#avr | Logs for 2012-09-08

Back
[03:05:00] <nickoe> Hello, I know this is not for arduino... I have a arduino uno, and I don't want to use it as an "arduino" anyway. And I have problems making it work. http://dpaste.com/798098/
[03:05:52] <nickoe> When I press the reset button on it a led blinks fast a few times and and one blink after a small dealy. I guess that is the bootloader doing that.
[03:07:37] <specing> nickoe: replace first PORTB with PORTB ^= 0xff;
[03:07:46] <specing> and delete the second loop and assignment
[03:10:57] <nickoe> specing, I am not quite sure why you wan't to delete the seceond loop, http://dpaste.com/798100/ ?
[03:11:15] <nickoe> my main problem is that I cannot program the device
[03:17:37] <specing> nickoe: because it is redundant
[03:17:56] <nickoe> ahh, yeah, with that ^= assignment :)
[03:19:31] <nickoe> specing, do you have any suggestion to me about programming it
[03:20:51] <nickoe> This is the device I have http://arduino.cc/en/uploads/Main/ArduinoUno_r2_front.jpg
[03:21:01] <specing> nope
[03:21:24] <specing> strace the arduino IDE, that will show you what it executes
[04:13:48] <RikusW> seems like the Saturday Silence again :-P
[04:14:20] <Corwin> problem with that?
[04:19:16] <CapnKernel> Lalalalala!
[04:19:48] <Corwin> teletubbies Laa-Laa ?
[04:30:57] <RikusW> http://www.youtube.com/watch?v=uvG8pfzT0m0&feature=g-logo-xit
[04:40:35] <nickoe> Hmm... I am confused now.
[04:40:51] <Corwin> get used to it... this is #avr :)
[04:41:52] <nickoe> Corwin, http://dpaste.com/798123/
[04:42:36] <nickoe> It is an arduino uno board, where I am trying to run plain avr-gcc... but I can't seem to see anything shifting on the pisn wven with a scope.
[04:43:22] <Corwin> shifting what where?
[04:43:51] <RikusW> nickoe: so you upload the hex file using avrdude ?
[04:44:01] <nickoe> yeah
[04:44:06] <nickoe> see the paste
[04:44:45] <nickoe> Corwin, I made that simple program in that paste and compiles with avr-gcc and uploads with avrdude via a ISP progammer (usbasp)
[04:45:00] <nickoe> but I dont see any activity on the ports
[04:47:18] <Corwin> that piece of code really compiled to 7028 bytes of code?
[04:48:51] <nickoe> Corwin, you see the same figures as I can see...
[04:49:06] <nickoe> du -h blink.hex
[04:49:06] <nickoe> 8,0K blink.hex
[04:49:32] <OndraSterver> the hex is 8kB, not the code
[04:49:54] <nickoe> yes
[04:50:23] <Corwin> OndraSterver, it wrote 7028 bytes of flash memory... i just cant believe that those lines can compile to such huge flash
[04:51:12] <Kevin`> it normally wouldn't. maybe the makefile or whatever he's using has a bunch of libraries pulled in
[04:51:12] <OndraSterver> hmm weird, does avrdude report "hex" file size or real data size?
[04:51:27] <OndraSterver> still 8kB hex is awful for that long program
[04:52:48] <Kevin`> when I build something like that it's only a few hundred bytes, most if it c init code and isrs and such
[04:53:04] <nickoe> Corwin, I compile it as it says in the paste, I just did it on commandline
[04:53:22] <nickoe> Any makefile I should try or?
[04:53:54] <nickoe> I compile like, avr-gcc main.c -I. -g -DF_CPU=16000000 -mmcu=atmega328p -Os -o blink.hex
[04:54:05] <nickoe> but I don't know if that is a good idea
[04:55:25] <OndraSterver> (I am no gcc expert) but shouldn't it run through gcc and then through ld?
[04:55:29] <OndraSterver> or am I thinking too much
[04:55:54] <Xark> OndraSterver: Unless you specify -c, it will compile then link by default.
[04:55:57] <OndraSterver> ah
[04:56:08] <Xark> nickoe: What does avr-size say about the executable?
[04:57:07] <nickoe> text data bss dec hex filename
[04:57:07] <nickoe> 176 0 0 176 b0 blink.hex
[04:57:15] <OndraSterver> that sounds ok
[04:57:17] <Xark> nickoe: One issue, I didn't think gcc made hex files. I suspect your "hex" file is really a .o file. :)
[04:57:21] <Xark> (or elf file)
[04:57:35] <nickoe> ahh, that may be the problem...
[04:57:53] <nickoe> I just did the -o blink.hex...
[04:58:04] <nickoe> so how do I make it a hex file?
[04:58:40] <Xark> nickoe: Something like: avr-objcopy -O ihex -R .eeprom input.elf output.hex
[04:59:10] <nickoe> is input.elf the a.out file from avr-gcc?
[04:59:21] <Xark> Yes.
[04:59:28] <nickoe> :D will try
[05:00:57] <nickoe> YEAHH! Now somehting happens. I have used all morning trying to figure this out. Thank you very much!
[05:03:06] <nickoe> Xark, thank you for pointing that exact thing about the hex out
[05:03:17] <Xark> NP. :)
[05:03:36] <Xark> I am surprised avr-dude didn't complain loudly about the non-hex file. ;)
[05:04:01] <nickoe> And that also made the hex 170 bytes
[05:04:14] <Xark> nickoe: Sounds much better too. :)
[05:04:34] <nickoe> Xark, it actually kind of did when I made the delay with _delay_ms() instead of _delay_cycle_2()
[05:05:04] <nickoe> But I did not think more about it, and just worked on with delay cycle 2
[05:06:02] <nickoe> Now I have to try to program a bootloader on it. Any suggestions for one, wiht fast programming speed preffearably
[05:07:37] <Xark> optiboot (the one used on Uno). Only .5K and uses 2X serial mode.
[05:07:45] <Xark> http://code.google.com/p/optiboot/
[05:08:28] <nickoe> seems nice
[05:08:48] <Xark> I have used it on bare AVRs and it works nicely.
[05:09:12] <Xark> (and upgraded some older Arduino clones with slower bootloaders with it).
[05:10:07] <RikusW> Flashing .o files is BAD :-P
[05:11:02] <Fleck> hi guys, how to you solder those chips with big ground connection to the board at back side? if I place solder on board - chip legs doesn't stick... too short... chip like this:
[05:11:03] <Fleck> http://i00.i.aliimg.com/img/pb/318/868/483/483868318_564.JPG
[05:11:28] <nickoe> RikusW, yeah I figured... :)
[05:11:31] <RikusW> thas qfn....
[05:11:41] <RikusW> Fleck: you'll need hot air
[05:11:49] <Fleck> got hot air
[05:11:53] <RikusW> and flux and preferable solder paste
[05:12:03] <RikusW> flux is essential
[05:12:04] <Fleck> solder paste?
