Back
[00:32:07] <jophish> hello!
[00:32:59] <jophish> i'm having a problem with programming an avr butterfly with a usb programmer.
[00:33:01] <jophish> I have the AVR Pocket Programmer, and i'm trying to flash an AVR Butterfly via ISP with it, using AVR Studio 4, but the program isn't recognizing the programmer. I installed all the drivers and workarounds, and can access the programmer just fine through command prompt, so i'm not sure what the problem could be. :|
[00:33:39] <Casper> the problem is: avr studio
[00:33:48] <jophish> alternative?
[00:36:31] <Casper> don't use it :D
[00:36:36] <Casper> programmer: avrdude
[00:36:56] <Casper> toolchain: avr-gcc
[00:37:11] <Casper> and your favorite text editor (programmer's notepad?)
[00:38:38] <jophish> alright, i tried that.
[00:38:51] <jophish> avrdude -c usbtiny -B 1 -patmega169 -U flash:w:main.hex
[00:39:17] <jophish> it apparently flashed through fine, but i'm not getting what i expected, or anything at all, on the butterfly
[00:41:13] <wollw> why the -B 1 ?
[00:42:37] <wollw> ah, seems the example usage for that programmer has it
[00:42:47] <jophish> yeah
[00:43:41] <wollw> avrdude flashed and verified it?
[00:45:00] <jophish> yeah.
[00:45:18] <wollw> then... it's probably programming fine
[00:45:28] <jophish> and there was a flash of garbled text on the lcd of the butterfly
[00:45:37] <jophish> hnngh
[00:45:40] <jophish> alright
[00:45:48] <jophish> thanks
[00:50:44] <Casper> you sure your program is fine?
[00:51:02] <Casper> and compiled for the right part?
[02:31:04] <killit2> hello everyone... I am having an issue what I can transmit from UART just fine but it never detects bytes comming in (I have verified the other circuit is functioning with my laptop and a TTL adapter).. are there fuse settings that might disable RX but still allow TX? I do have RXE0 set to 1...
[02:32:11] <ziph> killit2: Try shorting the RX and TX together at the RS232 level side (assuming you're using something like a MAX232) and seeing if it can detect characters that it transmits.
[02:32:49] <killit2> there is no rs232 involved.. its all ttl...
[02:33:36] <killit2> will that work if the chip is in async mode?
[02:33:46] <ziph> Disconnect the adapter and short it at TTL level then.
[02:34:30] <ziph> It should do.
[02:35:05] <killit2> alrighty
[02:35:10] <killit2> give it a go right quick
[02:36:08] <killit2> nada... the interupt never fires...
[02:36:22] <killit2> I tried programming a fresh chip too wondering if it was a damaged chip..
[02:36:47] <ziph> Doubt it.
[02:37:24] <ziph> Seeing as it can only be internal to the chip I'd use a JTAG/DebugWire debugger to look through all of the serial registers while carefully reading the datasheet.
[02:37:50] <killit2> dont have a programmer capable of that..
[02:38:16] <ziph> You could try running your code in the AVR Studio simulator then..
[02:38:34] <killit2> I do have an LCD on this circuit though.. I could have it write the values of the registers to the LCD..
[02:39:10] <ziph> Yeah, or do that.
[02:39:46] <killit2> ok... super happy fun times real quick
[02:40:06] <Roklobsta> i upgraded from Mythbuntu 11.04->11.10->12.04 today and it broke stuff. fresh install in the morning
[02:40:12] <Roklobsta> oh wrong channel
[02:40:20] <ziph> Linux sucks.
[02:40:37] <ziph> Oh god, why did you start another Linux flamewar?
[02:40:38] <Roklobsta> no way, i saw on hackaday linux running on an avr
[02:40:54] <ziph> Yeah, but it took hours to boot.
[02:40:58] <Roklobsta> slowest linux eva
[02:41:02] <ziph> Why would you do that when you could just run Windows?
[02:41:23] <killit2> running Linux now... :)
[02:41:26] <Roklobsta> it's so frickin cold in those northern climes you have to overlcok AVRs to keep warm.
[02:41:37] <Roklobsta> linux blow.
[02:41:53] <Roklobsta> i wasted 4 hours today fscking with linux. again.
[02:42:11] <specing> Ubuntu problems
[02:42:15] <Roklobsta> and another 4 tomorrow undoing todays mes
[02:42:22] <killit2> eh... its a tool really like any other... some people like different tools... at the end of the day its just preferance so long as the job gets done
[02:42:32] <Roklobsta> oh linux is a tool alright
[02:42:38] <killit2> lol
[02:43:40] <killit2> the problem with the Linux / Windows war is that typically people prefer whatever they started with... typically anyways... they are SOOOO diffrent.. like fish vs birds that people get agrivated by knowing how to do something in one, but being lost in the other..
[02:43:45] <Roklobsta> i am a linux advocate but win7 on my i7 laptop is actuallly pretty good.
[02:44:02] <Roklobsta> i am not moved ot make my laptop a linux machine.
