#avr | Logs for 2013-12-28

Back
[00:00:09] <rue_shop3> whats sn76496
[00:00:50] <_sirpatrick> https://www.google.com/search?btnG=1&pws=0&q=sn76496
[00:00:55] <rue_shop3> Epsilon-Auriga_, what you use for audio generation?
[00:01:18] <rue_shop3> ooooh
[00:01:18] <avrrock> is that the atmel tutorial ?
[00:08:25] <N2TOH> who here has a laser cutter?
[00:09:21] <rue_shop3> I dont!
[00:10:03] <Casper> N2TOH: I have a laser on my circular saw, does that count? :D
[00:10:25] <N2TOH> LOL, sure!
[00:33:04] <GreaseMonkey> rue_shop3: in case you're still wondering: http://www.smspower.org/Development/SN76489
[00:33:27] <GreaseMonkey> (i emulate the sega master system variant which has a different noise pattern)
[00:35:12] <N2TOH> does the SMS have the clock line available on the CART interface?
[01:35:33] <AndChat-111201> can you please explain why: bit_is_clear(PIND, 2) works, but bit_is_clear(PORTD, 2) does not
[01:36:01] <AndChat-111201> I just spend hours trying to debug my sfotware and hardware just to come to the above conclusion, but I don't understand it at all
[01:36:06] <AndChat-111201> PORTD should be R/W
[01:36:58] <AndChat-111201> apparently __SFR_BYTE(PORTD) is always equal to 0xFF ???
[01:49:43] <GreaseMonkey> i get the idea that PORTD is meant to be for output not input
[01:50:57] <GreaseMonkey> N2TOH: afaik yes the clock is available on the cart, there's even a mapper which requires it
[01:52:24] <AndChat-111201> in general if I want to monitor for a pin change on INT0 and respond by changing some pin on PORTB, is it faster to implement an ISR or to poll the pin continiously for changes?
[02:23:40] <GreaseMonkey> AndChat-111201: probably faster to poll, interrupts have to push stuff ontot the stack
[02:25:11] <AndChat-111201> GreaseMonkey: that is what I thought, but I did some experiements with a logic analyzer and it seems interupts are faster by about 4 cycles
[02:25:21] <AndChat-111201> 14 (polling) vs. 10 (int)
[02:30:37] <GreaseMonkey> oh hmm
[02:31:58] <AndChat-111201> so here is my thinking so far
[02:32:09] <GreaseMonkey> if it's, say, pin C0, and you're outputting on, say, pin B5 you could possibly IN r16, PINC \ LSR r16 \ SBC r16, r16 \ ANDI r16, 0x20 \ OUT PORTB, r16
[02:32:20] <AndChat-111201> I have while(1){if(!PIN){do something}}
[02:32:46] <AndChat-111201> while(1) has overhead attached to it, if() has overhead, checkign the pin state has overhead, and negating it has overhead too!
[02:33:23] <AndChat-111201> so if I manually write in like a hundred if statements and if I didn't have to invert the pin state then polling just might be faster
[02:33:45] <GreaseMonkey> if you need it to be fewer cycles then you'd be best to code a tight loop in asm
[02:34:22] <AndChat-111201> GreaseMonkey: yea that would be faster, but what you are doing there is just cbi(PORTY, PINY) = rbi(PINX, PINX)
[02:35:46] <GreaseMonkey> lp_wait: SBIC PINX, PX \ RJMP lp_wait \ SBI PINY, PY
[02:35:53] <GreaseMonkey> something like that
[02:36:23] <GreaseMonkey> replace PIN\zs[XY] with the appropriate thing and P\zs[XY] with the appropriate number
[02:36:35] <GreaseMonkey> maybe even SBIS
[02:36:39] <GreaseMonkey> instead of SBIC
[02:36:58] <GreaseMonkey> SBIS is if you want to wait until it's 1, C wait until it's 0
[02:37:36] <AndChat-111201> interesting idea
[02:37:45] <AndChat-111201> that looks like it could be faster then interupts
[02:37:56] <GreaseMonkey> interrupts have a 4 cycle minimum
[02:38:10] <AndChat-111201> mine are 10, how do you reduce?
[02:39:01] <GreaseMonkey> well, int is 4, SBI/CBI is 2, RETI is 4
[02:39:09] <AndChat-111201> I am already doing an ISR(INT0_vect, ISR_NAKED)
[02:39:10] <GreaseMonkey> so your interrupt handler is as fast as it's ever going to get
[02:39:37] <GreaseMonkey> and well, i guess that's a nice reminder that C compilers do produce good code
[02:39:53] <AndChat-111201> GreaseMonkey: when I say I have 10 cycle overhead I mean the time between INT.0 going low and my ISR routine responding
[02:40:17] <GreaseMonkey> oh right, so in this case you'll probably get a 6-cycle overhead
[02:40:34] <AndChat-111201> I see my ISR routine responing because it toggles a pin high at the start, sbi, 2 insturctions, so there are still 8 cycles unacounted for between the even the ISR routine executing
[02:40:55] <GreaseMonkey> avr-objdump -S yourcode.out
[02:41:18] <AndChat-111201> mhh, how do I do that, I am working in atmel studio
[02:41:32] <GreaseMonkey> i don't know how to use atmel studio
[02:41:37] * GreaseMonkey != windows user
[02:42:06] <GreaseMonkey> if you have the avr-gcc/binutils suite it MIGHT come in handy but i suspect atmel studio might not use ELF as an intermediate format
[02:43:00] <AndChat-111201> I have .elf .eep, .d, .hex, .lss, .map, .o, .srec
[02:43:08] <AndChat-111201> I guess .o is what you call .out ?
[02:44:25] <GreaseMonkey> .out, .elf, .o are usually all in ELF format, i think the .elf is what you want to do an objdump on
[02:44:39] <GreaseMonkey> .o is usually for intermediate code before it gets linked into the final binary
[02:44:51] <AndChat-111201> looks like the .lss file is the ASM file
[02:44:54] <GreaseMonkey> the ELF object is then converted to IHEX which is what the .hex is
[02:45:04] <GreaseMonkey> yeah you should probably read the .lss then
[02:45:14] <GreaseMonkey> look for jumps near the start of the program
[02:45:57] <AndChat-111201> ya
[02:46:05] <AndChat-111201> 4: 0c 94 81 00 jmp 0x102 ; 0x102 <__vector_1>
[02:46:07] <AndChat-111201> like that?
[02:46:12] <GreaseMonkey> yeah
[02:46:29] <AndChat-111201> 00000102 <__vector_1>:
[02:46:29] <AndChat-111201> 102: 2a 98 cbi 0x05, 2 ; 5
[02:46:36] <GreaseMonkey> anyhow, the SBIC \ RJMP \ SBI loop would probably take up 3 cycles when the rjmp instruction is not skipped, and 2 when it it is, leaving another 2 for the SBI
[02:46:57] <GreaseMonkey> and is the next thing an IRET?
[02:47:19] <AndChat-111201> no, lds, out, ldi, dec, brne, in , sts....
[02:47:47] <AndChat-111201> http://pastebin.com/G6TQ71GJ
[02:47:53] <AndChat-111201> this is what vector1 looks like
[02:48:46] <GreaseMonkey> well the cbi is the first thing that happens, are you trying to clear that pin ASAP
[02:49:29] <AndChat-111201> https://www.dropbox.com/s/s16fduxhm0fchci/screenshot.png
[02:50:08] <GreaseMonkey> 7 cycle delay
[02:50:18] <AndChat-111201> that is a pic of my logic analyzer, when BSY 1 is INT0, INT0 is setup to trigger on falling edge
[02:51:08] <GreaseMonkey> but yeah, i think the worst case scenario for the sbic thing is, assuming the check happens at the start and the write happens at the end, maybe 1+3+2+2 = 8 cycles?
[02:51:13] <AndChat-111201> how do you count 7? I count 9 in that picture 11 cycles
[02:51:32] <GreaseMonkey> 10 cycles, derp, was reading the time instead
[02:51:37] <AndChat-111201> I mean I count 11 cycles, minus 2 for the cbi instruction, that is 9
[02:52:01] <GreaseMonkey> what clock speed is that?
[02:52:27] <AndChat-111201> 16 MHz, time between T1 and T2 is 687.5 nS
[02:52:49] <AndChat-111201> 1/16MHz is 62.5 nS
[02:53:06] <AndChat-111201> 687.5 nS / 62.5 nS = 11 (so eleven cycles)
[03:20:07] <AndChat-111201> GreaseMonkey: http://www.embeddedrelated.com/usenet/embedded/show/1437-1.php
[03:20:21] <AndChat-111201> "The processor first synchronizes the external input to it's own clock, that's takes at 2 clocks. The processor also has to finish the currently executing instruction. It takes 3 cyles to go the interrupt vector, from where it executes a jump to your ISR, another 3 cycles. This is 8 cycles to 11 cycles total time, depending on the executing instruction; or 0.6875 us."
[03:20:28] <AndChat-111201> that is almost exactly what I am seeing
[03:50:45] <megal0maniac> zlog
[04:34:43] <overrider> Depending on how i write my program code, avrdude will fail with a verification error. Why please? My program reads a button on PB2 and blinks some LEDs. Adding a simple delay function like in the pastebin will cause avrdude to report a content mismatch error. http://www.pastebin.ca/2519198
[04:38:13] <overrider> Actually it seems its as soon the main.c goes over a certain size - but its really not large, like 100 lines when it doesnt work, and 73 when it works fine. How weird.
[05:10:23] <jerkey> wow awesome, now all you need is a sodium iodide scintillator to make a nuclear analyzer http://spaz.org/~jake/jerryNucAnal5a.PDF
[06:41:56] <Fleck> hey - is it possible to perm. change osc calibration value?
[06:45:50] <uv> re
[06:47:46] <Fleck> and how do I tell variable type int size?
[06:48:34] <Fleck> http://gcc.gnu.org/wiki/avr-gcc << ok, this helps! :D
[07:02:21] <timemage> Fleck, you don't often need to do that between sizeof, INT_MIN/INT_MAX, and the leastN types.
[07:02:49] <Fleck> ?
[07:08:10] <timemage> Fleck, confused?
[07:08:15] <Fleck> yep
[07:08:37] <Fleck> I just need to know variable sizes
[07:08:43] <timemage> Fleck, because?
[07:08:48] <twnqx> Fleck: iirc you can't, i usually store the calibration values in the internal eeprom and set them from software on startup
[07:09:07] <Fleck> twnqx: ok, thx, makes sense :)
[07:09:28] <Fleck> timemage: to decide - what type to use... ;)
[07:09:42] <twnqx> just use the smallest that suffices
[07:09:58] <Fleck> yeah and how do I know - suffices or not? :)
[07:10:00] <timemage> Fleck, right. are you familiar with the fast and least int types?