[05:12:08] <Fleck> got flux
[05:12:16] <RikusW> there is vids on the net
[05:12:28] <RikusW> the ic is a qfn pacakge
[05:12:40] <Fleck> its not like that, but similar
[05:13:23] <OndraSterver> check datasheet, the middle plate doesn't have to be always soldered on
[05:13:36] <Fleck> I did, it does!
[05:13:45] <OndraSterver> ok
[05:13:48] <OndraSterver> then hotair :)
[05:14:11] <Fleck> yeah but i dont have solderpaste...
[05:15:05] <nickoe> Xark, what is the .lst file in optiboot?
[05:15:06] <RikusW> you can do it with ordinary solder and flux
[05:15:16] <Fleck> how RikusW ?
[05:15:29] <RikusW> I don't have a link, but there is video's showing you how
[05:15:33] <OndraSterver> yep
[05:15:42] <OndraSterver> put a bit of tin on the midle pad and each pad
[05:15:44] <OndraSterver> -middle
[05:15:45] <Xark> nickoe: I believe that is an assembler listing (i.e., human readable output from assembler).
[05:15:50] <OndraSterver> put the chip on ti
[05:15:50] <nickoe> Hmm, I guess that is actually the source for the bootloader in asm
[05:15:51] <RikusW> basically put a bit of solder on the pads, and not too much on the center pad
[05:15:53] <OndraSterver> and reflow it
[05:15:56] <nickoe> ok
[05:15:58] <RikusW> and flux too
[05:16:07] <RikusW> the flux is really important
[05:16:19] <nickoe> Xark, so this should do the programming of ti, avrdude -p m328p -c usbasp -U flash:w:optiboot_atmega328.hex
[05:16:59] <Xark> nickoe: I don't think you can use a bootloader to load a bootloader. :) I believe you need an ISP programmer.
[05:17:20] <nickoe> yes, that is what I am doing -c usbasp
[05:17:33] <OndraSterver> nope, BL can't load a BL unless the first one "is capable of self-modyfing"
[05:17:37] <Xark> nickoe: Also, you probably need to tweak some fuses for optiboot (mostly to set the boot loader size to .5k).
[05:17:42] <OndraSterver> (I don't think anybody has done that on AVR, it is too much)
[05:18:30] <Xark> OndraSterver: Also, isn't that area is protected (depending on fuses)?
[05:19:25] <Xark> nickoe: Ahh, yes, with usbasp I think you are close. As I mentioned you may need to tweak the bootloader size fuse.
[05:21:11] <nickoe> size fuse... what exactly is that?
[05:22:55] <Xark> nickoe: See http://www.engbedded.com/fusecalc/ I believe you want the "boot start address=$3F00 [256 words]" for optiboot.
[05:22:55] <RikusW> Aself modifying bootloader ? .... could be done, but why ?
[05:24:03] <nickoe> RikusW, because I have none... because I just flashed a program with a isp over ot
[05:27:05] <OndraSterver> Xark, depending on fuses, yes :)
[05:27:07] <nickoe> Xark, ... how fast it it ... I cant find the required speed, but I think I read it somewhere.... grr
[05:27:56] <RikusW> bootloaders tend to be much faster than ISP
[05:28:04] <nickoe> really?
[05:28:07] <RikusW> yes
[05:28:22] <nickoe> my usbasp is pretty fast
[05:28:36] <nickoe> waht baud rate does optiboot use?
[05:28:38] <RikusW> bootloading is faster
[05:28:41] <RikusW> not sure
[05:28:46] <nickoe> it is not stated in the readme...
[05:28:47] <megal0maniac> 115200
[05:28:51] * Xark notes Atmel AVRISPMkII is much faster than his USBTiny (and I thought bootloaders...)
[05:29:22] <RikusW> I have https://sites.google.com/site/megau2s/home and its really fast
[05:29:26] <nickoe> yay, that was the figure.
[05:29:35] <nickoe> RikusW, and it went pretty fast
[05:29:40] <RikusW> even at 115200 with a bootloader its faster than ISP
[05:29:56] <nickoe> yeh
[05:30:01] <Corwin> RikusW, define "really fast" in term of writing speed
[05:30:01] <megal0maniac> Most other BLs use 57600
[05:30:15] <RikusW> Xark: I mean a bootloader using 115200bps or USB
[05:30:23] <Xark> nickoe: I notice Optiboot is noticeably faster with USB2Serial (Atmega16U2 based) vs FTDI serial for some reason.
[05:30:43] <RikusW> Corwin: faster than ISP
[05:30:46] <nickoe> ok
[05:31:02] <megal0maniac> Xark: The actual programming shouldn't be any faster, might have different timeouts though or more optimised firmware on the 16u2
[05:31:06] <Xark> RikusW: Interesting. I need to time a test, but I think my USB AVRISPMkII is significantly faster than optiboot...
[05:31:28] <RikusW> maybe running at 500kHz or more SPI speed
[05:31:44] <Xark> megal0maniac: Well, it is approx twice as fast uploading the same Arduino sketch on the exact same board (just switching USB 2 serial dongle).
[05:32:10] <RikusW> but then ISP wiring and AVR clock needs to be high enough, and CKDIV8 disabled if you want 2MHz ISP clock
[05:32:13] <Xark> megal0maniac: Like 11 sec vs 6 seconds on the 30K file I was testing at the time.
[05:32:21] <megal0maniac> ISP frequency doesn't default to maximum.
[05:32:32] <RikusW> for good reason
[05:32:43] <RikusW> high ISP clock cause problems
[05:32:55] <nickoe> well the bootloader works :)
[05:32:57] <megal0maniac> But it can go very quickly if you tell it to :)
[05:32:59] <RikusW> megal0maniac: I set my U2S default to 115k (actually 125k)
[05:33:11] <RikusW> but it usually works on 500k too
[05:33:12] <nickoe> I have not changed the fuse bits, do I need to when it works?
[05:33:28] <RikusW> 2MHz is unstable.... unless you setup it just right
[05:33:43] <Xark> nickoe: I don't think it is critical, you just waste 1.5K of flash space AFAIK.
[05:33:57] <megal0maniac> Took about 32s yesterday to verify 128kb...
[05:33:57] <RikusW> by that I mean, AVR clock is at least 8MHz and no extra wires connected to ISP wiring
[05:34:01] <nickoe> ok
[05:34:13] <RikusW> megal0maniac: by bootloader or programmer ?
[05:34:19] <nickoe> Xark, then I will wait to have a look at that. :)
[05:34:53] <megal0maniac> Programmer
[05:34:58] <RikusW> mine ? :)
[05:35:12] <RikusW> and what speed ?
[05:35:19] <megal0maniac> It seemed to write very quickly, but verification took ages.
[05:35:44] <megal0maniac> Yes, yours. And I suppose 125K
[05:35:54] <RikusW> yeah, ASt reads the entire chip unfortunately
[05:36:07] <RikusW> you could change it to 500k by default
[05:36:17] <megal0maniac> I think lufa avrisp took equally long
[05:36:54] <RikusW> the SCK speed in the settings file needs to be 1 or 2 I think, 0 1 2 3 4 is hardware SPI, higher numbers is slower soft SPI
[05:37:14] <megal0maniac> Not really complaining, 128kb is huge :)
[05:37:19] <RikusW> megal0maniac: my board and LUFA is faster than the dragon ISP.....