[02:44:10] <specing> then you are not a linux advocate
[02:44:15] <Roklobsta> yes i am
[02:44:30] <Roklobsta> have a stonking mythtv box which I buggered u ptoday
[02:44:37] <Roklobsta> and run some serves in the cloud
[02:44:43] <ziph> To be a real Linux user you must refuse to use anything that isn't GPL.
[02:44:44] <specing> a Linux advocate running around with a windows box? nope nope nope
[02:44:59] <ziph> Including microwaves.
[02:45:00] <killit2> I dual boot... I have windows for Office and such (school REQUIRES office format... - and my REQUIRED VB class.. useless programming class Ill never use... ) so I use windows when I NEED to...
[02:45:07] <nofxx> ubuntu just sucks... use archlinux
[02:45:08] <Roklobsta> i am not obese, smelly nor have a beard with dandruff so I can't possibly be a GPL only weenie.
[02:45:24] <specing> killit2: VM/Wine
[02:45:47] <Roklobsta> kubuntu runs fine in a virtualbox instance.
[02:45:47] <killit2> VM is SLOW caparitivly... and Wine does not work with everything...
[02:46:05] <ziph> Given that the slightest mention of Linux gets you all buzzing I'm guessing most of the channel is obese, smelly and bearded.
[02:46:14] <nofxx> and yea, you can only boot a windows machine to play some new game, apart from and that only reason, yr a desertor
[02:46:22] <Roklobsta> don't forget the beard dandruff
[02:46:38] <Roklobsta> ok, linux gui shits me
[02:46:47] <killit2> haha... such a pointless arguement... its like the "would you rather" conversations in the smoke pit
[02:46:50] <Roklobsta> win7 gui is what kde should be
[02:47:29] <Roklobsta> i have all the usual cygwin, mingw, quassell, GPL this that and the other running in win7
[02:48:12] <Roklobsta> and for some reason libreoffice runs and looks better in windows
[02:48:14] <killit2> I still prefer win2k's gui... barebones.. no flash graphics taking resources that are not needed... I wish they would put out a MS OS that is actually not crazy stupid bloated for no reason.. I don't need animations in my menus... give me back my clock cycles...
[02:50:07] <Roklobsta> win7 aero with the 3d bling is really nice
[02:50:52] <Roklobsta> it's not stupid over the top 3d bling like kde. it uses the 3d to make using the desktop smooth.
[02:51:23] <killit2> sure... looks wise.. but the extra strain I just don't feel is worth it... remember in space balls when they crash and she has all that luggage with the giant hair dryer and crap?! thats how I feel about the flashy graphics... sure it would be nice - but I really don't feel like carrying all this crap I don;t need
[02:52:02] <Roklobsta> eh, i reckon aero's smoothness is partly what's compelled me to stay with win7
[02:52:16] <killit2> yer a sucker for shiny eh?
[02:52:19] <Roklobsta> oh and the fact i don't need to flapping dist-upgrade eveyr 6 months
[02:52:44] <Roklobsta> it's not shiny
[02:52:53] <killit2> well you could always just do it slowly and call service packs... either way
[02:52:56] <Roklobsta> it's muted
[02:53:06] <mrfrenzy> does it really matter if the GUI uses 1% of one of your four cpu cores?
[02:53:14] <Roklobsta> 8 cores thanks.
[02:53:16] <mrfrenzy> I think it's totally worth it for a nice interface
[02:53:27] <mrfrenzy> that was directed to killit2 ;)
[02:53:29] <Roklobsta> ok 4 and hyper
[02:53:48] <Roklobsta> even bling with kubuntu in a virtualbox is fine
[02:54:16] <ziph> That'd be even more interesting, what kind of hardware is everyone using?
[02:54:22] <Roklobsta> the othe rmain reason i am sticking with win is Atmel in their infinite wisdom have their IDE's as win only
[02:54:33] <Roklobsta> i7-720QM
[02:54:43] <Roklobsta> shoudl be bitchin' for a few more years.
[02:54:44] <killit2> sure... if your video card has the extensions for it, your cpu has the extensions for it... and you have way more mem than is needed.. sure... eat it all up.. but I don't have four cores.. I still run a 3 year old dual core laptop.. and it runs GREAT without all that crap... but suddenly it runs like crap when it has to do all that shiny crap - but the end product of my work is the same... so I really don;t see the point
[02:54:59] <Roklobsta> you need a new laptop
[02:55:33] <Roklobsta> $599 for an acer with i5-2430, 8GB , 650GB HD and nvidia 520M i bought today.
[02:55:46] <killit2> so I should pay for a new tool ( lets face it... thats what a computer really is ) when the one I have works fine just because there is a new shiny one? consumerism if there ever was!
[02:56:14] <killit2> my truck is a 1995 c1500... should I replace that because I don't have GPS nav in it too?
[02:56:27] <ziph> Now talk about the politics of copyright.
[02:56:29] <Roklobsta> your dads old hammer is rusty and dated. You need a new hammer with a GrippySilicone(TM) grip
[02:57:14] <killit2> I think not.. I still have some of my grandfathers tools in the garage.. ww1 shit.. and it still works great... Illl keep my laptop until it dies or I start doing something that NEEDS more power.. until then I have no use for shiny that does nothing..
