Back
[01:32:22] <earthshine> Morning
[03:06:14] <inflex> hrmm
[03:06:27] <inflex> wonder why one of my T2313 based USBtiny programmers won't flash
[03:06:40] <inflex> was fine until I blew the fuses... *ponders bad xtal*
[03:06:59] <inflex> I know the settings are fine, because I just flashed one before it
[03:07:46] <ziph> You haven't slept with Murphy's wife any time recently have you? ;)
[03:09:57] <ziph> abcminiuser: What blog software do you use?
[03:11:34] <abcminiuser> ziph, wordpress
[03:12:20] <ziph> It doesn't offend you to be running PHP? :)
[03:13:58] <abcminiuser> No
[03:14:08] <abcminiuser> Hate the language, but it's defacto-standard
[03:14:14] <abcminiuser> More hole than swiss cheese
[03:14:21] <abcminiuser> And HORRIFIC standard lib
[03:14:29] <ziph> Actually running Mysql is probably the bit that would annoy me more.
[03:14:41] <ziph> Does dealing with the comments take up much of your time?
[03:19:35] <inflex> ziph: heh - well, changed the chip, reflashed, all works
[03:19:48] <inflex> ziph: *shrug* might have failed part way through the fuse flashing *shrug*
[03:20:14] <inflex> Anyhow, all good again, now have 5 programmers on hand again.
[03:20:24] <inflex> ( 2 x Tom's MkII's and 3 USBTinys
[03:43:11] <abcminiuser> ziph, no, not many people comment
[03:43:16] <abcminiuser> Or do you mean spam?
[03:43:26] <ziph> abcminiuser: With spam and retarded comments.
[03:43:43] <abcminiuser> Wordpress has an awesome "Akismet" spam detection plugin -- it heuristically detects them and puts them in a seperate category
[03:43:51] <abcminiuser> Only ever messed up once in three years
[03:43:56] <abcminiuser> One click, and it nukes them all
[03:44:02] <inflex> nice
[03:44:03] <ziph> How many do you get?
[03:44:04] <abcminiuser> As for retart comments, none so far
[03:44:15] <abcminiuser> But I approve any human-written comment, and respond
[03:44:24] <abcminiuser> ziph, few hundred a week or so
[03:44:26] <inflex> the nuke-all is a ncie thing, something that I greatly wish the simpler forum engines had :(
[03:44:32] <abcminiuser> All automated, all trying to push some scam site
[03:44:35] <ziph> abcminiuser: Even with captcha?
[03:44:46] <abcminiuser> ziph, no captcha on my site
[03:45:13] <abcminiuser> Since that could throw people off, and it really isn't much trouble to look over a list of junk and click the "Delete All" button
[03:45:28] <ziph> Maybe I just won't bother with comments. Do you find them useful? :)
[03:47:10] <abcminiuser> ziph, here's what I see:
http://i.imgur.com/fqKRT.jpg
[03:47:38] <abcminiuser> I just scan the poster's URL on the left under their name to detect spam, much easier - they're always really obviously spam sites
[03:47:45] <abcminiuser> Yeah comments are useful
[03:47:54] <abcminiuser> As I said, not many, but when I get them they're decent
[03:48:52] <abcminiuser> For example of the comments I get, see
http://fourwalledcubicle.com/blog/2011/11/the-final-print/#comments
[05:43:50] <ziph> Piece of s**t. PHP just segfaults apache.
[05:44:10] <Steffanx> So Apache is the piece of crap?
[05:44:23] <ziph> No, apache keeps running but the child handler dies.
[05:45:09] <ziph> Sigh.
[05:46:31] <ziph> And WordPress advises that none of their previous releases are safe to use.
[05:47:53] <mrfrenzy> are you running uptodate versions of php and apache??
[05:48:00] <mrfrenzy> if so I would say you have a hardware problem
[05:48:13] <mrfrenzy> or some third party modules causing a problem
[05:49:00] <ziph> Or the PHP script called something in PHP that is buggy.
[05:49:17] <Steffanx> PHP isn't THAT bad :P
[05:49:25] <ziph> Huh?
[05:49:30] <mrfrenzy> in which case someone would have reported it and it would have gotten fixed already
[05:49:51] <ziph> Unless they fired everyone responsible for the actual design of PHP and replaced them with far better programmers then I don't see why that would be unexpected.
