#avr | Logs for 2012-08-19

Back
[01:01:51] <bsdfox_> I thought they finally released a linux version of avrstudio
[01:17:03] <Kevin`> bsdfox_: would be nice, but it's not too much trouble since the entire toostack is availble in linux anyway
[01:17:18] <Kevin`> bsdfox_: on the contrary, current versions are based on embedded visual studio
[01:18:04] <theBear> screw avrstudio ! i'm glad it doesn't exist on my chosen os
[01:51:03] <inflex> heh
[01:51:07] * inflex sticks with vim
[03:14:48] <bsdfox_> yeah vim/avr-gcc/avrdude works great for me.. maybe atmel doesn't see a need for ICE support
[05:34:33] <Trieste> Hi, I have kind of a beginner question: I have an ATTiny45, and I couldn't get PWM to work at OC0B, so I wanted to check whether I can at least use it as a GPIO pin, so I just uploaded a program that cycles PORTB at 1 and 0 and plugged in a LED. However, the PB0 pin I used before is the only one working, and I have no idea why.
[05:35:32] <theBear> hmm, gotta go in a minute so i aint finding a datasheet, but does the pb1 pin have any other/alternate functions on tiny45 ? perhaps it's in a mode where that pin isn't really pb1
[05:37:21] <Trieste> Well that one doesn't work because I lef the PWM initialized, but none of the other pins seem to work, that's why I'm kind of stumped
[05:37:56] <specing> ddr
[05:38:02] <theBear> oh, i misread, thought you said cycles pb0/1 only
[05:38:08] <theBear> oh yeah, and what he said <grin>
[05:38:25] <theBear> not very effective twiddling bits on tristated/input pins
[05:38:38] <Trieste> DDRB is set to 1 too
[05:39:55] <grummund> set to 1, as in:
[05:39:56] <grummund> ATTiny45
[05:40:02] <grummund> DDRB = 1; /* config PB0 as output, PB1-PB7 as input. */
[05:40:19] <Trieste> !
[05:41:09] <Trieste> Well that solves my PWM problems too, I guess
[05:41:21] <Trieste> I can
[05:41:37] <Trieste> I can't find it in the datasheet, where did you find that setting DDRB to 1 does this?
[05:42:47] <grummund> in the section "I/O Ports"
[05:46:11] <specing> LOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOL
[05:48:01] <Trieste> grummund: link to the datasheet? I really have nothing about this in mine, just about how setting individual bits to 1 sets the pin as output
[05:48:27] <theBear> data direction is kinda one of the most basic and first things you learn in avr land... just shows how outta practice i am :)
[05:49:32] <Trieste> ah, wait, I'm an idiot.
[05:49:58] * Trieste shamefully crouches in a corner
[05:51:18] <Trieste> Didn't realize setting the register to 1 doesn't set all the individual bits to 1, heh.
[05:51:34] <grummund> i can kinda see how someone might think that setting DDRB=1 configures the entire port as output
[05:53:59] <Trieste> Well then, sorry, and thanks for the help :)
[05:59:52] <theBear> personally i'm fond of binary notation for setting stuff like DDR?
[06:00:14] <theBear> but i try to be a grownup and use a higher number system :)
[06:17:27] * grummund prefers hex (binary being not legal in standard C)
[06:18:22] <specing> ever used a different compiler than GCC for AVRs?
[06:18:40] <specing> ever used a different compiler than gcc for anything?
[06:18:54] <theBear> i used bascom avr the first few weeks i found out what avr's were... not even sure gcc did avr that early :)
[06:19:08] <theBear> but nah, gcc all the way
[06:19:38] <grummund> gcc doesn't accept 0b notation other than for AVRs
[06:20:09] <grummund> besides, hex is easier to read :P
[06:20:42] <theBear> like i said, i try, i'm just usually so outta practice i'm dirt slow thinking in bits and writing in hex <grin>
[06:29:20] <specing> grummund: sure it does
[06:34:01] <grummund> such as?
[06:46:45] <specing> gcc (GCC) 4.8.0 20120801 (experimental) does
[08:55:47] <Blecha> woot a library that might actually work without the arduino junk! http://en.radzio.dxp.pl/ks0108/
[08:56:45] <Tom_itx> lib for what?
[08:59:04] <smeding> for what the page says it's for, i presume
[09:00:30] <Blecha> Tom_itx ks0108 graphic lcd
[09:00:38] <Tom_itx> all i saw first was a big google ad that smacked me in the face so i closed it
[09:01:10] <Blecha> he has a main c file then seperated out low level drivers
[09:01:11] <Tom_itx> i hate sites that are pasted full of ads
[09:02:19] <Blecha> rather than arduino glcd library for it which will have the bootloader, all the bloat from going between the arduino pins and actual pins, then i have to go back to actual pins anyways because im not using an arduino...
[09:02:57] <Blecha> and the library uses too much arduino junk for me to want to port it somehow
[09:36:07] <OndraSter> do you bath PCBs in IPA to clean up all the flux residues or do it manually with some piece of antistatic cloth?
[09:36:26] <Tom_itx> depends
[09:36:37] <Tom_itx> scrub it with an old tooth brush if you need to
[09:36:57] <Tom_itx> i sometimes use my ultrasonic cleaner to speed things up
[09:37:59] <Tom_itx> if you submerge the board make sure all the components are ok with it
[09:38:11] <Tom_itx> i add those later if they're not
[09:41:14] <Blecha> I want to etch some boards soon
[09:41:26] <Blecha> Probably just do one with a sharpie for the experience
[09:41:49] <Tom_itx> make sure the etchant won't eat the sharpie
[09:42:36] <Tom_itx> i forget which one but sodium persulphate or ammonium persulphate isn't as nice on them as the other one
[09:42:53] <Tom_itx> and FECL3 is just nasty to use
[09:54:57] <Blecha> Tom_itx i was thinking of doing muriatic and peroxide
[09:55:08] <Blecha> then going for the reusable etchant
[09:56:26] <theBear> don't use a sharpie, get a proper dalo pen
[10:02:09] <Blecha> theBear might go to a local print shop then and just do toner transfer
[10:03:10] <theBear> buy a dalo for a couple bucks and get the experience
[10:03:19] <Tom_itx> :)
[10:04:30] <duinoob> Is it possible to hook up a weak laser diode directly to an avr without worrying of supplying it with constant current, as it will be used in nanoscale on/off bursts?
