#avr | Logs for 2012-11-29

Back
[00:53:31] <e2580> anyone here able to help with export law on crypto?
[00:53:42] <e2580> at32uc3a3256s, has crypto accelerator
[04:46:53] <theBear> there are no export laws regarding crypto.... OOOOOOOH, you assume we all know what stupid country you are in... tut tut tut
[04:48:23] <OndraSter> :D
[04:48:29] <OndraSter> the US.
[04:53:06] <theBear> i know that, but only 'cos they've been arrogant and assumptious that EVERYONE is like them for so long, that it's OBVIOUS
[04:54:13] <darknite> ot
[04:54:34] <darknite> theBear: i think the us is among the few countries that actually has laws regarding it
[04:54:40] <Malinuss> OndraSter, do you think it would be possible to emulate any kind of system on the Xmega?
[04:54:43] <darknite> maybe in comopany with china and other totalitarian states
[04:54:52] <OndraSter> Malinuss, I want to make small 286 emu someday, yes :D
[04:55:09] <darknite> but most western countries don't have absurd shit like that, it requires the megalomaniac us to establish those laws
[04:55:15] <OndraSter> they ran ARMv5 on Mega. If you port the few peripherals you can run ARMv5 at twice the speed (or more) on xmega
[04:55:45] <theBear> darknite, i think the only stupid one
[04:57:49] <Malinuss> OndraSter, what about something like the nes? or would that be out of reach? I heard the graphical unit was pretty complicated. Not saying I could ever do it, but would be cool if someone could ;P
[04:58:12] <OndraSter> NES has got issues emulating 1:1 clock and speed-wise even on current computers :)
[04:58:42] <OndraSter> I am sure somebody already tried NES emu for AVR
[04:58:56] <OndraSter> the beauty of xmega is that you can grab current projects and just change few registers and it works
[04:58:57] <OndraSter> even in asm!
[04:59:06] <OndraSter> same instruction set (with some additions)
[05:00:07] <Malinuss> OndraSter, so the difference between atmega/attiny programming and xmega isn't that big? compared to let's say any of the 32-bit uC?
[05:01:10] <OndraSter> there is 0 difference between mega and xmega
[05:01:17] <OndraSter> except that you have got more possibilities
[05:01:23] <OndraSter> differences*
[05:01:33] <OndraSter> *difference
[05:04:06] <Malinuss> OndraSter, and some of the registers were 16-bit? right?
[05:05:15] <OndraSter> if you are using C you do not care about that
[05:05:40] <OndraSter> the config registers are 8bit, the addressing registers are 24bit and the counter registers are 16bit. Gcc takes care about all the writing/reading
[05:06:44] <OndraSter> oh awesome, snow is not far from here and I need to go to school in 5 hours
[05:06:48] <OndraSter> for an important test.
[05:08:26] <Malinuss> ah I see
[05:15:20] <Malinuss> OndraSter, when I compile LUFA using the Atmel studio, do you think it should be a exacutable project or a static library project?
[05:16:21] <OndraSter> project
[05:16:24] <OndraSter> executable
[05:57:52] <Malinuss> OndraSter, well I expected compiling LUFA would be easier. I tried using the atmel studio but I just get "cannot find the path specified" each time, sigh
[05:58:48] <Malinuss> any ideas or should I just download the gcc and all the toolchains etc. and compile it that way?
[06:03:43] <day_> are all electrical memory devices based on flipflops? (ram,sd-cards...)?
[06:05:00] <Steffanx> Malinuss read the docs about how to export to/import it in to AS?
[06:05:29] <Malinuss> Steffanx, I sure did, I'm still one step before that though (compiling)
[06:06:57] <jadew`> damn shitty glossy paper
[06:07:02] <jadew`> ruined my pcb :/
[06:07:09] <Steffanx> poor you
[06:09:41] <Malinuss> Steffanx, or what?
[06:37:06] <Horologium> day_, ummm...not really.
[06:37:32] <Horologium> ram is.
[06:37:40] <day_> Horologium: sd cards too
[06:38:42] <Horologium> sd cards are flash.
[06:38:44] <Horologium> http://www.explainthatstuff.com/flashmemory.html
[06:38:50] <Horologium> that will explain how flash works.
[06:39:48] <day_> Horologium: what i dont understand is, why not just create a hugh array of small capacitors and make them accessable via an adress and allow me to but stuff in them.
[06:40:06] <Horologium> haha.
[06:40:10] <Horologium> caps leak
[06:40:25] <Horologium> they won't hold the data for long.
[06:40:51] <Horologium> sram is flipflops, basically.
[06:41:31] <Horologium> dram is capacitor based
[06:41:40] <day_> Horologium: dont take it literally. afair you could do it via tunnel injection like sdcards do. and your capacitor would keep the information for a long time
[06:41:42] <Horologium> but you have to constantly refresh it or it goes away.
[06:42:06] <Horologium> caps leak, they won't hold their charge very long.
[06:42:58] <Horologium> the large array of small caps is dram.
[06:43:02] <day_> well dram is basically what i suggested
[06:43:03] <day_> y
[06:43:25] <Horologium> but they won't hold data for long without constant refresh.
[06:43:32] <day_> yes i know
[06:43:49] <day_> but sdcards can keep the information, because?
[06:44:05] <Horologium> because they are not cap based..
[06:44:18] <Horologium> they are flash, a solid state semiconductor based device.
[06:44:29] <Horologium> different technology.....
[06:44:49] <Horologium> and they aren't permanent either,,,flash will eventually degrade.
[06:46:33] <day_> Horologium: maybe this is totally wrong but arent there tons of field effect transistors in them? and they use this tunnel injection thing to keep them closed/open? so the gate is basically a capacitor?
[07:07:40] <day_> i read the articles. and i still think the same :/ what am i doing wrong
[07:15:36] <Tom_itx> that sounds like the old ferrite memory beads
[07:16:10] <Tom_itx> that went out 30 yrs ago, get with the program
[07:16:38] <RikusW> get FRAM instead ;)
[07:19:25] <day_> T_T.. let me rephrase my question. why do i need a NAND element if the information itself is stored in one float gate?
[07:19:41] <theBear> ferrite memory beads ? like they magnetise/demag them ?
[07:19:48] <theBear> 'cos that'd be awesomely arcane
[07:19:56] <day_> i like magic
[07:20:27] <theBear> one floatey gate is fine if you want a single bit
[07:20:42] <theBear> i should know, i run a TWO bit operation <grin>
[07:21:13] <day_> but a nand element wont save more than one but needs many more fets
[07:21:43] <theBear> c'mon, that made less sense than what i just said :)
[07:23:28] <day_> is flash memory based on nand flipflops or nand gates? o0 please dont say the later one
[07:23:39] <Tom_itx> theBear, http://en.wikipedia.org/wiki/Magnetic-core_memory
[07:23:56] <Tom_itx> it's gonna make a big come back :)
[07:24:27] <OndraSter> literally "big"
[07:24:51] <day_> aaaand i just wasted 3h trying to understand something that doesnt even exists...yeayyyy
[07:24:52] <Tom_itx> imagine a 2gig stick of that
[07:25:14] <Tom_itx> day_, good experience
[07:25:39] <OndraSter> <day_> aaaand i just wasted 3h trying to understand something that doesnt even exists...yeayyyy
[07:25:42] <OndraSter> I am reading limits
[07:25:44] <OndraSter> derivations
[07:25:45] <OndraSter> and such
[07:25:48] <OndraSter> I have the same feeling :D
[07:26:55] * RikusW integrates all of those :-P
[07:27:09] <OndraSter> :D
[07:28:09] <day_> i love this. at first you try to understand every formular. then every passage. 5pages later you have no clue what you have read but at the end you will surely get it...the last page! the holy grail!
