itx> i use pins to locate it on 2 sided
[21:47:50] <jadew> and the PCB has to be hot, so it sticks as soon as it touches the board
[21:47:54] <Tom_itx> i've gotten lazy lately though and usually have em made now
[21:48:02] <jadew> placing the paper first and then ironing never worked for me
[21:48:53] <jadew> (I think that's the biggest secret)
[21:55:15] <jadew> damn, that buspirate thing seems awesome
[22:03:54] <DanFrederiksen> there are copy products of the roomba now for around 100$, not too bad to experiment with
[22:04:38] <DanFrederiksen> doesn't have a brand name. I hate that. goes by the name 'auto cleaner' or m-288