[02:57:24] <Roklobsta> i can stub my thumb much faster with a GrippySilicone(TM) gammer.
[02:57:41] <ziph> File sharing is stealing.
[02:57:53] <ziph> It's exactly the same as stealing a car.
[02:58:48] <killit2> no... not really... if I steal your car I deprive you of use... If I duplicate your car, you still have yours... I just created one for myself based of yours.. NOT THE SAME THING
[02:59:17] <Roklobsta> hehe you wouldn't steal a car....
http://www.youtube.com/watch?v=ALZZx1xmAzg
[03:01:00] <Roklobsta> haha truck. nono it's a UTE i tell you.
[03:01:19] <ziph> Compare and contrast: STTNG vs ST Enterprise.
[03:01:31] <killit2> good show vs crap...
[03:01:33] <killit2> comparison over
[03:01:52] <Roklobsta> pah, smelly green crap vs smelly brown crap
[03:02:01] <Roklobsta> they are both smelly
[03:02:14] <Roklobsta> you can argue about green vs brown
[03:02:49] <killit2> you must have a taste for scifi first... and starwars nerds dont count... starwars is the anti-scifi... its general fiction in disguise
[03:03:20] <Roklobsta> i used to watch star trek TNG when I was at uni but after I graduated and watched it I realised it was because I wanted to procrastinate at the time.
[03:03:48] <Roklobsta> i don't give a shit for starwars. anyway, best scifi is in books, not on the screen.
[03:04:47] <ziph> As opposed to now how you don't procrastinate on IRC?
[03:04:54] <killit2> reading is anti scifi.. a holographic projection or some type of immersion would be the most faithful scifi... until then Ill take a moving picture projection (TV) over words on paper
[03:05:00] <Roklobsta> and STDeepSleep9 and STVoyager ... oh don't get me started.
[03:05:47] <killit2> DS9 was a scifi freaking drama - more like a soap opera in space.. couldn't stand it
[03:06:19] <Roklobsta> anyway, to stay on topic I suspect Data was using an AVR Attiny overclocked to 2.6 terahertz.
[03:06:56] <killit2> external eeprom??? I knew alot for stuff for that small ammount of storage..
[03:07:05] <killit2> I == he
[03:08:58] <killit2> you are correct though... this conversation has been completely distracting reguardless of it quality...
[03:09:53] <ziph> What's the best IRC client?
[03:11:42] <ziph> mIRC?
[03:11:56] <killit2> XChat :)
[03:12:07] <killit2> ^ Linux client..
[03:12:46] <killit2> yeah - Mirc is pretty nice under windows though
[03:15:26] <Roklobsta> i am usinf quassel in windows with a quassel server on my mythbox
[03:15:36] <Roklobsta> that way i don't miss a word you guys say
[03:16:58] <Roklobsta> but the quassel client is a bit limited compared to mirc though it gets the job done. no dcc send
[03:17:19] <killit2> its been years since I xfered anything via irc anyways
[03:17:27] <Roklobsta> it's handy
[03:17:56] <Roklobsta> it's handy to transfer files in the context... like dcc with irc and files with skype
[03:18:15] <killit2> I imagine.. but if I do need to send something I have an FTP server so I just move it there and send the link - upload / dowload taken care of
[03:18:29] <Roklobsta> yeah me too. http though
[03:18:37] <Roklobsta> or scp
[03:18:46] <Roklobsta> coz compression rules.
[03:27:25] <Roklobsta> killit2: but just going back to earlier, i agree with you. my i7 laptop is a 5 year investement at least. i don't expect it'll be "crappy" for a long time.
[03:27:37] <killit2> alrighty... so I had it shit out the register values and everything looks ok.. UBRR0H: 0x00 UBRR0L: 0x77 (9600Bps) UCSR0B: 0x98 UCSR0C: 0x06
[03:28:08] <killit2> to 9600Bps with rx and tx enabled... also has the interupt enabled.. but nada...
[03:28:14] <killit2> Im missing something
[03:29:04] <killit2> anyone have any known working loopback test code?
[03:30:57] * Roklobsta hears crickets
[03:31:10] <killit2> I don't even hear that! :)
[03:31:26] <Roklobsta> they are outside my window. true story
[03:31:39] <killit2> I have a mouring dove..he's cooing.. thats all I got though
[03:32:38] <killit2> its a atmega328 (non pico-power) if that helps at all
[03:36:04] <Roklobsta> sorry ywre,/.' microwave oven kdnmflask ine&*(^# 2 interfering &*&^# WIFI (&* @ breaking up(*& @#( @!
[03:38:36] <killit2> wow... valid checksums on the packets and all!!! :P
[03:50:31] <nrdb> I am having trouble with the ADC on a at90usb1287 ... with how I have set it up, I appear to be getting impossible readings... please see
http://paste.debian.net/161518/ ... I need help in figuring out where I am going wrong?
[03:54:55] <killit2> EHHHHHH!! this makes no sence... my registers are correct...
[04:01:42] <DexterLB> hi guys, I have a little problem
[04:02:03] <DexterLB> whenever I enable the timer0 overflow interrupt, the AVR resets
[04:02:14] <DexterLB> other timers work fine
[04:04:50] <DexterLB> oh nvm - I had defined a wrong vector
[04:16:38] <nrdb> anyone! still stuck here.