[10:05:02] <Tom_itx> let us know how it goes
[10:07:09] <theBear> duinoob, just use a resistor and don't tempt fate
[10:14:16] <OndraSter> thanks Tom_itx
[10:42:39] <Kevin`> duinoob: yes, but you are relying on undocumented characterestics of the devices to have it not 'splode. and if you have a code bug and leave the laser on it may burn out
[10:43:42] <Kevin`> actually, that gives me a thought, can the protection diodes on these devices handle at least as much current as the drivers?
[10:44:17] <theBear> protection diodes ? erm, i think we're talking about different lasers
[10:44:34] <Kevin`> it wasn't entirely a related question
[10:45:07] <Kevin`> see, my thought would be to (for lighting use, not data use like this) hook a small inductor and led directly to the microcontroller pin
[10:45:11] <theBear> oh, what devices then ?
[10:45:18] <Kevin`> well, indicator use, not really lighting
[10:46:06] <Kevin`> that would mean, in the ideal case, half of the current flowing through the protection diode
[10:46:19] <theBear> it's common in various non-laser-led based devices to use an inductor to give the little suckers a kick and a joule-thief-esque high current short term kinda drive, but regular leds have much better low-duty-cycle pulse ratings than laser diodes
[10:46:58] <Kevin`> I mean to use it for smoothing current, not high output pulses
[10:47:20] <Kevin`> same configuration as a buck smps
[10:47:38] <theBear> oh, hmmm
[10:47:48] <Kevin`> avoid those nasty current limiting resistors :)
[10:47:52] <theBear> i think it's a bit late for me to imagine something like that
[10:48:06] <theBear> also, what devices are you thinking of that have protection diodes ?
[10:48:32] <Kevin`> almost all avr chips do
[10:48:36] <Kevin`> (all?)
[10:49:06] <OndraSter> all I presume
[10:49:08] <OndraSter> clamping diodes
[10:49:33] <OndraSter> you can simply grab DMM, hook it to "diode test" mode and the ground lead put to Vcc and the "power" lead into GND
[10:49:41] <OndraSter> you will see that it measures 400 - 600mV :)
[10:49:55] <OndraSter> same for any GPIO
[10:50:43] <theBear> oh, rely on the avr clamp diodes... NOW i get it
[10:51:09] <OndraSter> I have just joined the discussion, I have no idea what is being talked here about - I just saw "all avr chips have protection diodes"
[10:51:17] <OndraSter> NOTE that on RESET pin there is NOT clamping diode!
[10:51:27] <OndraSter> (maybe on xmegas there is, but on regular megas/tinys there is NOT!)
[10:51:43] <theBear> that's cool, i don't think he's gonna be sneaky-buck/boosting leds via the reset pin
[11:58:27] <Steffanx> Wha, what is Atmel doing? http://www.atmel.com/devices/ATTINY828.aspx?elqTrack=true&utm_campaign=August%205&utm_medium=email&utm_source=Eloqua&tab=parameters .. 512KByte SRAM :D
[11:58:50] <specing> wooo
[11:58:59] <specing> I want a sample of that!
[11:59:11] <specing> 0.o abcminiuser is back!
[11:59:43] <Steffanx> Shit happens specing
[12:00:04] <smeding> yeah it's bytes
[12:00:06] <smeding> someone just made a boo-boo
[12:00:16] <smeding> or maybe they couldn't enter 0.5 :p
[12:00:27] <Steffanx> They can
[12:04:22] <OndraSter> I told you guys about tiny828 few days ago :)
[12:04:25] <OndraSter> each pin = ADC
[12:13:04] <OndraSter> <OndraSter_> <OndraSter> I told you guys about tiny828 few days ago :)
[12:13:04] <OndraSter> <OndraSter_> <OndraSter> each pin = ADC
[12:13:32] <Steffanx> No need to repeat yourself OndraSter:P
[12:13:39] <OndraSter> I have no idea what comes through
[12:13:42] <OndraSter> because I got disconnected
[12:13:47] <OndraSter> every 2 minutes.
[12:13:48] <OndraSter> I get*
[12:13:55] <theBear> double-repeat
[12:14:05] <theBear> yeah, but WE don't get disconnected every 2 minutes
[12:14:07] <theBear> you see ?
[12:14:23] <OndraSter> but I don't know what message comes through before it disconnects me
[12:14:54] <OndraSter> you see?
[12:15:26] <theBear> hmm, i can't see why not, you should get a better grip on what your connection is doing, and either way you could at least edit out your nick from the quotes
[12:15:36] * Tom_itx quits messin with OndraSter's inet plug
[12:20:03] <Blecha> This guy has some cool stuff http://en.radzio.dxp.pl/avr-mcu/rc5+remote+control+infrared+assembly.html
[12:20:16] <Blecha> went for his lcd library stayed for the rc5 decoding
[12:20:29] <theBear> heh, i like yer turn of phrase
[12:20:38] <theBear> Tom_itx, hehehe
[12:21:19] <OndraSter> I feel like IN AN US MOVIE!
