#avr | Logs for 2012-03-11

Back
[00:00:19] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/atmega32u4/atmega32u4_1.jpg
[00:00:25] <Tom_itx> that was hand soldered
[00:00:35] <Tom_itx> i reflowed a few after that
[00:01:03] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/misc_stuff/5v_switcher.jpg
[00:01:17] <Tom_itx> the regs were stuck with an iron but it would be easier to reflow those too
[00:11:32] <Roklobsta> dpak has been easy with an iron that can pump the heat when you need it
[00:12:16] <Roklobsta> i tend to put a small blob on the big tag pad, preheat the blob and pad and sort of slide it on
[00:18:03] <Roklobsta> capn: what's the beer situation in China?
[00:18:30] <CapnKernel> Roklobsta: We're racking tonight
[00:18:53] <Roklobsta> and good local beers?
[00:19:10] <CapnKernel> Running the fermented beer from one barrel to the next. Then it'll sit there for another week, then we bottle it. Doing this means you end up with clearer beer, and less sediment at the bottom
[00:19:23] <CapnKernel> No, it's all horse piss which makes American beers look good.
[00:20:43] <Bird|lappy> Roklobsta, well, my tag pad is a good 6-7in2 on the top layer with 25 vias under the pad proper connecting it to the ground plane (no thermal relief on those vias, either)
[00:20:59] <Roklobsta> use a plumbers iron
[00:21:13] <Roklobsta> break out the blowtorch
[00:21:30] <Bird|lappy> don't think I'll have to push it /that/ far lol
[00:21:33] <Roklobsta> capn: that's desperate - making your own beer in china
[00:21:43] <Bird|lappy> whats worse is that it's a MSL3 part
[00:22:00] <Roklobsta> did you make from scratch or buy some cans of beer goo from coles before you left?
[00:22:30] <Roklobsta> what's the bread and dairy like?
[00:24:08] <CapnKernel> Roklobsta: It's concentrate I brought with me from Australia.
[00:24:25] <CapnKernel> The local hackers here love the idea of making beer. They are a very smart, very good bunch of guys.
[00:25:46] <nofxx> beer is waste of time, I prefer a good and old scotch
[00:25:56] <nofxx> in the lack of the former, or money, cachaça
[00:52:00] <Roklobsta> beer is scotch, just younger and more bubbly
[00:52:59] <Roklobsta> i am making solid beer at the moment - bread. I think I shall make a temperature and humidity contorlled proofing box with an AVR
[00:53:41] <Roklobsta> should be easy to place a couple of tmp75's in the box and have the AVR PWM some nicrome wire to maintain temp.
[00:58:00] <CapnKernel> Roklobsta: nice
[00:59:00] <CapnKernel> Roklobsta: The bread is shit. Plain bread is hard to come by. Most bready things have an awful glaze on them, and they can't decide whether to be sweet/savoury or both.
[00:59:22] <CapnKernel> Dairy is almost unheard of. I dream of cheese: http://capn.blogspot.com/2012/01/cravings.html
[01:01:54] <ureif> I think the Chinese will be offended by your body odor if you eat dairy products. (heard some such story once)
[01:03:57] <learningc> hi Roklobsta :)
[01:07:40] <learningc> CapnKernel: are you here?
[01:12:34] <kubuntu> Hello! What is the third fuse other than lfuse, hfuse ... ?
[01:14:14] <CapnKernel> welf0Va
[01:14:24] <CapnKernel> Oops time to change my password :-)
[01:14:47] <CapnKernel> learningc: Yes I'm here.
[01:15:12] <CapnKernel> kubuntu: efuse
[01:15:34] <CapnKernel> ureif: Big country, so I can't speak for all Chinese, but I don't think so.
[01:15:52] <CapnKernel> Hygiene is definitely very different here. I call China "The land that deodorant forgot"
[01:16:22] <ureif> CapnKernel, that's a different issue. When you're in a unfamiliar country the smells will obviously be strange.
[01:16:56] <CapnKernel> ureif: I've lived in China for a year.
[01:17:25] <ureif> ok.
[01:23:50] <CapnKernel> So the smells I smell are all familiar, but no less repulsive than the day I got here.
[01:24:43] <CapnKernel> This one takes the cake, because people actually pay money for it: http://en.wikipedia.org/wiki/Stinky_tofu
[01:25:42] <CapnKernel> I quote: " the odor of stinky tofu is said to resemble that of rotten garbage or manure, even by its enthusiasts."
[01:26:42] <CapnKernel> When I first went to Hong Kong in 1990, the streets in one district were so foul that I formed the theory that their sewerage system wasn't up to snuff. Well I was wrong, it was really just stinky bean curd.
[01:30:06] <OndraSter_> eya
[01:30:21] <CapnKernel> Indeed.
[01:38:48] <Alexer> hi, I heard there might be people here with a non-free version of Eagle, would anyone be willing to generate a few simple files for me?
[01:38:56] <Alexer> I'm writing a parser for the binary file formats, but I can't test some features myself since they're not available in the free version
[01:39:28] <Alexer> mainly, that means different sheets in schematics, I guess
[01:39:54] <ziph> Doesn't the new version default to XML?
[01:40:55] <Alexer> so I've heard (but only after I got the binary parser almost ready) :P
[01:41:32] <ziph> Or you could use DipTrace that has an export feature to a very nice text format.
[01:41:38] <Alexer> but yeah, I'd still like to complete the binary parser
[01:42:18] <ziph> The lexer for the DipTrace text format takes about 20 minutes to write and then you have a big tree of tuples to wander around.
[01:44:04] <Alexer> I've been thinkin about moving to KiCad (after it changes it's internal units to metric)
[01:44:59] <Alexer> but whatever software I'll choose, I want to finish that Eagle binary parser, just because :P
[01:50:33] <CapnKernel> Please, someone help him! :-)
[01:53:14] <ziph> Alexer: http://www.bitplantation.com/transfer/EagleTwoSheet.sch
[01:53:45] <ziph> Alexer: It's two sheets with a text area on each sheet.
[01:56:41] <Roklobsta> capn: being in W'bool it's dairy country here. Have you been to Cheeseworld?
[01:57:00] <Roklobsta> hallo leanrinc
[01:57:43] <Alexer> ziph, thanks :)
[01:58:36] <CapnKernel> No
[01:59:18] <CapnKernel> Roklobsta: Good country down that way.
[03:10:00] <Roklobsta> Chinese have stinky tofu, we have godawful Stilton cheese.
[03:10:19] <Roklobsta> http://en.wikipedia.org/wiki/Stilton_cheese
[03:30:27] * CapnKernel likes his Danish blue
[03:33:14] <Alexer> ziph, mind doing another one: an empty board with one internal layer added, first saved normally and then with the "Supply layer" checkbox checked?
[03:37:26] <Alexer> that should be the last one, since the only limitations in the free version should be the sheet count and layer count (and you can't mark top/bottom as a supply layer)
[03:39:50] <ziph> Alexer: http://www.bitplantation.com/transfer/eagle/
[03:41:29] <Alexer> ziph, awesome, thanks :)
[04:00:23] <nevdull> #falcon
[05:04:13] <mrfrenzy> CapnKernel: I'm intersted in primary CR123A and CR2, 100 quantities or more
[05:20:00] <CapnKernel> mrfrenzy: Panasonic?
[05:20:36] <mrfrenzy> any reputable brand will do
[05:20:56] <mrfrenzy> I also have good experience from Tenergy, but a "commonly known brand" is even better
[05:23:57] <brentbxr^t> Oh, i am in this channel :D
[05:24:15] <brentbxr^t> im on trillian so I dont really notice half the time CapnKernel
[05:25:02] <brentbxr^t> the '^t' is what I setup for that. In some setting im sure.
[05:26:28] <OndraSter_> https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!450&parid=40BF7833586103AB!111&authkey=!ANupAOyNB36KQ-E
[05:26:37] <CapnKernel> mrfrenzy: I can get you batteries that say Panasonic, w/ Japanese writing and stuff, but I can't vouch for them.
[05:27:34] <CapnKernel> OndraSter: Damn those unknown connector rotating fiends!