[05:50:23] <Steffanx> </rant> ?
[05:50:29] <ziph> </hellyes>
[05:51:01] <ziph> This happens any time I do anything remotely related to the web. ;)
[05:51:23] <Steffanx> I only had trouble with php once.. which was on windows + older version of php
[05:51:24] <Steffanx> :)
[05:51:43] <ziph> So you've never written a significant amount of PHP then? :)
[05:52:05] <Steffanx> Uhm, I pretty sure i wrote enough PHP
[05:53:42] <Regenschein> real programmers don't use PHP :-P
[05:53:45] <Regenschein> j/k
[05:53:59] <Steffanx> Agreed
[05:57:36] <inflex> oh lordy
[05:57:45] <Steffanx> Where?
[05:57:58] <inflex> real programmers use what is available and usable for the task and don't get into religious bs debates.
[05:58:45] <ziph> The problem is that some tools only get used for religious reasons.
[05:58:52] <ziph> Not on actual merit.
[06:00:58] <ziph> The Larry Elison owned frankenstein hack that is Mysql comes to mind. ;)
[06:02:41] <Steffanx> ziph please move to #php :P
[06:02:50] <ziph> No thanks.
[06:03:00] <ziph> I went in to #arduino this year, that's enough of that kind of thing for me.
[06:04:41] <Regenschein> for religious reasons? you probably mean Ruby on Rails
[06:04:50] <Regenschein> ;-)
[06:05:15] <ziph> Rails isn't so bad on its own merits.
[06:05:18] <ziph> Ruby though.......
[06:05:54] <ziph> It's about the most horrific homage to Smalltalk since Java. ;)
[06:06:34] <Regenschein> hehe..:D
[06:06:57] <ziph> Objective C is far, far more Smalltalk than Ruby.
[06:07:15] <inflex> oh hell... RoR is just something I looked at and went "Why?"
[06:07:17] <Steffanx> Objective-C .. don't talk about that please :p
[06:07:28] <inflex> Steffanx: how about some smalltalk :)
[06:07:59] <Steffanx> Never tried that
[06:08:01] <inflex> One language I did find very good was Borland Pascal
[06:08:07] <ziph> Smalltalk is nice, except for the way the runtime works and the lack of operator precedence on even simple expressions.
[06:08:14] <inflex> though I do find it a bit wordy to write in
[06:08:27] <inflex> ziph: some could still consider that a feature
[06:08:38] <ziph> 1 + 2 * 3 equalling 9? :)
[06:08:38] <inflex> ziph: but sure, others see it as "bug hell" :D
[06:08:56] <inflex> just lisp it up a bit with lots of parenthesis :p
[06:08:57] <karlp> inflex: was borland pascal any different to regular pascal?
[06:09:14] <inflex> karlp: it was, yes... they added the bits that made it actually useful
[06:09:24] <karlp> such as?
[06:09:29] <inflex> karlp: "traditional" Pascal was exceedingly limited in its IO
[06:09:47] <karlp> I think I only used turbo pascal, the borland ide, so I guess I was probably using borland pascal without noticing?
[06:09:54] <inflex> but I must say, I couldn't probably return to the limitations of write() / writeln()
[06:10:02] <inflex> yes
[06:10:07] * ziph finishes the WP install and waits for his server to be owned.
[06:10:40] <ziph> Borland had things like port I/O, more pointer support and OO features.
[06:10:42] <inflex> BP was quite nice too by the time it got to v7, with it breaking away the 64K segment limits
[06:10:44] <ziph> And interrupts..
[06:10:55] <inflex> and it's ASM integration was so lovely
[06:11:03] <ziph> Yeah, that was nice too.
[06:11:04] <ziph> inflex: Seen the Borland museum?
[06:11:29] * inflex was sad when they destroyed their name and changed to ... what ever it was... Intavision or something?
[06:11:37] <inflex> no, 've not seen the Borland museum
[06:11:48] <ziph> inflex: You can download all of the old Borland software.
[06:12:13] <karlp> all the borland pieces keep getting split and renamed.
[06:12:15] <inflex> ziph: oh hell... that could be bad for me :p My first commercial software was made with BP + TV
[06:12:36] <inflex> It back then that I realised I wanted to NEVER write accounting software
[06:12:50] <karlp> I had a lot of notfun at an earlier job trying to track down corba stuff, that had been borland, inprise, and ten thousand other names
[06:12:57] <ziph> Borland was clearly dead at the point that they had no captain to tell marketing they were frigging insane to want to change the name.