[05:37:35] <OndraSterver> I don't think I have ever used my Dragon as ISP lol
[05:37:40] <RikusW> and LUFA slightly faster on some speeds than mine
[05:37:44] <OndraSterver> well, as ISP yes, but through JTAG interface
[05:37:52] <nickoe> OndraSterver, what is the Dragon really for?
[05:38:02] <RikusW> OndraSterver: the dragon ISP and JTAG pins arre connected
[05:38:05] <OndraSterver> it is cheap debugger
[05:38:11] <OndraSterver> RikusW, some say so
[05:38:19] <nickoe> through JTAG interface?
[05:38:20] <RikusW> I measured it
[05:38:23] <OndraSterver> nickoe, yes
[05:38:36] <OndraSterver> RikusW, well, some websites say that too, some say that on the newer one not anymore
[05:38:47] <nickoe> Why is there all those holes in the borad then?
[05:38:54] <OndraSterver> which holes?
[05:39:09] <RikusW> Dragon -> MOSI + TDI MISO + TDO SCK + TCK Reset - Reset TMS....
[05:39:17] <nickoe> the coloums on the right, http://www.bfrdesign.com/blog/dragon/IMG_0333.JPG
[05:39:28] <OndraSterver> it would make sense for them to be connected
[05:39:35] <RikusW> nickoe: for HVPP programming and the 40pin DIP socket
[05:39:37] <nickoe> to rescue chips woth wrong fuses?
[05:39:37] <OndraSterver> yep
[05:39:48] <OndraSterver> you can put there ZIF socket or whatever and wire it up
[05:40:00] <RikusW> nickoe: but my U2S board can do HVPP/SP too
[05:40:06] <RikusW> its a STK500 clone
[05:40:15] <RikusW> and jtagice mki clone too
[05:40:28] <nickoe> RikusW, what is your U2S board0+
[05:40:40] <RikusW> megal0maniac: I uploaded the STK500 support files https://sites.google.com/site/megau2s/home/supporting-software/AS6.tbz?attredirects=0&d=1
[05:40:44] <megal0maniac> It's magical :P
[05:40:53] <RikusW> nickoe: https://sites.google.com/site/megau2s/home/
[05:41:06] <RikusW> Atmel was lazy....
[05:41:15] <RikusW> so I did the job instead...
[05:41:18] <nickoe> ohh
[05:41:45] <RikusW> nickoe: I got hold of the stk and jtag fw btw
[05:42:17] <megal0maniac> RikusW: I was online when you posted it :)
[05:42:18] <RikusW> made things much easier, and its quite compatible too :)
[05:42:37] <megal0maniac> Atmel was lazy?
[05:42:54] <RikusW> yeah, why didn't they put in those files ? :S
[05:43:07] <RikusW> I'll have to notify abcminiuser about it :-P
[05:43:08] <megal0maniac> Oh, for the files :P I thought you meant your board
[05:43:14] <RikusW> the files
[05:43:17] <megal0maniac> RikusW: Tell memoserv
[05:43:26] <RikusW> good idea
[05:44:09] <RikusW> just did
[05:44:36] <megal0maniac> RikusW: I still haven't figured out how to set this stuff up:https://sites.google.com/site/megau2s/home/uart
[05:44:56] <megal0maniac> Where is parameter 0x99 and how do I change it?
[05:44:59] <RikusW> its for advanced settings on the uart
[05:45:16] <RikusW> SetParameter(0x99,value);
[05:45:35] <RikusW> in bootloader mode before changing mode
[05:45:58] <megal0maniac> With the debug app or u2scli?
[05:46:28] <RikusW> no sw to do that exists yet, but SetParameter is in my source afaik...
[05:46:54] <megal0maniac> Ah, alright
[05:47:18] <RikusW> basically I load the last 16 bytes of eeprom straigh into the parameter values on powerup
[05:47:49] <megal0maniac> And then they live in RAM.
[05:48:04] <megal0maniac> But you could make them persistent by changing them in eeprom?
[05:48:21] <RikusW> and can be modified by SetParameter or GetParameter of the STK500 protocol
[05:48:25] <RikusW> yes
[05:49:24] <RikusW> afaik Parameter 0x99 is located at 0x117 (hopefully still is)
[05:49:44] <RikusW> that means you can change it from debug mode instead of SetParameter
[05:50:00] * RikusW needs to write some more support sw for U2S ;)
[05:50:19] <RikusW> megal0maniac: have you read AVR068 yet ?
[05:50:30] <megal0maniac> :$
[05:50:32] <RikusW> its a appnote on atmel.com
[05:50:37] <megal0maniac> Instruction set?
[05:50:45] <RikusW> it explains STK500v2 protocol
[05:50:51] <megal0maniac> Oh
[05:50:52] <megal0maniac> No
[05:51:10] <RikusW> thats only if you want to get into the details though
[05:51:19] <RikusW> not required reading
[05:51:47] <RikusW> I'll move all functionality over to the RavrProg lib in time
[05:52:01] <RikusW> the current U2S_Debug files is hackish...
[05:52:04] <OndraSterver> Rikus's AVR Prog? :)
[05:52:06] <RikusW> but it does work
[05:52:16] <RikusW> yes indeed :)
[05:52:24] <RikusW> or RaverProg :-P
[05:52:29] <OndraSterver> :D
[05:52:32] <OndraSterver> RoverProg!
[05:53:22] <RikusW> megal0maniac: the code in RavrProg STK500 protocol handling is more robust
[05:53:47] <megal0maniac> RikusW: I have so much stuff I need to read and figure out :/
[05:54:52] <RikusW> megal0maniac: the U2S.h and cpp files handles the protocol for Debug mode..
[05:55:06] <RikusW> for now anyways
[05:55:41] <megal0maniac> Are you planning on making a lot of changes in the future?
[05:56:19] <RikusW> IOPort Par99(0x117); that should allow you to do Par99 = 0x10; or whatever
[05:56:40] <RikusW> not to the U2S fw anymore, thats more or less stable now :)
[05:56:48] <RikusW> and bug free afaik
[05:57:08] <megal0maniac> Seems to be, so far
[05:57:17] * RikusW don't want to scratch where id don't itch, it just might start itching....
[05:57:21] <RikusW> *it
[05:57:38] <megal0maniac> DTR isn't implemented in UART, right?
[05:57:54] <RikusW> it is
[05:58:09] <RikusW> RTS/CTS isn't in the CDC spec unfortunately
[05:58:21] <RikusW> but HW RTS/CTS can be turned on
[05:58:25] <megal0maniac> Ah, but that's okay. Handshaking isn't necessary :)
[05:58:33] <megal0maniac> Not for me, anyways..
[05:59:00] <megal0maniac> Where is DTR then?