[05:28:05] <OndraSter_> yeah
[05:28:12] <OndraSter_> it is like.. each two pins are rotated
[05:28:16] <OndraSter_> not the whole connector
[05:28:18] <OndraSter_> :o)
[05:28:29] <CapnKernel> Those wily rascals...
[05:28:38] <OndraSter_> yeah
[05:28:45] <brentbxr^t> :3
[05:28:49] <OndraSter_> but in board layout it is correct
[05:28:54] <mrfrenzy> CapnKernel: I can do my own capacity test if I can be sure I get the same battery next time
[05:28:57] <OndraSter_> maybe somebody rotated my wires
[05:32:05] <CapnKernel> mrfrenzy: I could but I don't think I should.
[05:36:02] <OndraSter_> https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!451&parid=40BF7833586103AB!111&authkey=!AK5xR4DMvjTdxVs
[05:55:14] <Roklobsta> oh... ham and pineapple, and then bacon, salami, mushroom, shave onion and bits of english spinach leaves
[05:55:24] <Roklobsta> fsck
[05:55:30] <Roklobsta> wrong thingy
[05:55:45] * Roklobsta blames Carlton Draught again.
[05:56:48] <CapnKernel> And what was under all of that delectable stuff?
[05:57:25] <Roklobsta> well, cheese on top and a home made pizza base
[05:57:40] <Roklobsta> chu-heeze
[05:58:10] <CapnKernel> Screw you, you're doing it again :-(
[05:59:18] <CapnKernel> (Just kidding)
[05:59:34] <CapnKernel> But the cheese-free life I lead is so HARD!! <me sobs>
[06:04:20] <izua__> woo
[06:04:33] <izua__> .net can't open serial ports that will not start with "COM"
[06:04:40] <izua__> </random thought of the day>
[06:04:44] <CapnKernel> You have the source, fix it.
[06:08:19] <Roklobsta> what serial ports don't start with COM?
[06:10:11] <nevdull> /dev/ttyS0
[06:12:24] <CapnKernel> symlink?
[06:13:47] <Roklobsta> LOLdongs
[06:13:55] <Roklobsta> are you using m0n0?
[06:16:01] <Roklobsta> miguel de icaza must have missed that one
[06:28:26] <wollw> well that bit about mono and serial ports got my interest and i just tried. I'm reading serial data with mono on linux right now
[06:47:00] <izua__> well, for example, serial ports from com0com
[06:47:04] <izua__> (:although you can rename them)
[06:47:14] <izua__> or ports created by custom drivers
[06:47:19] <izua__> (which can't be renamed)
[06:50:25] <wollw> ah
[09:32:41] <Fluffi> hi
[09:32:55] <Fluffi> i need help with TWI and a display
[09:34:41] <Fluffi> i have that oled http://www.seeedstudio.com/wiki/Grove_-_OLED_Display_128*64
[09:34:55] <Fluffi> and i want to adress itwith an atmega32
[09:35:18] <Fluffi> but the controller of the oled wont give me any answers
[09:35:54] <Fluffi> and thats the point where you come in...
[09:38:10] <specing> Have you tried the demo code?
[09:38:31] <Fluffi> yes
[09:39:06] <Fluffi> but i dont use an ardunio system.
[09:39:47] <specing> Oh, and arduinos don't have AVRs inside or what?
[09:40:01] <Fluffi> yes but the code is different
[09:40:09] <Fluffi> I used the Peter Fleury TWI code
[09:40:17] <specing> Hmmm
[09:40:24] <Fluffi> the TWI code for ardunio is in cpp
[09:40:31] <specing> Does it work?
[09:40:34] <Fluffi> no
[09:41:08] <specing> meybe the unit is dead then...
[09:41:58] <Fluffi> oh no
[09:42:07] <Kevin`> maybe you don't have the pins configured correctly
[09:42:16] <Kevin`> maybe you are just using the library wrong
[09:42:19] <Kevin`> where's your scope
[09:42:32] <Fluffi> i used the TWI pins of the avr32. there is nothing to configure
[09:43:01] <Fluffi> i have no scope
[09:43:55] <Tom_itx> some OLEDs require you to power them up a certain way or you will destroy them
[09:44:02] <Kevin`> you said atmega32 before, not avr32
[09:44:13] <Fluffi> i mean atmega32 sorry
[09:44:18] <Kevin`> some modules require the pins to be configured as output as necessary
[09:44:24] <Kevin`> also, you need a pullup for i2c
[09:44:41] <Fluffi> i use the internal pullup
[09:45:15] <Kevin`> you just love your assumptions :)
[09:45:49] <Fluffi> no I dont
[09:46:56] <Fluffi> what is wrong there? what clock should I use?
[09:47:09] <jadew> the lowest for starters
[09:47:31] <jadew> oh, and try to change the mode
[09:47:32] <Fluffi> and that is?
[09:47:37] <Fluffi> what mode
[09:47:43] <jadew> nvm
[09:47:49] <jadew> that's SPI
[09:48:06] <Kevin`> and that sort of thing you are SUPPOSED to read right from the datasheet
[09:48:22] <Kevin`> compare timing graphs, select the correct mode, and enter the wrong registers for it ;p
[09:49:39] <Fluffi> but I used the same values like in the ardunio code. so parameters should be ok
[09:51:55] <jadew> hook up another avr to the i2c bus to figure out if the lines are switching
[09:52:10] <Tom_itx> or a logic analizer
[09:52:19] <Tom_itx> and read the data sheet
[09:52:58] <Tom_itx> things become alot clearer when you do that
[09:53:29] <jadew> yeah, you should write your own lib too, or try to understand the one you're using
[09:55:06] <jadew> oh, and use external pullups
[09:55:16] <jadew> I had issues with internal ones
[09:55:44] <Tom_itx> probably a good idea too. the internal pullups are pretty weak and may cause you problems
[09:58:34] <jadew> try different values as well, from what I could tell, as you add more devices to the bus you need to change the value of the pullups
[09:58:52] <jadew> because they control how fast your signal rises
[10:01:00] <jadew> and on the scope it looked like charging ca capacitor, I suppose all decives have some capacitance on their input to sample it
[10:01:17] <jadew> so higher resistance will mean those capacitors will get charged slower
[10:01:32] <jadew> (I'm just assuming here, don't know it for a fact)
[10:01:49] <Kevin`> I don't think the capacitance is intentional
[10:01:51] <Kevin`> it's just a property of things
[10:02:08] <jadew> but that's the only explanation I found for the rounded wave when the pullup resistor gets higher
[10:02:52] <Roamin> anyone has the original bootloader for atmega32u2 in hex form? I erased it on my dev board, thinking i could easily find it again and re-program it
[10:03:30] <Kevin`> isn't it on atmel's site? i've seen it on atmel's site
[10:04:10] <Roamin> i'll link the flip page, but the 3 bootloaders arent proper for atmega32u2
[10:04:18] <Roamin> http://www.atmel.com/tools/FLIP.aspx
[10:05:18] <Roamin> and on the atmega32u2 page is this : AVR231: AES Bootloader on tinyAVR and megaAVR devices but needs to be compiled and its way out of my league :(
[10:05:53] <Kevin`> Roamin: i'd probably just stick one of the lufa bootloaders on it, myself
[10:06:09] <Kevin`> also a good guy to ask where tf the original one is
[10:06:16] <Kevin`> when he's in
[10:06:29] <Roamin> will lufa bootloaders allow flip to be used? or will i need a different software?
[10:06:50] <Kevin`> depends which one you select. I think there's one compatible with flip
[10:07:16] <Kevin`> I personally don't use windows so i'm using different software even for the stock usb bootloader
[10:07:59] <jadew> why do you guys program them trough a bootloader?
[10:08:03] <jadew> they don't support ISP?
[10:08:14] <Roamin> faster than isp, easier too
[10:08:25] <jadew> ah
[10:08:52] <Roamin> i dont mind either, but i can easily just press reset, reflash and reboot without ever disconnecting anything
[10:09:10] <jadew> same with ISP
[10:09:29] <jadew> you just have to connect the programmer's header :P
[10:09:55] <Kevin`> jadew: the device is already connected via usb for normal communication though
[10:09:56] <Roamin> with usb, its powered, and just press reset to reprogram, and flip writes much faster than my gq-4x
[10:10:07] <Kevin`> and indeed the bootloader is several times faster than isp
[10:10:12] <jadew> ah, I see
[10:10:47] <Kevin`> if you were using hvpp/hvsp you could probably run it faster, but, no ;)
[10:20:51] <Sedric> can you start the watchdog timer using int0 on a tiny13?