[06:13:04] <inflex> Ah yes, Inprise
[06:13:30] <inflex> ziph: indeed, it basically fell to crap when they did that... all those students who had grown up on Borland products, now going "WTF?"
[06:13:39] <inflex> Delphi was a roaring success and they botched it
[06:13:59] <karlp> they did really screw up on that one didn't they.
[06:16:25] <ziph> Do people actually follow blog feeds?
[06:16:31] <inflex> ziph: ver very rarely
[06:16:42] <inflex> more chance of them following you on twitter
[06:16:45] <karlp> if I like the author,
[06:16:58] <inflex> http://dxp.me/i/oled.jpg <=- you guys see this?
[06:16:59] <karlp> I follow very few people on twitter, but lots of occasional blog posters
[06:17:10] <inflex> oh well, there you go then
[06:17:23] * inflex needs to update his wife's novelist page too
[06:17:39] <karlp> particularly for people that aren't writing articles every day, subscribing to a blog feed is a convenient of finding out _when_ they do update
[06:18:01] * inflex still likes having the updates emailed to him
[06:18:02] <ziph> So people don't use RSS readers any more? :)
[06:18:07] <inflex> unless that's what you mean karlp ?
[06:18:19] <Steffanx> An even smaller display inflex :)
[06:18:28] <inflex> Steffanx: yep, 96x16px OLED
[06:18:34] <Steffanx> Nice
[06:19:02] <inflex> the question is now, wth to use it for
[06:19:16] <inflex> will have to start using DFN parts
[06:19:22] <ziph> inflex: How's project LOM?
[06:19:33] <inflex> ziph: I'm peeved with it
[06:19:50] <inflex> ziph: when I saw the figures wiggling around on changing the power connectors I basically wanted to throw it out the door
[06:20:06] <karlp> ziph: what did you mean by blog feeds? I thought you meant the rss?
[06:20:17] <ziph> Yeah, I did mean that.
[06:20:36] <ziph> inflex: :(
[06:20:42] <ziph> inflex: I hate things like that.
[06:20:45] * inflex thinks he could do something with a Tiny85 and this OLED, since I can talk to the OLED with I2C... leaves me 2 or 3 pins for something like volts/current/timing sense
[06:21:11] <ziph> inflex: What about project-magic-battery-logger?
[06:21:13] <inflex> ziph: it's a buttload of money I've invested in it and I'm thinking it's time to just change the specs, sell what I can and get out
[06:21:37] <inflex> ziph: yeah, was thinking that - could have a "last 10 minutes" graph
[06:22:10] <inflex> ziph: wouldn't even need a lot of memory, since the display becomes the memory (I've seen I can page scroll it by pixels)
[06:22:37] <inflex> ziph: mostly see these display units as bling factor
[06:22:53] <ziph> :)
[06:23:40] <inflex> other idea was a sort of distribution graph, eg, show which level of current is used most
[06:25:39] <inflex> I have a 128x32 version which could work nicely as an altitude graph
[06:26:51] <inflex> trick though is finding a suitable sensor for barometric pressure which isn't a monster in size
[06:28:35] <karlp> BMP085?
[06:28:41] <karlp> that's pretty tiny isn' tit?
[06:28:47] <karlp> like on this:
http://jeelabs.com/products/pressure-plug
[06:31:41] <inflex> yeah, it's 5x5mm... it's a good sensor actually... but obscenely expensive
[06:31:46] <inflex> ($5+ in 100 units)
[06:32:24] <inflex> otoh, the MPL115 is a cheaper unit... but lacks resolution (10 bit only)
[06:34:37] <karlp> what are you talking about? we still have 100 feet of air beneath us....
[06:47:54] <ziph> inflex: Are you aware of the MCP3421?
[06:58:42] <amee2k> anyone know what kind of temperature coefficient to expect from shottky diodes?
[07:25:25] <inflex> ziph: ermm... *checks*
[07:26:08] <inflex> ziph: haven't used one myself but can't see it being a problem
[07:26:46] <ziph> inflex: It's small and useful looking, I thought it would be right up your ally.