[12:21:28] <OndraSter> I have just bought something on ebay 1 sec before it ended
[12:21:34] <OndraSter> (not ebay, it is local ebay-like website)
[12:21:44] <Corwin> :D
[12:21:53] <OndraSter> ((aukro))
[12:22:01] <Corwin> i know
[12:22:35] <Steffanx> Here local-ebay is owned by ebay :)
[12:22:48] <Steffanx> *the local ebay-like website
[12:51:12] <abcminiuser> specing, yup I'm back
[12:51:15] <abcminiuser> Our router got fixed
[12:51:25] <abcminiuser> Or rebooted, or whatever (finally)
[12:51:35] <abcminiuser> Now I'm large and in charge again
[12:53:13] <abcminiuser> !thislog
[12:53:14] <tobbor> This one: http://rueshouse.dyndns.org:82/~ircjunk/irclogs/html/%23avr-2012-08-19.html
[13:51:10] <specing> abcminiuser_: How did you cram 512Kbytes of SRAM into the ATtiny828? :D
[13:51:22] <abcminiuser_> specing, what now?
[13:51:29] <abcminiuser_> Bytes methinks :P
[13:51:32] <specing> http://www.atmel.com/devices/ATTINY828.aspx?elqTrack=true&utm_campaign=August%205&utm_medium=email&utm_source=Eloqua&tab=parameters
[13:51:37] <specing> I WANT SOME!!!
[13:51:44] <abcminiuser_> Our web team is a barrell of monkeys
[13:51:45] <RikusW> t828 ? since when does that exist ?
[13:51:55] <abcminiuser_> RikusW, last monday
[13:52:11] <specing> abcminiuser_: hahaha
[13:52:18] <specing> 28 analog pins 0.o
[13:52:25] <abcminiuser_> My old internal datasheet says 512 bytes :P
[13:52:25] <specing> 28 actual pins
[13:52:29] <abcminiuser_> Bubble: burst
[13:52:33] <specing> wait you can sample gnd and vcc too? :O
[13:52:53] <specing> ah wait, io 32
[13:53:05] <specing> I guess its tqfp :/
[13:54:10] <specing> Were there any other new additions in the past month?
[13:54:33] * specing makes holes in abcminiuser_
[13:54:35] <specing> abcminiuser_: leak some info dude :D
[13:55:07] <abcminiuser_> specing, you don't get the Atmel sales bulletins?
[13:55:11] <specing> no
[13:55:13] <abcminiuser_> There's been a few releases
[13:55:18] <RikusW> thats more like a mega828 ;)
[13:55:21] <specing> they are a bunch of spam I guess
[13:55:32] <abcminiuser_> specing, mostly, but the new devices are interesting
[13:55:39] <specing> it would be a mega if it had hw mul and 1kb sram
[13:55:43] <abcminiuser_> I had a support case on the Tiny828 a few weeks ago
[13:55:44] <specing> abcminiuser_: name a few!
[13:55:53] <abcminiuser_> Which made me say "Tiny828, what the hell is that?
[13:56:04] <abcminiuser_> specing, trying to think
[13:56:05] <specing> abcminiuser_: How so, wasn't it released last monday?
[13:56:11] <abcminiuser_> specing, publically, yes :P
[13:56:19] <OndraSter_> it was
[13:56:21] <OndraSter_> in the bulletin
[13:56:26] <OndraSter_> THAT'S WHERE FROM I LINKED IT TO YOU
[13:56:28] <OndraSter_> ...
[13:56:29] <abcminiuser_> I don't want to name any devices from my datasheets, as some are miscategorized as released
[13:57:34] <abcminiuser_> ATTINY1634 is public now according to the site
[13:57:53] <specing> you can't make the names public?
[13:57:56] <specing> :(
[13:58:16] <RikusW> specing: you can get the MLF version, time to try soldering something new :-P
[13:58:27] <OndraSter_> MLF is fun
[13:58:40] <abcminiuser_> specing, I can, but only if I don't want to work any more
[13:59:04] <specing> you would be fired just for telling us the name? :(
[13:59:09] <abcminiuser_> specing, actually yes
[13:59:11] <RikusW> abcminiuser_: why all the secrecy ?
[13:59:20] <specing> We can't do much with the name if there are no docs about it :X
[13:59:31] <specing> we can only guess what it does...
[13:59:39] <specing> s/does/has/
[13:59:55] <abcminiuser_> RikusW, depends on the device, but some have hefty budgets behidn them and it's not good to give competitors head-starts or insider info
[14:00:18] <abcminiuser_> Plus all the unreleased stuff has awesome codenames instead anyway
[14:00:27] <specing> :X
[14:00:37] <abcminiuser_> I wish more companies would release with cool names rather than HERPDERP1234 chip names
[14:00:49] <specing> Wait
[14:00:58] <specing> ATMEL has competition?
[14:01:05] <specing> who could possibly compete...
[14:01:16] <abcminiuser_> specing, you haven't seen our stock prices lately :(
[14:01:24] <Tom_itx> arm
[14:01:36] <abcminiuser_> Tom_itx, we make those now
[14:01:47] <abcminiuser_> Cortex M3 and M4, plus the old SAM7 and SAM9
[14:01:49] <Tom_itx> so i see
[14:01:56] <abcminiuser_> Hell they work in Studio now
[14:02:03] <abcminiuser_> (the Cortex, not the older ones)
[14:02:09] <Tom_itx> how you been?
[14:02:14] <abcminiuser_> For complete bullshittery, see: https://www.youtube.com/watch?feature=player_embedded&v=sdYSfsqCn64
[14:02:32] <abcminiuser_> Not bad, working VERY hard the last couple of weeks to close off as many ASF bugs as I could
[14:02:37] <abcminiuser_> We've just hit over 100 fixes
[14:02:40] <RikusW> abcminiuser_: does the JTAGICE support ARM ? how about the dragon ?
[14:02:42] <abcminiuser_> For this release
[14:02:50] <abcminiuser_> RikusW, none do, you need a SAM-ICE
[14:03:03] <abcminiuser_> Or other Segger debugger as of the latest Studio patch release
[14:03:15] <abcminiuser_> Which reminds me, everyone update your Studio 6 installs
[14:03:47] * RikusW still use 4.....