[07:28:25] <OndraSter> yeah I am writing the test today
[07:28:27] <OndraSter> in like 4 hours
[07:28:31] <OndraSter> I don't understand anything :D
[07:28:40] <OndraSter> "if you can't count it, derivate it"
[07:28:41] <OndraSter> all I know
[07:30:01] <theBear> Tom_itx, ooh just read, way more efficient than my mag/biaswipe idea <grin>
[07:30:11] <theBear> trust those scientists to beat me to it by a lifetime and a half :)
[08:11:10] <edmont> if i have two files that make use of variables with the EEMEM attribute, how can i tell the linker in which order i want the EEPROM to be allocated?
[08:12:43] <RikusW> you could force it by making segments
[08:13:05] <RikusW> you'll have to read the AVR LIBC docs for details
[08:17:11] <edmont> RikusW: you mean sections?
[08:17:56] <RikusW> yep
[08:18:16] <RikusW> not sure if you can force the address
[08:18:24] <RikusW> it might be possible without sections
[08:20:24] <OndraSter> well you can simply work with it as eeprom, not as "variable in eeprom" :)
[08:22:19] <day_> how does reading from a nor memory work? i understand how its possible to program the single elements but not how to read from them :/ http://en.wikipedia.org/wiki/File:NOR_flash_layout.svg
[08:22:25] <Malinuss> anyone every used LUFA+AS? I could use some guidance
[08:23:18] <Malinuss> Apperently I'm too retarded to simply follow step by step instuctions
[08:26:39] <edmont> OndraSter: what do u mean?
[08:27:35] <OndraSter> setting eeprom address ptr with your own address
[08:27:38] <OndraSter> you somewhere #define
[08:27:50] <OndraSter> anyway
[08:27:52] <OndraSter> I am off to the school
[08:27:53] <OndraSter> bb
[08:28:36] <edmont> ok, thanks
[08:31:22] <Malinuss> :(
[08:37:59] <edmont> RikusW: it doesn't seem to be possible to modify sections: http://www.nongnu.org/avr-libc/user-manual/mem_sections.html
[08:40:01] <RikusW> edmont: OndraSter is right, that would be easier
[08:42:33] <edmont> ok, i get it
[08:42:41] <yunta> edmont: you can play with sections and their addresses by modifying linker script
[08:42:48] <edmont> but i dont know how to define a pointer wiht an address :)
[08:43:09] <RikusW> just start at 0 going to eeprom size - 1
[08:43:10] <edmont> yunta: where can i find it?
[08:43:38] <RikusW> there should be some docs in AVR-LIBC docs that shows how to read eeprom
[08:44:06] <RikusW> http://en.wikipedia.org/wiki/Polycarbonate
[08:44:10] <RikusW> hardy stuff
[08:44:37] <edmont> RikusW: yes: http://www.nongnu.org/avr-libc/user-manual/group__avr__eeprom.html
[08:49:56] <yunta> edmont: defaults are somewhere in ldscripts/*.x , depending on your distro
[08:53:05] <karlp> edmont: are you planning on referencing the right eeprom address from a different application?
[08:53:13] <timemage> edmont, or you could use a struct type and let it create pointers/offsets for you.
[08:53:21] <karlp> or are you hoping to amke things stay in the same place over time so you can upgrade firmware without erasing eeprom?
[08:53:32] <karlp> (named sections is the way I've done that before)
[09:10:02] <RikusW> http://openbookproject.net/electricCircuits/
[09:19:21] <edmont> karlp: i use Contiki that uses some of the eeprom space to store network variables
[09:19:34] <edmont> such as MAC address
[09:20:08] <edmont> and i want to use my own variables for my program, without overwriting those contiki variables
[09:28:07] <karlp> well first and foremost, you'd better learn where contiki's keeping things .)
[09:28:42] <edmont> karlp: that i already know
[09:28:54] <edmont> at the beggining of eeprom space
[09:30:13] <edmont> karlp: http://pastebin.com/n5RYrrbT
[09:34:47] <megal0maniac> OndraSter: Have you finished the first one yet?
[09:48:32] <iSaleK> Can someone take a look at FT232RL on FT datasheet and one in the Arduino 2009 (DuoMilovane) schematic. Are the pinouts the same?
[09:49:33] <iSaleK> For example, FT232RL on DuoMilovane, pin13 is PIO02 but on datasheet pin13 is CBUS2
[09:51:07] <iSaleK> As far as I can see other pins (that have the same names) have the same pin numbers in datasheet and arduino schematic
[10:21:35] <megal0maniac> iSaleK: The more likely situation is that they're calling the same thing a different name, based on the application
[10:25:50] <iSaleK> megal0maniac: I guess that's the case... Since all other pins have the same names and same pin numbers except 4 or 5 pins that have diferent names :)
[10:26:06] <iSaleK> Also another question, SDIP is Small Dual Inline Package, right?
[10:26:45] * megal0maniac googles SDIP
[10:26:50] <iSaleK> Since in catalouge it says that they have ATMega328-PU that is in SDIP package
[10:26:53] <megal0maniac> Shrink Dual Inline Package
[10:27:26] <iSaleK> Also Altium's Atmel library is too old. It doesn't have ATMega328 or any other newer products :
[10:27:35] <RikusW> probably because its narrow and not wide
[10:27:53] <RikusW> iSaleK: copy the mega8 ?
[10:28:11] <megal0maniac> RikusW: The mega328s I have are PDIP
[10:28:13] <timemage> megal0maniac, or skinny dip. dunno if anyone still says that anymore.
[10:28:38] <megal0maniac> timemage: I was just pointing out that the info is readily available :)
[10:28:43] <iSaleK> RikusW: I will create my own footprint and schematic based on the datasheet... that way I'm sure it'll fit :)
[10:28:52] <timemage> megal0maniac, sure.
[11:12:50] <megal0maniac> Wait, isn't the mega8 a 40 pin chip?
[11:12:59] <megal0maniac> And mega88 the 28 pin?
[11:13:54] <Casper> should be the same package normally
[11:14:01] <megal0maniac> 88 and 48?
[11:14:06] <megal0maniac> I mean 8 and 88
[11:14:07] <Casper> the 8 and 88 is basically the same I think....
[11:14:43] <megal0maniac> Ah. No, 8 is 32 pin. And I think it's wide
[11:14:48] <megal0maniac> But definitely 32
[11:15:02] <Casper> 28 in dip
[11:15:16] <Casper> 32 in surface mount
[11:15:25] <megal0maniac> Ah, yes
[11:16:13] <Casper> ahhh the city big vaccum cleaner is near here
[11:16:23] <Casper> I was wondering what was that big noise
[11:17:18] <megal0maniac> 8 is the odd one, because 16/32 are 40 pin, but 48/88/168/328 are all 28
[11:17:26] <megal0maniac> Hence my confusion :)
[11:23:43] <megal0maniac> Anyone know what "SIGTERM" on IRC means?