[07:28:58] <inflex> aaah okay - tx
[07:42:51] <ziph> Heh. neat, a 36V 500mA switch mode with an LDO on the same die. Perfect for 5V + 3.3V
[07:44:24] <karlp> which part?
[07:45:37] <ziph> LT3645.
[07:45:43] <ziph> Only just out though.
[07:58:34] <inflex> yeah, I have to say, I do like Linear gear a LOT
[08:00:58] <inflex> okay, now I have to try adjust my firmware again to handle this 4MBit FLASH EEPROM
[08:01:04] <karlp> too many options, too confusing part numbers :)
[08:01:06] <karlp> but yeah, good gear
[08:01:28] <inflex> ja, usually I can take their design circuits/pcb's and go with it
[08:01:57] <inflex> but they can be damned expensive
[08:03:24] <inflex> whoops, forgotten which brand/model I used
[08:23:33] <soul-d> :') guy composes music for a antipiracy festival thingy it ends up behind anti piracy ad on dvd (without his knowledge) he sue's "started at 100k" then estimated estimated a mln in damages (since 2007) so now it's worth attention of the head of the copyright club and ofcourse for 1/3th of that mln " there would be made work of it " :') so glad i always download
[08:24:21] <Steffanx> Oh, soul-d is a dutchy too :P
[08:24:26] <soul-d> :)
[08:25:34] <soul-d> great news i wonder if people will still want to pay those clubs afther this
[08:41:54] <moe3> hi
[08:42:18] <moe3> Will enums in c use RAM or are they something like #define's
[08:43:54] <specing> moe3: enums are TYPECHECKED constants like defines
[08:45:35] <moe3> what does that mean
[08:46:04] <moe3> so they wont use memory?
[08:46:56] <specing> depends on how you use them
[08:47:22] <moe3> not special. i just use them
[08:52:03] <grummund> moe3: static const, enum, and #define all yield pretty much the same result.
[08:53:38] <moe3> ok but according to memory use they are different
[08:54:10] <grummund> have you tried?
[08:54:34] <moe3> i want to define variables which are stored neither in ram nor flash. but they wont be changed by the program. so how should i do that
[08:55:09] <moe3> yesterday you told me a const will be in ram
[08:55:56] <specing> moe3: How do you expect to use a variable that isn't stored anywhere?
[08:56:03] <grummund> the "variable" is in ram (even though it does not change)
[08:56:19] <grummund> the constant value obviously is defined in flash
[08:56:44] <grummund> and it is copied to the variable at startup
[08:57:35] <grummund> *But* if gcc knows the value of the variable at point of usage it won't bother with keeping it in ram.
[08:58:12] <grummund> assuming you use -Os
[08:58:26] <moe3> hmm
[08:58:44] <moe3> so i need to use the static const stuff
[08:59:39] <grummund> presumably you have const int foo = 123; in a .c file, but it is *used* in another file?
[09:00:02] <grummund> and perhaps extern const int foo; in a .h file?
[09:12:32] <keenerd> moe3: Sounds like you want a #define.
[09:13:18] <karlp> as grummund said, they still all take up flash or ram or both,
[09:13:37] <karlp> the differences are mostly due to what help the compiler can give you ahead of time
[09:15:43] <grummund> the problem is moe3 is porting code from a platform where const variables do "exist" in flash
[09:16:08] <grummund> yet avr doesn't allow that
[09:17:20] <karlp> well, we told him teh answer, pgmspace
[09:17:30] <karlp> and he said he didn't want to modify the code at all, he wanted a magic compiler flag
[09:17:35] <grummund> The workaround is to ensure that gcc knows the values at the point of usage - in effect to allow the optimiser to work so the 'variable' no longer requires storage.
[09:18:32] <buzzsaw> good morning
[09:18:42] <grummund> e.g. possible solution could be move all the const definitions to a header file and define them static const instead.
[09:19:25] <grummund> presumably a header file already exists with the variables declared extern const, so just replace that.
[09:20:19] <grummund> or use #define or enum, both of which likely yield the exact same code.
[09:42:12] <buzzsaw> I am using an atmega16. When I drive turn a pin on, what should be the voltage output?
[09:42:46] <grummund> same as Vcc whatever that is
[09:43:27] <buzzsaw> so then if I supply 5v in and turn 5 pins on, each output is 5v?