[05:59:06] <RikusW> changing DTR in Windows should make it alter on the board
[05:59:51] <megal0maniac> But where on the board?
[06:00:12] <RikusW> // D0 DTR // D1 RTS
[06:00:27] <RikusW> should put that on the site..
[06:00:56] <megal0maniac> Ah :)
[06:01:01] <megal0maniac> Yes.
[06:10:46] <megal0maniac> RikusW: What IDE do you use for the C source?
[06:12:40] <RikusW> megal0maniac: for example SetParameter(0x99,2); SetParameter(0x94,U&0xFF); SetParameter(0x95,((U>>8)&0xFF)); in boot mode for custom UBRR values
[06:12:46] <RikusW> AS4
[06:13:06] <RikusW> and Visual Studio 97 mostly but VS 2010 should work too
[06:13:21] <RikusW> VS 97 is fast, and I actually own it
[06:13:37] <megal0maniac> VS2010 (express) is giving me grief. Will try out 2008
[06:16:16] <RikusW> megal0maniac: SetParameter(0x99,2); SetParameter(0x94,U&0xFF); SetParameter(0x95,((U>>8)&0xFF)); SetMode(0x84); <<-----
[06:16:25] <RikusW> remeber SetMode
[06:16:29] <OndraSterver> bollocks linux having separate C and C++ compilers! lol
[06:16:33] <OndraSterver> and then wondering why it fails
[06:16:43] <megal0maniac> Will do :)
[06:16:51] * megal0maniac copy pastes
[06:17:03] <megal0maniac> (Into a text file for reference purposes)
[06:17:04] <Xark> RikusW: VS 97? Is that VC++ 6.0?
[06:17:16] <RikusW> 5
[06:17:32] <RikusW> it still works, but since XP SP2 the help is broken :(
[06:17:49] <RikusW> VS97 help is better than VS2010 imo
[06:18:43] <RikusW> megal0maniac: updated the site too, I'm on 4MBit for a while :-D
[06:19:00] <megal0maniac> IS DSL, I noticed :P
[06:19:11] <megal0maniac> You away from home?
[06:19:20] <RikusW> guess I'll notice the GPRS slowdown when I'm using that again :(
[06:19:23] <Xark> RikusW: I find Google is faster and better than any of the VS help systems (and 2008+ HELP2 is the slowest thing known to man).
[06:19:29] <megal0maniac> Not on IRC :P
[06:19:30] <RikusW> megal0maniac: in Bloemfontein
[06:19:48] <RikusW> Xark: VS97 help is fast
[06:20:19] <RikusW> I don't know why M$ version numbers these days indicate degrees degrees of crappyness
[06:20:27] <Xark> RikusW: Perhaps but too old for me to use (and the C++ compilers are poor quality until 2008 or so).
[06:20:35] <RikusW> hmm
[06:21:08] <RikusW> megal0maniac: fortunately IRC don't use much data :)
[06:23:17] <megal0maniac> Even my old phone can handle it
[06:23:31] <nickoe> Is there a unique identifier on an atmega, that can be accesed for the software you put on it?
[06:23:47] * megal0maniac looks forward to more supporting software
[06:24:36] <nickoe> OndraSterver, with some flag you can somehow make it automagically call the c++ compiler when needed I read the otehr day on wome gcc documentation...
[06:24:52] <OndraSterver> o_O
[06:25:40] <RikusW> megal0maniac: https://sites.google.com/site/megau2s/home/uart fnished updating
[06:26:01] <RikusW> nickoe: like serial number or signature of the AVR ?
[06:27:37] <nickoe> RikusW, yeah like serial number
[06:27:45] <RikusW> m32u2 does have one
[06:27:56] <RikusW> but not all AVRs does
[06:28:02] <RikusW> you can easily add one ?
[06:28:10] <RikusW> maybe in eeprom ?
[06:28:27] <megal0maniac> RikusW: While you have faster internet, have a look at this: http://www.kickstarter.com/projects/paulstoffregen/teensy-30-32-bit-arm-cortex-m4-usable-in-arduino-a
[06:29:29] <nickoe> RikusW, I don't know how to easyly dot it... and with eeprom.
[06:29:57] <mbertens> RikusW: solved de problem with the corrupted characters; i added a delay of 1/2 mS and that solved the problem, i dont know how or why yet, but it dud
[06:30:04] * Xark thought that new ARM Teensy looked interesting and tossed a few bucks Paul's way for an early one...
[06:30:12] <nickoe> it would have been easier if I could just read a specefic register on the device that was read only
[06:30:15] <megal0maniac> nickoe: What do you need it for?
[06:31:05] <megal0maniac> mbertens: Probably just means that the avr was sending commands too quickly for the glcd to receive them.
[06:31:24] <megal0maniac> Or at least to "hear" them
[06:31:31] <nickoe> Well it is not really very important, it could just be nice to be able to identify one unit by sending a command to it, and it could tell me the number, just for logging and knowing when the unit was made.
[06:31:46] <mbertens> megal0maniac: the corrution was on the serial port not the lcd
[06:32:17] <megal0maniac> nickoe: Well then eeprom is perfect, if you aren't using it for anything else. And if you are, then just use the high bits
[06:32:57] * mbertens is very happy with the font implementation on the gLCD.
[06:33:23] <nickoe> megal0maniac, but what would you suggest to do that easily. I am not particulary familiar with the eeprom
[06:35:47] <RikusW> megal0maniac: the teensy 3 does look nice indeed\
[06:36:15] <megal0maniac> nickoe: Not sure of the code, ask someone more familiar with avr C, or Google :)
[06:37:36] <nickoe> :)
[06:37:39] <Xark> nickoe: If you use avr-libc, I think you can just use the handy API -> http://www.nongnu.org/avr-libc/user-manual/group__avr__eeprom.html
[06:37:44] <nickoe> megal0maniac, what do you use then?
[06:38:02] <nickoe> looks good
[06:39:02] <megal0maniac> nickoe: Only worked with eeprom in arduino, so it's as simple as eeprom.write(addr, val)
[06:39:22] <nickoe> Xark, what does the eeprom_update_ do?
[06:39:29] <nickoe> different from the write
[06:40:57] <Xark> nickoe: I believe write does erase+write, update assumes already erased (but not positive about this).
[06:41:15] <megal0maniac> nickoe: This functions read each byte first and skip the burning if the old value is the same with new.
[06:41:15] <nickoe> ok
[06:41:21] <megal0maniac> Read the doc
[06:42:58] <megal0maniac> RikusW: I want to get a Teensy3, but I can't justify spending that money...
[06:43:38] <RikusW> hmm too bad
[06:44:02] <nickoe> megal0maniac, hmm, what is the price for it?
[06:44:16] <nickoe> I see no price on the kickstarter page
[06:44:18] <RikusW> $22 plus $6 shipping
[06:44:29] <RikusW> look at the funding on the right
[06:45:51] <nickoe> ahh, there, I just thought that was what the latest people have pledged
[06:54:26] <RikusW> megal0maniac: updated the USART page again
[06:54:43] <RikusW> its just too easy when using a 4MBit ADSL line :)
[06:57:31] <megal0maniac> Haha! I know the feeling. At work, in the evenings when it's quiet, I have 10mbps to myself. Funny thing is that most of the people here would be complaining if they "only" had 10mbps :P
[06:58:03] <megal0maniac> Richard_Cavell: You get your stuff yet?