[10:22:41] <jadew> should work, from an ISR
[10:24:24] <Sedric> I set it to system interrupt before my main loop, and the toggle between on/off when i use the hardware interrupt
[10:24:40] <Sedric> then*
[10:25:12] <Sedric> i can toggle it off, just not back on again
[10:25:46] <jadew> try debugging it
[10:26:15] <Sedric> i've only got an ISP MKII unfortunately
[10:27:31] <jadew> light a led when you turn it on
[10:27:41] <jadew> turn the led off when you turn watchdog off
[10:27:56] <jadew> that way you'll know the code gets ran
[10:36:19] <Kevin`> using the debug module disables the watchdog timer btw
[10:38:47] <OndraSter_> yeah
[10:38:52] <OndraSter_> otherwise it would reset quite often :P
[10:40:09] <Sedric> lol, i've done a hacky fix for now. Rather than turrning the WDT on/off I've toggeled the pin to input/output. It looks like it's turned off.
[10:55:06] <RikusW> So I recovered my supposedly dead m328p :)
[10:55:22] <RikusW> DWEN was on and external crystal selected....
[10:55:48] <RikusW> and my HVPP connection wasn't good....
[11:01:15] <RikusW> superglued the ic sockets to the protoboard, some got into the contacts :(
[11:08:46] <OndraSter_> I am getting frustrated
[11:08:53] <OndraSter_> it just won't lit up the pixels as it should :(
[11:09:11] <OndraSter_> I spent already bazilion hours on it
[11:16:50] <Tom_itx> RikusW, not a good idea
[11:17:02] <Tom_itx> OndraSter_, video?
[11:24:15] <RikusW> Tom_itx: the superglue ?
[11:33:01] <OndraSter_> Tom_itx, it should be drawing only 1pixel wide lines from top to bottom :(
[11:33:10] <OndraSter_> right now
[11:33:24] <cyanide> hello folks
[11:33:24] <OndraSter_> hello
[11:33:29] <cyanide> whats up OndraSter_
[11:33:34] <OndraSter_> I am frustrated
[11:33:38] <OndraSter_> IT JUST WOULDN'T WORK!
[11:34:04] <cyanide> :(
[11:35:20] <cyanide> one thing i like about programming is that you can keep trying until it works :)
[11:36:03] <cyanide> i dont have that liberty with engines
[11:36:09] <dirty_d> crap, i cant figure this out
[11:36:15] <dirty_d> i need to make a resistive signal mixer
[11:36:23] <dirty_d> otuput = 1/2v1 + 1/6v2
[11:36:39] <RikusW> unless you use uveprom.. then erasing takes some time :-P
[11:36:51] <cyanide> if it doesnt work, you're usually neck deep in shit and a load of money on broken parts
[11:36:53] <cyanide> lol
[11:37:15] <RikusW> or worse OTP prom
[11:37:17] <Tom_itx> OndraSter_, what is it doing?
[11:37:23] <OndraSter_> nothing
[11:37:27] <OndraSter_> that's the problem :D
[11:37:33] <OndraSter_> if it was showing something, I'd have something to work on
[11:37:43] <Tom_itx> so start probing
[11:37:47] <Tom_itx> with a meter etc
[11:37:53] <OndraSter_> well it all works fine
[11:37:57] <OndraSter_> parts on its own work
[11:38:00] <OndraSter_> but together it does not :(
[11:38:12] <Tom_itx> too much load?
[11:38:25] <Tom_itx> for the chip to support
[11:38:29] <Tom_itx> or the supply
[11:38:40] <Tom_itx> or the traces :)
[11:39:14] <OndraSter_> nope
[11:39:27] <OndraSter_> the high current goes only through LEDs from its own power supply
[11:39:29] <Tom_itx> i bet you find something
[11:40:18] <Tom_itx> are the modules wired in series or parallel
[11:41:15] <OndraSter_> I am trying to setup only one right now
[11:41:25] <OndraSter_> I have tested shifting data through all of them together, it works okay
[11:41:30] <OndraSter_> but I have soldered power supply to only one module
[11:43:02] <Tom_itx> all the chip enables being enabled as they should when in series?
[11:43:14] <Tom_itx> clock lines etc
[11:43:24] <OndraSter_> it all works!
[11:43:50] <OndraSter_> in theory.
[11:44:30] <OndraSter_> I don't have any working logic analyzer ffs
[11:44:34] <Tom_itx> theory doesn't always agree with the real world
[11:44:38] <OndraSter_> I know
[11:44:46] <OndraSter_> I've already found that out
[11:47:12] <OndraSter_> hmm and I am scared to run it with JTAG attached, because with JTAG attached watchdog is disabled
[11:47:19] <OndraSter_> and if it gets stuck => LEDs burn!
[11:56:49] <Tom_itx> you can't current limit them?
[11:57:43] <OndraSter_> doode
[11:57:48] <OndraSter_> you wont believe it
[11:57:57] <OndraSter_> it was borked crocodile clip I used here
[11:58:02] <OndraSter_> to connect to power supply
[11:58:03] <OndraSter_> lol
[11:58:41] <Tom_itx> 90% of the time it is something simple
[12:02:08] <learningc> Tom_itx: when I put and smd in bottom layer, I do mirror, right?
[12:02:15] <learningc> an*
[12:02:38] <Tom_itx> yes
[12:04:08] <learningc> Tom_itx: and the tnames becomes bnames, I suppose. But the problem is I bnames is now reading from right to left
[12:04:26] <learningc> is bnames*
[12:04:45] <learningc> is that normal?
[12:06:13] <Tom_itx> i think so yes
[12:06:46] <OndraSter_> 9v 0.5a in
[12:06:53] <OndraSter_> 3v3 ..A out
[12:06:59] <OndraSter_> per 24 LEDs :)
[12:07:44] <learningc> Tom_itx: but when I generate the gerber, it remains right to left even on the bottom solk screen
[12:08:08] <OndraSter_> what program?
[12:08:10] <Tom_itx> as you look 'thru' the board it will appear backwards
[12:08:14] <learningc> eaglecad
[12:08:16] <OndraSter_> damn, that coil is making high pitch sound
[12:08:23] <Tom_itx> if you flip the board over it will be right
[12:08:24] <OndraSter_> even though it is 650kHz supposedly
[12:08:32] <Tom_itx> lemme check one of my gerbers for sure
[12:08:37] <learningc> ok
[12:09:05] <Tom_itx> see one layer will be mirrored
[12:09:45] <OndraSter_> https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!452&parid=40BF7833586103AB!111&authkey=!AJk8gMkT6T0I7ak
[12:09:50] <OndraSter_> poor phone is having problems with brightness :P
[12:11:29] <learningc> Tom_itx: even the silk screen texts are reversed?
[12:12:03] <OndraSter_> there shouldn't be any silkscreen on top in gerber if you moved the part to the bottom
[12:12:09] <OndraSter_> it should be in bottom silkscreen ?!
[12:12:15] <OndraSter_> the same for copper layers
[12:13:11] <Tom_itx> the text appears backwards on my bottom gerber
[12:13:20] <OndraSter_> it should, yeah
[12:13:24] <OndraSter_> let me check my gerbers
[12:13:33] <Tom_itx> i just checked mine for him
[12:13:43] <Tom_itx> these are boards i've produced
[12:16:00] <Tom_itx> OndraSter_, so it's all working now?
[12:16:05] <OndraSter_> yeah :)
[12:16:09] <OndraSter_> dumb problems :(
[12:16:15] <OndraSter_> I knew there was one crocodile clip that was borked
[12:16:19] <OndraSter_> but I thoguht it was the black one..
[12:16:24] <OndraSter_> well, it wasn't lol
[12:16:28] <Tom_itx> so do you own a soldering iron?