[09:43:37] <grummund> yes
[09:43:38] <buzzsaw> sorry, have not messed with my controllers in over a year
[09:44:24] <buzzsaw> interesting... because I just measured and I am supplying 5.4v and it is only outputting 3.5v
[09:44:42] <buzzsaw> I must have done something wrong :-)
[09:44:51] <grummund> anything else connected to the pin(s) ?
[09:46:29] <buzzsaw> nothing connected to the pin I am measureing
[09:46:46] <soul-d> http://androidsecuritytest.com/features/logs-and-services/loggers/carrieriq/carrieriq-part2/
[09:46:47] <grummund> perhaps you failed to configure the pin as output?
[09:46:56] <soul-d> they are spying on us :(
[09:47:16] <buzzsaw> well when I plugged the led in that it should be driving it is just fine
[09:47:39] <buzzsaw> soul-d everyone is spying on everyone...
[09:48:09] <grummund> without seeing the code it's difficult to say
[09:48:20] <buzzsaw> i will look up the code :-)
[09:48:29] <buzzsaw> let me fire up my vm
[09:48:50] <Steffanx> soul-d .. who says it's true? Someone makes a nice app to check it, who says the app is OK?
[09:50:29] <Steffanx> It's is a nice way to make people install your own trojan horse on people's phone
[09:50:58] <Steffanx> Make them scared, give them a app to check if they have something and give them your own 'carrier iq' for free :)
[09:51:07] <buzzsaw> very simple code for very simple test ;-)
http://pastebin.com/twVHu6D8
[09:51:23] <soul-d> sadly it came from our ANP news service
[09:51:30] <soul-d> Ruim 140 miljoen mobiele telefoons zijn uitgerust met software die alle toetsaanslagen en binnenkomende sms'jes registreert. Het gaat om een programmaatje van het Amerikaanse bedrijf Carrier IQ.
[09:51:46] <grummund> buzzsaw: does it flash at 1s ?
[09:51:52] <soul-d> so how fake is that ?
[09:52:12] <Steffanx> ANP is a piece of crap news service :P
[09:52:12] <buzzsaw> a bit faster than that why?
[09:52:41] <grummund> ok but not so fast that the multimeter cannot respond to a stable voltage?
[09:53:10] <buzzsaw> well I am measuring the higher pins
[09:53:54] <buzzsaw> they stay on for about 10 seconds or longer
[09:55:07] <grummund> on = high?
[10:05:23] <buzzsaw> yes it stays on high
[10:05:45] <buzzsaw> i tried a different multimeater just to see.
[10:06:40] <buzzsaw> same results when the pin drives high it is only 3-4.5v
[10:06:43] <grummund> with that code, PB7 should be low (0V) for 127s after startup, then high (Vcc) after that.
[10:07:12] <grummund> measured between PB7 and Gnd
[10:09:52] <buzzsaw> oddly pb7 goes high at 10 seconds. I wonder if there is something wrong with the chip :-)
[10:10:15] <grummund> probably F_CPU is not defined correctly
[10:11:09] <buzzsaw> thats whats wrong i opened the wrong file
http://pastebin.com/jL1pb4F2
[10:11:22] <buzzsaw> this is why it runs faster. just a smaller ms#
[10:11:30] <buzzsaw> same code
[10:12:03] <grummund> then PB7 will toggle every 16s
[10:13:56] <buzzsaw> i measured directly from the pin an the ground with nothing else in circuit and it comes back at 3.8v this time
[10:14:18] <grummund> try also between pin and Vcc
[10:17:33] <buzzsaw> that wont return anything. from two hot pins?
[10:17:51] <grummund> should see 5V when the pin is low
[10:18:09] <grummund> and 0V when the pin is high
[10:20:36] <buzzsaw> okay i am getting 1.2v when high and 4.9 when low
[10:21:25] <grummund> one more test, well two actually...
[10:21:53] <grummund> try again with pin loaded by a resistor to gnd
[10:22:02] <grummund> and again with pin loaded by a resistor to Vcc
[10:22:44] <grummund> sounds like the avr output is shot, but that'll prove it.