[06:58:12] <Richard_Cavell> Yeah baby
[06:58:16] <Richard_Cavell> Sitting right here
[06:58:20] <megal0maniac> How's it going?
[06:58:23] <Richard_Cavell> Everything I need to power the Universe with AVR
[06:58:39] <Richard_Cavell> Well I'm not looking at it yet. Need to finish off my study of 68000 assembly and subversion
[07:00:46] <OndraSterver> is anybody selling DeLorean?
[07:00:48] <OndraSterver> I want one
[07:00:48] <OndraSterver> badly.
[07:00:58] <OndraSterver> I don't need flux caps at all
[07:01:02] <OndraSterver> I am fine with the car itself :D
[07:05:16] <RikusW> megal0maniac: look at https://sites.google.com/site/megau2s/home/u2s-settings
[07:05:29] <RikusW> there is the parameter to setting mapping
[07:10:06] <RikusW> megal0maniac: 0x90-0x9E. (0x10E-0x11C in SRAM)
[07:11:59] <RikusW> megal0maniac: and updated the USART page again..... :-P
[07:12:14] <RikusW> Think I've got it all now
[07:14:33] <RikusW> megal0maniac: You do know there is a clock generator on B7 right ?
[07:15:20] <megal0maniac> I figured that out a few seconds ago :P One of the few things that isn't going over my head...
[07:15:47] <RikusW> can be enabled using AS
[07:16:06] <megal0maniac> But must be off for HVPP/SP
[07:16:09] <RikusW> very usefull for incorrectly set clock fuses (usually lfuse)
[07:16:21] <RikusW> because it shares a HVPP line...
[07:16:35] <megal0maniac> I remember :)
[07:16:52] <megal0maniac> How does one use the "recovery clock"?
[07:17:12] <RikusW> just connect it to the XTAL1 pin
[07:17:33] <RikusW> and enable it in AS, it must be at least 4 times higher than the ISP clock
[07:17:38] <RikusW> preferably x5
[07:17:43] <RikusW> or more
[07:17:59] <RikusW> and remember CKDIV8 might be set....
[07:18:10] * megal0maniac remembers something about nyquist frequency
[07:18:30] <RikusW> its actually to do with the way AVR filters input pins
[07:18:38] <RikusW> it needs 4 clocks
[07:19:01] <OndraSterver> haha nyquist
[07:20:02] <megal0maniac> Ah, okay
[07:20:24] <RikusW> there is something about it in the datasheets
[08:17:11] <megal0maniac> zlog
[08:30:42] <megal0maniac> fuzzybugkilla: http://myxboard.net/boards.html
[08:30:58] <Corwin> ehm..... any sensible ideas how to utilize old IDE drivers for some project... except "speakers" ?
[08:31:08] <OndraSterver> Corwin, hook 'em up to AVR :)
[08:31:42] <Corwin> ... atmega328 with 40gb eeprom storage? :D
[08:31:51] <OndraSterver> with 40GB flash storage!
[08:32:02] <OndraSterver> although eeprom has more overwrites
[08:32:08] <OndraSterver> so yes, eeprom storage :D
[08:32:23] <OndraSterver> why not... you can make file server with it!
[08:32:28] <OndraSterver> ENC28J60 ethernet..
[08:32:32] <OndraSterver> $6.5 from ebay :D
[08:32:47] <Corwin> AVR powered NAS drive :D
[08:32:50] <OndraSterver> :D
[08:33:19] <OndraSterver> AVR datacenter!
[08:33:30] <OndraSterver> takes only 1kW to run 500k cores! :D
[08:33:45] <OndraSterver> that is actually a neat idea lol
[08:33:57] <OndraSterver> each core doing only single webpage :o
[08:34:23] <Steffanx> Warning: fantasy overload
[08:34:28] <OndraSterver> :(
[08:34:30] <Corwin> :D
[08:35:00] <Corwin> 7 floppy drives, 3.5inch
[08:35:13] <OndraSterver> mm
[08:35:17] <OndraSterver> you know what you have to do with those
[08:35:25] <Corwin> not really
[08:35:29] <OndraSterver> you do!
[08:35:36] <Corwin> throw them away ?
[08:35:45] <OndraSterver> http://www.youtube.com/watch?v=yHJOz_y9rZE
[08:35:47] <OndraSterver> but.. 7!
[08:35:59] <OndraSterver> but... with 7! *
[08:37:42] <raven> hi
[08:37:55] <OndraSterver> hi
[08:38:11] <Corwin> hmm... 3 pins only are required to controll the floppy if i read it right
[08:38:43] <OndraSterver> .. wrong button
[08:38:59] <raven> Erlkoenig?
[08:39:16] <Erlkoenig> raven?
[08:39:16] <OndraSterver> I have got here some peripheral connected to my PC that randomly gets stuck CTRL key... or it is maybe my keyboard ... :D
[08:39:45] <Steffanx> Your fingers
[08:39:52] <Steffanx> Time to clean your keyboard
[08:40:08] <raven> Erlkoenig would you like to try avrdude with avrispmkii and linux?
[08:40:14] <OndraSterver> better my fingers than salad fingers
[08:40:22] <Erlkoenig> i don't have an avrispmkii
[08:40:27] <raven> you asked yesterday about my problem ;)
[08:40:30] <megal0maniac> Corwin: Yip, only 3
[08:40:52] <Erlkoenig> ah your mkii didn't work on linux right?
[08:41:05] <raven> Erlkoenig yes
[08:41:26] <raven> perhaps you have additional ideas
[08:41:26] <Erlkoenig> what error message does avrdude display?
[08:41:42] <raven> do you speak german?
[08:42:01] <megal0maniac> Jawol!
[08:42:03] <Erlkoenig> jopp
[08:42:12] <Erlkoenig> gip mal die fehlermeldung vom avrdude
[08:42:27] <Corwin> 3 wires per floppy drive, it makes it 6 drives connectable directly to arduino board.... i have 3 boards.... need more floppy drives
[08:43:23] <megal0maniac> You could actually control two of them with an attiny25/45/85 if you use the reset pin as an IO
[08:44:46] <Corwin> back to digging in stuff
[08:52:10] <Corwin> weird... i just found cartdrige from Nintendo gameboy with Super Mario Land 3
[08:55:04] <CapnKernel> W00t!
[08:55:19] <Corwin> made in 1989... with battery backup inside
[08:55:45] <CapnKernel> If the battery's no good, go ask for your money back.
[08:55:50] <Corwin> :D
[08:56:13] <Corwin> cant... since i never paid for this
[08:56:33] <Corwin> i actually never owned nintendo
[08:56:39] <Corwin> no idea where i got this
[08:59:53] <megal0maniac> Battery backup??