[11:25:06] <AR_> megal0maniac, it means the process has been killed
[11:25:32] <megal0maniac> Strange. I was disconnected from freenode. Reason being "SIGTERM"
[11:25:43] <AR_> weird
[11:25:51] <AR_> oh, there was a netsplit
[11:25:55] <AR_> maybe a server got killed
[11:26:02] <megal0maniac> Ah. Yeah, likely
[11:26:13] <megal0maniac> Netsplit would explain it
[11:32:18] <Malinuss> okay LUFA kinda sucks, I hoped there would be a noob friendly virtual COM setup for the USB, with functions like "read_byte" and "send_byte"... but no :(
[11:32:38] <megal0maniac> Malinuss: It isn't really for beginners
[11:32:45] <karlp> there's pretty decent example code,
[11:32:49] <megal0maniac> That's arduino ;)
[11:32:56] <Malinuss> yes, but none with what I need
[11:32:57] <karlp> I managed to copy and paste my way pretty well using lufa
[11:33:02] <karlp> using dual endpoitns and everything
[11:33:05] <karlp> try harder :)
[11:33:38] <Malinuss> megal0maniac, okay so what would you do?
[11:36:44] <Malinuss> karlp, haha well that wasn't very helpfull, but at least encouraging
[11:42:03] <karlp> https://github.com/abcminiuser/lufa/blob/master/Demos/Device/ClassDriver/VirtualSerial/VirtualSerial.c isn't good enough?
[11:45:00] <megal0maniac> Malinuss: I probably wouldn't try, given my current skill level. But if I did, I'd learn a bit about terms which keep coming up (endpoints for example) and then I'd tackle the demo program, paying attention to the comments
[11:45:34] <Malinuss> karlp, yeah this is suposed to be the "easy demo" of the full-blown USBtoSerial project.... So I can simply use fputs and get it via some port listner?
[11:46:25] <Steffanx> yes
[11:46:32] <Malinuss> oh
[11:46:46] <Malinuss> well then I might figure it out ;P
[11:46:49] <Steffanx> Didn't try the demo did you? :P
[11:47:46] <karlp> horses don't like water or something.
[11:48:10] <Malinuss> Steffanx, still not sure how to compile it into my own AS project, well I compiled the main library and tried adding the VirtualSerial.c/.h and whatever else there were, but then I get some errors
[11:50:38] <megal0maniac> Some errors? Sounds serious...
[11:50:56] <Malinuss> <.<
[11:51:18] <Steffanx> Too bad you don't sound serious megal0maniac
[11:51:19] <Malinuss> Error 1 unrecognized argument to -mmcu= option: 'ATMEGA32U4' C:\Users\XXX\Documents\Atmel Studio\LUFAtest\LUFAtest\USBtoSerial.c 1 0 LUFAtest
[11:52:14] <megal0maniac> compiler incompatibilities? Or path stuff?
[11:52:53] <Malinuss> so instead of making a static library I did like the step-by-step tutorial said and copied some files into the project and added garbage stuff... I tried compiling, everything okay. Then I added the files from the demo, and figured it should work, but no
[11:53:17] <megal0maniac> Have you got win avr installed?
[11:53:37] <Malinuss> yes
[11:53:53] <Malinuss> thats how I've made the .tar thinge to copy to the project
[11:53:56] <megal0maniac> Try compile with avr-gcc
[11:54:04] <megal0maniac> instead of studio
[11:55:24] <Malinuss> okay I'll try
[11:58:32] <megal0maniac> The LUFA code was designed to work with avr-gcc, iirc
[11:59:46] <Malinuss> well there was this nice tutorial http://www.fourwalledcubicle.com/files/LUFA/Doc/120730/html/_page__exporting_library.html
[12:01:52] <megal0maniac> Read all the things :)
[12:02:03] <megal0maniac> prpplague: Pandaboard looks very cool
[12:02:54] <Malinuss> megal0maniac, I did, and it compiled fine... I just can't get the demo to work with it
[12:03:19] <Malinuss> megal0maniac, propably something about where the makefiles are, and what folders are named as etc. urgh
[12:03:56] <Malinuss> megal0maniac, there should just be a "usb_to_serial.h" ;D hehe
[12:04:15] <megal0maniac> OndraSter!
[12:04:22] <megal0maniac> Is that first board finished?
[12:05:09] <TechIsCool> megal0maniac: He was working on it yesterday I think
[12:05:26] <megal0maniac> He was, that's why I'm asking if it's finished :)
[12:08:23] <Malinuss> hahaha "MCU ATMEGA32U4 supported for assembler only"... Well go fuck yourself LUFA
[12:08:37] <megal0maniac> Using avr-gcc?
[12:08:40] <Malinuss> yes
[12:08:49] <megal0maniac> Heh :)
[12:09:07] <Malinuss> welp time to make my own stacks, see you in 2014
[12:09:19] <Malinuss> but seriously, what the fuck, haha
[12:09:25] <megal0maniac> You just need a newer version. It's not a LUFA problem, it's the compiler
[12:09:50] <megal0maniac> You can write your own stack, try again in 2 years when you're done, and you'll have the same problem
[12:09:50] <Malinuss> downloaded it 20min ago "get newest versio -here-" from the sourceforge
[12:10:01] <megal0maniac> Look at the filename.
[12:10:05] <megal0maniac> It has 2010 in it
[12:10:33] <Malinuss> so WinAVR-20100110 is not the newest version?
[12:10:46] <megal0maniac> No, it is
[12:10:56] <Malinuss> so?
[12:12:02] <megal0maniac> Maybe this? http://www.atmel.com/Images/avr-toolchain-installer-3.4.1.1195-win32.win32.x86.exe
[12:12:28] <megal0maniac> I can't remember whether that helps or not, but it's dated August 2012
[12:13:10] <Malinuss> also, if I compile it with the standard makefile setting, it compiles just fine
[12:13:54] <prpplague> megal0maniac: thanks
[12:14:39] <megal0maniac> Malinuss: What device does the default compile for?
[12:15:09] <megal0maniac> Also, http://www.nongnu.org/avr-libc/user-manual/
[12:15:18] <megal0maniac> 32u4 is supported by the latest version
[12:15:49] <Malinuss> megal0maniac, at90usb1287
[12:16:51] <megal0maniac> It's an older chip. Look at that avr-libc link. All the devices with a [1] are only supported by assembler. The version that comes with winavr has a whole bunch more chips with [1] next to them
[12:17:21] <megal0maniac> You simply have a dated version of the avr toolchain
[12:17:36] <Malinuss> okay I updated now
[12:17:42] <Malinuss> lets try again with the new toolchain
[12:19:06] <Malinuss> yay megal0maniac <3
[12:19:13] <megal0maniac> No, none of that
[12:19:21] <Malinuss> >no homo
[12:20:27] * megal0maniac wonders if arduino has updated their toolchain, with the Leonardo using a 32u4...
[12:20:47] <megal0maniac> It was using that same version of winavr before
[12:20:56] <Malinuss> now I just need to make it work wish AS5, haha
[12:31:39] <Malinuss> megal0maniac, okay I think I'm getting closer, so now I get some confilct and undercleard errors.. http://pastebin.com/Shx7iFiz
[12:32:28] <megal0maniac> I really have no idea. My knowledge of C is almost nothing
[12:33:06] <Malinuss> just makes no sense heh
[12:46:44] <Malinuss> oh shit
[12:46:46] <Malinuss> it compiled
[12:49:47] <Steffanx> Help!
[12:49:54] <Steffanx> What now to do Malinuss ?