[14:03:52] <Tom_itx> i would if i had it installed
[14:04:05] <OndraSter> I have found out that I make better soldering job at QFN rather 0603 resistors :/
[14:04:26] <specing> Holyshit
[14:04:35] <specing> the atmega3290a datasheet is 40MB
[14:04:42] <RikusW> 0603 is easy, if you have fine tweezers anyways
[14:04:43] <specing> -rw-r--r-- 1 specing specing 39M Aug 19 20:53 doc8284.pdf
[14:05:31] <specing> abcminiuser_: atmel is failing on the stock market because it is being so closed about stuff
[14:05:47] <specing> If people know what is coming they can plan ahead...
[14:06:18] <abcminiuser_> specing, the idea is that people buy what we have now, rather than planning on something that may or may not be released (or released on time)
[14:07:20] <specing> :S
[14:07:34] <specing> Are there any plans on developing an ethernet AVR?
[14:07:37] <abcminiuser_> ^ specing that datasheet is so freaking big because of the electrical charactersistics section
[14:07:51] <abcminiuser_> It's one datasheet for 10 devices
[14:08:00] <abcminiuser_> Which means graphs * 10 = 300 pages of graphs
[14:08:11] <specing> oh I see
[14:08:39] <specing> any pointers on an ethernet (x)mega?
[14:08:48] <specing> this channel need a bot, you know? :D
[14:08:53] <specing> needs*
[14:08:56] <abcminiuser_> I cannot comment on future products :P
[14:08:57] <RikusW> AVR32 does support it
[14:09:28] <specing> avr32...
[14:09:30] <Tom_itx> aww c'mon..
[14:09:38] <specing> I ain't as cool as an bit bot!
[14:09:41] <specing> 8*
[14:10:45] <RikusW> abcminiuser_: it would be nice if current AVR tools get ARM support too
[14:11:06] <Tom_itx> studio 6 is current
[14:11:13] <RikusW> another thing, will OpenOCD stuff work with AS6 + ARM ?
[14:11:13] <Tom_itx> the rest are obsolete
[14:11:42] <Tom_itx> unless you mean hardware
[14:11:57] <RikusW> I mean hardware tools
[14:12:55] * RikusW think we should create ASL, a portable win/lin AS clone :-P
[14:13:10] <abcminiuser_> RikusW, I believe that's on the cards however
[14:13:22] <RikusW> that would be very nice
[14:13:43] <abcminiuser_> (ARM support for AVR devices - a dev said they would do it and said I could release that with the usual "may or may not happen" caveat)
[14:13:54] <abcminiuser_> Man I'm still pissed about the Synaptics video
[14:14:43] <RikusW> why ? I haven't seen it...
[14:15:13] <abcminiuser_> https://www.youtube.com/watch?feature=player_embedded&v=sdYSfsqCn64
[14:15:30] <abcminiuser_> Basically, they pit their brand new touch technology against of the oldest chips we offer
[14:15:53] <abcminiuser_> In what is either a completely fucked up phone hardware/software/both design, or one that's been deliberately gimped
[14:16:24] <RikusW> sue them ;)
[14:16:35] <abcminiuser_> In the biggest piece of horse-shittery since the unfortunate horse laxative tests of 2012
[14:16:52] * specing is confused
[14:18:17] <abcminiuser_> specing, summary: SOMEONE IS LYING ON THE INTERNET
[14:20:38] * specing doesen't have these problems as he doesen't visit youtube
[14:20:44] <Tom_itx> abcminiuser_ is that allowed?
[14:21:30] * RikusW don't want to visit youtube with a GPRS connection....
[14:21:52] <specing> It would take like 30s just for the UI to load!
[14:22:46] <abcminiuser_> Tom_itx, https://www.youtube.com/watch?v=Wp8Vx73M2FE
[14:24:15] <RikusW> specing: 30s ? more like 30m :-/
[14:24:55] <RikusW> well maybe not that loong
[14:25:36] <RikusW> Atmel AVR32027: Mixing Assembly and C with AVRGCC http://www.atmel.com/Images/doc32195.pdf
[14:29:03] <OndraSter> abcminiuser_, the other day I ordered samples from atmel
[14:29:06] <OndraSter> they came in 5 days or so
[14:29:07] <OndraSter> via DHL
[14:29:09] <OndraSter> nice :)
[14:35:04] <OndraSter_> did anyone write anything since I devcoder_ joined?
[14:35:07] <OndraSter_> bloody interwebz
[14:35:34] <OndraSter_> the log throws me 403 :P
[14:35:37] <RikusW> no
[14:35:40] <OndraSter_> k
[14:35:51] <RikusW> zlog
[14:36:00] <OndraSter_> ah this one works
[14:36:01] <OndraSter_> thanks
[14:39:02] <abcminiuser_> OndraSter_, didn't I tell you we have actual samples now?
[14:39:09] <abcminiuser_> New initiative
[14:39:11] <OndraSter_> you told it here, yes :)
[14:39:20] <abcminiuser_> Ja, well good to see it working :)
[14:39:29] <abcminiuser_> Also, everyone file your ASF bugs
[14:39:46] <OndraSter_> when I do run into some - I will :)
[14:40:03] <OndraSter_> there is no release notes to the patch2 on atmel's website though
[14:40:07] <OndraSter_> so I am not sure what is fixed and what not
[14:43:44] <abcminiuser_> One sec
[14:43:57] <abcminiuser_> http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&t=123707&sid=25b0a0cae79f376dfae65ec1780fec1c
[14:44:39] <abcminiuser_> That
[14:44:53] <abcminiuser_> *That's an old SID, before you get excited
[14:45:07] <OndraSter_> ah, thanks
[14:48:39] <abcminiuser_> Hurrah, I'm useful
[14:49:06] <OndraSter_> :D
[15:19:11] <Posterdati> hi
[15:48:01] <molavy> hi
[15:50:19] <OndraSter> hihihi
[16:02:03] <molavy> i want create custom webcam using tcm8240md
[16:02:16] <molavy> toshiba image sensor
[16:02:22] <molavy> https://www.sparkfun.com/products/8668
[16:02:25] <molavy> i have this
[16:03:03] <molavy> but i don't know how to connect it to MC
[16:04:18] <specing> molavy: neither do we
[16:04:35] <specing> Another satisfied customer! Next!