[07:10:39] <Fleck> nope
[07:12:12] <Fleck> I need 24bit counter
[07:13:46] <Epsilon-Auriga_> http://www.nongnu.org/avr-libc/user-manual/group__avr__stdint.html
[07:14:17] <timemage> Fleck, you may like to read on them. the various integer types have minimum ranges in the standard. generally they're as last as they can be accomading that minimum range. the c99 standard introduce a bunch of typedef int_least16_t and int_fast16_t (for example) that chose the smallest and fastest representation for an integer of the given width they're not required to provide them for 24bit types, but there mightbe one on avr. least_int32_t may
[07:14:17] <timemage> be.
[07:14:50] <timemage> Fleck, s/as last as/as fast as/
[07:15:28] <timemage> Fleck, "accommodating". sorry for the mangling.
[07:16:04] <Epsilon-Auriga_> use uint32_t...that will give you a 32bit unsigned integer.
[07:16:50] <timemage> using the least type is slightly more portable.
[07:18:17] <Epsilon-Auriga_> or you could do an array of 3 characters and deal with the 3 bytes yourself directly to get 24 bits...that's messier though.
[07:20:57] <Fleck> I think I have one spare byte :D so I'll use uint32
[07:22:29] <Fleck> thank you guys! You are helpful!
[07:52:44] <Arrowsmaster_> hi there!
[07:55:35] <twnqx`> sorry, i am not shopping with arrow. only digikey and mouser.
[07:56:03] <Arrowsmaster_> :D is it possible to use hardware SPI in FatFs?
[08:01:59] <carabia> um. doesn't it use it?
[08:02:13] <carabia> or wait
[08:02:25] <carabia> heh. tired. Arrowsmaster_ yeah should be possible?
[08:02:44] <carabia> aren't there like example codes with fatfs using hw spi
[08:02:57] <carabia> sounds like sd cards
[08:02:58] <Arrowsmaster_> it uses bitbanging (software spi)
[08:03:06] <carabia> oh it does? rofl
[08:03:10] <Arrowsmaster_> yea, SD cards indeed
[08:03:28] <Arrowsmaster_> yes it does and it sucks :P
[08:03:30] <carabia> i only glanced at fatfs and deemed it bloat
[08:03:42] <carabia> so i wrote some of my own... not too hard
[08:03:47] <Arrowsmaster_> i'd like to use h/w spi but i dont know how to re-write some funcs
[08:04:20] <carabia> like?
[08:04:21] <Arrowsmaster_> i will stick to farfs :)
[08:04:52] <Arrowsmaster_> like xmit_mmc();
[08:05:14] <Arrowsmaster_> it justs sends bytes via spi
[08:05:17] <carabia> need a pastebin or something...
[08:05:30] <Arrowsmaster_> copy that base :D standby ;p
[08:05:52] <carabia> affirmative
[08:06:50] <Arrowsmaster_> http://pastebin.com/JqTPyvW3
[08:07:58] <Arrowsmaster_> line [15, 31] - i tried with this:/* Start transmission */
[08:07:59] <Arrowsmaster_> SPDR = cData;
[08:08:01] <Arrowsmaster_> /* Wait for transmission complete */
[08:08:03] <Arrowsmaster_> while(!(SPSR & (1<<SPIF)))
[08:08:04] <Arrowsmaster_> ;
[08:09:01] <carabia> yes, and?
[08:10:08] <carabia> seems legit if cData is the byte
[08:10:20] <twnqx`> unless spi hasn't been set up before
[08:10:34] <carabia> yes true
[08:11:11] <carabia> ...and then you read the response by sending 0xff to the spi bus
[08:11:37] <carabia> 0xff's
[08:11:41] <carabia> ";D"
[08:12:15] <twnqx`> sigh, i need half a GB of free disk space
[08:12:20] <twnqx`> to uninstall avr studio 5.1 ...
[08:12:21] <Arrowsmaster_> well i dont really know what im doing, all i know is that i have to replace this with hardware spi commands
[08:12:46] <twnqx`> but that only works with arduino, not real avr programming :P
[08:12:58] <carabia> Arrowsmaster_: that's basically how you send data over the spi bus
[08:13:15] <carabia> so you can make a function that sends a byte
[08:13:30] <Arrowsmaster_> wait, wait
[08:14:43] <carabia> yes?
[08:14:45] <Arrowsmaster_> http://pastebin.com/T3cZjBBf
[08:14:51] <Arrowsmaster_> am i right?
[08:16:40] <carabia> test it
[08:17:00] <carabia> but i suspect it will get complicated
[08:17:16] <Arrowsmaster_> but is it okay if i leave the other part as it is now?
[08:17:34] <carabia> should be okay
[08:18:10] <Arrowsmaster_> what about the SPI config? where should i put it?
[08:19:46] <carabia> before it inits the card
[08:19:48] <carabia> but
[08:20:06] <carabia> i'm not sure how the bitbanging is implemented and will it interfere with the hw spi...
[08:20:57] <carabia> maybe someone has done this before, i suggest you search around. Or if not, then if i were you i'd write my own or use a lib with hw spi
[08:22:13] <Arrowsmaster_> or i can replace everything with hw spi and then test
[08:22:26] <carabia> or that
[08:22:48] <carabia> there shouldn't actually be too much that you need to replace
[08:22:54] <Arrowsmaster_> yea
[08:23:12] <carabia> actually if it's done clever, i think the only thing you would have to replace is the write function
[08:24:36] <Arrowsmaster_> btw: SD pin "DO" goes to uC pin "MISO", correct?
[08:26:59] <carabia> that's some breakout or?
[08:27:09] <carabia> I think so. I never learned the labels of them, i just had a pinout :D
[08:27:19] <Arrowsmaster_> :D ok, testing now ;p
[08:27:32] <Arrowsmaster_> and thank you for your time
[08:30:02] <carabia> I dropped a pullup on miso
[08:30:35] <carabia> not sure if it's absolutely necessary
[08:33:36] <Arrowsmaster_> ive got it too
[08:34:29] <carabia> and um
[08:34:34] <carabia> the part where it gets tricky is
[08:34:45] <carabia> for the init you need to have the spi bus on low freq
[08:35:00] <Arrowsmaster_> oh, btw
[08:35:04] <carabia> and after you have the card in the readystate then you can push higher freq
[08:35:44] <Arrowsmaster_> oh, nvm :)
[08:37:04] <carabia> so you need to meddle with the register
[08:39:04] <carabia> i believe the freq for the init had to be somewhere in the kilohertz range...
[08:39:19] <carabia> i think i had something like 125 or 400 or the likes
[08:41:11] <Arrowsmaster_> i think you mean I2C
[08:41:47] <carabia> no it's the card itself
[08:42:16] <carabia> the clock
[08:42:17] <Arrowsmaster_> lol i thought that SPI here is >1MHz
[08:42:55] <carabia> yeah it is
[08:43:09] <carabia> but for the initialization of the card you need it lower. at least i did.
[08:43:27] <carabia> once you have the card ready then you can ramp up the clock speed
[08:43:30] <carabia> of the bus
[08:43:32] <Arrowsmaster_> so its not really worth using hw spi with sd card
[08:43:49] <carabia> how so. the same procedure would apply to bitbanging
[08:44:41] <Arrowsmaster_> i thought i can get faster reading/writing
[08:45:36] <carabia> um. well what was it. You can get up to fsck/2 or /4, can't remember?
[08:46:07] <carabia> so it's quite fast alright
[08:46:47] <carabia> but the initialization, is a procedure which prepares the card to accept data. First you basically tell it to ready itself, it goes to sleep, and then you wake it up
[08:46:52] <Arrowsmaster_> ist the bitbanging in FatFs already at its max speed?
[08:47:03] <Arrowsmaster_> isnt*
[08:47:18] <carabia> I don't know. Could be
[08:48:15] <carabia> i'd imagine you'd have slightly faster transfers using the hw spi
[08:48:19] <Arrowsmaster_> hmm it says " * Low Speed - The data transfer rate will be several times slower than hardware SPI."
[08:48:33] <Arrowsmaster_> "several times" :p
[08:48:50] <carabia> yeah, i have no idea
[08:49:09] <Arrowsmaster_> i think its not worth it
[08:49:28] <carabia> depends, what's your application?
[08:50:57] <Arrowsmaster_> sd card is supposed to store python script images for a virtual machine that runs on AVR
[08:51:44] <carabia> okay, what's the application in a broader sense?
[08:52:19] <Arrowsmaster_> some kind of smartphone/tablet
[08:52:38] <carabia> okay :)
[08:52:58] <Arrowsmaster_> based on sim900 :P
[08:53:25] <carabia> well it all comes down to how much data you'll be moving
[08:54:33] <carabia> it's really easy to make read-functionality from scratch for fat16/32... it's the writing part that gets tricky
[08:54:49] <carabia> that is if you're concerned about the speed
[08:55:01] <carabia> in any case i gotta run now. Sorry i couldn't be of much help
[08:55:21] <Arrowsmaster_> i dont even know if fatfs really has SDHC support
[08:55:27] <Arrowsmaster_> alrighty, thanks a lot :)
[08:56:27] <carabia> sdhc only has a few minor differences
[08:57:12] <carabia> first of all in the init you need to tell the card that "host supports sdhc"
[08:57:39] <carabia> and as for reading, you read one sector at a time whereas in the standard sd you specify the length
[08:58:27] <Arrowsmaster_> im an engineer, i dont want to know how it works. i just want it to work :D
[08:59:27] <carabia> sounds like all the software tech engineers out there :D
[09:00:09] <carabia> but i'm really running late ;c sorry gtg
[09:00:30] <Arrowsmaster_> cya!
[09:00:35] <Epsilon-Auriga_> if you want high speed from an SD card you don't use the spi interface for data transfer as I recall.
[09:00:47] <carabia> shi-
[09:01:08] <carabia> Epsilon-Auriga_: yep but the high speed transfer mode is commercial and you really can't find examples for it
[09:01:10] <Arrowsmaster_> i know that
[09:01:16] <Epsilon-Auriga_> yup.
[09:01:19] <carabia> ->
[09:01:43] <Arrowsmaster_> all i want is to have FatFs running with HW SPI, period
[09:02:26] <Epsilon-Auriga_> so, make it happen.
[09:03:03] <Arrowsmaster_> well... im trying
[09:10:11] <Epsilon-Auriga_> aahh..I see now why fatfs is done bitbanged.
[09:10:15] <Epsilon-Auriga_> it is multiplatform.