[10:23:17] <buzzsaw> lol no telling :-) I could have wired soemthing wrong and screwed it up a year ago and jsut forgot about it
[10:24:06] <buzzsaw> no sure what you mean by pin loaded by a resistor to gnd
[10:24:29] <grummund> connect a resistor between pin and gnd
[10:24:45] <buzzsaw> got it
[10:31:12] <buzzsaw> if i do pin resistor ground i get the same thing
[10:31:48] <grummund> right, and Vcc?
[10:34:30] <buzzsaw> it stays at .8
[10:34:52] <buzzsaw> hold on it slipped i try agin ;-
[10:36:29] <buzzsaw> yeah it just stays at .8-.9
[10:37:03] <grummund> how is the R connected and where do you measure between?
[10:37:54] <buzzsaw> i did vcc resistor ground and measured from vcc to pin
[10:38:43] <grummund> so measure pin to gnd, with resistor pin to vcc.
[10:41:17] <buzzsaw> its ~3.7 when high and .001 when low
[10:41:41] <grummund> no change then. that's odd
[10:42:17] <grummund> if the output was blown you would expect the resistor to pull the pin to Vcc.
[10:45:25] <buzzsaw> funky chip i guess :-(
[10:50:42] <buzzsaw> so if all the pins are high does it drop the voltage across the board?
[10:52:51] <grummund> if a pin is high then you should end up with 0V between pin and Vcc
[10:54:10] <buzzsaw> okay
[10:54:20] <grummund> what val resistor did you use?
[10:54:25] <buzzsaw> it was a 220
[10:54:48] <grummund> i find it strange that it made no difference
[10:55:11] <buzzsaw> hum it could be operator error...
[11:11:53] <buzzsaw> okay i went back and and double checked. with vcc resistor to ground i measured pin to ground. 0v when off and 3.9v when on
[11:12:30] <RikusW> and vcc itself ?
[11:13:51] <buzzsaw> 4.98v
[11:14:56] <RikusW> an avr ?
[11:15:06] <buzzsaw> atmega16
[11:15:08] <grummund> So... with no resistor fitted, the high pin is 3.5V measured to gnd...
[11:15:12] <grummund> and with resistor to gnd the high pin is 3.9V measured to gnd.
[11:15:21] <grummund> that make no sense at all!
[11:16:16] <RikusW> buzzsaw: the same on other io pins too ?
[11:22:49] <buzzsaw> vcc to ground is 4.9v, ping to ground is 3.5-3.8v vcc to pin is 1v high and 4.9 low
[11:23:42] <buzzsaw> pb6 is running at 3.8v as well
[11:25:19] <RikusW> buzzsaw: did you connect vcc->avcc ?
[11:26:34] <buzzsaw> my avcc is not connected to anyting
[11:26:54] <grummund> aha
[11:30:24] <buzzsaw> i am guessing mr newbie did something wrong ;-)
[11:31:29] <grummund> check the datasheet - avcc needs to be connected to vcc
[11:31:43] <grummund> good catch by RikusW ;)
[11:32:34] <RikusW> :)
[11:32:48] <RikusW> buzzsaw: working better now ?
[11:33:05] <buzzsaw> let me find a way to connect it real quick :-)
[11:36:24] <buzzsaw> lol much better :-)
[12:14:01] <buzzsaw> well thanks for the help everyone :-) I need to head out and get a starter for my sruck
[12:17:33] <RikusW> truck ?
[12:19:43] <buzzsaw> yeah that too
[12:19:50] <buzzsaw> cold hands dont type well...
[12:20:19] <RikusW> ah of course
[12:48:30] <RikusW> hi vectory
[12:50:35] <amews_aj> RikusW, Hi. Thanks again for the help a few days ago :) It works fine now :)
[12:51:09] <RikusW> nice
[12:51:46] <RikusW> amews_aj: what was the problem again ?
[12:52:14] <amews_aj> RikusW, wrong clock frequency - difficult getting the right frequency for an IR LED
[12:52:26] <RikusW> ah
[13:48:48] -kornbluth.freenode.net:#avr- [freenode-info] please register your nickname...don't forget to auto-identify!
http://freenode.net/faq.shtml#nicksetup
[14:38:34] <sabesto> yeah, i dont think the datasheet is going to solve this one
[14:38:51] <sabesto> i am able to open a terminal with the bootloader and such
[14:38:54] <RikusW> avr@atmel.com ?
[14:40:07] <sabesto> "A pre-installed Linux image on the enclosed 256 MB SD card ensures that the user can boot Linux and start program development directly after power up."