[13:02:27] <nickoe_> hello, I am having a hard time to figur out how to set up a pin change interrupt routine
[13:02:44] <nickoe_> I am using an Atmega2560 with avr-gcc
[13:03:44] <nickoe_> I want the pin PCINT23/PK7 to be the pin, and wan't it to interrupt in faling edge
[13:04:22] <nickoe_> *on faling edge
[13:04:47] <nickoe_> I have been reading http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&t=89843&start=all&postdays=0&postorder=asc but I cant seem to figure how to really do it
[13:06:53] <damjan> anyone knows whats the generic name of this connector http://uk.farnell.com/jsp/search/productdetail.jsp?SKU=143129&MER=baynote-143129-pr ?
[13:09:48] <nickoe_> damjan, a molex connector
[13:10:19] <nickoe_> or KK interconnection system
[13:10:30] <nickoe_> as the datasheet says
[13:11:10] <Malinuss> Steffanx, hehe, feed the chip with it and check if I get the extra port
[13:15:54] <Malinuss> of course, it dosn't work, silly me to think it would ;D
[13:17:00] <Malinuss> heh "things, why don't they just wrok?"
[13:19:37] <Malinuss> ;(
[13:29:07] <nickoe> woops I ran out of power
[13:29:35] <Malinuss> this might be crazy thinking, but when I compiled the LUFA "VirtualSerial", with all the right settings and uploaded it to my uC I expected a COM port would pop up when I plug it in. but nooo why should it work
[13:29:44] <Malinuss> that would be too easy!
[13:36:18] <Malinuss> the worst thing is, that it compiles with no problems, so there is no way for me to figure out what is wrong, haha
[13:51:40] <Malinuss> finally
[13:51:43] <Malinuss> ysy
[13:51:45] <Malinuss> yes
[13:56:42] <OndraSter> Malinuss, winavr ! atmel's gcc
[13:56:50] <OndraSter> !=
[14:01:01] <OndraSter> <Malinuss> welp time to make my own stacks, see you in 2014
[14:01:01] <OndraSter> <Malinuss> but seriously, what the fuck, haha
[14:01:08] <OndraSter> if you know what to do - you need just one week :)
[14:01:17] <OndraSter> about 20 hours is my dev time on it
[14:01:22] <OndraSter> and I have got almost it completely working
[14:01:24] <OndraSter> including CDC serial
[14:27:02] <RikusW> anyone want tools ? http://buy1.snapon.com/catalog/item.asp?P65=&tool=hand&item_ID=641844&group_ID=675881&store=snapon-store&dir=catalog
[14:33:21] <jadew`> haha
[14:34:20] <Malinuss> ?
[14:35:34] <jadew`> RikusW, found a must have toolbox
[14:35:41] <jadew`> with out the comma
[14:35:45] <jadew`> [22:14.58] <RikusW> anyone want tools ? http://buy1.snapon.com/catalog/item.asp?P65=&tool=hand&item_ID=641844&group_ID=675881&store=snapon-store&dir=catalog
[14:37:18] <Malinuss> what a cheap-ass
[14:37:23] <Malinuss> only 40k
[14:37:24] <RikusW> heh
[14:38:16] <Malinuss> yay I can into virtual serial COM... well can't get it to work with AS yet but well ;P
[14:39:03] <RikusW> Malinuss: what are you building ?
[14:40:38] <RikusW> jadew`: that tools are actually meant for the #AutomotiveVehicleRepair channel.... ;)
[14:40:39] <Malinuss> RikusW, my long term project is a few-part avr-only driven, AM-transmitter-playing wav files from SD and displaying info on 5510 nokia display - device.... right now I'm just playing around with the usb peripheral
[14:41:24] <RikusW> hmm interesting
[14:42:54] <Malinuss> I can transmit a 4sec long wav file from the flash memory (stored as a array of course), about ~10m, at the 533khz frequency, *only* using my avr and two transistors (and two resistors :P)
[14:43:59] <RikusW> nice
[14:44:21] <RikusW> whats the AVR clock ?
[14:44:26] <Malinuss> 16mhz
[14:44:38] <RikusW> some can go to 20
[14:44:47] <Malinuss> doesn't really matter
[14:45:16] <Malinuss> almost everything is done in the hardware anyway
[14:45:21] <OndraSter> aye, he will get xboard soon :D
[14:45:24] <OndraSter> with 32MHz
[14:45:29] <OndraSter> which ran run at 48MHz.. or more :D
[14:45:40] <Malinuss> do you need crystal for that btw.?
[14:45:43] <OndraSter> no
[14:45:51] <Malinuss> so you can set it to 48?
[14:45:57] <Malinuss> fuses?
[14:45:59] <OndraSter> upto 62MHz when using USB
[14:46:00] <OndraSter> no fuses
[14:46:10] <OndraSter> no fuses related to clock, that is
[14:46:15] <Malinuss> why only when using usb? I don't get it
[14:46:41] <OndraSter> because there are two clock sources (well, two main):
[14:46:44] <OndraSter> 2MHz and 32MHz
[14:46:46] <RikusW> no fuses, builtin PLL
[14:46:51] <OndraSter> the 32MHz one has to be used for USB
[14:46:57] <OndraSter> the 2MHz can be used for core
[14:47:02] <OndraSter> and with PLL you can multiply it by upto 31 times
[14:47:26] <Malinuss> PLL?
[14:47:33] <RikusW> so why do Atmel provide overclocking ability ?...
[14:47:40] <nevdull> phase lock loop
[14:47:40] <RikusW> Malinuss: Phase Locked Loop
[14:47:50] <RikusW> used to multiply a frequency
[14:47:56] <specing> Because they are not Intel/AMD
[14:48:08] <Malinuss> This is something for the Xmega only, right? no atmega can do that?
[14:48:15] <RikusW> hmm 32MHz * 31.... fast xmega :-D
[14:48:23] <OndraSter> RikusW, 32MHz RC is /4
[14:48:26] <OndraSter> before being fed to PLL
[14:48:50] <OndraSter> but if you want to use USB you have to dedicate the 32MHz one to the USB module
[14:48:57] <RikusW> Malinuss: some atmegas got a clock divider
[14:49:12] <RikusW> so you can change clock frequency on the fly
[14:49:13] <Malinuss> RikusW, yeah, but that only allows me to go lower
[14:49:14] <OndraSter> only divider
[14:49:14] <nevdull> might be able to on the atmega32u4 family and the at90usb* family
[14:49:22] <OndraSter> doesn't one attiny have PLL?
[14:49:24] <Malinuss> nevdull, really? how?
[14:49:25] <OndraSter> for timer/PWM
[14:49:51] <RikusW> OndraSter: t26 does have 64MHz PWM
[14:50:41] <RikusW> nevdull: I think those got a divider, and maybe PLL for USB only
[14:50:50] <RikusW> s/maybe//
[14:50:53] <nevdull> Malinuss: using a reference clock to get 48MHz via PLL for USB interface
[14:51:06] <nevdull> yeah could be
[14:51:27] <Malinuss> 2deep4me
[14:55:40] <nickoe> Hello, I got the interrupt to work somewhat... but... I am having trouble still. The problem is that it seems like the processor crashes somehow, atleast the device stops blinking. this is my main code, http://dpaste.com/839107/
[14:56:11] <nickoe> I am having three uarts running peter pleurys UART library also wiht interrupts.
[14:58:26] <nickoe> Running on an ATmega2560
[14:58:57] <Malinuss> nickoe, can't help you, but I'm wondering, what are you using for the USB?
[14:59:20] <RikusW> nickoe: is your WDT off ?