[09:13:52] <solexious> Anyone had this issue? When I run avrdude it hangs for a good 30 seconds - a min, then runs normally. I'm using a tinyisp on a tiny13 on its own
[09:14:38] <solexious> on ubuntu
[09:16:20] <Arrowsmaster_> Epsilon-Auriga_: yeah it is
[09:16:48] <Epsilon-Auriga_> solexious: you might check in #avrdude on that one.
[09:16:59] <Epsilon-Auriga_> or not.
[09:17:02] <Epsilon-Auriga_> guess it's not there.
[09:21:54] <solexious> cool, will have another google, pita as I have 100 ic's to program today :/
[09:30:16] <Epsilon-Auriga_> the fatfs site has some samples...one of them has hardware spi for avr.
[09:30:30] <Epsilon-Auriga_> http://elm-chan.org/fsw/ff/ffsample.zip
[09:30:41] <Epsilon-Auriga_> it's under avr complex
[09:44:24] <Arrowsmaster_> yeah i have noticed
[09:46:11] <Arrowsmaster_> its too difficult for me :P
[10:01:35] <carabia> :( i was so chuffed about finally founding a nice little dac+amp to go with my s4. turns out i'll have to wait weeks
[10:02:21] <carabia> founding, err, finding
[10:06:52] <Tom_itx> http://www.digikey.com/product-search/en?pv7=2&k=LTC1669CS5%23TRMPBF&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25
[10:07:03] <Tom_itx> i got a couple of those but appears they're non stock now
[10:09:05] <carabia> nah i mean a finished product... uses some ti dacs and ti opamps
[10:10:27] <Tom_itx> finish one then you'll have a finished product
[10:10:30] <carabia> all in all i couldn't design it pretty enough to be carried around!
[10:10:40] <carabia> i have some cirrus dacs... but no amps at hand
[10:10:59] <Tom_itx> put it in a little box
[10:11:35] <carabia> yeah spose
[10:12:05] <carabia> and i'd need to take apart my lithium packs and use a step up to 5v
[10:12:48] <carabia> but, stock on the fiio e18 will be up on jan 2nd in a UK store so i'll just... be lazy :)
[10:13:41] <Tom_itx> http://www.digikey.com/product-search/en?x=19&y=12&lang=en&site=us&KeyWords=1551FTBU+
[10:14:52] <carabia> with the holidays and all it'll be probably as fast to get the whole pack from UK than order parts from digikey
[10:15:07] <Tom_itx> others stock it
[10:15:16] <carabia> alright
[10:15:17] <Tom_itx> http://www.hammondmfg.com/1593HAM.htm
[10:15:22] <Tom_itx> you can get one for your pi too
[10:15:24] <Tom_itx> or bbb
[10:15:25] <Tom_itx> etc
[10:15:47] <carabia> don't have any i'm afraid
[10:16:04] <Tom_itx> better add that to your list too then :D
[10:18:25] <carabia> pis and shit?
[10:18:38] <carabia> i don't have any use for such nonsense
[10:20:12] <carabia> i don't get such huge kicks as others seem to do out of blinking leds with python...
[10:20:44] <Tom_itx> just with dacs?
[10:20:53] <Tom_itx> and noises
[10:20:59] <Tom_itx> no light show to go along with it?
[10:21:00] <megal0maniac> carabia: What about http://www.kickstarter.com/projects/214379695/micro-python-python-for-microcontrollers
[10:21:42] <carabia> i stopped at hello my name is damien and i look like (and am) a software engineer
[10:22:19] <carabia> Tom_itx: yeah. and a laser diode light show synchronized to that
[10:22:38] <Tom_itx> can't wait to see the youtube
[10:22:49] <carabia> there will be blindness
[10:23:50] <carabia> i'll be using some diodes out of cnc machines
[10:23:59] <carabia> the good stuff
[10:24:45] <Tom_itx> you have a cnc?
[10:25:27] <carabia> not me. some people. apparently now they don't cause i nicked the diodes
[10:25:48] <Tom_itx> teach them to let you near it
[10:26:08] <Tom_itx> i think it would be cool to have a cnc laser
[10:26:17] <Tom_itx> all i have is a mill
[10:26:32] <carabia> but it's even cooler to have a laser disco ball
[10:26:41] <Tom_itx> pfft
[10:26:58] <Tom_itx> those went out in the 80's
[10:26:59] <carabia> i'm sure it will go ballistic when i play some skrillex through it
[10:27:29] <carabia> i guess it will cut nice patterns in the walls
[10:31:02] <_1_0_0_0> Where can I buy 666 timer?
[10:31:40] <Arrowsmaster_> xD
[10:31:59] <carabia> :D
[10:32:12] <MrMobius> megal0maniac, i hesitate to ask but what is the big deal about python on mcus? it seems like an aweful idea
[10:32:43] <carabia> because python is cool amongst software devs yeah?
[10:32:55] <_1_0_0_0> There's 555 timer, so 666 must be improved version of it.
[10:33:12] <carabia> one of these days it will be micro nodejs and so on mcus
[10:33:38] <carabia> "hello my name is damien and this is nodejs for mcu"
[10:33:55] <carabia> "you can blink leds with it super efficiently, i swear"
[10:34:30] <megal0maniac> MrMobius: It's more the fact that he re-wrote python. And it's a comfortable scripting language that works pretty efficiently on that particular set-up
[10:35:02] <carabia> and it begs the question what does it do that c doesn't?
[10:35:17] <carabia> most likely more efficiently
[10:35:58] <Arrowsmaster_> are you talking bout programming avrs using python? lol
[10:36:26] <MrMobius> i understand it is "comfortable" but why would i use it if it runs at least 10x slower than C and needs 60k of flash for a runtime?
[10:36:33] <megal0maniac> Arrowsmaster_: Well that does happen https://github.com/abcminiuser/gdp
[10:36:37] <MrMobius> and also takes up more ram
[10:36:40] <_1_0_0_0> You need to write a python int3rpr3t3r for avr, don't you?
[10:36:55] <MrMobius> its like taking your mcu specs and dividing them
[10:37:33] <megal0maniac> MrMobius: I wouldn't use it in production, but I think it's a cool concept. Don't see any practical use for it personally. I'd use it to learn Python
[10:39:13] <carabia> im thinking javascript on mcu is good
[10:39:20] <carabia> let's do that
[10:39:26] <MrMobius> It seems like an awful idea to me. (Im not trying to troll.)
[10:39:48] <megal0maniac> carabia certainly is :P
[10:39:51] <Arrowsmaster_> http://code.google.com/p/python-on-a-chip/
[10:40:01] <MrMobius> carabia, we might as well make visual basic 6 for it too :/
[10:40:04] <MrMobius> :P
[10:40:19] <carabia> "Requires roughly 55 KB program memory"
[10:40:21] <carabia> :D
[10:40:24] <Arrowsmaster_> xD
[10:40:29] <Arrowsmaster_> but its a VM
[10:40:33] <carabia> yeah it is
[10:40:38] <carabia> awesome
[10:41:03] <carabia> megal0maniac: no-one has done that yet, no?
[10:41:43] <carabia> Initializes in 4KB RAM; print "hello world" needs 5KB; 8KB is the minimum recommended RAM.
[10:41:55] <carabia> hella efficient hello worlding
[10:42:31] <carabia> megal0maniac: so let's do it. call it avr.js?
[10:42:35] <MrMobius> thats ridiculous
[10:42:36] <Arrowsmaster_> hahahaha
[10:43:01] <MrMobius> the farthest i would go is a python to C converter so you dont need a VM
[10:43:33] <Arrowsmaster_> are there any VM for C?
[10:44:11] <Arrowsmaster_> >The NanoVM is a java virtual machine for the Atmel AVR ATmega8 CPU
[10:44:13] <carabia> https://code.google.com/p/talktic/wiki/HowToBuildOnMacOSX
[10:44:14] <Arrowsmaster_> what XD
[10:44:16] <carabia> really...
[10:44:41] <MrMobius> why would you want a VM for C?
[10:44:51] <carabia> :D
[10:45:23] <carabia> i'm all out of ideas what's more ridiculous than a javascript vm on an avr
[10:45:44] <MrMobius> you could make a PIC VM for AVR so that you can run compiled code without having to recompile
[10:47:28] <carabia> or you can make a javascript vm inside a python vm on an avr
[10:47:45] <Arrowsmaster_> so i can drive while i drive, dawg
[10:48:13] <Arrowsmaster_> i herd you like VMs xD
[10:48:36] <carabia> oh, and must include arduino somewhere in there
[10:49:35] <carabia> a js vm on top of python vm on top of arm vm on top of x86 vm on an avr
[10:50:05] <carabia> and hello, my name is damien...
[10:52:02] <Arrowsmaster_> xD
[10:56:16] <Arrowsmaster_> http://www.clifford.at/embedvm/
[10:56:19] <Arrowsmaster_> this looks good
[10:56:20] <_1_0_0_0> What is something between attiny and its big brother?
[10:57:05] <carabia> tiny mega xmega err?
[10:57:10] <Epsilon-Auriga_> carabia: there is a C interpreter...picoc
[10:57:26] <Epsilon-Auriga_> _1_0_0_0: atmega.
[10:57:58] <_1_0_0_0> I don't want that much pins
[10:58:02] <_1_0_0_0> many
[10:58:08] <carabia> how many do you want then
[10:58:35] <twnqx`> 8pin arm from NXP :]
[10:58:48] <carabia> twnqx`: does it run javascript
[10:59:22] <megal0maniac> attiny841
[10:59:55] <Epsilon-Auriga_> _1_0_0_0: atmega8....28 pins
[11:00:16] <twnqx`> carabia: probably
[11:00:42] <carabia> 8 pins is more than enough then. i can blink 4 leds probably
[11:01:06] <carabia> wonder if i have to use transistors...
[11:04:22] <Arrowsmaster_> but... will it blend?
[11:04:52] <carabia> even the 3310 blended... :( blendtec too stronk
[11:05:22] <carabia> i always liked 3210 better though
[11:10:01] <megal0maniac_afk> carabia: You won't need transistors to blink LEDs with the attiny841
[11:13:20] * twnqx` has two chains of 50 RGB LEDs with ws2811 conrollers on a small at :P
[11:13:26] <twnqx`> only using ONE pin!
[11:22:13] <carabia> megal0maniac_afk: no the 8 pin arm
[11:22:39] <Tom_itx> _1_0_0_0 you mean a 556?
[11:22:52] <_1_0_0_0> ?