[16:05:14] <OndraSter> to MC?
[16:05:17] <OndraSter> aka microcontroller
[16:05:22] <OndraSter> I presume?
[16:06:05] <OndraSter> molavy, it has got I2C for control interface and 8bit parallel for data
[16:06:14] <OndraSter> including HBLK/VLBK/DCLK signals
[16:06:22] <specing> atmel is down!
[16:06:31] <OndraSter> specing, atmel is down every odd 5 minutes :P
[16:06:45] <OndraSter> in the past week or so
[16:07:05] <specing> < abcminiuser_> Our web team is a barrell of monkeys
[16:07:10] <specing> :D
[16:07:59] <OndraSter> molavy, anyway, you will be feeding 6 - 20MHz into EXTCLK
[16:08:13] <OndraSter> reading data period is 6us
[16:08:32] <OndraSter> also, weren't you here already askhing about this?
[16:08:38] <hyperkineitc> molavy: Why go custom when there is *som* much cheap hardware with *drivers* already?
[16:09:18] <Amadiro> Is the 8u2 the smallest AVR (in terms of pin count) with a usb device?
[16:09:30] <OndraSter> what about V-USB?
[16:09:37] <OndraSter> that can fit on tiny4x... I think
[16:09:38] <Amadiro> OndraSter, not counting that
[16:09:40] <OndraSter> ok
[16:09:42] <Amadiro> Yeah, I know
[16:09:45] <hyperkineitc> While an AVR can control the i2c, none of the AVR family is really up to the task of procesing the video.
[16:09:52] <molavy> because they don't have good resolution
[16:09:56] <OndraSter> hyperkineitc, xmega!
[16:10:20] <OndraSter> you can always burst download one image, stop clock, process it and then resume the clock :P
[16:10:26] <OndraSter> but regular mega... meh
[16:10:44] <OndraSter> hell, you could probably set up DMA to read the data from the sensor
[16:10:58] <hyperkineitc> Ging to need external RAM for a frame buffer.
[16:11:00] <OndraSter> (I have not studied what is the DMA capable of on xmegas yet...)
[16:11:14] <OndraSter> yep
[16:11:27] <OndraSter> 1300x1040 = 1.29MB
[16:11:36] <OndraSter> xmega128a1u :)
[16:11:44] <OndraSter> I should design my XBoard Ultra soon appearantly :D
[16:12:04] <OndraSter> aka xmega128a1u + 16MB SDRAM
[16:12:50] <hyperkineitc> Personally, for a project of that scope I'd go with an ARM processor
[16:13:03] <OndraSter> you'd still need external memory for that :)
[16:13:07] <OndraSter> aka get some bigger board
[16:13:11] <OndraSter> and those are not cheap
[16:13:57] <molavy> i seen a lot of breakout board for this sensor but i can't understand why they use under 10 pin's of this sensor
[16:14:02] <hyperkineitc> The R-Pi could easily handle that thing.
[16:14:11] <OndraSter> right, RPI
[16:15:15] <specing> rip-off pi
[16:16:02] <hyperkineitc> Same price as an Arduino
[16:16:04] <hyperkineitc> If you can get your hands on one. Still quite a wait.
[16:17:19] <molavy> i need full mega (1,3MP),but frame rate is not important,because i want make picture from live image's on PC,5 frame per second is good,i think
[16:17:49] <hyperkineitc1> That's a tall order.
[16:17:57] <molavy> here are some breakout board:
[16:17:58] <molavy> http://sigalrm.blogspot.com/2011/03/tcm8230md-breakout.html
[16:19:17] <hyperkineitc1> That's 20 MB/sec! Still think an AVR is up to the task?
[16:19:39] <molavy> http://robosavvy.com/store/product_info.php/manufacturers_id/17/products_id/1263
[16:20:22] <hyperkineitc1> Raw transfer rate anyway. If it's already compressed, it might be doable.
[16:20:39] <OndraSter> I see 18 pins, not 10, molavy :)
[16:20:55] <Amadiro> hyperkineitc1, the rpis CSI firmware is unfortunately not publicly released
[16:21:06] <Amadiro> But you could do USB
[16:22:18] <hyperkineitc> If you could find a commercial device using the same part, you could copy the firmware image, but you're still on the hook for figuring out how to talk to it.
[16:23:05] <Amadiro> probably more trouble than it's worth...
[16:23:46] <hyperkineitc> I've got a fun problem. I need to come up with 5V, up to 15A from a wearable battery pack, in the next couple of days!
[16:24:07] <Amadiro> Also, I think there is only one other device using the bcm2835, it's a settop-box
[16:24:27] <Amadiro> so it probably doesn't use the CSI either
[16:25:59] <specing> hyperkineitc: get a *very efficient* smps chip
[16:26:25] <OndraSter> duh
[16:26:25] <OndraSter> <Amadiro> hyperkineitc1, the rpis CSI firmware is unfortunately not publicly released
[16:26:32] <OndraSter> I thought it was some CSI zooming feature
[16:26:38] <OndraSter> and you were joking around since he was buying camera :D
[16:26:50] <Amadiro> heh
[16:28:08] <tmpvar> So, I've been staring at a barebones board with an atmega32u4, fixed some VCC issues and was able to flash adafruit's lufacdc bootloader on the device. The bootloader led is flashing so I'm confident the chip is running, but it doesnt enumerate a device on my computer. I'm not even sure what question to ask here, every thing looks sane. Are there any common gotchas with this type of thing?
[16:28:31] <OndraSter> are you connecting it to lin or win box?