[14:59:21] <nickoe> one of the UARTS with a FTDI device
[14:59:31] <nickoe> RikusW, what is WDT?
[14:59:39] <nevdull> watchdog timer
[14:59:46] <nickoe> hmm, dunno
[15:00:02] <nickoe> but if I try to disable the reading of the uarts it runs forever
[15:00:51] * nickoe don't quite remember the purpose of the watchdog timer
[15:01:30] <nevdull> it'll reset your uC if you reach brownout voltages, etc
[15:01:42] <nevdull> check the fuses if you need to know if its set
[15:01:47] <nickoe> okay
[15:01:53] <RikusW> did you maybe enable some irq without giving it an ISR ?
[15:02:18] <nickoe> RikusW, I don't think so, you can see the code
[15:02:35] <nickoe> that was the only "new" interrupt I made (I am new to interrupts)
[15:02:59] <nickoe> though I know that the UART lib I am using i is interrupt drive
[15:05:36] <nickoe> I only assigned PCICR |= 1<<PCIE2; and PCMSK2 |= (1<<PCINT23); and using the ISR(PCINT2_vect) {...} RikusW
[15:07:06] <nickoe> can an UART interrupt interrupt an pin change ingterrupt?
[15:07:18] <RikusW> not sure
[15:07:40] <nickoe> when I remove the two lines 23 and 24 it keeps running
[15:07:46] <RikusW> while inside and irq the I flag is off
[15:08:18] <nevdull> i think using the standard ISR macro it sets a prologue and eplogue of cli/sei i think
[15:08:21] <nickoe> so why could it stop?
[15:08:32] <nickoe> nevdull, yeah, that was what I read
[15:08:33] <nickoe> also
[15:08:40] <RikusW> I think you're doing to much inside the ISR
[15:09:06] <nickoe> maybe, it will take a long time ..
[15:09:58] <nickoe> hmm, how could I make that faster? If I just put those two lines in another function it won't help right?
[15:10:44] <RikusW> put it in the main loop
[15:11:12] <nickoe> ahh, and then have a counter in the ISR and then check that in the main loop?
[15:12:23] <RikusW> yes
[15:12:35] <nickoe> RikusW, like this? http://dpaste.com/839116/
[15:14:53] <RikusW> I'll add volatile char reset = 0; and check that
[15:15:07] <RikusW> and set to 1 inside ISR
[15:15:21] <nickoe> instead of int adis_ready_counter=0; outside main?
[15:15:33] <RikusW> that won't quite work
[15:16:00] <nickoe> so as the variable to check in main?
[15:16:15] <nickoe> should the reset variable also be global?
[15:16:15] <RikusW> yes
[15:16:20] <RikusW> yes
[15:16:28] <RikusW> and set to 0 in main again
[15:16:41] <RikusW> it must be volatile...
[15:17:09] <nickoe> why is that better than just using the other counter?
[15:17:34] <RikusW> because using the counter that way won't work
[15:17:51] <RikusW> you'll most likely miss it...
[15:18:20] <RikusW> unless you set it to 0 in main and not the ISR
[15:18:22] <nickoe> hmm, okay, the interrupt is coupled on a 800 Hz signal
[15:18:31] <nickoe> ok
[15:18:48] <RikusW> and I'll use >= 100 rather than == 100
[15:19:08] <RikusW> and then maybe -= 100 instead of =0;
[15:20:44] <nickoe> RikusW, so like this http://dpaste.com/839121/?
[15:21:11] <nickoe> hmm, yeh, that could possibly be better
[15:21:56] <RikusW> I'd still use >= 100 and then -= 100;
[15:22:07] <nickoe> yes
[15:22:16] <RikusW> you could miss 100......
[15:22:36] <RikusW> well maybe not, but I don't like taking chances...
[15:22:47] <RikusW> its a bug waiting to happen ;)
[15:23:22] <nickoe> RikusW, hehe I thin I just experinced that, maybe
[15:23:29] <nickoe> :)
[15:23:47] <RikusW> heh
[15:24:33] <RikusW> since it only count to 100 you can use -> volatile char adis_ready_counter=0;
[15:24:56] <RikusW> or unsigned char...
[15:25:03] <RikusW> or maybe just leave it int ;)
[15:25:29] <nickoe> I might want to count higher than a char
[15:27:43] <nevdull> anybody remember the flag to switch between intel and at&t syntax with avr-as or avr-gcc?
[15:28:46] * OndraSter didn't know that it supports both :D
[15:29:40] <nevdull> hehe yah unless i totally dreamed it, i think they added support. or maybe i'm remembering it backwards and atmel's asm supports both?
[15:32:04] <RikusW> at&t is ugly :S
[15:32:57] * nevdull agrees
[15:49:42] <tomatto> hi
[15:49:43] <tobbor> Hello tomatto
[15:52:05] <tomatto> please, why this line void sendChar(char data) cause this error usart.c:140:20: error: expected ';', ',' or ')' before numeric constant
[15:54:14] <Tom_itx> google finished mapping earth, now it's moving on: http://www2.electronicproducts.com/Google_partners_with_NASA_European_Space_Agency_for_out_of_this_world_project-article-fajb_google_stars_nov2012-html.aspx
[15:57:06] <tomatto> i don't see what is going on in this error. can you help me please?
[15:59:05] <tomatto> oh, i see
[15:59:07] <tomatto> sorry
[15:59:32] <Tom_itx> i knew he'd get it
[16:03:22] <nickoe_> RikusW, heh, run empty of power for the third time today
[16:03:42] <nickoe_> RikusW, thank you for your help, it seems to work, now some more testing :)
[16:07:51] <RikusW> good :)
[16:11:33] <Malinuss> Tom_itx, I asked you yesterday but you didn't really respond - where do you live - aka. where do you ship from?
[16:13:14] <Malinuss> if it isn't secret of course ;P
[16:13:30] <RikusW> USA ?
[16:18:11] <OndraSter> of course the big US
[16:21:21] <nevdull> if you're just dying to find out, you can ballpark it with a /ctcp <name> time request. i think that returns host local time of the connected client. ;)
[16:28:59] <Malinuss> us and a?
[16:39:32] <karlp> nevdull: which may or may not be the timezone of the person you're talking to....
[16:40:37] <nevdull> yah i dunno. always thought it got the time from the client, but i suppose someone could be ssh'd into a unix account in another country or something
[16:41:50] <yunta> like me..
[16:45:51] <karlp> (or me)
[16:46:19] <karlp> or even, running a web client, and ctcp may be server time, maybe user time, maybe who knows what
[16:49:28] <nevdull> i prefer to use my wrist watch for the time. the batteries are dead but it's no longer *slightly* fast or *slightly* slow, it's always precisely the time twice a day.
[16:49:38] <theBear> hehe
[16:49:41] <OndraSter> hehe
[16:49:45] <nevdull> ;)
[16:50:53] <theBear> i've got an old fob watch that had some 'repairs' done before i got it, like the superglue i had to carefully chisel off the front glass, but anyway, i don't have/know what fancy lubes you use in a mechanical watch, so i haven't messed with that side yet, but damned thing regulates a little slow or a little fast depending on amount-of-wind
[16:52:30] <nevdull> hehe could be picking up your kinetic motion and translating that energy into the cogs, if you happen to be a particularly kinetic sorta person :)
[16:55:57] <yunta> how tf should I select opamp, there is millions of them :(
[16:58:14] <nevdull> yunta: slew rate, CMRR, voltage gain, etc
[16:58:29] <Malinuss> okay okay okay. Let's assume I would like to use gcc instead of AS, which editor would you guys sugst tne?