[11:23:02] <Epsilon-Auriga_> 556 is a dual 555
[11:23:09] <Tom_itx> at least get the numbers right if you're gonna troll
[11:23:10] <carabia> 666 > 555
[11:23:32] <Tom_itx> gawd, they don't even have decent trolls anymore
[11:23:49] <Epsilon-Auriga_> 666 is a satanic processor
[11:23:56] <carabia> timer
[11:24:05] <_1_0_0_0> http://imgs.xkcd.com/comics/circuit_diagram.png
[11:24:08] <carabia> 666 is the doomsday clock
[11:24:15] <_1_0_0_0> Look at the upper right corner
[11:24:17] <avrrock> on avrbeginners.net , the link on top of the page
[11:24:27] <avrrock> is it good for 8 bits and 32 bits avr ?
[11:24:31] <_1_0_0_0> xkcd must be serious. There MUST BE a 666 timer!
[11:24:58] <Tom_itx> don't think i'll click on one of your links
[11:25:04] <Tom_itx> you haven't proven worthy yet
[11:25:43] <Epsilon-Auriga_> it's a very old pseudo-humor circuit diagram.
[11:25:46] <Tom_itx> now, if you're trying to decide on an avr i suggest you look at the parametric table on atmel's site
[11:25:58] <carabia> you have been weighed you have been measured and you have been found wanting
[11:26:37] <Tom_itx> otherwise sit down, shut up and enjoy the ride
[11:26:47] <carabia> i really want one of those batman components
[11:26:56] <Tom_itx> oh and buckle your seatbelt, it could get bumpy
[11:41:51] <Tom_itx> twnqx` did you put those leds on your christmas tree and blink random patterns?
[11:42:02] <twnqx`> no... window
[11:42:08] <twnqx`> but random patterns, yes :3
[11:42:30] <dsaj> does anyone know that if atmel publishes eagle cad schematics (.sch) of the development kits?
[11:42:41] <Tom_itx> no
[11:43:03] <Tom_itx> why would they want somone else publishing their products?
[11:44:01] <dsaj> on the user manual there are schematics on the development kits
[11:44:10] <dsaj> *of
[11:44:16] <Tom_itx> but not a ready made file
[11:44:35] <Tom_itx> it's a fine line to publish a schematic of a product
[11:44:40] <dsaj> isn't the purpose of the development kits to use them as a reference project and modify for your own need
[11:44:55] <twnqx`> either way they wouldn't use eagle
[11:45:04] <Tom_itx> i suppose so, but i haven't found the source for their xplain yet
[11:46:07] <Tom_itx> it's like.. look what it can do, now you go figure out how to do it
[11:46:46] <Tom_itx> he was underqualified
[11:49:59] <aep> is there an avr with a usb-device side that lets me implement a usb audio device?
[11:50:11] <Tom_itx> i suppose so
[11:50:17] <aep> how would i find it?
[11:50:30] <Tom_itx> look at the LUFA library
[11:50:32] <aep> probably avr32 have them, but those barely have any ready code
[11:50:46] <aep> nice
[11:50:53] <Epsilon-Auriga_> any of the USB AVR chips could do it...it's a matter of implementing the usb audio protocol.
[11:50:54] <Tom_itx> you would need an external dac
[11:51:18] <Tom_itx> pretty sure dean has an audio example
[11:51:28] <aep> yeah, i've got the dac.
[11:51:44] <aep> this lib looks awesome
[11:52:10] <dsaj> could anyone tell me the meaning of the symbol inside the red circle http://oi43.tinypic.com/6td3c6.jpg
[11:52:26] <Tom_itx> Audio In Device
[11:52:26] <Tom_itx> Audio In Host
[11:52:26] <Tom_itx> Audio Out Device
[11:52:26] <Tom_itx> Audio Out Host
[11:52:47] <Tom_itx> there you go
[11:52:50] <Epsilon-Auriga_> dsaj: it is 2 things.
[11:52:55] <Epsilon-Auriga_> a capacitor and a ground.
[11:53:07] <Epsilon-Auriga_> likely an electrolytic capacitor even.
[11:53:33] <dsaj> Epsilon-Auriga_, thank you
[11:53:38] <Tom_itx> no + - designation
[11:53:42] <Tom_itx> may not be
[11:53:58] <Epsilon-Auriga_> Tom_itx: no, but, it does have a | and ]
[11:54:12] <Epsilon-Auriga_> so might be polarized...not necessarily though.
[11:54:50] <Epsilon-Auriga_> from the schematic I would say it's a bypass cap.
[11:55:04] <dsaj> why would you need capacitor on the VDD connection?
[11:55:09] <Tom_itx> from the schematic i would say it's a filter cap for the regulator
[11:55:31] <Epsilon-Auriga_> hmm..yeah...didn't notice that was a regulator.
[11:55:43] <Epsilon-Auriga_> dsaj: to smooth out power fluxuations.
[11:55:56] <Tom_itx> dsaj because it's an unregulated input to a regulator
[11:56:27] <Tom_itx> left off the regulator could go into oscillations and pop like a popcorn kernel
[11:56:37] <Tom_itx> just don't eat it
[11:56:43] <dsaj> :)
[11:56:49] <dsaj> thanks for the explanation
[11:56:56] <Epsilon-Auriga_> betting that's a regulator inside a chip or something from the other connections...like a cpu or fpga.
[11:57:46] <dsaj> it is a schematic to show how to power sam3u processor
[11:58:09] <dsaj> I am trying to understand it :)
[12:02:59] <Epsilon-Auriga_> might I suggest learning basic schematics first?
[12:04:08] <dsaj> yeah but I already bought the development kit
[12:04:42] <carabia> that's a good reason not to learn anything
[12:04:54] <Arrowsmaster_> lol
[12:04:55] <carabia> this is often referred to as "the beaky approach"
[12:06:20] <Epsilon-Auriga_> trying to run before learning to crawl will often lead to a broken leg....
[12:06:24] <dsaj> my last question, here is the powering schematic of the board http://oi40.tinypic.com/sy2hhi.jpg there is +5V symbol on the left right schematic. does it mean that here is where it gets the 5V? J11 is the power connector (5V, in) MN13 is a circuit protector, MN14 is a EMI filter
[12:08:02] <Tom_itx> the left or the right?
[12:08:10] <dsaj> does the flat line on 5V symbol mean different than the arrow symbol ( >)
[12:08:41] <dsaj> upper left one when you look through the screen
[12:09:10] <Epsilon-Auriga_> looks like that is connected to the 5V rail in the circuit..
[12:09:24] <Epsilon-Auriga_> assuming that is a voltage regulator then that is regulated 5V output.
[12:09:24] <Tom_itx> it means all those +5v are connected
[12:10:09] <dsaj> now it makes sense to me, thanks a lot
[12:10:18] <Epsilon-Auriga_> aahh..the bnx002-01 is a ferrite bead.
[12:10:25] * Tom_itx gives dsaj an attiny10 to play with until he learns a bit more
[12:11:08] <Arrowsmaster_> Epsilon-Auriga_: damn, i really like your nick
[12:11:21] <Epsilon-Auriga_> thanks...picked it randomly from a list of star names.
[12:11:33] <Arrowsmaster_> quite an interesting star
[12:11:35] <Epsilon-Auriga_> it means old goat or something like that.
[12:11:43] <Arrowsmaster_> misspelled on purpose? ;p
[12:11:58] <Epsilon-Auriga_> different spellings out there..this one kinda just fit.
[12:12:09] <twnqx`> why do you all have _ at the end of your nicks? is that a new trend?
[12:12:09] <Epsilon-Auriga_> should be Aurigae or some such I suppose.
[12:12:29] <Epsilon-Auriga_> my irc client does it when I reconnect and my old nick is still online..I just didn't change it.
[12:12:49] <Tom_itx> twnqx` some ppl use that as an 'away' indicator
[12:12:51] <Epsilon-Aurigae> there, better
[12:12:55] <twnqx`> ... i guess that's how i got my `
[12:12:58] <Tom_itx> or their client does it for them
[12:13:17] <Arrowsmaster_> yeah, because of the declension in latin
[12:13:17] <Duality> I want to toggle a rts pin on a serial device how would i do that ?
[12:13:42] <Epsilon-Aurigae> Duality: all depends on the device.
[12:15:33] <Epsilon-Aurigae> I should go do something productive...it's a pretty day here for a change.
[12:15:46] <Epsilon-Aurigae> supposed to be up to 50F today...and a high of 0F tomorrow.
[12:16:38] <twnqx`> what's that in normal measures?
[12:18:29] <carabia> dsaj: the circuit is self explanatory, simple enough, you add a jumper wire between +5v and ground in the voltage reg
[12:18:40] <aep> oh. lufa is only usb1.0 :/
[12:18:58] <Duality> Epsilon-Aurigae: it's an bluetooth serial
[12:19:21] <carabia> or well, this can be done in any part of the circuit. voltage reg is just a preferable choice since the pins are closeby
[12:19:40] <Casper> aep: yes, don't expect more than that on bit banged usb
[12:20:03] <aep> right. i guess i need a separate chip, but then it wont support composite :/
[12:52:29] <megal0maniac_afk> Casper: LUFA is native USB only
[12:52:33] <megal0maniac_afk> You're thinking of VUSB
[12:52:50] <TLoFP1> hi guys, I have a problem with for loops and _delay_us() not completing. I suspect it is becuase they are interupted by an ISR and then somehow reset back to zero, but I don't understand this at all. how could I work around it?
[12:57:14] <avrrock> light up a led after _delay_us()
[12:57:51] <TLoFP1> avrrock: that is what I am doing, well, I toggle an IO Pin and hook it up to a logic analyzer
[12:58:27] <TLoFP1> it seems that once the for loop, or delay takes longer to execute then the time between interupts then it never finishes!
[12:58:28] <Epsilon-Aurigae> inside the () for _delay_us() are you using a variable or constant?
[12:58:34] <TLoFP1> constant
[12:58:45] <TLoFP1> I code like this "_delay_us(100);"
[12:59:13] <aep> megal0maniac_afk: wouldnt that mean it should run at whatever speed the chip usb thing does?
[13:00:02] <megal0maniac_afk> aep: Yes. You're at the mercy of the transceiver
[13:00:25] <avrrock> TLoFP1: have you checked the code for _delay_us ?
[13:00:27] <megal0maniac_afk> The point is that "don't expect more than that on bit banged usb" is wrong
[13:00:59] <TLoFP1> avrrock: yes it uses __builtin_avr_delay_cycles()
[13:01:07] <aep> yeah, i got that
[13:01:16] <avrrock> TLoFP1: have you checked the code for that function ?
[13:01:19] <TLoFP1> I tried to do __builtin_avr_delay_cycles(1000) too but that also fails above a magic number of cycles
[13:01:19] <avrrock> ^^
[13:01:28] <TLoFP1> I could not find code for __builtin_avr_delay_cycles()
[13:01:30] <aep> just wondering how much speed i'll get out of an avr with lufa. i dont think usb1 is going to be enough
[13:02:19] <Epsilon-Aurigae> is that usb1.0 a limitation of the software or the hardware?