[12:16:28] <OndraSter_> anyway, thanks
[12:16:36] <OndraSter_> duh... sure?
[12:16:39] <Tom_itx> fix it
[12:16:55] <OndraSter_> I will put it into "gotta fix it someday" drawer
[12:16:59] <Tom_itx> naw
[12:17:01] <Tom_itx> fix it now
[12:17:02] <OndraSter_> ... if I manage to close it
[12:18:24] <Tom_itx> OndraSter_, so what is this for?
[12:18:47] <OndraSter_> my school project
[12:18:53] <OndraSter_> with which I will end middle (high, for US) school
[12:19:12] <Tom_itx> final project
[12:19:45] <OndraSter_> yeah
[12:20:25] <Tom_itx> now you need to make a scope front end for it
[12:20:30] <Tom_itx> display the scope on it
[12:20:33] <OndraSter_> heh
[12:20:39] <OndraSter_> I mostly need to solder the remaining 3 segments
[12:20:41] <OndraSter_> this is 1/4th of it
[12:20:48] <OndraSter_> it is already glued together...
[12:20:54] <OndraSter_> I should've first soldered and then glued... oh well
[12:21:20] <OndraSter_> it is this big
[12:21:20] <OndraSter_> https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!451&parid=40BF7833586103AB!111&authkey=!AK5xR4DMvjTdxVs
[12:21:33] <OndraSter_> (hardwired one line on each segment)
[12:21:39] <OndraSter_> segment = 3x2 matrices
[12:21:45] <OndraSter_> or w/e to call it
[12:22:22] <OndraSter_> btw, doing 16bit math on 8bit atmega in assembler is not that much fun I've learned :D
[12:23:40] <OndraSter_> btw, no bodge wire so far in my project
[12:23:44] <OndraSter_> I call that success
[12:24:08] <OndraSter_> oh, and FYI, I am running those MOSFETs only at -3V3 on -Vgs
[12:24:13] <OndraSter_> they are rated 4.5V
[12:24:17] <OndraSter_> to be "fully opened"
[12:24:26] <OndraSter_> but I didn't get any drop even at 3.3V
[12:24:30] <OndraSter_> SP8J3... P-channel
[12:24:47] <OndraSter_> they are fire saled on farnell right now, about 1k? 1k5? pieces left
[12:24:52] <OndraSter_> at 25c/piece
[12:25:02] <OndraSter_> and one piece has 2 MOSFETs inside
[12:25:06] <Tom_itx> what about ebay?
[12:25:25] <OndraSter_> I have 32 in my drawer as a safety for future projects
[12:25:32] <OndraSter_> they are great
[12:25:47] <OndraSter_> something around 3 - 5A constant I think... 14A pulsed load even
[12:25:59] <OndraSter_> SO-8 package
[12:26:18] <OndraSter_> anyway, if you are looking for TTL-usable P mosfets, grab those
[12:26:21] <OndraSter_> I recommend them :P
[12:26:44] <OndraSter_> eya CapnKernel
[12:26:51] <OndraSter_> CapnKernel, https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!452&parid=40BF7833586103AB!111&authkey=!AJk8gMkT6T0I7ak
[12:26:54] <OndraSter_> I can't not link it :P
[12:27:10] <Tom_itx> digikey or mouser doesn't stock them
[12:27:28] <OndraSter_> 1774 on farnell still left
[12:27:37] <OndraSter_> so like I said.. if you need some, grab them, till they last
[12:27:45] <CapnKernel> OndraSter_: You're really cooking now!
[12:27:49] <OndraSter_> yeah
[12:27:51] <OndraSter_> literally lol
[12:27:55] <Tom_itx> SP8J3
[12:27:57] <CapnKernel> Oh?
[12:27:58] <Tom_itx> CapnKernel?
[12:28:08] <OndraSter_> I bet they could get hot after a while
[12:28:16] <Tom_itx> put fans behind em
[12:29:33] <OndraSter_> more like behind those LED drivers
[12:29:37] <OndraSter_> some small 40mm fan(s) :P
[12:33:59] <OndraSter_> duh I am feeling like spammer
[12:34:25] <Tom_itx> you are but we'll forgive you
[12:34:47] <OndraSter_> thaaaaaanks
[14:22:49] <skorket> I'm confused about what speed my ATMega328 chip is running at. I checked the low order fuse bits and it claims it's 62 which, from looking at the datasheet, is the default internal 8Mz clock with CKDIV8 set. So I believe it should be running at 1Mhz. But the following code makes the hooked up LED from PB0 blink at a rate of about 49KHz: http://pastebin.com/SPUZJ5u7 . Does anyone have a clue what's going on? What am I missing?
[14:23:57] <Tom_itx> did you define F_CPU?
[14:26:03] <skorket> Tom_itx, when compiling with avr-gcc I specify 'atmega328' and when programming with avrdude, I specify 'm328'
[14:26:21] <Tom_itx> did you define F_CPU?
[14:26:28] <skorket> Tom_itx, I'm not sure it matters, but I copied the config from avrdude's entry for the ATMega328P into the ATMega328.
[14:26:33] <skorket> Tom_itx, if you mean in the code, no
[14:26:38] <Tom_itx> try that
[14:27:15] <Tom_itx> http://www.engbedded.com/fusecalc/
[14:27:22] <Tom_itx> then use that as a fuse guide
[14:29:44] <skorket> Tom_itx, F_CPU should just be a define in my code '#define F_CPU 1000000'?
[14:30:24] <Tom_itx> yes
[14:32:15] <skorket> Tom_itx, just a quick look around but it looks like F_CPU is used for 'delay.h' functions, which I don't use. At any rate, I've added the #define and no change. New code here: http://pastebin.com/YtFhaDp4
[14:33:00] <Tom_itx> wrong caps on the crystal?
[14:33:15] <OndraSter_> internal oscillator
[14:33:16] <skorket> Tom_itx, no crystal (yet).
[14:33:23] <Tom_itx> mmm
[14:33:47] <Tom_itx> wdt isn't resetting it is it?
[14:36:32] <skorket> Tom_itx, high fuse byte is 'd9' and bit 4 is the 'watchdog timer always on' bit, which is in it's default '1 (unprogrammed)' state. Brown out detector should also be disabled. It's fresh from Mouser, I'm just trying to run through a few 'Hello World' programs to familiarize myself with the chip
[14:42:52] <skorket> ahg! I'm sorry, I mispoke about the above. It looks like my _chip_ is running at about 49KHz. The LED attached to PB0 is blinking about once every two seconds (on for 1sec, off for 1sec).
[14:43:17] <OndraSter_> that's weird
[14:45:17] <OndraSter_> I should've set up 3 screens when I was doing the boards, not when I am programming...
[15:15:14] <learningc> how should I drive a dc motor from the microcontroller output?
[15:15:38] <learningc> 12V dc motor
[15:15:58] <learningc> any recommendations?
[15:17:08] <specing> by using a motor driver, silly
[15:26:38] <jimlay> Hey, got a problem I've been hacking around for months and I can't figure out where to get the answer in the docs.
[15:27:35] <jimlay> How long do pins have to be set to high or low before it's safe to assume they actually have that value?
[15:28:22] <jimlay> When I set an output pins state from high to low and call delay, it doesn't always actuall toggle.
[15:29:12] <jimlay> The problem never happens if I send debug info over the serial, but calling delay, even for long periods of time seem sto act like it doesn't try to finish setting the pin state until after it finishes the delay.
[15:29:25] <jimlay> There isn't a concept of flushPinState() is there?
[15:31:12] <skorket> jimlay, my bet is what you're seeing has more to do with what the pin is connected to than anything else
[15:32:09] <jimlay> That's my suspicion as well, but I've put it on a scope to try and verify this and not had any luck.
[15:33:44] <skorket> jimlay, what's connected to it?
[15:33:50] <learningc> specing: any recommended motor driver?
[15:33:56] <jimlay> I've also tried using very long delays, delays longer than what the serial port write is causing.
[15:34:13] <jimlay> Directly to the inputside of a dual H bridge.
[15:34:32] <jimlay> How long should the io output take to go high?
[15:36:13] <jimlay> I think what's really confusing me is that I can't find the bandwidth for digital IO.