[16:28:56] <tmpvar> osx, and I've flashed a breakout board with the same bootloader and it enumerates
[16:28:59] <hyperkineitc1> Lots of modules like that don't 'just work'. The require licensing some proprietary code to make all the features work.
[16:29:56] <tmpvar> pretty lost here, I can send pictures / the .brd if that is helpful!
[16:30:05] <OndraSter> you can try :)
[16:31:46] <hyperkineitc1> Ha! I'm still talking about that camera module.
[16:32:21] <hyperkineitc1> tmpvar: You're doing a soft USB device?
[16:32:52] <Posterdati> hi
[16:33:08] <Posterdati> is anyone interested in my little RTOS for ATmega2560?
[16:34:09] <hyperkineitc1> Post the link.
[16:34:25] <Posterdati> link of what?
[16:34:26] <hyperkineitc1> How is it different from the myriad of others?
[16:34:26] <molavy> tmpvar,yes please
[16:36:19] <tmpvar> molavy & OndraSter: http://tmpvar.com/tmp/tpad4-mega/ -- board and a picture of some of the minor fixes I had to perform.
[16:37:42] <Posterdati> hyperkineitc1: ?
[16:38:08] <Amadiro> tmpvar, that's a lot of empty space you got there
[16:39:08] <tmpvar> Amadiro, yeah, I'll probably worry about all that after this iteration works :)
[16:39:32] <Amadiro> well, not saying it's a bad thing, I don't know what you're doing.
[16:42:56] <hyperkineitc1> Didn't you offer to share your RTOS?
[16:44:48] <tmpvar> Amadiro, yeah its there for a reason.. for now :)
[16:44:52] <OndraSter_> zlog
[16:44:58] <OndraSter_> !botsnack
[16:45:52] <hyperkineitc1> Damn. Still on Eagle 5.
[16:45:53] <tmpvar> oh yeah, did I mention I'm _really_ new to this?
[16:46:21] <OndraSter_> tmpvar, homemade 2 layers board?
[16:46:31] <hyperkineitc1> You get as far as getting a proto run made and stuffed. You're not doing too bad.
[16:46:58] <tmpvar> OndraSter_, it's a barebones board from advanced circuits
[16:47:39] <tmpvar> previous iterations dropped down into a breakout board, which I'm trying to integrate in this revision
[16:48:19] <hyperkineitc1> I've never done a USB device, but have read up on what it takes.
[16:48:33] <OndraSter_> where is vbus going to the MCU?
[16:49:28] <hyperkineitc1> USB is really fussy about termination, symmetry in the D+ and D- lines.
[16:50:17] <OndraSter_> you should have rotated the chip by 180° or put the USB on the other side
[16:50:24] <OndraSter_> to avoid vias :)
[16:50:36] <OndraSter_> USB is the thing that I design the board around
[16:50:48] <OndraSter_> aka to have the shortest path without bends if possible etc
[16:50:51] <tmpvar> pins: 2,7,34,24,44 and 33 via a 10k resistor
[16:50:55] <tmpvar> ah
[16:51:14] <hyperkineitc1> Layout is very critical.
[16:51:20] <OndraSter_> yep
[16:52:28] <tmpvar> so are we thinking that is the problem?
[16:52:38] <OndraSter_> tmpvar, do you have the original schematic file?
[16:53:16] <hyperkineitc1> I'd start there. Termination should be as close to the I/O pins on the AVR as possible.
[16:53:28] <OndraSter_> termination is inside the AVR itself
[16:53:45] <OndraSter_> or do you mean those R1 R2?
[16:53:55] <tmpvar> OndraSter_, I didn't design using a schematic, but it is based off of the adafruit atmega32u4 breakout. schematic for that: https://github.com/adafruit/Atmega32u4-Breakout-Board/raw/master/atmega32u4bb.sch
[16:54:07] <OndraSter_> awww no schematics = possible problems
[16:54:08] <OndraSter_> believe me
[16:54:23] <OndraSter_> I always draw schematics, from that generate board with just parts, lay them, connect it all up
[16:54:23] <hyperkineitc1> What value are you using for termination?
[16:55:01] <tmpvar> R1/R2 are 22ohm
[16:55:11] <tmpvar> if thats what you mean
[16:55:28] <OndraSter_> tmpvar, when you were programming it via the ISP, did you power it via USB?
[16:55:33] <OndraSter_> or via external PSU?
[16:55:41] <tmpvar> OndraSter_, ah, yeah I opted out of paying for that since I wasnt sure if i'd use it
[16:56:00] <OndraSter_> ugh? opted out of paying for that?
[16:56:04] <OndraSter_> I thought you made this yourself
[16:56:22] <tmpvar> OndraSter_, i mean paying for eagle's schematic thing
[16:56:28] <OndraSter_> huh?
[16:56:30] <OndraSter_> since when?
[16:56:46] <OndraSter_> eagle is free for schematics upto 100x80 or something, just as for brd
[16:57:01] <hyperkineitc1> http://www.usb.org/developers/usbfaq#sig6
[16:57:32] <Amadiro> KiCAD is also free and pretty good, so there you go
[16:57:32] <OndraSter_> hyperkineitc1, the termination is inside Mega, there are 22Ohm series resistors though
[16:57:54] <OndraSter_> I tried Altium, I couldn't get used to its way of selecting parts and its way of controlling it
[16:57:58] <OndraSter_> I am die hard user of Eagle :D
[16:58:27] <hyperkineitc1> That might do. Is it a straight shot from the USB connector? Bends? Vias?
[16:59:05] <tmpvar> hyperkineitc1, yes, bends and vias
[16:59:05] <OndraSter_> check the board image - yes, there are two vias on it :/
[16:59:05] <OndraSter_> on each
[16:59:06] <OndraSter_> this is USB that works:
[16:59:07] <OndraSter_> http://clip2net.com/s/2dLUh
[16:59:07] <OndraSter_> on my project
[16:59:20] <hyperkineitc1> Have you verified that D+ and D- are connected to the right polarity the software expects?