[16:58:29] <nevdull> really depends on what configuration you're having your op amp in and what you need it for
[16:59:00] <nevdull> Malinuss: what OS? i usually use vi or avr-eclipse
[16:59:10] <Malinuss> nevdull, win
[16:59:21] <nevdull> there's winAVR
[16:59:31] <Malinuss> nevdull, yeah the editor looks like puke
[16:59:52] <nevdull> i've written some instructions online on how to use the regular Visual Studio 2011/12 to produce AVR hex files, too
[17:00:00] <Malinuss> really?
[17:00:02] <Malinuss> interesting
[17:00:04] <nevdull> yep
[17:00:11] <Malinuss> link
[17:00:13] <Malinuss> pls
[17:00:46] <yunta> nevdull: any suggestion for opamp for current metering with shunt resistor on gnd side ?
[17:00:46] <nevdull> http://www.instructables.com/id/Use-Visual-Studio-2010-to-Compile-AVR-Hex-Files/
[17:01:39] <yunta> Malinuss: I recommend emacs. Takes some time to learn, but it's one of the best.
[17:02:08] <Malinuss> I'm just so used to msvc, hehe
[17:02:26] <nevdull> yunta: is this for a differential amp or? i'd probably look into offset current and bias current
[17:02:41] <yunta> yes, diff
[17:03:03] <yunta> I guess rail-to-rail, as I need one inputs to be at gnd.
[17:03:12] <yunta> (is it still considered differential then?)
[17:03:37] <nevdull> i'd take a look at some of the tools ti.com offers to pick between op amps. also, if you register you can have them ship you 3 each of 5 different op amps as samples. i do that about once every 2-3 weeks :)
[17:03:58] <nevdull> yunta: hmm sounds kinda like an inverting amp
[17:04:08] <yunta> interesting
[17:04:29] <yunta> do you have any opamp you personally like to use for many applications?
[17:04:59] <nevdull> haha i keep a spread sheet just for that occassion of op amps i have in my parts stock ;)
[17:05:08] <yunta> lol
[17:05:24] <yunta> I see there is no easy way out of my problem.
[17:05:32] <yunta> I'll just keep learning then....
[17:05:42] <OndraSter> Malinuss, why would you do not want to use AS6?!
[17:05:49] <OndraSter> it is the best IDE for embedded developing around :)
[17:05:53] <OndraSter> let alone free!
[17:06:04] <nevdull> OPA2134/4134 SoundPlus hi-performance audio amp, OPA4131 Quad FET op amp, TL074 low cost JFET op amp (good for impedence matching)
[17:06:26] <Malinuss> OndraSter, I know hehe, it just not so cooperative around LUFA
[17:06:27] <OndraSter> TL072/074/082 are like the most used opamps everywhere :)
[17:06:41] <nevdull> yunta: do you have very rapid sampling periods/windows with your app?
[17:07:03] <OndraSter> AVR has got internal gain for differential measurements
[17:07:15] <nevdull> yeah i bought like 50 LM741's on ebay for $15 or something and use those a lot too
[17:07:28] <OndraSter> I wouldn't trust ebay opamps :D
[17:07:33] <OndraSter> to match their specifications
[17:07:35] <yunta> nevdull: your nick is hard to type. it's hard not to start with dev instead of nev :)
[17:07:39] <nevdull> hehe
[17:07:50] <nevdull> yunta: that's the secret ;)
[17:08:20] <yunta> I use tl074 now, but it can't see below 3V I guess, so I had to place shunt on high side
[17:08:27] <nevdull> i do answer to /dev/null, null, {d|n}ev, and "hey you with the pencil behind your ear"
[17:09:04] <OndraSter> what about yo dawg
[17:09:42] <yunta> OndraSter: I'm measuring 12V and 6V rails. no-can-do for xmega :(
[17:09:45] <nevdull> lol i haven't been called dawg since maybe the 1990's ;)
[17:09:50] <OndraSter> yunta, too bad
[17:09:58] <yunta> well, on low side
[17:10:01] <yunta> maybe
[17:10:07] <OndraSter> 12V is too lot for any regular MCU :D
[17:10:11] <OndraSter> too much
[17:10:24] <nevdull> yunta: you can put a voltage divider on it and sample with xmega and multiply by the dividor factor
[17:10:30] <OndraSter> aye
[17:10:31] <yunta> true
[17:10:43] <OndraSter> xmega has got 12bit ADC so if you get .1% resistors..
[17:10:49] <OndraSter> you can get fairly accurate readings
[17:10:52] <yunta> good enough
[17:11:10] <yunta> yeah, I think I'll go that route for power measurement
[17:11:16] <yunta> thanks guys (girls?)
[17:11:23] <OndraSter> hmm do we have any girls here?
[17:11:56] <nevdull> yeah, i've found (in playing around and making various resistor ladder adcs), that it's less important to have the resistors match a given resistance but more accurate to have all resistors within the same standard deviation
[17:12:22] <OndraSter> resistor ladder ADCs? With comparators?
[17:12:32] <OndraSter> aka backwards R2R DAC? :D
[17:13:01] <nevdull> well, it's not uncommon to put an op amp or comparator at the end of the adc for impedence but yah ;)
[17:13:14] <OndraSter> hehe
[17:13:50] <Malinuss> nevdull, so getting 3 different op-amp trough sampling is ok? canI I try to get more then that and expect to get it?
[17:13:52] * OndraSter sings how he does not need R2R DAC nor R2R ADC with xmega
[17:14:20] <nevdull> Malinuss: at ti.com? you can have 5 items with quanitity of 3 each per sample order.
[17:14:35] <OndraSter> hmm I remember getting 5 pieces of one SMPS controller
[17:14:37] <OndraSter> in one order
[17:14:37] <Malinuss> nevdull, and they don't bg check me?
[17:14:41] <nevdull> well, some higher priced items only allow 1 or 2
[17:14:54] <OndraSter> 31bit ADC :P
[17:15:17] <nevdull> Malinuss: nope. i get samples from ti.com, maxim, ST micro, NXP almost every week (i rotate which company i sample from every couple of weeks :)
[17:15:29] <Malinuss> oh okay
[17:15:33] <Malinuss> thanks for the tips
[17:15:33] <OndraSter> lol nevdull
[17:15:39] <nevdull> sure thing
[17:15:40] <OndraSter> ST denied my sample request
[17:15:46] <OndraSter> and they were for legit project
[17:15:46] <nevdull> no way? really?
[17:15:48] <OndraSter> so screw ST
[17:15:54] <OndraSter> I am not going to use their parts
[17:15:59] <OndraSter> it was $3 chip :(
[17:16:00] <nevdull> that sucks. i've never had that happen to me before.
[17:16:10] <nevdull> yah that's chump change to ST
[17:16:13] <OndraSter> 3band audio + i2c equalizer
[17:16:19] <Malinuss> lol you can't sample the TL072/074/082 op-amps :(
[17:16:25] <nevdull> ooh nice
[17:16:37] <OndraSter> I can't find any similar cheap filters..