[09:00:16] <Corwin> it has SRAM chip
[09:00:30] <Corwin> needs backup power when cartridge not in use
[09:03:11] <megal0maniac> What is it for? Didn't know that NES cartidges had SRAM...
[09:03:18] <megal0maniac> Especially persistent SRAM
[09:05:19] <timemage> some of the cartages had extra hardware on them. zelda is another example.
[09:05:38] <Corwin> http://gameboyprogramming.blogspot.cz/2010/01/update.html
[09:05:42] <Corwin> ^ this shows whats inside
[09:06:19] <timemage> sorry, i thought you were talking about nes.
[09:07:40] <megal0maniac> So did I :P
[09:08:06] * megal0maniac re-reads
[09:08:07] <timemage> megal0maniac, ah, well, in that case my statement stands =)
[09:08:36] <Corwin> nintendo gameboy, not NES
[09:12:37] <OndraSterver> I have got here two games for NES
[09:12:44] <OndraSterver> but I am afk
[09:49:12] <nickoe> Xark,
[09:49:16] <nickoe> woops
[09:49:55] <nickoe> ohh, hmm, it was you. But shouldn't optiboot be able to return the device signature?
[09:56:43] <megal0maniac> It does
[10:01:22] <nickoe> megal0maniac, hmm, I have some trouble then
[10:02:39] <nickoe> This is me programming throught the bootloader (optiboot) http://dpaste.com/798193/ , it works, but it still complains about the device signarute is 0.
[10:03:59] <megal0maniac> Then it's just a quirk of your bootloader
[10:04:38] <megal0maniac> If it works, then I wouldn't worry
[10:04:56] * nickoe trying to flash bootloader again
[10:05:06] <megal0maniac> Optiboot, for me, returns a signature.
[10:05:50] <nickoe> megal0maniac, how do you test? avrdude -p m328p -P /dev/ttyACM0 -c stk500v1 -b 115200 ?
[10:07:38] <megal0maniac> Can't remember. Think I used something else for -c
[10:08:07] <nickoe> ok
[10:08:27] <megal0maniac> -c arduino
[10:08:42] <megal0maniac> Not sure if that's a standard option with avrdude, but try it
[10:09:29] <nickoe> hmm, that gives me a device signature...
[10:09:52] <megal0maniac> Then use that. It's probably more compatible
[10:09:57] * megal0maniac shrugs
[10:10:05] * megal0maniac is AFK
[10:10:31] <nickoe> :)
[10:11:32] <megal0maniac> Pleasure
[10:41:52] <xaxes> hey all.. I have problems with my jtag mkII clone .. it seems that something is wrong with it. avrdude (wine) is able to synchronize after startup of the clone. But then it can't synchronize again.. similar behavior with avarice http://paste.frubar.net/15105 .. but using avrdude I have no problems (on 19200 baud).. any ideas?
[10:42:36] <xaxes> using avrdude I can set fuses/program etc, but I would like to use the debugging mechanism
[10:57:01] <nickoe> xaxes, why do you use avrdude in wine?
[10:57:30] <xaxes> nickoe: oh, sorry, I meant avr studio ^^
[10:57:41] <xaxes> avr studio in wine, others on linux
[11:05:32] <nickoe> xaxes, no idea, sorry
[11:05:38] <xaxes> pitty
[11:22:20] <xaxes> mhh, what's the default baudrate of jtagice mkI ? I have to use 19200 .. shouldn't it be more than that? maybe here is an error cause
[11:34:41] <nickoe> don't know :/
[11:35:51] <xaxes> mhh, Im using the firmware provided by avrstudio, so I think software should be okay.. maybe I havn't soldered a part correctly so the communication isn't stable.. but in this case, the avrdude stuff should fail too..
[11:36:02] <xaxes> s/shouldnt/should/↵
[12:13:58] <orlyyy> hello
[12:14:07] <orlyyy> anti retro virals
[12:30:27] <tomatto> hi
[12:31:02] <tomatto> is vusb using any timer in avr or i can use all of them for myself?
[12:38:59] <JyZyXEL> so i was trying to compile "owslave" (https://github.com/smurfix/owslave) to ATMega8, the makefile had MCU=atmega168 by default but it also had MCU=atmega8 commented out, so i uncommented it and tried to "make" but got this: onewire.c:170: error: ‘EIFR’ undeclared (first use in this function)
[12:39:26] <JyZyXEL> someone happen to know off the bat what might be wrong?
[12:42:19] <RikusW> EIFR register is not present on the m8 or the name is a little different
[12:42:42] <Erlkoenig> JyZyXEL: the ATmega8 doesn't have the EIFR register, because it doesn't have external interrupts (only INT0 and INT1). The ATmega88 has it
[12:43:16] <JyZyXEL> well tahts weird o.o
[12:43:30] <JyZyXEL> this is supposed to be m8 compatible
[12:44:13] <Erlkoenig> seems to be not
[12:44:36] <Erlkoenig> maybe the program has some #define's you have to adjust so it uses INT0/1
[12:44:52] <Erlkoenig> or maybe it's just mega88 compatible, because the m88 is the new version of the m8
[12:45:39] <JyZyXEL> #define OWPIN PIND
[12:45:43] <JyZyXEL> #elif defined (__AVR_ATmega8__)
[12:45:51] <JyZyXEL> it had defined for the mega8
[12:58:58] <RikusW> specing: http://en.wikipedia.org/wiki/Wayland_%28display_server_protocol%29 heard of this ?
[12:59:48] <specing> yep
[13:03:39] <specing> it will increase 3D and 2D performance by atleast 5% and 10%, resp.
[13:04:11] <specing> Though I don't like the idea of window managers talking directly to the underlying hardware
[13:04:23] <specing> unless they wrap it all in the kernel (DRM)
[14:31:58] <tomatto> can i use 18pF capacitors with 12-25MHz crystal?
[14:32:08] <tomatto> crystals
[14:32:21] <specing> tomatto: whatever works
[14:32:47] <specing> 15pf seem to work with pretty much anything
[14:32:49] <specing> 27pf not much
[14:32:52] <learningc> where do you guys buy parts cheaper?
[14:33:16] <specing> learningc: ebay, mouser,....
[14:33:29] <learningc> mouser is cheap?
[14:37:36] <tomatto> specing: i don't know if 15p/18p works, i haven't that, but i want to buy it, when you tell me, that it is good for these crystals
[14:38:03] <specing> 15pf works for me (tm)
[14:40:20] <tomatto> specing: and what range of crystal did you used with?
[14:40:33] <Rikus1> 20MHz
[14:41:21] <specing> tomatto: 12-20M
[14:42:23] <tomatto> so it must work with 18pF too
[14:42:34] <Roklobsta> mega0maniac: have you tried http://www.codelite.org/ ?
[14:46:26] <tomatto> so 15p or 18p?
[14:48:07] <Rikus1> 18p or 22p
[14:48:26] <Rikus1> I consider 18p to be better fo 16 - 20MHz
[14:54:55] <tomatto> and for 12 and 1.7456?