[13:02:22] <Casper> megal0maniac_afk: lufa isn't the bitbanged? woops
[13:02:39] <megal0maniac_afk> aep: Well the 32u4 has "USB2 full speed"
[13:02:53] <aep> whatever this thing is, it says it has usb 2.0 http://www.minimususb.com/
[13:03:26] <Casper> TLoFP1: your reset issue is probably because you ran out of ram
[13:03:26] <megal0maniac_afk> That's a 32u2
[13:03:26] <aep> full speed is usb1
[13:03:32] <megal0maniac_afk> No
[13:03:47] <megal0maniac_afk> USB2.0 Full Speed
[13:04:07] <megal0maniac_afk> Which is 12mbps, but still USB2.0
[13:04:10] <aep> right. i mean, full speed was supported by usb1
[13:04:12] <aep> yeah
[13:04:22] <TLoFP1> Casper: how can I test that?
[13:04:34] <Casper> good luck
[13:04:35] <aep> i think i need like 50-100
[13:04:51] <megal0maniac_afk> Which will be plenty for audio
[13:05:04] <Casper> seriously, there is no real way to test beside taking out your pen and paper and estimate the memory consumption
[13:05:20] <Casper> remember that each function depth use about 6 bytes
[13:05:24] <aep> i think i'll try a device with lufa anyway. so far this is the least horrible lib i found
[13:05:33] <Casper> also
[13:05:39] <Casper> be sure that you use the right signal
[13:05:45] <TLoFP1> damn :(
[13:05:56] <Casper> an unhandled interrupt because you used the wrong signal name will cause a reset
[13:08:27] <TLoFP1> I have a total of 2000 bytes of RAM (328P) my ISRs are not that complicated :( I estimate at most 128 bytes for the ISR
[13:09:42] <TLoFP1> Casper: I also just checked for the chip rebooting and it does not reboot/reset
[13:10:16] <TLoFP1> http://pastebin.com/Atdduf2B here is a code example that does and doesn't work
[13:10:44] <TLoFP1> the only difference in the code is that the dealy changes from 40 uS to 50 uS
[13:11:11] <DarkSector> Hahaha! So check this out. Arduino Bootloader uses _BV() and the other stupid functions.
[13:11:29] <Casper> have you forgot to set the F_CPU ?
[13:12:01] <TLoFP1> Capser: F_CPU is set
[13:12:27] <Casper> are you using an old version of the toolset? (avr-gcc avr-libc avr-binutils) ?
[13:12:39] <TLoFP1> Capserlatest version of AtmelStudio
[13:13:24] <Casper> I don't use atmel studio, so I don't know what it use
[13:14:34] <Casper> actually
[13:14:53] <Casper> there might be a problem
[13:15:01] <Casper> is F_CPU=16000000UL ?
[13:15:29] <TLoFP1> yes! how did u know, that is first line of my code
[13:15:29] <Tachyon`> that's the speed in hertz
[13:15:35] <TLoFP1> "#define F_CPU 16000000UL"
[13:15:42] <Casper> there is a small issue...
[13:15:48] <TLoFP1> so that is 16MHz
[13:15:54] <Casper> void _delay_us ( double __us )
[13:15:55] <Casper> The maximal possible delay is 768 us / F_CPU in MHz.
[13:16:08] <Casper> If the user requests a delay greater than the maximal possible one, _delay_us() will automatically call _delay_ms() instead. The user will not be informed about this case.
[13:16:09] <Casper>
[13:16:30] <Casper> so the max is 48us
[13:16:39] <TLoFP1> right, I have tried _delay_ms() as well (eventually I want to delay 100 mS, but that doesn't work
[13:17:00] <TLoFP1> mhhh, this is very interesting
[13:17:05] <Casper> I wonder...
[13:17:15] <Casper> if you use 16MHz, it mean you use an external crystal
[13:17:18] <TLoFP1> I was using naked ISRs, I stopped doing that and now it works
[13:17:20] <TLoFP1> what gives?
[13:17:27] <Casper> have you set the ckopt fuse?
[13:17:35] <TLoFP1> Casper: yes I am using external osc
[13:17:36] <Casper> to give the full swing oscillator?
[13:17:43] <TLoFP1> yes
[13:18:26] <Casper> double check that, with non-full swing the avr can stop, also... be sure that you did not enabled the watch dog timer
[13:19:22] <TLoFP1> watch dog is disabled
[13:20:28] <TLoFP1> I got it to work now! what!!!
[13:21:07] <TLoFP1> I was using naked ISRs, when I went to just ISR(INT1_vect) instead of ISR(INT1_vect, ISR_NAKED) it suddenly worked
[13:21:48] <twnqx`> lol naked
[13:22:02] <twnqx`> did you save the registers at all? :P
[13:22:33] <TLoFP1> no :(
[13:22:38] <TLoFP1> I am just learning about that lol
[13:22:38] <Casper> lol on this chocolate box...
[13:22:41] <TLoFP1> :(
[13:22:56] <Casper> "May contain all allergens." ← lol
[13:49:20] <aep> hm, so the teensy3 has a cortex M4 MK20DX128, the website lists support for LUFA, but lufa itself doesnt actually list the chip
[13:49:41] <aep> that doesnt even sound like an avr
[13:54:03] <specing> its not
[13:54:30] <aep> blarp, freescale, who wants that :/
[13:55:13] <aep> is there any avr with usb 2.0 high speed?
[13:55:27] <aep> the selector on the atmel site doesnt really list usb speed
[13:59:10] <specing> aep: the AVR core can't handle high speed
[14:00:07] <aep> oh
[14:02:32] <specing> granted, the xmegas could probably do ~50 Mbit USB if you devnull it all with DMA
[14:02:59] <specing> but the host would start sending in data at 480 Mbits!
[14:05:08] <aep> uhm, it would?
[14:05:09] <aep> damn
[14:05:18] <aep> 50mbit would be ok with me :/
[14:05:35] <specing> (if you devnull it)
[14:05:39] <aep> right
[14:05:46] <specing> processing it is another story...
[14:05:49] <aep> maybe the avr32's can handle it? those are arm, right?
[14:06:07] <specing> no, those are not ARM
[14:06:25] <specing> the SAM series are ARMs
[14:06:40] <aep> oh yeah, those are the only ones i found to have high speed usb listed
[14:06:57] <aep> but they dont support the LUFA lib, and i didnt find any other usable stack
[14:07:26] <specing> aep: olimex sells A13's in small quantities afaik
[14:07:44] <specing> those are TQFP-208 cortex-a8
[14:07:53] <aep> yeah, but software :/
[14:07:57] <specing> Linux...
[14:08:01] <aep> oh. right
[14:08:03] <aep> a8..
[14:08:19] <specing> https://www.olimex.com/Products/Components/IC/
[14:08:31] <Epsilon-Aurigae> if you are needing that kind of throughput, avr is not the toy for you.
[14:08:38] <aep> whut, they are cheap
[14:08:52] <Epsilon-Aurigae> I meant the 480Mb/s usb
[14:08:53] <specing> yes.
[14:08:58] <aep> i didnt know cortex-a8 are that cheap
[14:09:16] <specing> aep: on the other hand, intel sells some of their chips as low as $30
[14:09:27] <aep> intel arm?
[14:09:30] <specing> Intel x86
[14:09:45] <aep> werf? crazy
[14:10:04] <specing> http://ark.intel.com/products/35469/Intel-Atom-Processor-Z510-512K-Cache-1_10-GHz-400-MHz-FSB
[14:10:07] <specing> $20
[14:10:51] <specing> If you know where to get them, of course
[14:11:45] <aep> oof. A8 needs a ton of external stuff
[14:12:08] <specing> yes -.-
[14:12:14] <specing> as all microprocessors do
[14:12:15] <Epsilon-Aurigae> higher end processors often do.
[14:12:20] <specing> System on a chip my ass.
[14:12:33] <specing> Epsilon-Aurigae: cortex a8 is not higher end lol
[14:12:46] <Epsilon-Aurigae> compared to AVR it is.
[14:12:53] <specing> I guess.
[14:13:18] <specing> aep: you can just skip this and go for intel atom
[14:13:35] <aep> uh, i dont see how these are easier to build
[14:13:47] <aep> x86 is probably a nightmare to boot
[14:13:54] <Epsilon-Aurigae> someone said they were easier?
[14:13:57] <specing> http://ark.intel.com/products/78474/Intel-Atom-Processor-E3825-1M-Cache-1_33-GHz
[14:14:20] <aep> lots of nice to have power, but unrealistic that i can build a board with this
[14:14:28] <specing> $34 dual-core with AES-NI and hardware virtio
[14:14:32] <aep> i'd have to buy a PC, and that has bios, which is horror
[14:14:57] <Epsilon-Aurigae> bios is just a bootloader!
[14:15:19] <aep> i really dont think i want to go there :P
[14:15:48] <aep> maybe i should just get a dedicated usb dma chip thing
[14:15:54] <Arrowsmaster_> ok guys, i gotta go. see you!
[14:15:59] <specing> aep: bios is going to stay as long as windows does
[14:16:10] <specing> get rid of windows and you'll get rid of bioses
[14:16:21] <aep> sure, but uhm, i dont think i can build that stuff myself
[14:19:32] <carabia> armchair detector activated
[14:19:39] <carabia> beep beep beep
[14:20:55] <carabia> you guys total pussies i perfboard x86
[14:21:12] <megal0maniac_afk> HOLY CRAP
[14:21:16] <Epsilon-Aurigae> deadbug em!
[14:21:44] <Epsilon-Aurigae> I've seen x86 on solderless breadboard...but that was an 80286.
[14:21:58] <megal0maniac_afk> LUFA Audio device using pwm sounds amazing
[14:21:59] <specing> carabia: where do you buy BGA perfboards? :D
[14:22:16] <aep> megal0maniac_afk: yeh duh :P
[14:22:28] <megal0maniac_afk> No seriously
[14:22:43] <megal0maniac_afk> There's no DAC in this thing, but it sounds brilliant
[14:22:47] <megal0maniac_afk> mega32u4
[14:23:01] <aep> i use an external dac
[14:23:12] <carabia> specing you don't steady hand is what you need
[14:23:12] <megal0maniac_afk> Of course that's the better way of doing it
[14:23:36] <aep> th32u4 only has high speed 12mbs :/
[14:23:41] <aep> err full speed
[14:26:20] <megal0maniac_afk> One sec
[14:27:10] <aep> specing: apparantly linux doesnt even have usb client mode
[14:27:19] <aep> or at least the A10 can't do it
[14:28:55] <specing> aep: it does
[14:28:59] <specing> aep: and it works
[14:29:02] <specing> on the A10
[14:29:07] <specing> (have one here)
[14:29:13] <aep> maybe the guy was BSint
[14:29:22] <aep> i have one too. A20, cubieboard
[14:29:30] <aep> but i have no idea how linux works at all
[14:29:33] <specing> definetly bullshitting
[14:29:39] <aep> what do i google? linux usb client mode?