[17:16:38] <OndraSter> audio filters
[17:16:42] <OndraSter> with i2c controll
[17:16:43] <nevdull> try TL071 or TL064
[17:16:43] <OndraSter> or SPI
[17:17:07] <OndraSter> only the one from ST :(
[17:17:11] <nevdull> oh i use the TI audio filter studio and build my own active spectrum filters
[17:17:18] <OndraSter> I need it as cheap as possible
[17:17:20] <Malinuss> nevdull, those neither - you have out-sampled them ;P
[17:17:26] <OndraSter> it was supposed to be small pocket equalizer with small OLED display :)
[17:17:27] <nevdull> lol@malinuss
[17:17:36] <nevdull> oh it's that OLED
[17:17:39] <OndraSter> aye, 1"
[17:17:43] <OndraSter> 128x64
[17:17:48] <nevdull> good resolution
[17:17:51] <OndraSter> aye
[17:17:58] <yunta> Malinuss: I think you got your personality mixed somehow. It was *me* looking for opams, not you :)
[17:18:02] <OndraSter> does TI have some audio cheap band filters?
[17:18:12] <nevdull> oh yah
[17:18:12] <Malinuss> like you can only get the TL072-EP but not the TL072... which is much more higher voltage :(
[17:18:26] <Malinuss> yunta, haha but I need them too, for my am-transmitter ;D
[17:18:28] <OndraSter> let's check audio section
[17:18:49] <nevdull> TBA820M, MAX4364, LM4819M, LM4871M, OPA2/4131
[17:19:00] <nevdull> TEA2025
[17:19:05] <nevdull> those are the ones i've smapled before
[17:19:56] <nevdull> and if you're really wanting to do some spectral analysis you can buy a cheap MSGEQ7 7-band filter at SparkFun for a few bucks
[17:20:14] <OndraSter> I don't want spectral analysis
[17:20:19] <OndraSter> I want audio pocket equalizer :)
[17:20:40] <Malinuss> I just sample some different types, I can worry about how they actually work later ;P
[17:21:23] <nevdull> OndraSter: check out TI's Filter Pro studio app. it's really useful for building low/high-band pass, notch filters, active/passive filters, and sourcing all the parts in one app
[17:25:16] <Malinuss> nevdull, are those chips normally DIP?
[17:26:21] <nevdull> Malinuss: mine are a mix of PDIP and SOIC8 or 16 (for dual/quad op amps). i like to use PDIP for prototyping and SOIC for my production boards
[17:26:37] <Malinuss> but you can't choose when you sample, can you?
[17:26:42] <nevdull> yes you can
[17:26:59] <nevdull> just scroll thru all the part numbers and look under hte 'package' column
[17:27:24] <nevdull> and select "Sample" from the row with IC in the package form you need
[17:28:02] <OndraSter> nevdull, http://www.st.com/internet/imag_video/product/75615.jsp
[17:28:03] <OndraSter> this is the chip
[17:28:59] <nevdull> oh wow, that's heavy duty
[17:29:07] <nevdull> looks like a full SoC
[17:29:17] <Malinuss> nevdull, oh thanks. So "Pkg D" would be DIP?
[17:29:18] <OndraSter> heh
[17:29:23] <OndraSter> no, just audio filters
[17:30:10] <nevdull> Malinuss: often D or N is PDIP, but to make sure i always pull down the PDF tech sheet at the top of the page and double check the model number with the package form in the datasheet to be sure
[17:30:28] <Malinuss> oh thanks, you've been helpfull ;D
[17:30:31] <OndraSter> yeah, ALWAYS TWICE CHECK PACKAGING
[17:30:37] <OndraSter> I have ordered some 0603 caps last time
[17:30:39] <OndraSter> but..
[17:30:40] <OndraSter> 0603mm
[17:30:43] <OndraSter> instead of 0603"
[17:30:48] <OndraSter> bloody imperialistic world!
[17:30:54] <OndraSter> so I got 0201 caps :D
[17:30:56] <Malinuss> lol
[17:31:12] <nevdull> OndraSter: check out the OPA2/4134 SoundPlus(tm) hi-performance audio amps. i got 8 of them in my parts bins all from samples.
[17:31:24] <Malinuss> not using SI units? in 2012, what is this? do you measure weight in stones too OndraSter ?
[17:31:32] <Malinuss> ;D
[17:31:41] <OndraSter> Malinuss, I live in the europe so I use the proper units
[17:31:44] <nevdull> oh those are small...i don't tend to go smaller than 0805
[17:31:53] <OndraSter> I use 0603 everywhere if possible
[17:31:59] <OndraSter> higher capacities/voltages require higher sizes
[17:32:03] <OndraSter> bigger*
[17:32:08] <nevdull> oh, and you're very welcome Malinuss ;)
[17:32:08] <OndraSter> thanks, will check out the 4134
[17:32:38] <OndraSter> oh
[17:32:40] <OndraSter> I know these
[17:32:43] <OndraSter> regular opamps
[17:32:47] <OndraSter> quite different type of "chip"
[17:32:54] <OndraSter> than what I want
[17:32:55] <Malinuss> nevdull, I've buyed their TI-83 back in 05, and it costs like 100x times more then it should. now it's PAYBACK TIME TI!!!
[17:33:15] <OndraSter> haha
[17:33:37] <nevdull> yeah they have an outrageously high slew rate, good supply range, i think around 8 or 10MHz bandwidth, and low noise (cuz they're a fET)
[17:33:47] <OndraSter> I don't need those properties
[17:33:53] <nevdull> Malinuss: bleed 'em dry, buddy ;)
[17:33:58] <OndraSter> I need something:
[17:34:01] <OndraSter> a) power efficient (batteries)
[17:34:05] <OndraSter> b) compact (pocket!)
[17:34:11] <OndraSter> c) "embedded"
[17:34:56] <nevdull> the OPA2134 can run from 2.5V to 18V with a 120dB open-loop gain in a PDIP or SOIC8 package
[17:35:32] <OndraSter> but how much space would take me to make separate 3band pass filter for both channels, then filter them out and then put back together AND DIGITAL CONTROL of those!
[17:36:09] <OndraSter> damnit I can't type anymore even
[17:36:22] <Tom_itx> better than the average irc'r
[17:36:36] <OndraSter> my grammar fell from the table and now is partially brain damaged
[17:36:49] <Tom_itx> if you spell your code wrong, do it everywhere and nobody cares
[17:37:00] <OndraSter> KERNAL! KERNALS EVERYWHERE!
[17:37:13] <OndraSter> I saw it on too many places.
[17:37:16] <nevdull> i built something very similar on a 50x70mm belt buckle (a portable spectrum analyzer on your belt buckle using SMD leds, a slew of op amps, and two stereo jacks for ipod ;)
[17:37:51] <nevdull> lol brain damaged grammar?
[17:37:51] <OndraSter> that is not that similar - I need to disassemble the signal into the bands, manually set their attenuation and then put it back together and let it shine into headphones
[17:37:55] <OndraSter> yes :(
[17:37:59] <Malinuss> " 8-Bit Serial-In/Parallel-Out Shift Register" rating: MILITARY ... w-w-what?
[17:38:10] <OndraSter> you can shoot it with a 9mm
[17:38:11] <nevdull> i suppose it's true that i'd rather have a bottle in front of me than to have a frontal lobotomy
[17:38:39] <OndraSter> meh I don't have the time for the project anyway right now
[17:38:47] <OndraSter> XBoard with custom USB CDC stack it is right now!
[17:38:47] <Tom_itx> sure you do
[17:38:50] <OndraSter> :D
[17:39:31] <Tom_itx> you're on the path of another here
[17:39:37] <Tom_itx> writing a usb stack n all
[17:39:46] <Tom_itx> maybe there's room in the xmega division for you
[17:39:48] <OndraSter> :D
[17:39:51] <OndraSter> I am so close!