[15:36:18] <jimlay> I thought it was 50khz.
[15:37:02] <skorket> jimlay, what is this H-bridge driving?
[15:37:16] <RikusW> there should be rise/fall time specs
[15:37:22] <jimlay> I feel like I must be using the wrong terminology, what's the maximum rate that the digital IO's can be toggled?
[15:38:18] <jimlay> The H-bridgen are connected direct to steppers. (Rechecking all my specs here quickly before I put my foot in my mouth, I've checked and double checked all this but that doesn't meen I didn't make a mistake)
[15:38:54] <RikusW> the avr io pins cannot supply a lot of current, so you might need buffering
[15:39:17] <RikusW> it can be toggled at half the avr clock
[15:39:47] <jimlay> No shit. Thank you. That "it can be toggled at have the avr clock" is what I've been looking for.
[15:40:02] <jimlay> What's the name of that specification?
[15:42:12] <jimlay> I don't know where I got the idea that it was 50khz and couldn't go over that. I thought I had to leave a 1 on the pin for at least 1/50ks and use a delay to ensure that it actually settled.... don't know why I made that up.
[15:42:22] <specing> learningc: Amps?
[15:43:56] <learningc> specing: less than 1amp
[15:44:21] <learningc> @ 12V
[15:44:35] <jimlay> I was worried for a second there, thought I just read 100mA input current to my H-bridge. That's for the whole thing. I'm driving CMOS gates that have a max rating of 100uA's when being driven high.
[15:44:40] <learningc> perhaps more like 6W
[15:44:53] <learningc> so 5mA...
[15:44:59] <jimlay> learningc: What're you powering with the motors?
[15:45:10] <learningc> jimlay: small pump
[15:45:17] <learningc> for ink
[15:45:36] <jimlay> 5mA? Less than an LED to pump ink? sure that's correct?
[15:45:49] <learningc> sorry, 500mA
[15:46:04] <jimlay> so a small DC motor? uni-directional?
[15:46:12] <learningc> or .5A is what I intended to write
[15:46:25] <jimlay> you need to control speed or just on off?
[15:46:58] <specing> learningc: a mosfet then, or three ULN2003 channels together
[15:46:58] <learningc> yes, probably unidirectional (not decided of the final specification yet)
[15:47:52] <learningc> 3 uln2003 channels together? in parallel you mean?
[15:48:03] <jimlay> specing: Why 3? I just did this last week with 1 darlington.
[15:48:58] <jimlay> ... what was it... 12V, 2A.. we hooked an accelerometer up to a attiny and used a single Darlington to toggle the taillight according to deceleration.
[15:49:07] <specing> jimlay: one is rated for 500mA and learningc said 1 amp
[15:49:21] <jimlay> (he keeps changing how many amps :)
[15:49:37] <specing> Well then
[15:49:48] <specing> learningc: just stick a big ass mosfet right there and be done with it
[15:50:04] <learningc> won't it have trouble in parallel? (because the pn junction drops is not the same for all 3?)
[15:50:36] * jimlay was questioning the same thing... doesn't mean I haven't done it though. :)
[15:51:36] <jimlay> specing: We spent a couple hours looking at Mosfets (it was a friends project and I tossed out, I think I have a mosfet that'll be perfect, otherwise I have some darlingtons) but after I started digging I couldn't actually find a mosfet that lookekd good.
[15:51:41] <specing> learningc: MOSFET MOSFET MOSFET MOSTE MOFSTERM MONSTER MOSFET
[15:51:47] <learningc> ok, if I use a power mosfet? so any resistors here and there? but where?
[15:51:50] <jimlay> Do you have a "go to Mosfet" or family of them?
[15:52:15] <specing> I use a FQU11P06
[15:52:33] <jimlay> for loads in the 500mA-10A? .... actually... I'm running a 40A load in the garage but I have some 50A SSR's I'm using for that.
[15:52:39] <specing> learningc: Do you need me to wipe too?
[15:53:05] <jimlay> :P
[15:53:08] <learningc> jimlay: I have just ordered that 2n7002 mosfet
[15:53:34] <jimlay> specing: thanks on the FQU11P06
[15:53:50] <learningc> specing: yes, full service is nice, I could give you some tips too :)
[15:54:06] <OndraSter_> screw PNP/NPN transistors... nowadays it is all about mosfets!
[15:54:09] <OndraSter_> or any FETs!
[15:54:29] <OndraSter_> at school we still use "bipolar as a switch in computers" or something like that :(
[15:54:34] <OndraSter_> instead P/N channel MOSFET
[15:54:41] <specing> Yes, normal transistors are just vacuum tubes today
[15:54:44] <OndraSter_> haha
[15:54:47] <OndraSter_> yeah
[15:54:47] <jimlay> I prefer Josephson Junction's.
[15:54:55] <jimlay> Very nice switching times. :)
[15:54:59] <learningc> ok, well, actually I do not really know how to drive the mosfet
[15:55:31] <learningc> any voltage divider to bias the gate?
[15:55:52] <learningc> the uC is 3.3V output
[15:56:02] <OndraSter_> 3V3 won't do .5A
[15:56:06] <OndraSter_> looking at datasheet
[15:56:11] <OndraSter_> http://www.fairchildsemi.com/ds/2N/2N7000.pdf
[15:56:28] <OndraSter_> at 3V Vds they got barely 100mA Id
[15:56:31] <OndraSter_> err
[15:56:33] <OndraSter_> 3V Vgs
[15:56:51] <OndraSter_> (why czechs and probably germans only use U for voltage and the rest of the world uses V?)
[15:57:01] <OndraSter_> as in... Uds, Ugs...
[15:57:04] <jimlay> The FQU11P06 looks like it's either cheeper or the same price as the 2n7002 and has much nicer specs.
[15:57:32] <OndraSter_> you should've gotten some TTL-logic (or how they call them) MOSFET
[15:58:08] <learningc> OndraSter_: where do you see that 3Vgs?
[15:58:25] <OndraSter_> http://clip2net.com/s/1GECY
[15:58:26] <OndraSter_> here
[15:59:06] <OndraSter_> or here
[15:59:06] <OndraSter_> http://clip2net.com/s/1GEDv
[15:59:25] <learningc> ah
[15:59:30] <learningc> then it's useless
[15:59:51] <OndraSter_> I always take those typical characteristics as the worst case scenario
[16:00:12] <OndraSter_> I am pushing few amps through my MOSFET, at -3V3 Vgs, where it was stated to be pushing this current with -4V5 or so
[16:00:19] <learningc> any other mosfet that can do the job?
[16:00:36] <OndraSter_> N or P channel?
[16:00:46] <OndraSter_> you posted N, jimlay suggested P
[16:01:12] <OndraSter_> so.. which one :)
[16:01:40] <learningc> p channel
[16:01:42] <learningc> no, n channel
[16:01:51] <learningc> the one that pull high to enable
[16:01:58] <OndraSter_> yes, N
[16:02:56] <OndraSter_> what Vds voltage?
[16:04:30] <learningc> 12 volt supply
[16:04:36] <OndraSter_> and what package :)
[16:04:45] <learningc> smd package smallest as possible
[16:04:59] <OndraSter_> sot23 okay?
[16:04:59] <learningc> (I'm doing a tiny board
[16:05:02] <learningc> yes
[16:05:06] <OndraSter_> and 1A current?
[16:05:15] <learningc> no, perhaps 500mA max
[16:06:36] <OndraSter_> <OndraSter_> 3V Vgs
[16:06:37] <OndraSter_> damn
[16:06:41] <OndraSter_> IRFML8244TRPBF
[16:06:46] <OndraSter_> http://uk.farnell.com/international-rectifier/irfml8244trpbf/mosfet-n-ch-25v-5-7a-sot23/dp/1857298RL
[16:07:27] <learningc> do they all(all mosfets in sot23) have the same pins configuration?
[16:07:43] <OndraSter_> hmm most likely
[16:07:45] <OndraSter_> but I'd check the datasheet always :)
[16:07:52] <learningc> ok thanks
[16:07:56] <learningc> yeah
[16:08:21] <learningc> just in case I want to change to another mosfet later if that one does not suit the purpose
[16:08:40] <OndraSter_> ah
[16:08:44] <OndraSter_> what are you gonna be driving it with?