[14:29:47] <specing> Linux usb gadget
[14:30:05] <aep> nice
[14:30:30] <aep> http://linux-sunxi.org/USB_Gadget
[14:30:45] <aep> so yeah, that might be a viable option if avr just can't go fast enough
[14:31:05] <aep> although all the stuff that the a8 needs kind of holds me back :/
[14:31:16] <aep> external nand, ram and all that
[14:31:26] <aep> avrs have it all in one package. neat for hobbyists
[14:32:41] <megal0maniac_afk> Noise floor is high as shit, but still
[14:32:51] <megal0maniac_afk> ima upload a recording
[14:33:50] <aep> what, you already built something? :D
[14:34:28] <megal0maniac_afk> Of course
[14:34:40] <megal0maniac_afk> http://megal0maniac.dyndns.biz/32u4.mp3
[14:35:15] <megal0maniac_afk> There's 5s of silence. And at 1m I turned the volume down because it was clipping
[14:35:28] <aep> uhm
[14:35:33] <aep> so this is through an avr?
[14:35:37] <megal0maniac_afk> Yes
[14:35:40] <megal0maniac_afk> With PWM
[14:35:57] <aep> pwm output to a speaker or what?
[14:36:09] <aep> its actually quite good for that
[14:36:24] <aep> i'd have expected PWM to be close to inaudible bad
[14:36:43] <megal0maniac_afk> It's incredible. I expected the same
[14:36:55] <megal0maniac_afk> Laptop > 32u4 > PC line in
[14:37:20] <aep> uh, with all the usb stuff going on, how does it even drive the pwm fast enough?
[14:37:31] <megal0maniac_afk> Ask Dean
[14:37:33] <Epsilon-Aurigae> hardware pwm maybe?
[14:37:35] <aep> yeah
[14:37:44] <megal0maniac_afk> Oh yeah, it is hardware pwm
[14:38:07] <megal0maniac_afk> Using OC3A as output
[14:38:48] <megal0maniac_afk> There's also a mode where you can use the whole of PORTC as your output
[14:38:59] <Epsilon-Aurigae> build an r2r dac?
[14:39:11] <megal0maniac_afk> Yeah, or use an opamp
[14:39:15] <aep> specing: do you have an a10 circuit or a ready board?
[14:39:24] <megal0maniac_afk> (summing amp)
[14:44:26] <specing> aep: cheapshit chinese tablet
[14:56:09] <aep> specing: the only working circuit i know is the cubieboard, which has parts i can't even order
[14:56:28] <aep> except if i find someone who speaks .... err whatever http://detail.1688.com/offer/1200514843.html
[14:57:02] <specing> the one who makes cubieboards speaks that
[14:57:18] <specing> you can probably also order components from him
[14:57:56] <aep> good point. might just ask that guy to build me a circuit. ugh this is all going way over budget. all for usb2 :/
[15:01:35] <specing> you could also buy an ARM9 from atmel
[15:01:52] <specing> all it needs is 3.3 V
[15:02:53] <megal0maniac> Yeah, maybe go with that. This all seems far too complicated
[15:03:12] <aep> no lufa :/
[15:03:27] <aep> and between lufa and linux there isnt much
[15:03:31] <specing> why do you need lufa?
[15:03:35] <specing> when you have Linux?
[15:03:46] <aep> oh, you said arm9
[15:03:51] <specing> yes, arm9
[15:04:07] <specing> ATMEL makes 475 MHz ones
[15:04:13] <specing> I have a 192 MHZ one
[15:04:17] <aep> cool. with all the stuff on board? ram and nand
[15:04:22] <specing> SAM9260
[15:04:35] <megal0maniac> Why USB2?
[15:04:38] <specing> you don't need nand, just RAM
[15:04:44] <specing> ram is in tssop
[15:04:54] <megal0maniac> Or rather, why high speed USB2?
[15:05:00] <specing> because
[15:05:03] <specing> high
[15:05:04] <specing> speed
[15:05:07] <specing> HIGH SPEED
[15:05:07] <aep> megal0maniac: because i plan to add more bandwith than 12mbs
[15:05:16] <specing> *HIGH* *SPEED*
[15:05:31] <specing> neeed4speed usb drift
[15:05:31] <carabia> doge. much high very speed. wow
[15:05:50] <specing> many 2.0
[15:05:50] <megal0maniac> specing: It was more a question of why he needs that bandwidth
[15:06:15] <carabia> i think it's porn
[15:06:42] <aep> megal0maniac: 12 stero audio streams
[15:06:58] <aep> i havent done the math actually. but i'm sort of sure that is more than 12mb
[15:07:22] <carabia> high speed overclock ln2 cooled avr multiplexing porn streams
[15:07:48] <carabia> 1.21 gigawatts
[15:07:59] <Roklobsta> jiggawatts
[15:08:04] <carabia> niggawatts
[15:08:17] * Epsilon-Aurigae prefers jigglewatts
[15:08:28] <Roklobsta> there goes the tone.
[15:09:20] <aep> specing: all the boards i find with that chip have both external nand and ram
[15:10:50] <aep> maybe i want a cortex m3 or something
[15:11:35] <specing> aep: you don't need nand, trust me
[15:12:44] <specing> doubt there is any cortex m3 with USB2.0
[15:13:19] <aep> yeah, incredible
[15:17:53] <Roklobsta> specing: you sure? what about LPC family?
[15:30:24] <specing> Roklobsta: the problem is that m3 can't handle such data rates
[15:31:04] <megal0maniac_afk> And there's no USB2.0 in the m3 core, one vendor added it though, iirc
[16:26:44] <Fleck> hey - any good tutorial/description about timer waveform generation modes? No clue what does fast pwm mean and how does it differ from normal or phase correct mode...
[16:27:33] <GreaseMonkey> i read a thing on the interwebs but i'll give you the lowdown anyway
[16:28:09] <GreaseMonkey> phase correct PWM ensures that it starts at 0, goes to 1 at the very middle, and then goes back down to 0 at the end again
[16:28:38] <GreaseMonkey> e.g. one cycle might look like this: __""__
[16:28:47] <GreaseMonkey> fast PWM is more like this (two cycles): __"__"
[16:29:31] <GreaseMonkey> Fleck: ^ does that explain anything
[16:30:11] <GreaseMonkey> oh yeah apparently phase correct is good for if you're controlling motors
[16:30:18] <GreaseMonkey> fast PWM is perfectly OK for audio though
[16:32:09] <Fleck> thx GreaseMonkey
[16:32:17] <Fleck> I need audio, so, I need fast PWM
[16:32:28] <Fleck> and what about normal mode?
[16:33:48] <Fleck> even more - two fast pwm modes for attiny13
[16:37:42] <GreaseMonkey> well i'm not sure what the other mode is for, i guess you should probably read the specs a bit more if you want to know the diff
[16:37:48] <GreaseMonkey> i have an atmega328p btw
[17:20:09] <Tom_itx> Fleck, http://tom-itx.dyndns.org:81/~webpage/abcminiuser/articles/avr_timers_index.php
[17:20:15] <Tom_itx> pwm is toward the bottom
[17:20:22] <Tom_itx> but i'd recomend reading it all
[17:21:15] <Fleck> ok, thx Tom_itx
[17:24:53] <Fleck> http://tom-itx.dyndns.org:81/~webpage/abcminiuser/articles/Fast_PWM.gif << this simple image explains it... :/
[18:17:49] <carabia> Tom_itx: i envy your sunglasses are they oakleys?
[18:19:21] <carabia> no wait it's not you it's abcminiuser
[18:19:45] <abcminiuser> Hva?
[18:21:36] <aep> oh lucky me
[18:21:56] <aep> abcminiuser: you'e the lufa dude, right?
[18:22:03] <abcminiuser> Indeedy
[18:22:12] <aep> is there any chip that does high speed usb?
[18:22:58] <aep> i've seen some defines related to high speed, but i cant find a chip that actually runs it
[18:23:36] <abcminiuser> UC3A3/UC3A4 only
[18:23:55] <avrrocks> uc3c2 too no?
[18:25:07] <DarkSector> How do I pack my data in frames to be sent over serial?
[18:25:12] <aep> uuh, all of them? because http://www.atmel.com/products/microcontrollers/avr/32-bitavruc3.aspx doesnt mention anything
[18:25:26] <DarkSector> I need some sort of markers before and after sending the stream?
[18:26:18] <aep> oh yeah, there.
[18:26:40] <aep> awesome, i am saved after all. /me drops working on the crappy cortex 19
[18:27:24] <carabia> http://tom-itx.dyndns.org:81/~webpage/abcminiuser/articles/dean.jpg yeah. abcminiuser , oakleys?
[18:27:34] <carabia> liek, gief
[18:27:44] <abcminiuser> Bahahaha no
[18:27:49] <abcminiuser> $20 Target cheapies
[18:27:56] <carabia> very crop much burr
[18:28:04] <carabia> ;D
[18:28:15] <carabia> so is it back to uni then?
[18:28:15] <abcminiuser> :P
[18:28:24] <abcminiuser> It's about one million years old (~8)
[18:28:26] <abcminiuser> Probably need new pics
[18:28:36] <abcminiuser> Nah, working at www.lifx.co soon
[18:28:38] <carabia> still fly!
[18:29:15] <carabia> lifx sounds like a true ripoff for hipsters
[18:29:20] <carabia> no offense
[18:29:42] <abcminiuser> Don't care they're willing to pay me
[18:29:50] <abcminiuser> Also the product actually seems pretty good
[18:30:07] <aep> ow. so i have one of these blue uc3-a3 things. docs say no workie. damn
[18:30:16] <carabia> *cough*
[18:30:50] <aep> abcminiuser: is that because you dont have one, or is it broken?