[04:21:29] <cyanide> hello folks
[04:31:32] <j4cbo> does anyone know how the atmega32u4's maximum pll and high-speed timer clock speed relate to vcc?
[04:32:04] <j4cbo> the datasheet specifies maximum core clock vs vcc, but doesn't seem to indicate for the pll
[04:34:43] <j4cbo> "The High Speed Timer maximum frequency input depends on the power supply voltage and reaches its maximum of 64 MHz at 5V."
[04:34:52] <j4cbo> but no more detail than that is given
[05:21:05] <nrdb> j4cbo, is there a graph?
[05:27:42] <j4cbo> nrdb: no, unless i've gone blind
[05:33:01] <nrdb> j4cbo, section 15.2.1 says to reduce speed if voltage below 4 volts.
[05:33:07] <j4cbo> yeah
[05:33:15] <j4cbo> i saw that, but it doesn't say what to reduce it to :P
[05:33:27] <nrdb> j4cbo, what voltage you working at?
[05:34:10] <nrdb> j4cbo, that is one complex timer there.
[05:43:51] <nrdb> j4cbo, I can't find anything else.
[05:45:07] <j4cbo> well, i'm running at 3v, and half of the maximum is plenty for my application, so that should be fine
[06:01:56] <nrdb> j4cbo, I would think so.. bit of a lack in the datasheet there.
[06:36:50] <killit2> ehhhhhhhh.... I got it to start reading but for some reason its either not reading the bytes right or its not calculating the checksums right
[07:36:03] * Casper wonders what chupas is not
[07:36:29] <chupas> im not a cat
[07:36:55] <Casper> you know what chupas would mean in french?
[07:37:44] <chupas> in french? no
[07:37:49] <chupas> in spanish, perhaps
[07:37:59] <chupas> in english, yes
[07:38:19] <Casper> chu is a "contraction" for "Je suis" = I am, and "pas" == not
[07:38:25] <Casper> so "I am not"
[07:39:38] <chupas> I learned something today
[07:40:22] <Casper> and in spanish it mean what? :D
[07:48:27] <Steffanx> Google translate knows
[08:07:21] <OndraSter> chupacabra?
[08:21:16] * specing just looked at xmega parametric shet
[08:21:19] <specing> sheet*
[08:21:26] <specing> Wow, these things pack a punch
[08:23:36] <mrfrenzy> compare that with the stm32 sheet ;)
[08:24:09] <Steffanx> Ha beat that specing :P
[08:26:17] * specing looks
[08:27:06] <specing> Im thinking if it would be better to drag 10 UARTS over 20m, or have one rs485 bus
[08:28:40] <OndraSter> or one wireless! :P
[08:29:08] <OndraSter> 10 UARTs over 20m is 20 wires over 20m... at minimum
[08:29:21] <OndraSter> afk
[08:29:52] <Valen> a 5v uart over 20m is a bad idea
[08:30:06] <Valen> a statically charged cat walking past it will fry stuff
[08:31:28] <specing> Valen: with line drivers and ESD protection ofcourse
[08:31:59] <specing> Getting a dog to protect those cables from evil cats is doable
[08:32:57] <specing> Nah, im more for the 485 option
[08:33:17] <specing> half duplex is 2 + 1 cables while full duplex is 2* 2 + 1 cables, right?
[08:35:04] <specing> Or maybe some xmegas with 8 UARTs acting as load-balancing UART routers
[08:35:16] <specing> That would certainly be a fun project
[08:36:20] <Valen> specing: line drivers and ESD and gee wouldn't a bus be easier ;-P
[08:43:59] <RikusW> the dog will chew up your cables ;)
[08:48:07] <specing> the router network idea sounds good
[08:48:32] <specing> with it I can just keep adding stuff to the network and never worry about congested links
[08:48:59] <specing> on the downside, the code to do it would be massive
[08:49:25] <specing> (atleast compared to my code size up to now)
[09:15:27] <cyanide> whats up my peeps
[09:15:46] <Steffanx> Tha sky bro
[09:16:20] <cyanide> you don't say
[09:16:39] <Steffanx> Say what?
[09:17:22] <cyanide> something relevant to this channel
[09:17:38] <cyanide> ive found arrow to be cheaper than mouser which is usually taken as the cheaper source of components
[09:18:58] <cyanide> just a heads-up to those looking for components
[09:28:23] <learningc> cyanide, how does digikey compare to them?
[09:31:34] <specing> what about shipping?
[09:32:02] <learningc> shipping is $8 flat at digikey
[09:34:41] <Steffanx> To the other side of the world.. sure?
[09:36:45] <cyanide> digikey is usually more expensive than either
[09:37:37] <specing> Is there any component supplier that is based in china?
[09:37:47] <specing> that one would be the cheapest, then
[09:37:50] <cyanide> shipping via arrow is pretty cheap too. 4 big hall effect sensors in a box the size of the usps-flatrate small. about 30 usd via dhl from us to india, 4 days
[09:37:53] <specing> + free shipping!