[16:59:43] <hyperkineitc1> I know I'm asking the obvious, but sometimes.....
[17:00:58] <tmpvar> hyperkineitc1, yeah
[17:01:07] <tmpvar> OndraSter_, the last bootloader flash I used external power. I can try with usb power if that would reveal anything
[17:01:20] <OndraSter_> measure Vcc with USB plugged in
[17:01:22] <OndraSter_> directly on the mega
[17:02:14] <hyperkineitc1> If you flashed it, have you verified that it is indeed running your code? Any kind of diagnostic for signs of life?
[17:02:31] <hyperkineitc1> Serial output, LED, etc
[17:02:48] <tmpvar> OndraSter_, 4.98v
[17:02:51] <OndraSter_> k
[17:04:00] <tmpvar> hyperkineitc1, led is throbbing as per the bootloader
[17:04:23] <hyperkineitc1> Great. Do you have a scope?
[17:05:58] <tmpvar> ok, so usb power causes the avrispmkii to think I have the isp header plugged in backwards
[17:05:58] <tmpvar> oh, nevermind.. reset the programmer and its flashing now
[17:06:32] <tmpvar> hyperkineitc1, Yeah, i have a low end scope what were you thinking?
[17:07:14] <hyperkineitc1> Checking the rise time on D+ when you plug the USB in.
[17:08:34] <OndraSter_> do you realize how fast the stuff is? :D
[17:08:57] <OndraSter_> also that USB is not floating
[17:09:20] <OndraSter_> in theory USB should be earthed
[17:09:43] <OndraSter_> but... the sparks say otherwise
[17:12:13] <tmpvar> OndraSter_, is there a way to confirm that the vias/bends are the cause of the problem?
[17:12:28] <hyperkineitc1> Did you space your traces as per the spec? There could be crosstalk.
[17:12:45] <OndraSter_> the spec says "as close as possible" :)
[17:12:47] <OndraSter_> since it is diff pair
[17:12:50] <hyperkineitc1> Yes, use the scope to look for reflections
[17:13:16] <OndraSter_> tmpvar, hook up USB wire right at the mega pins
[17:13:29] <hyperkineitc1> You should see a ghost in your signal like with old analog TV
[17:15:49] <hyperkineitc1> Something to consider on your next round should this not work for you: http://www.digikey.com/product-detail/en/USBUF01P6/497-3282-6-ND/1848586
[17:16:46] <OndraSter_> I am not using anything like that
[17:16:50] <OndraSter_> just 22R network and ESD
[17:17:39] <tmpvar> OndraSter_, in your design what are D1/D2?
[17:17:50] <OndraSter_> ESD
[17:17:57] <OndraSter_> diodes
[17:18:53] <tmpvar> aha
[17:19:14] <tmpvar> will this work without the 22R's?
[17:19:23] <hyperkineitc1> Nope.
[17:19:26] <tmpvar> (in response to "hook up the usb wire directly to the chip")
[17:19:51] <tmpvar> ok
[17:20:17] <hyperkineitc1> Do you have the USB connector shield connected to ground?
[17:20:52] <tmpvar> yes
[17:21:16] <hyperkineitc1> "it is best to avoid directly connecting the USB shield and signal ground on the PCB" - http://www.ftdichip.com/Support/Documents/AppNotes/AN_146_USB_Hardware_Design_Guidelines_for_FTDI_ICs.pdf
[17:21:56] <hyperkineitc1> This is targeted at USB 2.0 designs, but much still applies to 1.0
[17:22:36] <tmpvar> "is best" doesnt sound like a hard requirement, but ill try it!
[17:23:56] <tmpvar> thank you for the help hyperkineitc1 and OndraSter
[17:24:05] <OndraSter> zlog
[17:24:29] <OndraSter> np
[17:24:35] <OndraSter> I don't have mine shield tied up at all
[17:24:46] <OndraSter> for quite simple reason
[17:24:47] <hyperkineitc1> No prob. Keep futzing with what you have. It may not be what we were fussing about.
[17:24:54] <OndraSter> the footprint I used didn't have it ;D
[17:27:52] <hyperkineitc1> I'd focus on making sure you have the right trace width and spacing for the PCB material you're using, and consider that termination part. It eliminates a bunch of design and parts selection, gives you ESD protection etc.
[17:31:45] <hyperkineitc> http://www.usb.org/developers/docs/hs_usb_pdg_r1_0.pdf
[17:53:19] <hyperkineitc2> Anyone mess with the STM32?
[18:15:54] <tmpvar> it has been confirmed! the via/bends are the problem
[18:18:18] <hyperkineitc> Bummer!
[18:19:02] <tmpvar> http://tmpvar.com/tmp/tpad4-mega/hack.jpg
[18:19:19] <tmpvar> im pretty sure thats not what OndraSter meant, but thats how it happened, haha
[18:19:59] <OndraSter> wow
[18:20:01] <OndraSter> crazy :D
[18:20:09] <OndraSter> I ment actually stripping the USB cable right to the wires..
[18:20:20] <OndraSter> or soldering micro/miniUSB connector right at the mega pins...
[18:20:23] <OndraSter> but that would be dangerous
[18:20:23] <tmpvar> yeah, worked for ~3 seconds before the cable ripped every thing off ;)
[18:20:42] <OndraSter> heh
[18:20:43] <OndraSter> so much capton!
[18:20:45] <OndraSter> or mylar
[18:20:47] <OndraSter> or whatever it is
[18:20:51] <OndraSter> kapton-
[18:20:52] <OndraSter> *
[18:20:54] <tmpvar> kapton :)
[18:21:10] <hyperkineitc> Looks great. SHIP IT!
[18:21:14] <tmpvar> ok, so i think with a slight redesign this should be good :)
[18:21:26] <OndraSter> I saw some PWM LED drivers there
[18:21:32] <OndraSter> is it going to be USB blinkey? :D
[18:21:42] <tmpvar> yep!