[14:55:00] <tomatto> <4.7456
[14:55:04] <tomatto> 14.7456
[14:57:19] <Rikus1> sligthly larger for lower clocks, try 22p ?
[14:57:20] <Rikus1> there is some appnotes about it on atmel.com
[14:59:10] <Rikus1> tomatto: try reading AVR040 and AVR042
[14:59:33] <Xark> tomatto: The capacitor value is generally specified with the crystal (and can vary).
[15:01:22] <Rikus1> When using the clock option “ext. crystal oscillator”, crystals with a nominal frequency
[15:01:23] <Rikus1> from 400kHz and up can be used. For these standard “high” frequency crystals the
[15:01:23] <Rikus1> recommended capacitor value is in the range 22-33pF.
[15:52:52] <tomatto> can LUFA works without hardware usb in avr?
[15:55:08] <Xark> tomatto: AFAIK, you need V-USB for that.
[15:55:34] <OndraSterver> no built-in V-USB into LUFA yet? :P
[15:55:41] <OndraSterver> DUSB!
[15:55:43] <OndraSterver> Dean's USB
[15:56:52] <tomatto> so lufa is only for usb based avrs?
[15:56:59] <specing> yes
[15:57:27] <OndraSterver> note that with V-USB you must use 12MHz xtal
[15:58:28] <Xark> OndraSterver: Or an ATTiny with internal xtal for some uses (quite a hack how it works, using USB clock to calibrate internal osc). :)
[15:59:06] <tomatto> any way to use 8MHz clock for V-USB?
[15:59:11] <OndraSterver> no
[15:59:15] <OndraSterver> not enough clocks available
[15:59:54] <tomatto> someone familiar with this linux kernel usb warning http://pastebin.blesmrt.net/2654/13471361/ ?
[15:59:56] <Xark> tomatto: ATTiny gets away with it because it has a PLL unit on its internal osc (most AVRs don't).
[16:00:42] <OndraSterver> Xark, tinys have internal xtals? I thought that only RC
[16:00:47] <OndraSterver> oh
[16:00:53] <OndraSterver> I re-read it whole now :)
[16:01:03] <tomatto> Xark: which it is exclyding attiny2313, isn't it?
[16:01:25] <Xark> tomatto: Correct. ATtiny25/45/85 are the ones I have seen it on.
[16:04:00] <Kevin`> tomatto: you can on some chips adjust the internal oscillator to an appropriate frequency
[16:04:44] <Kevin`> tomatto: using the normal offset control. out of spec, but worth trying for one-off projects
[16:05:32] <tomatto> Kevin`: it isn't attiny2313 right? on what chips?
[16:07:01] <Xark> tomatto: You can try it. Basically you use the internal osc (8Mhz) but then "crank up" the clock calibration register and you can sometimes get a nice speed up (~12Mhz). I would be dubious of it being stable enough for VUSB though...
[16:08:46] <tomatto> Xark: what is name of that register?
[16:09:20] <Xark> OSCCAL
[16:09:35] <Kevin`> tomatto: attiny2313 should be able to do it
[16:10:20] <Kevin`> tomatto: you want to use the "12.5mhz" low accuracy selection for building of course (and that frequency)
[16:10:58] <Kevin`> 12.8 rather
[16:13:46] <tomatto> i will try
[16:14:14] <tomatto> thx
[16:16:27] <tomatto> Kevin`: it is by percentage?
[16:16:36] <Kevin`> ?
[16:17:10] <Kevin`> for the low-accuracy options, the clock needs to be 1% accurate
[16:19:46] <tomatto> how can i calculate value to OSCCAL register?
[16:28:03] <Kevin`> tomatto: use the code that calibrates it based on incoming usb traffic
[16:28:40] <tomatto> oh, ok
[16:34:37] <tomatto> Kevin`: i can't find virtual serial in vusb, can you?
[16:34:56] <Kevin`> I do remember it being there
[16:35:22] <Kevin`> tomatto: no cdc serial?
[16:35:53] <Kevin`> http://vusb.wikidot.com/usb-device-classes
[16:37:23] <Tom_itx> inflex you around?
[16:37:34] <Tom_itx> what did you use to hold your heatsinks on your chips?
[16:37:56] <tomatto> Kevin`: i tried this http://www.recursion.jp/avrcdc/cdc-232.html with attiny2313, but while receive/transmit bytes cause some weird waring in linux kernel http://pastebin.blesmrt.net/2654/13471361/
[16:38:54] <Kevin`> tomatto: could just be notifying you that bulk endpoints aren't allowed for low speed usb devices
[16:40:25] <tomatto> Kevin`: so it is ok? i can't do anything with it?
[16:40:49] <Kevin`> well you could use a different type of device, of course
[16:41:29] <Kevin`> but I wouldn't worry too much, yes
[17:36:23] <megal0maniac> Roklobsta: I've seen it, and it's familiar, but I don't think I've used it. Looks nice
[17:37:13] <Roklobsta> yeah next embedded project i will try it out
[17:38:22] <megal0maniac> Does it come with a compiler? Sorry, don't have internet on pc and phone can't multitask :P
[18:17:03] <Roklobsta> mega: well it does optionally come with mingw so you can make windows programs
[20:14:30] <inflex> Tom_itx: adhesive heat-transfer paste, you can get it from DX
[20:16:09] <inflex> Tom_itx: http://dx.com/p/adhesive-solder-heatsink-plaster-for-pc-hard-parts-5g-37514
[20:19:09] <OndraSterver> can I eat it?
[20:19:23] <OndraSterver> I do wonder
[20:19:26] <inflex> wow, DX is expensive for replacement ipod/iphone parts
[20:19:36] <OndraSterver> if we stuck xmega into liquid nitrogen
[20:19:40] <OndraSterver> put it onto 5V
[20:19:44] <OndraSterver> how much would we overclock it? :D
[20:20:15] <OndraSterver> or that doesn't work that way on micros? :(
[20:20:28] <OndraSterver> small LN2 pot...
[20:41:35] <Tom_itx> inflex thanks
[20:51:21] <inflex> np
[21:08:08] <CapnKernel> !seen learningc
[21:08:09] <tobbor> learningc is here
[21:08:21] <CapnKernel> learningc: hi!
[21:25:49] <Casper> maybe learningc is busy helping someone unflood their basement...
[22:16:30] <Casper> Does anyone know of some opensource buck-boost board? one that can do 12V SLA to 12.00V at 4+A? ideally synchronous rectification...
[22:21:21] <learningc> Hi CapnKernel! Sorry I was busy cooking, but not PCBs! :P
[22:22:23] <learningc> Hi Casper
[22:22:42] <learningc> Casper: did you get your SMPS working yet?
[22:26:13] <Casper> learningc: no, still part hunting
[22:26:39] <Casper> learningc: how's CapnKernel's board? he said you order lots from him?
[22:27:48] <CapnKernel> (LOL, so public!)
[22:28:18] <learningc> the quality and finish is great! but I have yet to order 4 layers PCBs :P
[22:28:39] <learningc> How's it going CapnKernel?