[09:38:24] <cyanide> ebay lol
[09:38:48] <Steffanx> Most of the time that doesn't mean free shipping
[09:38:53] <specing> cyanide: you are in india? Why did you purchase from us which gets the stuff from china anyway and chine is like right next to you
[09:39:53] <cyanide> im not sure how you expect me to buy from china if there's no component supplier that's based in china like you said :)
[09:40:03] <cyanide> shouting in chinese and expecting someone from china to hear me? lol
[09:40:19] <cyanide> someone in china*
[09:40:49] <cyanide> besides, this was a hamlin electronics product. i couldn't even find it at most big component suppliers in the us
[09:41:05] <cyanide> mouser didn't carry them, digikey, arrow and maybe 2-3 others who were charging quite a bit
[09:41:21] <cyanide> digikey was charging about 25 for each sensor, arrow just 9.90
[09:42:44] <cyanide> till the time CapnKernel was in china, i was getting my stuff through him
[09:43:07] <cyanide> ebay is hit-or-miss depending on the mood of the postman here in india
[09:43:15] <cyanide> half of my parcels aren't delivered
[09:43:33] <specing> lol
[09:43:44] <specing> I guess that is one hell of an UDP postman
[09:44:18] <cyanide> lol
[09:45:06] <specing> I wonder how many postmen they have to replace each year?
[09:46:04] <cyanide> i want a honda cbr1000rr :(
[09:46:11] <specing> What?
[09:46:29] <cyanide> a bike
[09:46:34] <Steffanx> Ebay cyanide :)
[09:46:55] <cyanide> for a bike?
[09:47:00] * Tom_itx breaks cyanide's leg and saves him the expense of a bike
[09:47:41] <cyanide> then i will make a motorised wheelchair
[09:47:51] <cyanide> ive always wanted to make one with a relatively big engine
[09:48:22] <Steffanx> Yeah, you can find enough of them in on ebay
[09:49:08] <cyanide> Steffanx, not in my country
[09:49:21] <cyanide> we have a ~140% import duty on bikes
[09:49:43] <cyanide> a 1000rr roughly costs about 30k usd
[09:49:52] <Steffanx> And more mobile phones than toilets :P
[09:50:16] <cyanide> heh
[09:50:59] <cyanide> this isnt fair,
http://cgi.ebay.com/ebaymotors/NEW-2012-HONDA-CBR1000RR-ALL-COLORS-WHITE-BLACK-RED-12-SALE-/290688915055?pt=US_motorcycles&hash=item43ae68d66f
[09:51:05] <cyanide> 12k for a brand new one
[09:51:40] <cyanide> and that isnt even the cheapest auction on ebay
[09:58:09] <Steffanx> Are the roads in inda good/safe enough for a motorbike like that?
[09:59:43] <specing> Steffanx: that is a one time use bike, if you know what I mean
[09:59:56] <Steffanx> :P
[10:00:57] <cyanide> :)
[10:01:02] <cyanide> yes, they are safe enough
[10:01:28] <cyanide> i live in mumbai/bombay. has its fair share of idiots like every other big city in the world. but things improve
[10:30:51] <learningc> CapnKernel is out of china?
[10:36:57] <Steffann> Yes
[11:24:13] <ama2er> when interfacing a MAX232 to PC should the RS232 cable be in NULL modem connection or should it be a straight wiring. I'm connecting the PIN 2 of cable to PIN 14 of MAX232 and 3 of cable to 13 of MAX232.
[11:27:57] <specing> Tx goes to Rx and vice versa
[11:28:32] <Casper> ama2er: straight
[11:29:48] <ama2er> specing:, Casper: which is correct i get different answers
[11:33:23] <Kevin`> Casper: the null cable is the one with crosssed connections
[11:33:49] <Kevin`> although I suppose it depends how you are using the max232
[11:34:09] <ama2er> can you be more elaborate
[11:34:47] <ama2er> i am trying to connect PC to ATMEGA32
[11:34:58] <Casper> Kevin`: hence why you use straight
[11:35:05] <Casper> null is for pc-pc
[11:35:11] <Casper> while straight is pc-device
[11:35:38] <Kevin`> I would consider the microcontroller a host, but I suppose it really doesn't matter how you wire it
[11:35:40] <Casper> but anyway, since you wire your stuff, you can use any of them, but better use straight
[11:35:55] <Casper> I would consider the microcontroller as a device, not a host
[11:36:08] <Casper> unless you connect the avr to a modem or alike
[11:36:08] <Kevin`> most embedded devices are wired as hosts (eg, networking equipment)
[11:36:16] <Casper> Kevin`: are you sure?
[11:36:47] <Kevin`> well all the one's i've seen with db9 ports. it stops mattering once they have logic-level ports since you just have to wire it up sanely
[11:37:08] <Kevin`> anyway, either way will work as long as the device matches ;p
[11:37:42] <ama2er> in the RS232 standard it says use NULL for DTE and DTE and use straight for DTE and DCE.
[11:38:02] <Casper> nobody follow the standards
[11:40:09] <Casper> time to go dig up a tree...