[18:22:05] <OndraSter> I have got my own blinkey :)
[18:22:05] <OndraSter> http://www.youtube.com/watch?v=YwhEUHvmGSw&list=UUPUODrGRAQkIIvn0u_HCH0w&index=1&feature=plcp
[18:22:07] <OndraSter> with USB
[18:23:32] <hyperkineitc1> Don't feel too bad. My last proto run, I swapped VSS and VDD when I made the part in eagle. EVerything else was 100% once I figure out why the CPU was getting hot!
[18:24:23] <OndraSter> heh
[18:26:08] <OndraSter> I have accidentaly used LM2936 footprint instead 1117-3V3
[18:26:08] <tmpvar> OndraSter, here's what im going for http://tmpvar.com/tmp/tpad4-mega/final.jpg
[18:26:09] <OndraSter> the pins are swapped absolutely
[18:26:09] <OndraSter> nice :)
[18:26:10] <tmpvar> yep, I've made 4 so far.. looking to get the price down a bit to do a larger beta run :)
[18:26:10] <OndraSter> this is how I have solved the issue
[18:26:10] <OndraSter> http://clip2net.com/page/m12051/32701130
[18:26:10] <tmpvar> btw: the aluminum / buttons / wood were all made in my garage :)
[18:26:10] <OndraSter> I have got 10 boards
[18:26:11] <OndraSter> that will be fixed this way :)
[18:26:12] <tmpvar> haha
[18:26:21] <tmpvar> atleast it is fixable
[18:26:45] <OndraSter> and you'd be surprised but I really enjoy soldering the QFN xmega :)
[18:26:46] <OndraSter> http://clip2net.com/page/m12051/32701168
[18:26:53] <OndraSter> it is easier than it look
[18:26:54] <OndraSter> s
[18:27:15] <tmpvar> i hate qfn's.. atleast the a6281
[18:27:45] <tmpvar> i have burnt out so many boards working/reworking/reworking again those damn things
[18:30:03] <OndraSter> heh
[18:30:05] <OndraSter> I found a sweet spot
[18:30:12] <OndraSter> around 4
[18:30:25] <OndraSter> (I don't have digital temperature gauge, only manual settings 1 - 9)
[18:33:11] <OndraSter> I am not sure if I could do it with solder paste without stencil
[18:33:15] <OndraSter> so I always use a lot of flux
[18:33:18] <OndraSter> tin the pads
[18:33:25] <OndraSter> (with properly heated up iron)
[18:33:29] <OndraSter> then clear the flux
[18:33:34] <OndraSter> put the chip on the tinned pads
[18:33:43] <OndraSter> and hotair it to the board :)
[18:34:32] <OndraSter> I always check the pads before soldering the chip on it under magnifying glass (7x zoom I think)
[18:34:38] <OndraSter> to make sure there are no shorts
[18:35:05] <tmpvar> hrm, ill have to try that. im pretty sure the technique changes when you go from no soldermask to solder mask
[18:35:14] <OndraSter> this has soldermask
[18:35:23] <Tom_itx> OndraSter i had one i couldn't see until i got it under the binocular scope
[18:35:37] <Amadiro> tmpvar, what is that?
[18:35:50] <OndraSter> haven't done any sub-1mm without solder mask yet, I have one 0.8mm waiting here, but that is going to be regular "solder one pad and then the rest" style, because it is homemade PCB with crappy printer
[18:35:55] <OndraSter> Tom_itx, huh
[18:36:14] <OndraSter> well, flux helps a lot to make sure that if there are shorts that they will be big
[18:36:22] <tmpvar> for instance the best success ive had, was heating up the area where the chip is going to go, drop a bunch of solder paste *everywhere* and heat it.. the solder is allowed to run down the traces so it *usually* kills the shorts
[18:36:37] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/scope/DSC_0001.JPG
[18:36:40] <Tom_itx> 1.9M
[18:37:26] <OndraSter> where was the short?
[18:37:36] <Tom_itx> where i scraped the mask off
[18:37:40] <OndraSter> oh
[18:37:49] <OndraSter> I thought it was somewhere during soldering
[18:37:54] <OndraSter> was it professionaly made PCB?
[18:37:57] <Tom_itx> yes
[18:38:01] <OndraSter> interesting
[18:38:05] <Tom_itx> it happens once in a while
[18:38:29] <OndraSter> one would expect them that when they use etest that it shouldn't happen... although they check for WORKING lines and not shorted lines
[18:38:31] <OndraSter> I think
[18:38:32] <OndraSter> right?
[18:38:41] <Tom_itx> i dunno
[18:38:49] <Tom_itx> i got those from somebody else
[18:38:58] <Tom_itx> but they came from goldphoenix iirc
[18:39:18] * Tom_itx goes back outside to finish up
[19:11:43] <Lt_Lemming> http://i.imgur.com/EtMoJ.jpg <--- yays! all the channel traces done
[19:15:46] <OndraSter_> what IC would that be?
[19:25:19] <Lt_Lemming> OndraSter_, TLC5947
[19:25:24] <Lt_Lemming> 24 channel PWM controller
[19:25:30] <Lt_Lemming> designed for LED's
[22:34:37] <aarossig> Hi all, I thought you might enjoy my Sunday project. I managed to create a simple software defined AM transmitter using an AVR Butterfly. I sample the ADC at audio frequencies and perform AM modulation in software which is outputted through an R/R2 DAC into an antenna. Sample Audio: http://gtabfans.com/~andrew/upload/radio/premadonna%20girl%20cut.mp3 (received using a multi-band radio)
[22:35:12] <aarossig> the effective range is about 2m which I think is great considering I don't buffer the output of the R/R2 DAC (no op-amps available in my collection that are fast enough)
[22:58:25] <theBear> aarossig, that's pretty cool !