[14:40:17] <sabesto> i just wish i had that SD card :P
[14:43:18] <sabesto> there is a lot of info at avrfreaks wiki
[14:44:47] <RikusW> email atmel ?
[14:48:21] <sabesto> i've only tried for 6 hours :)
[14:48:34] <sabesto> every hour i get a step closer :)
[14:50:44] <RikusW> ok
[15:27:51] <sabesto> meh, getting the same error
[16:25:41] <amee2k> so.. any ideas how to get rid of some 2W from a 2x2mm thermal pad on the bottom of a part?
[16:26:02] <Tom_itx> thick copper
[16:26:17] <Tom_itx> thru holes to the other side and more thick copper
[16:27:03] <amee2k> okay, i get the thick copper part. how do you mean "thru hole"?
[16:27:56] <amee2k> oh, almost forgot. the thermal pad is not electrically neutral so it needs to be isolated from the heatsink
[16:29:33] <Tom_itx> if i can find my server i'll show you
[16:29:37] <amee2k> i've considered putting a large mica washer on the heatsink with a *thin* layer of epoxy, then putting a large ground plane and traces from 300um copper foil on it
[16:29:59] <amee2k> but i have no idea how to make the foil cut-outs with enough precision
[16:30:46] <amee2k> voltage is only like 30V max though and a small leakage current is acceptable
[16:30:47] <Kevin`> I think he means vias
[16:30:59] <Tom_itx> yeah
[16:31:37] <amee2k> hmm
[16:31:49] <amee2k> single sided board :/
[16:32:10] <amee2k> but not a bad idea... using a ground plane on the back as heatspreader
[16:33:00] <Tom_itx> isp has been screwed up all day
[16:33:06] <Tom_itx> can't find my server atm
[16:33:48] <amee2k> i can poke you tomorrow about it
[16:34:12] <Tom_itx> it'll show up here in a sec
[16:37:33] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/boards/atmega32u4/atmega32U4_3.jpg
[16:37:36] <Tom_itx> like those
[16:38:06] <amee2k> i see
[16:39:28] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/3995_hbridge/A3995_Hbridge_bottom_pour.jpg
[16:39:41] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/3995_hbridge/A3995_Hbridge_top_pour.jpg
[16:43:10] <Steffanx> Fancy
[16:43:37] <Tom_itx> that's just the kind of stuff we do here
[16:43:55] <Steffanx> You :P
[19:17:58] <landonf_> Random question, but anyone know offhand why LUFA's serial driver sets PORTD bit 2 and DDRD bit 3 when initializing the UART on AVR8 targets? Head scratching at the code
[19:17:58] <landonf_> http://code.google.com/p/lufa-lib/source/browse/trunk/LUFA/Drivers/Peripheral/AVR8/Serial_AVR8.h#145
[19:18:27] <landonf_> (And related, can anyone suggest other projects that might serve as good examples of development/architectural style?)
[19:24:15] <vectory> landonf_: one might be the TX and the other the RX pin
[19:26:13] <landonf_> That was my first thought, but the RX pin is PD0, TX pin is PD1
[19:27:48] <vectory> then the other two might still be related
[19:27:56] <vectory> rs232 uses more than two lines
[19:28:38] <vectory> with automatic baud detiction these can be ignored, maybe, i dont know
[19:29:17] <inflex> hmmm?? other than rx/tx/gnd, you don't need much else for RS232
[19:29:23] <vectory> wrong
[19:29:34] <vectory> see wikipedia, it lists 4 or 5 wires
[19:29:44] <inflex> well, damn, tell that to my project sitting here running over 3 wires...
[19:29:48] <landonf_> you mean flow control?
[19:30:59] <vectory> i dont know
[19:31:06] <vectory> probably yes
[19:31:23] <inflex> vectory: so, what other wires are -needed- for serial over wire?
[19:31:50] <landonf_> Landon: Are you my arch nemesis?
[19:31:59] <Landon> :|
[19:32:05] <Landon> landon, fair enough
[19:32:08] <Landon> but you had to take my last name too
[19:32:31] <Landon> (well, initial)
[19:32:40] <vectory> you have
[19:32:44] <vectory> request to send
[19:32:48] <landonf> If you're also 'fuller' than I'll be so amused.