[11:40:14] <Casper> actually 2
[12:10:27] <RikusW> http://quantumaniac.tumblr.com/post/18733391757/the-great-beards-of-physics-for-many-physicists
[12:25:14] <RikusW> http://hackaday.com/2012/03/28/building-the-worst-linux-pc-ever/
[12:28:03] <Steffanx> Yeah, an oldy :P
[12:28:05] <Steffanx> ie
[12:30:03] <RikusW> I don't see the use of even trying to build that :-P
[12:32:32] <Steffanx> "Because I can" is a good reason
[12:32:43] <Steffanx> "because it's possible"
[12:59:04] <clOI> hello everybody. Where is __AVR_ATmega48P__ defined? Is this defined by the avr-gcc because of the mmcu flag?
[13:00:13] <specing> probably
[13:00:39] <clOI> thanks
[13:03:38] <clOI> does anybody use kdevelop for avr-projects?
[13:08:13] <specing> nope
[15:32:42] <gkwhc> hmm i wonder why buzzing audio noise is introduced when doing ADC oversampling to 16bit, rather than direct/default 12 bit conversions
[16:16:10] <rue_house> oversampling?
[16:16:33] <rue_house> do you mean you start the next sample before the converstion of the last one is finished?
[16:30:56] <Casper> gkwhc: which avr is that?
[18:38:03] <spizzen2012> ciao ciao
[18:38:16] <spizzen2012> !list
[20:59:38] <nrdb> why do most op-amp datasheets not mention min. voltage? I can easily find max. voltage.
[21:27:34] <Roklobsta> min power ?
[21:27:46] <Roklobsta> or min input?
[21:28:52] <nrdb> Roklobsta, few mention min. supply voltage
[21:30:00] <nrdb> looking for amplifying some voltages, need 5V op-amps any recommendations
[21:33:46] <Casper> nrdb: well, the minimum voltage depend on your usefull output
[21:34:07] <Casper> they DO mention how many volts the opamp need from the rail for the output
[21:34:12] <Casper> usually around 3V per rail
[21:36:26] <nrdb> Casper, I don't know. but hiding the figure like that certainly doesn't make things clear.
[21:36:42] <Casper> you probably just don't know what to look for
[21:38:32] <nrdb> Casper, that may be so.. but why do most say very clearly what the max is with out also what the min is just as clearly?
[21:38:54] <Casper> there is no direct answer for minimum
[21:39:02] <Casper> it all depend on your needs
[21:39:15] * nrdb oh!
[21:39:22] <Casper> if you need +/-5V then the minimum is different from +/-1V
[21:39:45] <Casper> hi Richard_Cavell
[21:40:25] <Richard_Cavell> hiya
[21:41:34] <Casper> sup?
[21:44:23] <nrdb> Casper, One thing I need to do is amplify a current shunt... the NE5532 datasheet says voltage range of +-3V to +-20V ... do you think it would work fine from 5v rail? ... I was thinking of using 1 opamp as a diferential amp, on for to generate a 2.5V signal as ground.
[21:44:59] <Casper> +/-3V = 6V rail to rail
[21:45:47] <Richard_Cavell> Casper: I'm studying anatomy. Doing some video editing at work, so looking for RAID storage solutions.
[21:46:04] <nrdb> Casper, I know. 5V is a little low.
[21:47:12] <nrdb> Richard_Cavell, I have been using the Linux , mdadm system for a long time now, no problems.
[21:47:17] <Casper> nrdb: might not work well
[21:47:25] <Casper> Richard_Cavell: how redundant must it be?
[21:47:30] <nrdb> Casper, ok.
[21:47:49] <Casper> raid5 is usually good enought
[21:48:06] <Richard_Cavell> Casper: RAID 5 is good enough for me
[21:48:29] <nrdb> Richard_Cavell, I have 5 x 1TB drives, in raid 5.
[21:49:06] <Casper> nrdb: small
[21:49:27] <Casper> I have 6x 3TB in raid5 :D
[21:50:02] <Guest60319> i need a terminal program (windows) that will interface with the serial port SPI interface that ponyprog uses
[21:50:21] <Richard_Cavell> Casper: what kind of box do you have them in?
[21:50:25] <Richard_Cavell> I'm thinking Netgear Ultra 6
[21:50:25] <nrdb> Casper, I don't trust 3TB drives yet! ... plus when I built the system they didn't exist.
[21:51:33] <Casper> color warning!
[21:51:37] <Casper> OS: Gentoo | Up: 2 days, 12:52 Record: 62 days, 06:15:03 | CPU: Intel(R) Core(TM)2 Quad CPU Q9650 @ 3.00GHz (load: 0.04, 0.07, 0.16) | Gfx: nVidia Corporation GT200b [GeForce GTX 275] (rev a1) Res: 1680x1050 | Ram: 2409/5952MB | Hdd: /mnt/data 6764GB/15001GB (46%), /home 46GB/85GB (58%), /tmp 4GB/22GB (19%)
[21:52:17] <Casper> be warned about hardware solution: if the controller fail for whatever reason, you most probably lost ALL the data
[21:52:33] <Casper> as you need the exact same controller and firmware to be sure to recover
[21:53:03] <Richard_Cavell> Casper: I'm unsure whether to buy an off the shelf solution, or build my own with a hardware RAID card
[21:53:39] <Casper> I'ld say: go software raid!!