[22:28:42] <Casper> learningc: order to delivery, how long exactly?
[22:29:31] <learningc> Casper: Standard order takes 5 days to manufacture and shipping through DHL is a speedy 3 days!
[22:29:31] <CapnKernel> learningc: Fine thanks. Enjoying being with my youngest son for his birthday :-)
[22:29:50] <CapnKernel> learningc has realised the awesome power of DHL :-)
[22:30:00] <Casper> learningc: ok, and standard shipping?
[22:30:30] <learningc> Casper: haven't yet tried "normal" shipping.
[22:30:39] <Casper> :(
[22:30:49] <CapnKernel> Generally about 16 days to USA/Canada
[22:31:27] <learningc> Casper: since the boards are for the company I work for, I cannot really afford waiting for a few weeks, unless it's really not urgent.
[22:31:42] <Casper> yeah
[22:31:58] <Casper> CapnKernel: working days or including weekends?
[22:32:14] <CapnKernel> Calendar days
[22:32:15] <learningc> Casper: but perhaps in the future for my own boards I'll take the standard shipping
[22:33:31] <learningc> CapnKernel: ah, his birthday today? how old?
[22:35:47] <Casper> learningc: do you know of a good smps IC that do buck-boost? 10-15Vin 12V 4+A out?
[22:36:01] <Casper> ideally synchronous rectification
[22:36:39] <learningc> Yes, but only boost, not sure it can do buck, but it might
[22:36:54] <Casper> need both
[22:36:58] <Casper> at the same time :/
[22:37:13] <learningc> And if you don't mind a TQFN 16..
[22:37:14] <CapnKernel> I have two sons, 13 and 14.
[22:41:43] <Casper> in hand solderable format...
[22:42:52] <CapnKernel> Here's the youngest son hand-soldering, for the first time ever: http://capnstech.blogspot.com.au/2011/06/playpause-3-assembly-and-programming.html
[22:43:02] <CapnKernel> 1208 + SOIC-8
[22:43:17] <CapnKernel> 1206...
[22:44:06] <learningc> CapnKernel: Like father like son. :P
[22:44:43] <learningc> Casper: it's still hand solderable..
[22:45:05] <learningc> Casper: but you need the right technique
[22:45:51] <CapnKernel> Read the last line.
[22:46:28] <Casper> first thing to do: hack the ups so I gain back a ups until I make my DC one...
[22:46:28] <CapnKernel> Most of the soldering on these boards was done by people who had never done any kind of soldering before, thus proving that surface mount soldering doesn't require any particular skill or dexterity, but just some instruction and guidance.
[22:49:00] <learningc> Casper: TPS55340
[22:49:10] <learningc> It's the part I use
[22:49:24] <inflex> I think dexterity is important, though -knowledge- about what they're soldering, less so
[22:51:02] <Casper> maybe in a sepic arrangement....
[22:51:08] <learningc> Casper: You can also go with linear technology: lt1170
[22:51:28] <Casper> do you know the efficiency of sepic?
[22:52:32] <learningc> It depends on many factors
[22:52:52] <learningc> the diode you use
[22:53:17] <CapnKernel> learningc: Now I know who to come to when I need to know about this stuff :-)
[22:53:25] <learningc> the voltage difference between vout and vin
[22:54:02] <learningc> also the switch voltage drop
[22:55:53] <learningc> CapnKernel: I'm no expert in power electronics :)
[22:56:24] <Casper> I'ld like to be in the 90+% efficiency...
[22:56:26] <learningc> but who knows, maybe in... 20 years..
[22:58:36] <learningc> Casper: you want an all-in-one solution?
[22:58:55] <Casper> just something hand solderable
[22:59:07] <learningc> You can also have just the controller and choose your own switch
[22:59:17] <Casper> I was going to go with ltc3789, but the BOM is quite high
[23:02:11] <Casper> the ltc3789 is ssop28 narrow
[23:03:14] <learningc> Casper: if you want high efficiency, you need to choose low forward voltage diode, low fet RDon, low inductor resistance, etc
[23:04:06] <learningc> also low esr output cap
[23:18:16] <CapnKernel> learningc: There's a particular circuit I've been after for a long time
[23:18:20] <Casper> might go sepic
[23:18:30] <CapnKernel> I want to drive an ATmega168 at 5V
[23:18:36] <CapnKernel> 5V can come from USB
[23:18:49] <CapnKernel> But I also want a 3.4V LiPo battery
[23:19:15] <CapnKernel> I want to be able to charge the LiPo from USB
[23:19:33] <CapnKernel> And when USB is not connected, generate 5V from the LiPo for the ATmega
[23:23:34] <Casper> that shouln't be that much of an issue
[23:23:43] <Casper> there is some nice chip for that
[23:24:45] <Casper> what is the phase margin for smps?
[23:27:12] <CapnKernel1> I've been looking at a two-chip solution: MCP73831 to charge the battery, and NCP1400 boost chip for generating 5V from the battery
[23:27:55] <CapnKernel1> But I'd like to do it with one chip
[23:28:06] <Roklobsta> op capn. what is your url now?
[23:28:49] <CapnKernel1> Curiously enough, this is what's required for USB OTG: To be able to charge a battery from 5V, and supply 5V when acting as a host.
[23:28:56] <learningc> CapnKernel1: what about using a 3.3V nominal uC?
[23:29:14] <CapnKernel1> For various reasons, I want to run the uC at 5V
[23:29:21] <learningc> ah ok
[23:31:31] <CapnKernel1> Circuit like this: https://docs.google.com/open?id=1L1w2EMsKrhPcsEpG1df1vseTwz9LRlWuziW_u97g_6y1ZToCFSad6PTmbJQx
[23:33:13] <CapnKernel1> I know that circuit will work, but I'm interested to know how to do it better/smaller/cheaper
[23:33:33] <CapnKernel1> 1 IC would be a big advantage, as long as the BoM was less.
[23:35:17] <Casper> also
[23:35:20] <Casper> with lithium
[23:35:25] <Casper> you want proper charging
[23:35:54] <Casper> so going 5V -> 3.x -> 5V might not be a good idea
[23:36:00] <learningc> There might be an IC that can both charge regulate the LiPo and boost, but it would cost more
[23:36:24] <Casper> I've seen some of 'em
[23:36:37] <Casper> but can't recall how I found them or how they were called
[23:37:44] <learningc> TI has a bunch of power management ICs
[23:39:17] <Casper> yeah that's it!
[23:39:23] <Casper> now, have fun :D
[23:39:29] <Casper> linear does too
[23:41:12] <learningc> CapnKernel1: Take a look at this one: TPS65217
[23:41:49] <learningc> It might be overkill, but I believe it would satisfy all your fantasies :P
[23:54:35] <CapnKernel1> My goodness there's a lot of stuff in that chip!
[23:54:43] <CapnKernel1> The question then becomes whether it can be used without I2C
[23:57:56] <CapnKernel1> I can't see any sign of it boosting the 3.4V from the battery back up to 5V