[18:32:11] <abcminiuser> It it a bitch to set up, but should work
[18:32:23] <abcminiuser> The UC3 port was experimental, but I concluded it's not viable
[18:32:24] <carabia> you turn your microwave on the lights go b0nk3rZ
[18:32:37] <abcminiuser> The LUFA design was for the really tiny USB AVR8 devices
[18:32:44] <aep> oh. good to know
[18:32:46] <abcminiuser> Which really, really doesn't translate up to the larger parts
[18:32:53] <abcminiuser> The official ASF stack works tho
[18:33:02] <aep> yeah but its ugly as hell and no audio
[18:33:13] <abcminiuser> Ja, but performant
[18:33:17] <abcminiuser> LUFA is dog slow on them
[18:33:21] <aep> damn
[18:33:30] <aep> its the only readable lib i found :(
[18:33:41] <aep> everything else looks like... uhm yeah
[18:33:47] <abcminiuser> Well you could have a crack at it
[18:33:56] <abcminiuser> But big-endian arch means a lot of conversion macros in the descriptors
[18:34:00] <aep> nah. if you say it wont scale, it probably wont
[18:34:08] <abcminiuser> Not really
[18:34:09] <aep> ah
[18:34:18] <abcminiuser> Polled is fine for the 8-bits, but sucks on the 32-bits
[18:34:33] <abcminiuser> On those interrupt wins by a huge margin, which means a complete re-write
[18:35:06] * abcminiuser emergency shower
[18:35:13] <aep> back to working on the arm thing then *groan*
[18:35:58] <aep> looks like the uc3 needs an external nand anyway, which kinda takes the incentive of using atmel away
[18:55:32] * abcminiuser is back
[18:55:40] <abcminiuser> Youwah? No it doesn't
[18:55:47] <abcminiuser> It can have ext mem, but not required
[18:56:09] <carabia> joowot
[18:57:59] <aep> oh
[18:58:12] <aep> cant find any example of it being used standalone
[18:59:00] <abcminiuser> The Xplained boards have the ext mem, because why not
[18:59:09] <abcminiuser> But you can just plonk in a chip and use the internal RAM
[18:59:16] <Tom_itx> abcminiuser does atmel release source to their dev boards?
[18:59:31] <abcminiuser> Ja, look for ASF "Demo" applications for the respecitve boards
[18:59:44] <Tom_itx> hmm i didn't realize they did
[18:59:47] <abcminiuser> "XMEGA B1 Xplained Demo Application - B1 Xplained" and similar
[19:01:03] <aep> abcminiuser: awesome. i wont need much anything except massive usb bandwidth, so is there any documentation on how to actually build a circuit around them?
[19:01:16] <abcminiuser> Err
[19:01:25] <abcminiuser> IIRC the datasheet has some stuff in it somewhere
[19:01:27] <abcminiuser> Hrm
[19:01:50] <aep> yeah it probably has all the info. i was more looking for something straight out of the box
[19:02:00] <aep> there tons of that stuff for the 8bit things
[19:02:06] <aep> with drawings and all :D
[19:02:11] <abcminiuser> Schematic checklist for UC3s will have all the info
[19:02:14] <abcminiuser> Wherever that is
[19:03:04] <abcminiuser> http://www.atmel.com/Images/doc32130.pdf
[19:03:30] <aep> found this: http://www.core.st/projects/AVR32_Starter_Kit/images/AT32UC3B1256_CORE_Board_17_m.jpg
[19:03:41] <aep> if this is any acurate, i basically need nothing
[19:04:10] <Roklobsta> what's better, AVR32 or ARM Cortex
[19:04:41] <aep> eh. easy for me, arm needs tons of external crap that i dont know how to make work :D
[19:04:55] <abcminiuser> Sep: see http://www.atmel.com/Images/doc32130.pdf
[19:06:21] <aep> abcminiuser: looks really trivial. all i care about is the power circuit and usb right?
[19:06:25] <aep> and then i can go use it
[19:06:28] <abcminiuser> Jupp
[19:06:32] <aep> amazing
[19:10:55] <aep> so, i have no idea how to program this xplain board thing. do i just throw it out of the window, or do i actually need to learn it because other chips work the same way?
[19:11:55] <Tom_itx> learn it
[19:12:17] <Tom_itx> xmega use PDI
[19:12:25] <Tom_itx> interfaced thru the jtag connector
[19:12:34] <Tom_itx> all in the pdf
[19:12:56] <Tom_itx> i made an adapter for it for my 6pin programmer
[19:13:52] <aep> uuh
[19:13:59] <aep> sounds horrible
[19:14:08] <aep> i was hoping for something usb + reset pin
[19:14:14] <Tom_itx> rather easy really
[19:16:41] <Tom_itx> which xplain do you have?
[19:17:38] <abcminiuser> The UC3s should ship with a USB bootloader pre-installed
[19:17:46] <abcminiuser> Needs the FLIP software from Atmel to program it via USB
[19:18:06] <abcminiuser> Works, but crappier than using an external programmer (slower, no debug)
[19:18:11] <aep> yeah the blue UC3 board something. the manual is whining about some windows program
[19:18:15] <aep> i never got it to work
[19:18:27] <abcminiuser> BATCHISP, from FLIP
[19:18:32] <aep> yeah
[19:20:20] <Epsilon-Aurigae> doncha hate it when you need some program to make something do what you want it to do?
[19:22:26] <abcminiuser> In the case of FLIP, yes
[19:22:38] <abcminiuser> If the devil could program, he'd still say FLIP sucks
[19:23:37] <aep> sooo, this is not what you use ISP, do you?
[19:23:48] <Tom_itx> it should
[19:24:03] <aep> eh you said something about PDI, is this the same thing?
[19:24:21] <Epsilon-Aurigae> similar but different...kindasorta
[19:24:25] <Tom_itx> there are 3 basic protocols depending on the chip you use
[19:24:30] <Tom_itx> ISP PDI and TPI
[19:24:35] <aep> oh great, more confusion :D
[19:24:55] <Tom_itx> fortunately my programmer does all 3
[19:24:59] <aep> any recomendation for a programmer?
[19:25:04] <Tom_itx> :D
[19:25:10] <Epsilon-Aurigae> the one Tom_itx sells is great.
[19:25:24] <aep> i have an olimex ... erm mk2 whatever. i never got it to do anything with the xplained
[19:25:36] <Epsilon-Aurigae> I dunno if it works with the xplained boards but it works great for all the AVRs I have tried it on.
[19:25:44] <aep> Tom_itx: url?
[19:25:47] <Tom_itx> it does
[19:25:53] <Tom_itx> i test PDI with the xplain
[19:26:12] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/USBTiny_Mkii/USBTiny_Mkii_index.php
[19:26:24] <Tom_itx> there's a 'purchase' menu link
[19:26:50] <aep> works with avrdude i guess?
[19:26:56] <Epsilon-Aurigae> yup.
[19:27:10] <aep> i know who to yell at if it doesnt
[19:27:11] <Tom_itx> works on just about anything afik
[19:27:25] <Tom_itx> windows, studio, avrdude, linux, osx....
[19:27:32] <Tom_itx> even avrs
[19:27:47] <aep> bought
[19:28:02] <Tom_itx> it won't go til monday
[19:28:12] <aep> it wont arive until january anyway :D
[19:28:22] <Tom_itx> where are you?
[19:28:25] <aep> post delivery dudes have holidays too
[19:28:27] <aep> .de
[19:28:38] <Tom_itx> oh, no it won't make it till next year
[19:28:57] <aep> no worries. neither will my other stuff
[19:29:52] <Tom_itx> i'll get you a customs number in a bit
[19:30:01] <Tom_itx> once i get off my butt and grab one
[19:30:24] <aep> customs?
[19:30:47] <aep> at least for importing stuff, everything below Eur20 is fine. just needs to say it on the package
[19:31:11] <aep> and they never check packages from china, only from the US. :(
[19:39:02] <aep> ok, decent programer: check. next some avr32's and schmartboards
[20:09:26] <aep> uh, they make cortex m4 with just 48 pins
[20:09:34] <aep> i wonder if they need any external components
[20:09:59] <aep> doesnt look like it
[20:10:19] <aep> Tom_itx: does your programer work with those SAM4 things?
[20:10:39] <Tom_itx> i don't think so no
[20:10:43] <aep> darn
[20:10:47] <Tom_itx> no arm
[20:10:53] <aep> what do i need for them?
[20:10:57] <Tom_itx> those are probably jtag
[20:11:00] <Tom_itx> i'm not sure
[20:11:04] <aep> k
[20:22:57] <Tom_itx> abcminiuser, what is 'production file' in the Device programming dialog for?
[20:23:10] <Tom_itx> does it contain fuse information as well as program hex?
[20:27:52] <abcminiuser> Jupp, ELF with EEPROM, FLASH, fuses etc
[20:28:02] * abcminiuser semi-AFK
[21:47:10] <aep> gah, the only high speed avr32 is actually the AT32UC3A3128 and that is qfp144
[21:47:29] <aep> schmartboard doesnt even have adapters that small
[21:49:40] <N2TOH> have you checked the other companies?
[21:49:53] <N2TOH> Aries, seeedstudio to name a few
[21:50:12] <aep> for breakouts? yeah i can probably find some
[21:50:22] <aep> but its ridiculously small anyway
[21:50:30] <aep> i'll need an oven or something
[21:50:46] <N2TOH> do you need an adapter, or just a way to connect to the device?
[21:51:06] <N2TOH> have you tried the spray and mop method?
[21:51:08] <aep> just prototyping
[21:51:14] <aep> uh what? sounds new
[21:52:04] <N2TOH> dose everything in flux, and run a pencil iron over the pins, then use desolder braid to clean up the solder bridges
[21:52:38] <aep> heh
[21:52:55] <aep> yeah, i can do that with 62
[21:53:01] <aep> but 144 is hardcore
[21:53:02] <N2TOH> I use that method for leaded devices smaller then 50mil pin spacing
[21:53:45] <N2TOH> start trolling the thrift stores for used toaster ovens
[21:53:58] <aep> i have one.
[21:54:10] <aep> its so crap i dont use it for toast
[21:54:27] <N2TOH> lol
[21:55:15] <aep> i'm not sure how how it gets though
[21:55:27] <aep> i guess lead doesnt need that much heat
[21:56:30] <N2TOH> the lead allow paste has a lower melt temp then common solder, the ROHS stuff needs a bit more heat
[21:56:45] <aep> amazing, basically the AT32UC3A3128 is the only chip on earth that includes both nand and high speed usb
[21:57:05] <aep> and its twice as expensive as the one with lower speed usb
[21:57:51] <N2TOH> I've done the toaster oven trick before, you can even bypass the thermostat if your willing to sit there and watch the boards bake
[21:58:31] <aep> how long does it take?
[21:58:34] <N2TOH> I don't know if anybody has tried adding insulation to the outside of the oven
[21:58:46] <aep> i have a heat gun
[21:59:03] <N2TOH> a few minutes, depends on how many watts the oven is
[21:59:16] <aep> bleh, the thing has a timer
[21:59:31] <aep> people make it sound all complicated
[21:59:38] <N2TOH> the oven has a more uniform temp
[21:59:46] <aep> err, not mine
[21:59:56] <aep> pretty sure a heat gun works better at toasting too
[22:00:23] <N2TOH> I just cramked it up to "11" and watched for all the solder to reflow, then removed it from the oven
[22:01:03] <aep> so how does this work? you just paste everything dirty and then it magically moves up to a clean joint?