[21:54:15] <nrdb> I second that... can use any hardware
[21:54:39] <Richard_Cavell> and what, keep the drives internally and attach them to my motherboard?
[21:55:14] <Casper> how many hds?
[21:55:20] <Richard_Cavell> at least 4
[21:55:24] <Richard_Cavell> so I can do RAID 5
[21:55:34] <Casper> raid5 need minimum 3 drives btw
[21:55:44] <Richard_Cavell> I have a Netgear Duo but 2 bays isn't enough
[21:55:45] <nrdb> I am testing a Dreamplug atm.
http://www.globalscaletechnologies.com/t-dreamplugdetails.aspx ... the esata port can use a port multiplier, so you can connect upto 5 HDD to it.
[21:56:12] <Casper> NEVER USE A PORT MULTIPLIER ON SATA!!!!!
[21:56:24] <Casper> none are reliable
[21:56:32] <Casper> they all disconnect under heavy use
[21:56:37] <Casper> like rebuild
[21:56:46] <Casper> or big file transfert
[21:57:37] <nrdb> Casper, I transfered several movies to it no problem. But I will keep that in mind.
[21:57:42] <skynetishere> i need a terminal program (windows) that will interface with the serial port SPI interface that ponyprog uses
[21:57:49] <skynetishere> repeated cuz i changed nicks
[21:59:17] <Casper> skynetishere: to flash? or to issue random data?
[22:01:30] <skynetishere> just random
[22:01:43] <Casper> I do not think that any exists
[22:02:18] <skynetishere> i am going to use it in the project later i just need to be able to basically type in chrs and read them (like on terminal+serial+max223+uart)
[22:03:13] <skynetishere> how do you guys test SPI?
[22:03:34] <Casper> actually trying to talk to the device?
[22:03:38] <Casper> that's how!
[22:03:47] <skynetishere> with what?
[22:03:56] <skynetishere> another SPI device?
[22:04:18] <Tom_itx> wire up an spi device and make it work
[22:04:31] <Tom_itx> read the data sheet to see what it wants
[22:04:36] <nrdb> I need an opamp to use as a differential amp, using a 5V rail, optionally generate a 2.5V signal as ground. any recommendations?
[22:06:12] <skynetishere> next step is to crank out another device that interfaces serial (or maybe usb) ... guess i will put this part of the project down and build up that part
[22:06:30] <Roklobsta> nrdb: what are the max and min volatge levels you are trying to amplyfy and what is the gain?
[22:07:04] <Roklobsta> because dodgy crap can happen if your signal goes near the rails on the input and output.
[22:07:25] <skynetishere> (serial or usb) -> (avr SPI_master) -> (lots of avr slaves)
[22:07:56] <Roklobsta> even if the #*#%^ data sheet says they are rail to rail
[22:08:52] <Roklobsta> my last employer used a small signal amp frontend to the ADC that was from 0 to 3.3V and it was a problem...
[22:09:11] <Roklobsta> stupid twit never did THD testing to see they didn't introduce harmonics.
[22:09:36] <Roklobsta> he was a 65 year old electronics engineer and didn't know what THD was.
[22:09:36] <nrdb> Roklobsta, I am still designing is but about 25mV @ 100x gain or so ... but figures up for change in I need to.
[22:09:49] <Casper> Roklobsta: that is actually one of the issue I have with one project...
[22:10:05] <Casper> the signal is so low, and the noise so high that the result is useless
[22:10:19] <Roklobsta> i discovered after over a hundred devcies were in the field that it had "issues" at the low end.
[22:10:25] <Roklobsta> he didn't care
[22:11:03] <Roklobsta> </gripe>
[22:11:26] <Roklobsta> casper: are you measurung the signal differentially?
[22:11:31] <Roklobsta> what is the signal?
[22:12:21] <Roklobsta> if you can try and use differential measuring and use opamps fed with say -6V to +6V or more so you don't have to do stupid DC offsets.
[22:13:07] <Roklobsta> and make sure your input stage of the ADC doesn't overload the opamp's output stage.
[22:13:24] <Roklobsta> some opamps go stilly when loaded near the rails. data sheet = LIES DAMNED LIES
[22:14:24] <Casper> Roklobsta: yes, but the signal was a few mV, and the noise was awefull (rought smps)
[22:15:24] <Roklobsta> hm... smps is fine but you always need to use a low dropout regulator to power things that are sensitive
[22:16:24] <Casper> I had instability for some reason
[22:16:34] <Casper> was hacking an ATX psu
[22:16:45] <Casper> to try to add current limiting
[22:16:53] <Casper> to sum up, I basically failed
[22:17:43] * Casper stills need some GOOD smps expert for advices
[22:25:42] <Casper> Roklobsta: are you good with smps?
[22:25:59] <Roklobsta> er i have done a few... they worked..
[22:26:30] <Casper> ok, I have issue with one I try to make... I do not get the power out that I expected
[22:26:42] <Roklobsta> the last bunch i did a while back i actually used linear chips and used the switchercad from linear to design them and analyse noise and deal with it in simulation. Lo and behold the real thing worked well.