[16:08:45] <OndraSter_> some micro?
[16:09:16] <learningc> yes, microcontroller output, probably using PWM for speed control
[16:09:18] <OndraSter_> ah
[16:09:34] <OndraSter_> so also watch out for gate charge
[16:09:42] <OndraSter_> and what current can the micro sink/source
[16:09:45] <learningc> gate charge?
[16:09:51] <OndraSter_> yeah
[16:09:55] <OndraSter_> http://www.irf.com/product-info/datasheets/data/irfml8244pbf.pdf
[16:09:57] <learningc> not sure what that is
[16:09:59] <OndraSter_> page 4
[16:10:08] <OndraSter_> well, the gate on the MOSFET acts as a capacitor
[16:10:35] <OndraSter_> but bah, I wouldn't care about that unless you are going to hundreds of kHz switching speed s:)
[16:10:37] <OndraSter_> speeds*
[16:10:50] <OndraSter_> I just don't care about those parameters myself... and it works at 2kHz switching just fine
[16:10:55] <OndraSter_> the ones I use
[16:11:25] <learningc> ah, I see
[16:12:22] <learningc> and, any bias I need to drive the mosfet, like I used to do with bjts?
[16:12:39] <OndraSter_> bias?
[16:13:23] <learningc> like any resistors on the gate, source, drain pin?
[16:13:29] <OndraSter_> oh
[16:13:32] <OndraSter_> nope
[16:13:43] <OndraSter_> or at least I wouldn't use it
[16:13:44] <learningc> none?
[16:13:50] <OndraSter_> that doesn't mean that there shouldn't be any...
[16:13:52] <OndraSter_> but I never used any for MOSFETs :)
[16:14:05] <OndraSter_> they are not BJTs...
[16:14:18] <OndraSter_> MOSFETs are good as switches...
[16:14:23] <learningc> I have seen people put a resistor between gate and drain
[16:14:29] <OndraSter_> either fully open with Rds about 0.00nothing or fully closed with many meg/tera ohms
[16:14:35] <OndraSter_> pullup
[16:14:37] <OndraSter_> probably
[16:14:45] <learningc> ah
[16:14:48] <OndraSter_> there is pullup in the micro though
[16:14:56] <OndraSter_> on 99% there is :)
[16:15:06] <learningc> so it means the mosfet is on by default?
[16:16:09] <OndraSter_> what does "default" mean?
[16:16:17] <OndraSter_> either the gate is charged... or not
[16:18:45] <learningc> like when the microcontroller is just turned on and the output pin is high impedance?
[16:19:07] <OndraSter_> oh... I never tried to do that..
[16:19:11] <OndraSter_> that's what the pullup/down is for
[16:19:42] <OndraSter_> (amongst other things)
[16:22:23] <learningc> in what case I would need an optocoupler to drive the mosfet?
[16:22:40] <OndraSter_> if there was dangerously high voltage around the MOSFET
[16:22:54] <OndraSter_> something like... wall warts :)
[16:22:58] <OndraSter_> 230V in, few volts out
[16:23:04] <OndraSter_> the controller works at 5 - 20V usually
[16:23:50] <OndraSter_> so it is rather good idea to safely cross some creepage (in case of problems) with optocoupler... to completely make sure that there can't be any problems
[16:24:36] <cjoe> Hi all. I compiled myself an avr-gcc toolchain (4.5.1 with AVR Studio 5.1 patches) and works fine. It compiles fine any AVR source code and all avr-libc examples. But I can't use this toolchain with the Eclipse plugin, because the avr-gcc --target-help output (here: http://pastebin.com/qh6fiN6S) doesn't list the Known MCU names section. But the avr-gcc -mlist-devices --target-help shows all of supported MCUs (here: http://pastebin.com/qwJeGbrf)
[16:24:36] <cjoe> . I don't know that the avr-gcc --target-help why doesn't print out the Known MCU names section. Is anybody know this issue?
[16:26:28] <specing> NO, because we either use vim or regular AS
[16:26:38] <OndraSter_> :D
[16:27:02] <specing> and you should be complaining somewhere else... like in #eclipse
[16:27:23] <learningc> OndraSter_: but isn't wall warts low voltage at secondary and also transformer isolated?
[16:27:32] <OndraSter_> yes
[16:27:38] <OndraSter_> they need to be completely isolated
[16:27:42] <OndraSter_> that's why there is optocoupler
[16:28:10] <learningc> oh, you mean in wall warts itself
[16:28:27] <OndraSter_> yes
[16:29:14] <learningc> what if the mosfet operates from 12V supply?
[16:29:24] <OndraSter_> no problem
[16:29:29] <OndraSter_> there is Vgs voltage, not Vds :)
[16:29:34] <OndraSter_> btw
[16:29:34] <OndraSter_> http://pastebin.com/6KfrN8pJ
[16:29:46] <OndraSter_> I rewrote today whole DISPLAY_TIMER part
[16:29:51] <OndraSter_> to the RETI instruction
[16:29:55] <learningc> is there any chance 12V breaks into the gate and burn my microcontroller output?
[16:30:00] <OndraSter_> no
[16:31:34] <learningc> what about ESD since the motor will connect to the board to outside world by a cable?
[16:32:16] <OndraSter_> I am not sure you are asking the right guy, I am n00b in electronics, really
[16:32:32] <learningc> oh ok
[16:32:48] <learningc> maybe some expert is reading here?
[16:35:00] <johjoh_4> Hi! I'm trying to compile a simple c++ program with avr-gcc/avr-g++, but it compiles with this error message: "... fatal error: iostream: No such file or directory
[16:35:00] <johjoh_4> compilation terminated." I have searched the web for a while now on how to solve the issue with the compiler not able to find #include iostream. Is there anyone that has any experience with this or knows how to solve this? Any tip?
[16:35:34] <OndraSter_> iostream for AVR? hmm
[16:35:46] <OndraSter_> haven't used AVR C pretty much at all :)
[16:36:17] <johjoh_4> OndraSter_: Ok
[16:36:32] <OndraSter_> do you use anything from iostream?
[16:37:03] <specing> johjoh_4: Oh, and where do you expect stdin, stdout and stderr go?
[16:37:45] <specing> johjoh_4: If you want to use C++, go love some #arduino
[16:37:57] <OndraSter_> :D
[16:38:17] <johjoh_4> OndraSter_: Yeah, just compiling a hello_world program. Just using cout, cin for stf
[16:38:19] <johjoh_4> std
[16:38:22] <specing> johjoh_4: If you want to become an uber(excuse me for not having umlauts) nerd/pro/younameit, stick here and use C
[16:38:46] <OndraSter_> but,,, where would the stdin/out/.. go to?
[16:38:55] <jadew> why use C? he can use ++
[16:39:02] <jadew> *c++
[16:39:07] <specing> He can
[16:39:20] <jadew> if he can take into account ram and stuff like that, he'll be fine
[16:39:24] <specing> At the expense of huge code
[16:39:28] <jadew> not true
[16:39:33] <specing> true.
[16:39:36] <johjoh_4> OndraSter_: to console and from keyboard
[16:39:42] <specing> I've tried C++ myself
[16:39:45] <OndraSter_> and where would that be connected to
[16:39:47] <specing> its massive
[16:39:54] <specing> LOL
[16:39:55] * OndraSter_ still writes in assembly
[16:39:57] <jadew> the same code compiles the same
[16:40:12] <jadew> it's not massive
[16:40:14] <OndraSter_> can't go much lower than assembly lol
[16:40:22] <specing> OndraSter_: you can, machine code!