[22:01:04] <N2TOH> keep in mind some parts need to be prebaked to rif them of moisture or else the parts will explode
[22:01:17] <aep> uh
[22:01:46] <N2TOH> the water vapor can boil off and the resulting steam pops the parts
[22:03:18] <N2TOH> as for the paste yeah basically, I didn't use a stencil I just swabbed the board with paste and dropped the parts in place. then baked it.
[22:03:40] <aep> wow. sounds easy
[22:03:46] <N2TOH> a Q-tip with some 90% iso to mop up the crud afterwards
[22:04:07] <N2TOH> if you can use real lead based solder paste.
[22:04:35] <N2TOH> the only crappy parts about paste is that it has a 6 month shelf life
[22:04:40] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/USBTiny_Mkii/Stencils/stencil_form3.jpg
[22:04:49] <Tom_itx> stencil application jig
[22:05:12] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/toaster_oven/toaster_oven_index.php
[22:05:13] <aep> these look machine milled
[22:05:14] <Tom_itx> cooker
[22:05:19] <Tom_itx> they were
[22:05:31] <aep> i have nothing :D
[22:05:33] <Tom_itx> not the stencil, the plexi
[22:05:40] <aep> oh
[22:05:42] <N2TOH> I have seen stencils made from beer cans
[22:05:47] <aep> how do you make the stencils?
[22:05:48] <aep> haha
[22:06:13] <Tom_itx> http://ohararp.com/stencils/
[22:06:17] <Tom_itx> send them to him
[22:06:54] <aep> US. takes forever
[22:07:15] <N2TOH> you cut open a can and flatten it, use acetone to remove all coatings. then with a toner transfer you can etch them just like DIY PC boards
[22:07:27] <aep> i can find someone with a laser cuter here i guess
[22:07:52] <N2TOH> why laser cut when you can etch them in a cat box?
[22:07:58] <inflex> aaah, good ole toaster oven
[22:08:05] <aep> heh
[22:08:43] <aep> bah, someone should just make packages with less pins
[22:08:46] <aep> and more stuff in them
[22:10:13] <voxadam> Does anyone know of a library or even a protocol description for a multidrop EIA/RS-485 serial "network"?
[22:10:32] <N2TOH> http://lowpowerlab.com/blog/2013/01/23/diy-home-made-metal-stencils-step-through/
[22:11:38] <N2TOH> voxadam, RS-485 protocol is a misnomer. it's an industry free forall
[22:12:06] <inflex> I'll tell you this much about metal stencils... they suck
[22:12:11] <inflex> (unless they're framed)
[22:12:18] <aep> huh why
[22:12:31] <voxadam> I realize that 485 is an electrical only specification, I was just wondering if there were any good protocols for doing single master serial networks over 485 on AVRs.
[22:12:32] <N2TOH> warpage
[22:12:57] <inflex> aep: when you go to clean them, you'll invariably ruin them by accidentally folding them back
[22:13:03] <N2TOH> sorry no idea about the RS485, maybe build a sniffer?
[22:13:08] <aep> ah
[22:13:25] * inflex ended up just making stencils using his vinyl cutter and mylar
[22:13:33] <inflex> a lot more durable, a lot less work
[22:13:36] <N2TOH> hmm
[22:14:08] * N2TOH knows a guy with said vinyl machines
[22:15:06] <inflex> N2TOH: needs to be done in a certain way, you have to have the mylar backed with tape/padding and you need to change the HPGL files to make the cutter perform 2~4 cycles of the same polygon without lifting
[22:15:27] <avrrocks> voxadam: rs232 ?
[22:15:48] <avrrocks> you build a rs232 to 485 converter
[22:16:00] <voxadam> avrrocks: rs232 doesn't support multidrop
[22:16:25] <avrrocks> yes you need to convert it to 485
[22:16:35] <avrrocks> i know its pint to point
[22:17:11] <avrrocks> but there must be a way to boost the signals to allow >1 receiver
[22:17:13] <inflex> N2TOH: This is the software I wrote to do the job - http://pldaniels.com/dxf-hpgl/
[22:18:12] <N2TOH> thank you inflex, I will pass the info along.
[22:18:18] <voxadam> I forgot, I can use Modbus.
[22:20:06] <N2TOH> Modbus?
[22:21:43] <voxadam> It's an old PLC communications protocol.
[22:26:33] <N2TOH> ah
[23:22:31] <aep> whats the difference between ATSAM3U4CA-CU and ATSAM3U4CA-AU
[23:22:39] <aep> A-CU vs A-AU
[23:23:41] <Casper> iirc, CU = commercial temperature range in (whatever package)
[23:23:55] <Casper> while AU is commercial temperature range lead free in whatever package
[23:24:10] <Casper> i.e. C and A, the U is the package
[23:25:13] <aep> thanks
[23:25:13] <Casper> iirc, you have commercial (0-85C), industrial (-40 to 85C) and millitary (-55 to 125C) or something like that
[23:25:18] <aep> why would i want the lead thing?
[23:25:21] <Casper> in leaded and leadfree...
[23:25:29] <Casper> lead one would be the older package
[23:25:44] <aep> oh ok thats why its out of stock
[23:25:46] <Casper> which basically come from their older line that has not yet been converted to a leadfree process
[23:25:58] <aep> cool, thanks
[23:26:12] <Casper> or old new stock :D
[23:27:32] <N2TOH> militarily electronics are exempt for the lead free nonsense
[23:28:08] <N2TOH> they like the stuff they buy to last at least 20 years...
[23:29:14] <aep> i doubt my soldering joints last that long
[23:29:59] <N2TOH> I have stuff older then that, that still works 100%
[23:31:03] <aep> now why is the AT32UC3A3256 twice as much as the ATSAM3U4C, although slower
[23:31:17] <aep> because everyone wants it because it was in the xplained board?
[23:31:34] <N2TOH> perhaps it's a volume issue
[23:32:34] <mitc0185> I'm trying to get a basic blinky LED program to work on a sparkfun atmega32u4 breakout board. The best I can do is to have the LED on solid on a given pin
[23:33:02] <mitc0185> can't figure out why it's not working
[23:33:23] <Casper> how do you make it flash?
[23:33:57] <aep> ooh, the AT32UC3A3256 is an AVR, the others are arm
[23:34:04] <aep> i wonder which i want :/
[23:34:42] * N2TOH wonders why the atmega1284p keeps getting ignored...
[23:34:44] <mitc0185> Casper: are you asking me?
[23:34:55] <avrrocks> nonsense?
[23:34:57] <Casper> yes
[23:34:58] <avrrocks> they pollute!
[23:34:59] <jadew> freaking e-bay... showing me "what other people looked at" after I finished paying for the item
[23:35:12] <jadew> turns out there were other options, 3 times cheapper, same item
[23:35:23] <mitc0185> http://pastebin.com/wt9QFy9b
[23:36:05] <Casper> #define F_CPU 8000000UL
[23:36:29] <N2TOH> jadew, yeah ebay needs to offer the "WHOOPS" button
[23:36:34] <mitc0185> hmm. read somewhere that didn't matter, but I'll give it a try
[23:36:45] <Casper> not sure if it will fix, but it used to matter I think
[23:37:05] <jadew> N2TOH, hehe, yeah
[23:37:16] <mitc0185> same result -- LED is on solid
[23:37:41] <N2TOH> maybe it is blinking, but just too fast
[23:38:03] <N2TOH> can you hang a scope on the GPIO pin in question?
[23:38:15] <mitc0185> N2TOH: unfortunately no
[23:38:23] <mitc0185> Santa did not bring me a scope
[23:38:29] <N2TOH> :(
[23:38:38] <mitc0185> but you're right, that could be what's happening
[23:38:39] <Casper> mitc0185: you've been naughty then! bad mitc0185 !
[23:38:44] <jadew> santa did not bring me anything
[23:38:49] <N2TOH> try altering your code to make the LED blink much much slower
[23:39:03] <Casper> but it should be fine that way
[23:39:14] <Casper> wait
[23:39:19] * N2TOH got a lump of coal for Christmas
[23:39:30] <Casper> do you use 5V to power the led? or the GPIO pin as the positive?
[23:39:40] <mitc0185> GPIO pin
[23:39:45] <mitc0185> PF0
[23:40:02] <Casper> ok... so like PF0 -> res -> led -> gnd ?
[23:40:06] <mitc0185> here's something I don't understand -- I set both lines in the while loop to 0x00
[23:40:12] <mitc0185> and the LED is still on
[23:40:30] <mitc0185> PF0 -> led -> res -> gnd
[23:40:30] <jadew> wrong pin?
[23:40:38] <N2TOH> are you sure your on the correct pin?
[23:40:44] <Casper> does the led stay on when you keep the reset pin low?
[23:40:46] <N2TOH> jinx!
[23:42:09] <mitc0185> when I push the reset button the LED turns off
[23:42:21] <mitc0185> jadew: I am pretty sure I'm on the right pin
[23:42:30] <jadew> double check
[23:42:44] <N2TOH> LEDs on all the pins!
[23:42:51] <N2TOH> lol
[23:43:05] <Casper> N2TOH: I already put over 150 white leds on my breadboard
[23:43:13] <Casper> that was.... blinding
[23:44:03] <mitc0185> jadew: just checked
[23:44:07] <mitc0185> all IO pins are low
[23:44:13] <mitc0185> except for PF0
[23:44:20] * N2TOH once made the mistake of gazing into a Luxon 5 watt LED.... a mistake I will never repeat!
[23:44:46] <Casper> N2TOH: I'm sure it's less of a mistake than glazing into my camera flash at full power
[23:44:53] <Casper> mistake that I did not do
[23:44:58] <N2TOH> same difference
[23:45:28] <Casper> if you put your hand on it and trigger it, you feel a burn
[23:45:39] <Casper> and if you smell your hand, it smells a bit the burned
[23:45:41] <N2TOH> it burned a hole in my vision for the better part of 15 minutes
[23:46:08] <Casper> look like my flash is worse
[23:46:30] <Casper> someone I flashed it to at 5ft... said she was still seeing the spot almost an hour later :D
[23:46:55] <N2TOH> yikes!
[23:47:14] <Casper> yeah
[23:47:25] <Casper> let's just say that I did not do that again
[23:47:47] <N2TOH> so who here has noticed the increased price of lead products?
[23:50:40] <Casper> not me
[23:51:09] <Casper> and the steam client love to fragment the files... man... awefull..