[22:27:00] <Roklobsta> do you have a schamtic?
[22:27:13] <Casper> somewhere
[22:27:53] <Casper> . . . look like I didn't made the final one
[22:28:23] <Casper> basically, I used a salvaged core from an ATX psu and adjusted the turn ratio for the voltages I used in and wanted out
[22:28:41] <Casper> the main problem is: the duty cycle is WAY too high
[22:30:19] <Roklobsta> do you know what the inductance is?
[22:30:35] <Casper> Roklobsta: sadly nope
[22:30:53] <Roklobsta> that would be your problem. you can't use any old inductor or transformer
[22:31:07] <Casper> I adjusted the turns
[22:31:19] <Roklobsta> the ferrite material might be wrong
[22:31:30] <Roklobsta> what is the SMPS switch?
[22:31:45] <Casper> basically, same volt/turn, same frequency
[22:31:51] <Casper> TL494
[22:32:27] <Casper> I used the suggested tip31, and then made a nchan fet drive... both gave the same results
[22:32:39] <Casper> with barelly any heat in the transistors/fets
[22:32:46] <Casper> and high duty cycle
[22:33:06] <Roklobsta> is the feed back ok?
[22:33:10] <Casper> yes
[22:33:13] <Casper> voltage is ok
[22:33:43] <Casper> output: 5V, input 15V
[22:33:52] <Casper> at around 800mA out the output start to drop
[22:34:14] <Casper> and then I get near 100% duty cycle
[22:34:36] <Casper> and of course, since I didn't set the max duty cycle I start to get cross conduction
[22:35:03] <Roklobsta> you might be saturating the ferrite
[22:35:37] <Casper> http://img815.imageshack.us/img815/5598/72921324.jpg ← don't mind the turn ratio on the xformer, it's wrong
[22:36:20] <Casper> and the schem might contain some other errors, that's one early design, that didn't worked well at all, I did more change (mainly rewind) and got way better results, yet not enought
[22:36:58] <Casper> basically, I took an ATX psu, mesured it's switching frequency, then took out the transfo and parts
[22:37:18] <Casper> rewinded the transfo to adapt to the voltages, same with the output inductor
[22:37:26] <Roklobsta> you have to be sure any inductor and transformer ferrites you use are suited for the switching frequency and can handle the currents you want - by that I mean the peak current/
[22:37:30] <Casper> based on the oscope reading I did
[22:38:17] <Casper> I used multi-strands of enamel wire for the transfo, so no skin effect can occur
[22:38:43] <Roklobsta> like i said, you might be saturating the ferrous material with the currents you are using
[22:38:56] <Casper> which isn't the case
[22:39:01] <Roklobsta> are you sure?
[22:39:04] <Casper> transfo is good for atleast 350W
[22:39:05] <Roklobsta> how do you know?
[22:39:27] <Roklobsta> was the transformer from the mains side or the DC side?
[22:39:29] <Casper> I have about 20W passing throught only max
[22:39:35] <Casper> main transfo
[22:40:00] <Casper> http://img809.imageshack.us/img809/1436/img3061p.jpg ←
[22:40:03] <Roklobsta> yeah well mains is higher voltage so doesn't need as much current.
[22:40:05] <Casper> actually 300W transfo
[22:40:17] <Casper> it'S the left pcb
[22:40:22] <Casper> rightmost transfo
[22:40:31] <Casper> with the big toroidal core
[22:40:36] <Roklobsta> the toroid
[22:40:37] <Roklobsta> ok
[22:40:52] <Casper> main transfo EI core for the main transfo
[22:40:58] <Casper> and toroidal as the output inductor
[22:41:02] <Roklobsta> oh
[22:41:15] <Roklobsta> the HIPOT?
[22:41:18] <Casper> basically, the only difference is the drive and the turn ratio
[22:41:28] <Casper> for the main xfrm yes
[22:41:46] <Roklobsta> can you use the toroid instread?
[22:42:06] <Casper> why?
[22:42:19] <Roklobsta> just to be sure you aren't in saturation
[22:42:22] <Casper> I wanted to make an isolated one
[22:42:24] <Roklobsta> it doesn't hurt
[22:42:44] <Roklobsta> toroid can be isolated.
[22:42:52] <Casper> shouln't be in saturation, beside, saturation would make the input current peak and make the transistor overheat
[22:42:57] <Casper> which do not occur
[22:42:57] <Roklobsta> true
[22:43:05] <Roklobsta> it's just that you are getting to max duty
[22:43:19] <Casper> it's really a lack of power transfert that I just can't explain
[22:43:20] <Roklobsta> befor eyou want to
[22:43:54] <Roklobsta> whis the feedback path tuned right?
[22:44:22] <Casper> I think so
[22:44:36] <Casper> output is stable until I reach the max duty
[22:44:41] <Roklobsta> like that one doesn't have any ressitor devider
[22:44:42] <Roklobsta> in the diagram
[22:45:01] <Casper> possible, I think I didn't drew it
[22:45:23] <Casper> it was a quick draw