[16:40:28] <OndraSter_> well asm = machine code
[16:40:30] <OndraSter_> just human readable
[16:40:36] <specing> Hmm
[16:40:49] <jadew> and, no there's no difference between C and C++, it's just that in C you might do stuff the shitty way
[16:40:54] <jadew> which is also possible in C++
[16:41:20] * specing wonders what planet jadew fell from
[16:41:48] <jadew> give me an example of how C++ code will result in bigger code please
[16:42:09] <specing> simple, make a class and call some functions
[16:42:17] <OndraSter_> if you start writing in C++ rather C and compiling with C++ compiler
[16:42:22] <OndraSter_> you GET bigger code
[16:42:29] <johjoh_4> OndraSter_: It will by now not be connected to anything. I just have to verify if I can compile c++ with avr-gcc/avr-g++
[16:42:29] <OndraSter_> C++ = making use of all new stuff such as classes
[16:42:30] <jadew> specing, that's the same as making a structure and calling some functions passing the pointer to it
[16:42:51] <OndraSter_> johjoh_4, that dosesn't explain my question, where would keyb/console be connected to :)
[16:42:57] <specing> jadew: not really, no
[16:42:58] <OndraSter_> ergo stdin/out are NOT implemented
[16:43:00] <OndraSter_> and do not exist
[16:43:02] <jadew> specing, yes, really
[16:43:05] <specing> jadew: C++ uses double pointers
[16:43:13] <specing> base + offset
[16:43:28] <jadew> that's the same with C
[16:43:37] <jadew> you have base (the pointer to the structure)
[16:43:42] <jadew> + offset to the value you're looking for
[16:43:50] <specing> While I havent checked what avr-g++ produces, on amd64 the pointers are 16 bytes
[16:44:10] <jadew> don't compile for 64bits then
[16:44:17] <specing> >_>
[16:44:28] <OndraSter_> 64bit AVR!!! WHERE!!
[16:44:29] <OndraSter_> :P
[16:44:37] <Steffanx> 8 in parallel
[16:44:41] <OndraSter_> haha
[16:44:59] <OndraSter_> what about running some heavy duty math on AVR cores :P
[16:45:04] <OndraSter_> parallel processing...
[16:45:14] <OndraSter_> many, many, many AVR cores in one 1U 19" rack
[16:45:25] <OndraSter_> or using ARM for that purpose
[16:45:28] <OndraSter_> way higher clocked :P
[16:45:36] <specing> check tilera
[16:45:59] <specing> 512-core Mips-like 1U at 1.5Ghz
[16:46:12] <OndraSter_> yay
[16:46:15] <OndraSter_> sounds sexy
[16:46:19] <specing> indeed
[16:46:29] <specing> Im still looking where to buy myself one
[16:46:55] <Fleck> http://vimeo.com/36579366
[16:47:22] <johjoh_4> OndraSter_: That is not relevant. I just wan't to compile a simple hello_world.cpp program with avr-gcc/avr-g++. Any tip?
[16:47:31] <OndraSter_> remove #include <iostream>
[16:48:26] <johjoh_4> OndraSter_: That does not take me any further
[16:48:32] <OndraSter_> how come?
[16:48:37] <johjoh_4> ?
[16:48:38] <OndraSter_> it doesn't report "cannot find iostream" anymore :P
[16:48:44] <specing> :D
[16:48:57] <johjoh_4> :-D
[16:49:09] <OndraSter_> cout, cin, .. do NOT exist
[16:49:14] <OndraSter_> you have to make your own ones
[16:49:33] <johjoh_4> OK
[16:58:05] <specing> ... And johjoh_4 proceded to inventing his own iostream
[17:11:13] <OndraSter_> specing, haha I thought the asame thing LOL
[17:11:20] <OndraSter_> same*
[17:32:28] <jimlay> specing: Thanks man! My CNC machine is finally cutting GCODE nicely! -- I've made a full stack machine from scratch, motors out of printers, custom controller (3 H-bridges, some diodes, 1 microcontroller), my own firmware, and a gui controller that I've written... the only part I didn't make was the CAD software I'm using to do the drawings.
[17:33:12] <jimlay> Total cost < 70$ - 36" by 36" by 12" with an accuracy around 1/100" -- I'm satisfied.
[17:34:39] <OndraSter_> WOW
[17:36:43] <Tom_itx> jimlay, did you write the gcode translator?
[17:55:21] <jimlay> I did. avr c
[18:15:28] <Tom_itx> jimlay, got any pics of your setup?
[18:15:32] <jimlay> Tom_itx: I tried to use grbl but it would randomly crash on me and I couldn't get it to crash repeatedly so I was stabbing in the dark trying to figure out why.
[18:20:54] <Tom_itx> jimlay, what avr are you using for yours?
[18:21:09] <Tom_itx> care to show/share the code?
[18:26:28] <jimlay> I'm using a teensy 2.0 ++
[18:27:16] <Tom_itx> what is that a 1287?
[18:27:23] <Tom_itx> 1286..?
[18:27:57] <Tom_itx> is it a takeoff of the brbl code?
[18:36:56] <learningc> do one needs to add any pull-up resistors on SPI lines?
[18:43:19] <Tom_itx> i doubt it
[18:53:28] <jadew> learningc, only on the slave select line
[18:55:42] <jadew> also, if you're acting as master and you don't expect to get turned into slave, make it an output
[18:56:01] <jadew> otherwise noise can kick you of master mode and into slave mode
[18:56:10] <jadew> *out of
[19:03:22] <cyanide> folks
[19:03:33] <cyanide> my first printed board
[19:03:40] <cyanide> WORKS :D
[19:03:46] <Tom_itx> you etched?
[19:03:50] <cyanide> itead
[19:04:02] <Tom_itx> did you expect other results?
[19:04:03] <cyanide> next batch on the way from capn
[19:04:24] <cyanide> not really. but my first time
[19:04:32] <Tom_itx> you're a pro now
[19:04:36] <cyanide> and hand soldered 0805 smds
[19:04:36] <cyanide> lol
[19:04:38] <cyanide> haha
[19:05:06] <cyanide> i wrecked my first board, learnt a few things and soldered this one in a different order
[19:05:26] <cyanide> just blinked an led
[19:05:29] <Tom_itx> solder tall stuff last
[19:05:30] <cyanide> :)
[19:05:33] <cyanide> yea
[19:05:46] <cyanide> i went for the header pins first, previously
[19:06:11] <cyanide> and even otherwise. im really new to all this
[19:06:30] <cyanide> i just messed with an arduino, then designed something myself and sent it off to itead for kicks
[19:07:08] <Tom_itx> fun stuff to do
[19:11:27] <jadew> hah, nice idea: http://iteadstudio.com/store/index.php?main_page=product_info&cPath=19_20&products_id=423
[19:11:37] <jadew> you get 2 random boards for 10 cents
[19:12:33] <jadew> cyanide, so you ordered 10 boards?
[19:13:00] <Tom_itx> or 5
[19:13:44] <jadew> ah true, they can do 10 as well if they're larger than 10x10
[19:14:06] <jadew> sorry, 5
[19:14:20] <Tom_itx> they don't list their board deals very well
[19:17:23] <jadew> damn, my order from dealextreme hasn't been shipped yet
[19:17:37] <jadew> and I placed it 7 days ago
[19:21:13] <specing> Ugh itead should brush up on their english
[19:21:55] <Tom_itx> they probably use translate.google to make their pages
[19:24:43] <specing> ...
[19:46:06] <jadew> so their engrish is great?
[20:05:25] <learningc> when draw a power pas in eagle package, is it better to draw it as a surface mount pad or as a rectangle?
[20:05:56] <learningc> pas = pad
[20:11:36] <Roklobsta> pad
[20:11:41] <Roklobsta> give it a pin number
[20:12:09] <Roklobsta> woooop wait up. is this for a SMT diode or tranny?
[20:12:39] <learningc> no, it's a power pad under smps IC
[20:13:12] <learningc> I'll need to add vias for thermal dissipation to the bottom layer
[20:13:14] <Roklobsta> usually it will connect to vcc or gnd so make it a pin.
[20:13:31] <learningc> ok, thanks. I was thinking the same too
[20:13:48] <learningc> otherwise, how is it going?
[20:25:36] <Roklobsta> ok playing with avrdude
[20:32:04] <Tom_itx> doing what with it?
[20:32:36] <jadew> you know..
[20:33:26] <Tom_itx> i'd like to find a windows exe of it for 5.11
[20:33:57] <Tom_itx> i tried to compile one but don't have libusb installed right or something
[20:34:45] <Roklobsta> oh, testing the FTDI chip and mucking about with incarnations of libusb-win32
[20:35:29] <Roklobsta> OMG! have a look at helix.air.net.au I have a windows bin there I am using.