[17:40:02] <OndraSter> in my test code every data received I send back
[17:40:08] <nevdull> OndraSter: they are sorta similar. in both examples we're taking two channels, breaking them into N-bands (i used 10), then recombining them back into two channels. :)
[17:40:08] <OndraSter> but PC picks up only the first packet :/
[17:40:19] <OndraSter> nevdull, well I can do it in 1 chip
[17:40:20] <Malinuss> OndraSter, so it works as a virtual com now?
[17:40:25] <OndraSter> Malinuss, since beginning
[17:40:31] <nevdull> OndraSter: wlel you have me sorely beat there :)
[17:40:50] <Malinuss> nice
[17:41:44] <Malinuss> I've did the same using LUFA today, am I skilled too?
[17:41:55] <OndraSter> that depends, how much code of the LUFA you wrote? :D
[17:42:01] <OndraSter> in LUFA-
[17:42:19] <Malinuss> I deleted a couple of lines, and added some. Does that count?
[17:42:33] <Malinuss> I changed some names too!!
[17:42:33] <OndraSter> were they in .c or in makefile?
[17:42:44] <Malinuss> both actually ;D
[17:43:05] <Malinuss> yeeeaaahh
[17:43:13] <OndraSter> how big is the binary image?
[17:43:19] <OndraSter> as in, how much flash does it occupy?
[17:43:44] <nevdull> OndraSter: seriuosly check out the MSEGQ7. it breaks out a stereo signal into 7 bands equally spaced automagically for yo
[17:44:27] <OndraSter> well, I need to fit into something like 5x5cm board tops :)
[17:44:28] <Malinuss> OndraSter, I think it was 11% of my atmega32u4... so some 1k's
[17:44:42] <OndraSter> 11% is actually over 3.3kB
[17:44:47] <nevdull> https://www.sparkfun.com/products/10468
[17:44:51] <Malinuss> no idea why so much
[17:45:15] <Malinuss> I tried really stripping it to the point where I could only use fputs to send strings trough it
[17:46:30] <OndraSter> nevdull, but I don't want the chip for showing me peaks
[17:46:57] <nevdull> right right, i see what you're saying
[17:48:04] <nevdull> i'm off for a little bit. nice chatting!
[19:27:06] <slidercrank> what capacity should a cap have connected to GND and Vcc pins of atmel microcontroller?
[19:35:31] <jadew`> 100nF
[19:46:09] <Tom_itx> it varies but generally .1uf is adequate. some controllers specify several values across them for better filtering
[19:46:29] <Tom_itx> i've seen up to 3 different values across each GND VCC connection
[20:48:39] <inflex> mmm
[20:48:53] <inflex> been a while since I ordered parts from E14.... nice to see they managed to deliver the next day as always :)
[20:49:34] * inflex isn't really going to start full production again, just trying to run out the last of my boards
[20:49:57] <inflex> postage to overseas is too insane now to make it worth sending anything
[20:49:59] <Tom_itx> don't kid yourself. you may get back into it
[20:54:29] <Casper> I still wants some 100 white leds, can't find them cheaply anymore on dx :(
[21:09:48] <inflex> Tom_itx: really don't want to ... design, sure, I love that... but not manufacturing
[21:19:09] <TechIsCool> alright so I have already own dallas 1 wire temp sensors. Should I buy TMP36 sensor or just use the dallas ones.
[23:17:48] <eadthem> so where would be getting started info for gcc C++ with avr 8bit
[23:18:02] <eadthem> for someone who is use to doing avr with asm
[23:20:13] <nevdull> eadthem: there are some good tutorial posts on avrfreaks re: using c++ on atmegas/tinys
[23:20:45] <eadthem> ya mostly need info on how IO works in c/c++
[23:22:04] <nevdull> well, you won't have cout, of course, but ports and registers should be familiar from your asm
[23:22:24] <eadthem> ya cout is trash in any system
[23:22:44] <eadthem> my big question is how easy are character LCD's via c or c++
[23:23:03] <nevdull> very easy.
[23:23:08] <eadthem> is there a library that gives printf or even echo type functionalty ?
[23:23:25] <eadthem> if there is then its time i moved to c/c++ for avr
[23:23:28] <eadthem> else ile stick with asm
[23:23:30] <nevdull> you can abstract the lcd or usart or other peripheral, overload the insertion/extraction operators (<< and >>) and use it basically like cout
[23:24:08] <eadthem> mmm ya dont supose there are normal IO operatuions avalable for that are there? ie not cin cout trash
[23:24:32] <eadthem> doubting printf is avalble
[23:24:37] <nevdull> none that i'm aware of, at least not in any standardized sorta way.
[23:24:42] <nevdull> printf is available but it's bloated
[23:24:50] <eadthem> ya i had a feeling
[23:25:17] <eadthem> a puts() putc() putchar() would be perfect
[23:25:23] <eadthem> or like
[23:25:51] <eadthem> also had feeling that cin cout would be 2x more bloated than printf as is on x86
[23:26:09] <nevdull> to what are you putting these bytes? there is a way to setup default streams so that you map, say, stdout/stdin to the USART or SPI or whatever
[23:26:18] <eadthem> tbh beond basic objects not even sure c++ is very useful for avr atm
[23:26:39] <eadthem> say a standard character lcd 4 or 8 bit interface
[23:26:57] <nevdull> i use c++ and asm on attinys to avr32uc3 and get good mileage with it. you just have to be smart about using it.
[23:27:25] <nevdull> namespaces pollute, templates are shunned, multiple inheritance is frowned upon, at least where heavy virtual tables are required
[23:27:45] <nevdull> do away with exceptions and rtti especially
[23:28:26] <nevdull> create guard methods for new and delete
[23:28:45] <eadthem> you can acctualy do new and delete
[23:29:31] <nevdull> yep, just watch your heap space with RAMEND or overload new to create in PROGMEM etc
[23:31:25] <eadthem> ok wow
[23:31:26] <eadthem> DDRA=0b11111111;//set all pins as output
[23:31:26] <eadthem> PORTA=0b10101010;//write ones and zeros to pins
[23:31:39] <eadthem> some reason i figured IO would not be that simple
[23:32:35] <nevdull> yah it's pretty straight forward
[23:33:34] <eadthem> So whats the best way to convert numbers?
[23:33:53] <eadthem> numbers to text for lcd output
[23:35:42] <nevdull> heh i use some custom asm routines to do hex/dec to char * but with an lcd like hd4480 or whatever you can just pulse the characters serially to, say, a parallel-out shift register to use the 4/8-bit interface without conversions
[23:36:32] <eadthem> ya its stuff like that that i was kinda hopeing libarys existed from atmel for
[23:37:40] <nevdull> there are several around for the most common lcds
[23:38:26] <nevdull> check out the new asf library. it might have some new goodies in it.
[23:45:18] <eadthem> dont know kinda was hopeing atmel would have a page or manual that would list everything there GCC librarys contained
[23:45:47] <inflex> there's the gcc/gnu page eadthem
[23:49:36] <eadthem> http://www.nongnu.org/avr-libc/user-manual/modules.html
[23:49:37] <eadthem> found it
[23:50:02] <inflex> Sorry, was going to get it for you, but had a customer turn up
[23:58:03] <eadthem> k think first thing ile do is make a short lib that can do the LCD work and convert numbers