[19:32:49] <vectory> clear to send
[19:32:52] <landonf> then
[19:32:56] <Landon> no, fowles :P
[19:32:57] <vectory> data set ready
[19:33:01] <landonf> Landon: whew
[19:33:05] <vectory> data carrier detect
[19:33:13] <vectory> 2 more
[19:33:17] <Landon> crisis averted, huh
[19:33:37] <landonf> for now ^_^
[19:33:43] <inflex> yeah, they're modem type lines
[19:34:10] <inflex> but serial works just fine with RX, TX and GND and nothing else
[19:34:12] <vectory> i have no idea about lufa though, so im the wrong guy to ask
[19:34:22] <theBear> serial CAN work on just gnd and a single wire per direction, but it can use more signals regardless of hardware involved... just slightly different protocols, if you will
[19:34:39] <theBear> but if yer hang around long enough, someone who knows all about lufa will turn up eventually
[19:34:58] <landonf> LUFA isn't making use of those pins the implement flow control, so still confused and not sufficiently sure of myself to declare it a bug.
[19:35:02] <landonf> Sounds good to me.
[19:35:05] <theBear> maybe look at the rest of the code, does it reference those pins anywhere ? does the serial function in use perhaps mention them
[19:35:16] <inflex> vectory: Three wires (transmit, receive, and ground) are usually the minimum recommended. A fully implemented RS-232 connection can have as many as 25 wires between each end.
[19:35:34] <vectory> landonf: i think one of these is used for interrupt signal
[19:35:39] <vectory> so flow controle makes sense
[19:35:54] <Landon> I might have to revise my highlites now :P
[19:36:19] <theBear> if you think about it, flow control is a lot easier than just 'listening to the wire' all the time
[19:36:22] <vectory> i had to change nick
[19:36:31] <vectory> from vector to vectory
[19:36:37] <Landon> hah
[19:36:54] <vectory> mayybe you go with LFowler now
[19:37:13] <Landon> nah, I'm not in channels with landonf that often
[19:37:17] <inflex> vectory: basically, sure, you can have a lot of lines on a RS232 setup... but a 2 or 3 wire setup is still permitted and is what is used quite frequently these days
[19:37:27] <Landon> I think maybe I've been highlighted from someone talking to him once before
[19:37:58] <inflex> eg, I'm using a 2-wire setup here, TX and GND only, as the device is strictly 1-way data
[19:38:00] <vectory> inflex i know
[19:38:11] <vectory> at least i think i do
[19:38:24] <inflex> anyhow, back to work
[19:38:36] <Tom_itx> aww
[19:38:58] <vectory_> uart chips do only use 2, anyway, isnt it?
[19:39:20] <Tom_itx> depends
[19:39:41] <Tom_itx> more if you use hardware handshaking
[19:41:10] <vectory_> and i think, lufa implemented just that
[19:41:41] <vectory_> s/abcminiuser/lufa
[19:42:19] <theBear> pc uarts use lots of lines, IF you tell them to
[19:42:43] <theBear> you know, 16?50 and derivitives
[19:43:15] <vectory_> nope
[20:42:37] <buzzsaw> I have an item that is triggered when the pin is driven low cant change them because the board is premade. would it be better to use a transistor and use an atmega and turn it high or... turn the port to a pin and drive it low that way?
[20:43:08] <buzzsaw> I dont want to mess up my controller if its not good to use it that way.
[20:43:49] <buzzsaw> This is what I am controlling
http://www.doityourselfchristmas.com/wiki/images/e/ef/SSRez-sch.pdf
[20:54:19] <landonf> AVR devkits: always a great christmas present.
[20:55:05] <Essobi> :D
[21:00:52] <buzzsaw> controlling christmas lights with avr better ;-)
[21:44:13] <buzzsaw> I am connecting to a board similar to this
http://designbybuzz.com/schematics/remoteLED.jpg (not exactly but this gives an idea) pin 1 is 5v and pin 7 is run to ground (to show that the board is getting power. pin 2,4,6,8 are run low to enable the led's
[21:45:12] <buzzsaw> now i want to use my atmega on the other end to drive this board. should I put a npn in line and drive my atmega high, or is it safe to drive the pins low?
[22:13:47] <buzzsaw> good night :-)
[23:05:51] <Gorroth> hi
[23:07:07] <Casper> low