[20:35:51] <Roklobsta> I am testing the C232HM cable. It's nice.
[20:36:47] <Roklobsta> I have avrdude-5.11svn1068-Patch7729-win32.zip you can try
[20:37:09] <Roklobsta> even though i built this to use with the ftdi chips in mind it should work with every other device avrdude supports
[20:37:24] <Roklobsta> so try your programmer with it too.
[20:37:44] <Tom_itx> i don't use dude much but like to have a current ver handy
[20:38:07] <Roklobsta> AFAIK you only need libusb if you plan to use the ftdi parts.
[20:38:23] <Roklobsta> please try it and see if my binary works for your programmer.
[20:38:38] <Tom_itx> i will here in a bit
[20:39:34] <Roklobsta> i don't expect you will have any problems.
[20:40:47] <Roklobsta> i am using the um232h module in the pic as an FTDI serial port as ftdi seems to be the only serial usb chip that doesn't lock up or BSoD in Win7. The prolific chip and some other USB serial cables have caused me too much grief.
[20:42:04] <Tom_itx> what's the patch 7729 about?
[20:43:09] <Roklobsta> bloody libftdi 0.19 missed putting in the VIDPID for the 232H part, so I have added a patch to support it. When libftdi.20 comes out the patch will be integrated into avrdude
[20:44:02] <Roklobsta> it also enabled the 30MHz SPI clock in the newer FTDI parts. The defaul tis 6MHz.
[20:44:43] <Roklobsta> basically a 128KB rom can be programmed and verified in 4 secs.
[20:46:13] <Roklobsta> how quick is your programmer?
[20:46:26] <Tom_itx> pretty quick
[20:46:36] <Tom_itx> faster than the original mkii i think
[20:46:46] <Roklobsta> c'mon lets have a speed shootout
[20:46:53] <Tom_itx> failed to start because of pthreadGC2.dll was not found
[20:46:58] <Roklobsta> OH!
[20:47:00] <Roklobsta> interetsing
[20:47:01] <Roklobsta> ok
[20:47:10] <Roklobsta> WTF
[20:47:17] <Roklobsta> you are the first to report that
[20:47:18] <Tom_itx> it's slower than it was
[20:47:36] <Tom_itx> dean slowed it down for reliability issues with various hardware
[20:48:23] <Roklobsta> oh, crap, some was telling me some progeammers need pthreads. the ftdi chip doens't which is why i haven't hit this bug.
[20:48:48] <learningc> Tom_itx: is there a way to change the size of the solder mask for a pad in eagle?
[20:48:51] <Tom_itx> how do you propose timing the download?
[20:49:03] <Tom_itx> not for just one pad
[20:49:04] <Roklobsta> oh i just use the -v option
[20:49:16] <Tom_itx> what's -v get you?
[20:49:40] <Tom_itx> what chip do you wanna test with?
[20:49:40] <Roklobsta> more debug info.
[20:50:00] <Roklobsta> no i am just using the timings reported by acrdude like http://helix.air.net.au/helix/files/8313/2883/8614/avrdude-dump.txt
[20:50:01] <Tom_itx> i'll just use 5.10
[20:50:10] <Roklobsta> i only have an atmega 1280
[20:50:24] <Tom_itx> what code do you wanna test?
[20:50:45] <Tom_itx> i got a 128 i can play with
[20:50:47] <Roklobsta> arduino bootloader
[20:51:00] <Tom_itx> no thanks
[20:51:04] <Roklobsta> why not?
[20:51:11] <Tom_itx> i shudder to think i'd mess with anything arduino
[20:51:15] <Roklobsta> it's a hexfile that is readily available
[20:51:39] <Tom_itx> where's it at?
[20:51:46] <Roklobsta> i'll email it to you
[20:51:56] <Tom_L> for a 128?
[20:52:03] <Roklobsta> yes
[20:52:06] <Roklobsta> 1280
[20:52:06] <Tom_L> i got a 328
[20:52:10] <Roklobsta> oh
[20:52:16] <Tom_L> and a 128
[20:52:17] <Roklobsta> hmmm
[20:52:19] <Tom_L> or t10
[20:52:21] <Tom_L> or....
[20:52:28] <Tom_L> xmega
[20:52:49] <Tom_L> i gotta find some of em though, still in boxes after the move
[20:53:02] <Roklobsta> curse you pthreads
[20:54:40] <Roklobsta> no atmega?
[20:54:46] <Tom_L> sure
[20:54:51] <Roklobsta> 1280
[20:54:55] <Roklobsta> nono
[20:54:55] <Roklobsta> ok
[20:54:59] <Tom_L> 32 48 168 328 128
[20:55:07] <Roklobsta> i read you
[20:55:08] <Tom_L> 1280 but not on a board
[20:55:15] <Tom_L> 2560
[20:55:19] <Tom_L> not on a board
[20:55:23] <Roklobsta> can a 328 and 1280 be compared for flash write times?
[20:55:35] <Tom_L> flash is flash i guess
[20:56:17] <Tom_L> gimme a file for a 328
[20:56:31] <Tom_L> i dug it out
[20:57:54] <Roklobsta> hrm, i need to go. try later
[20:58:01] <Tom_L> aww
[20:58:14] <Roklobsta> it's sunny out
[20:58:26] <Tom_L> won't take 5 min
[20:58:59] <Tom_L> why would i have -vvvv on a avrdude line??
[20:59:07] <Tom_L> does that do more than just -v?
[20:59:25] <Roklobsta> yeah
[20:59:29] <Roklobsta> burps out way more crap
[20:59:34] <Tom_L> k
[21:00:57] <Roklobsta> what's your email
[21:02:24] <Tom_L> pm
[21:06:01] <Tom_L> i was wrong, this is a 168
[21:06:06] <Tom_L> i'll still load it
[22:13:06] <kickit2> hello everyone, I have a basic program which does NOTHING but turn a single LED on and off. The flash size being programmed is still over 7k (as reported by AVRDUDE). The whole reason I made this program was to see if the program size was being bloated and it seems like it is. What is a typical flash size when programming an mcu w/o bootloader or anything?
[22:14:50] <Tom_itx> there is no 'typical' as each program is different
[22:15:16] <kickit2> http://pastebin.com/Yr35Gk37 <---- I just find it difficult to accept this is 7k flash
[22:17:49] <Tom_itx> PORTC |= (1<<PC3);
[22:17:55] <Tom_itx> doesn't need to be in the loop
[22:18:12] <Tom_itx> errr, nevermind
[22:18:17] <inflex> kickit2: likely you're just including something like a delay() call that's generic, so it has to pull in a big lump of code
[22:18:21] <Tom_itx> i read that as DDR
[22:18:35] <Tom_itx> must be tired.
[22:18:52] <kickit2> inflex: I thought of that too, but even removing the delay_ms from the code only takes out a few bytes
[22:19:01] <inflex> well, just do while (1) { PINC = (1<<PC3); delayms(30); }
[22:19:48] <inflex> (setting a bit via PINx toggles the output state - convenient trick)
[22:20:08] <inflex> Something like that shouldn't be more than 100 bytes of flash uploaded, more like 50
[22:20:24] <inflex> if you have a lot more - then it's likely something in the linker stage still forcibly loading
[22:20:44] <inflex> usually, if it's not 'time critical', it's more compact to write your own delayms() routine
[22:20:53] <kickit2> Im diging though the makefile now
[22:21:21] <inflex> eg, delayms( uint16_t d ); while (d--) { asm volatile ("nop"); }
[22:24:24] <kickit2> well pruneing the makefile made the diff..
[22:24:30] <kickit2> 216 Bytes now :)
[22:26:41] <CapnKernel> jimlay: Regarding your home built CNC machine, if there are no pictures, it doesn't exist!
[22:30:21] <kickit2> yes... that was it... I was still using an old nerdkits makefile and it was forcing uart and some other goodness in there as well.. 270bytes with delays now :) I KNEW 7k could not be right
[22:31:54] <kickit2> anyways... thanks everyone... night