Back
[00:07:32] <jadew> bought it
[03:18:47] <molavy> i have camera and want when take a pic
[03:19:09] <molavy> camera have sdcard and save data to sd card
[03:19:30] <molavy> but i want instead of save data to sd card
[03:19:54] <molavy> route data to my computer usb port
[03:20:12] <molavy> how can i implement this
[03:20:14] <molavy> ?
[03:20:57] <molavy> i want have pic on my computer on taking picture
[03:22:44] <theBear> not really the right channel, or even close, depends on the camera, a lot of them there is already an app
[03:23:55] <Xark> Yeah, even if it had something to do with AVR, AVR is probably not ideal for any image processing applications.
[03:24:12] <theBear> also depends a lot on os
[03:25:25] <jadew> Xark, it's not image processing, it's more like file moving :)
[03:26:00] <theBear> several mb file moving, depending on what interfaces the camera provides
[03:26:01] <Xark> jadew: Yes, if it is just coping then it is perhaps suitable. :)
[03:26:11] <Xark> copying*
[03:28:47] <jadew> molavy, I think you could emulate a sd card with an avr
[03:29:20] <jadew> so when the camera sends data to the virtual sd card, you send it over usb to the pc (to an application you'll have to write)
[03:29:27] <jadew> and then that data can get in a folder
[03:29:29] <jadew> easy
[03:29:31] <jadew> :)
[03:30:09] <Xark> Just a SMOP (small matter of programming). :)
[03:30:19] <jadew> yeah :P
[03:31:06] <jadew> google, maybe there's a project already doing that
[03:31:22] <jadew> if not, I'm sure I could do it for 2k :D
[03:32:24] <Xark> Sounds generous of you. :)
[03:32:48] <jadew> was actually thinking higher, but I guess I'd enjoy the project :P
[03:33:08] <jadew> (there's a lot of reading involved)
[03:47:51] <specing> molavy: install digicam and press the appropriate button
[03:48:02] <specing> (Or was it digikam?)
[04:19:46] <Malinuss> morning avr
[04:30:00] <OndraSter> hey Malinuss
[04:38:36] <Malinuss> I'm still thinking about how my RF data transfer thing should work... Maybe I should just use the library or the software usart heh
[04:40:52] <OndraSter> uh?
[04:40:58] <OndraSter> software USART?
[04:40:59] <OndraSter> why?
[04:43:48] <Malinuss> OndraSter, because the RF link is going to be between two attiny13.. (one way only though)
[04:49:13] <OndraSter> the tiny13a hasn't got UART?
[04:49:22] <OndraSter> why not use attiny85? it is cheaper on mouer :)
[04:52:54] <Malinuss> OndraSter, nope, no uart. I'm pretty sure my method will work though ;P. On mouer I have to pay this crazy 20€ shipping price... The attiny13A was 1$ (with shipping) on ebay, while the attiny85 was 2,5$ :)
[04:54:30] <OndraSter> eh
[04:54:31] <OndraSter> for one ?!
[04:55:27] <Malinuss> for one what? The shipping is 20€ yeah...
[04:55:50] <OndraSter> attiny13a: .73 eur, attiny85: .688 eur
[04:56:05] <OndraSter> I have got local reseller from mouser, $6 incl. VAT shipping :)
[04:56:53] <RikusW> only t2313 /4313 got uart
[04:57:07] <OndraSter> oh
[04:57:11] <RikusW> some others got usi....
[04:57:31] <OndraSter> 85 has got USI
[04:57:48] * RikusW don't want to use USI for UART....
[04:58:00] <RikusW> its yuck...
[04:58:20] <OndraSter> it is :)
[04:58:50] * RikusW borrows one of OndraSter's xmega uarts and put it on a ATtiny :-P
[04:59:23] <OndraSter> :)
[04:59:26] <OndraSter> you wish :D
[04:59:27] <RikusW> *an
[04:59:41] <OndraSter> puts* (if you're at it)
[04:59:47] <RikusW> not so sure silicon grafting is possible :-P
[04:59:59] <RikusW> heh
[05:00:31] <RikusW> that one was a typo...
[05:00:38] <Malinuss> OndraSter, I'm hopeing the high for 20msec for logic 1, and high for 10msec for logic 0 (and let's say 5msec low between each bit) will work... because even with a lot of noise, the propability of a noise exactly that lengh, shouldn't be very big
[05:00:48] <Malinuss> and the DC balance should be fine too
[05:00:57] <OndraSter> what was wrong on manchester? :(
[05:02:06] <OndraSter> oh well
[05:02:09] <OndraSter> I suppose I could get up
[05:02:12] <OndraSter> although I do not want..
[05:02:30] <RikusW> ?! its almost bedtime already :-P
[05:02:41] <RikusW> well not quite yet/// ;)
[05:03:41] * RikusW give OndraSter an epi shot :-D
[05:06:06] <Malinuss> haha my seides
[05:06:08] <Malinuss> *sides
[05:06:43] <Malinuss> I just realised that "manchester" OndraSter was refering too, is a encoding type...
[05:07:21] <Malinuss> First I thought "damn that was a pretty random statement, I never said anything about manchester"'
[07:01:10] <Malinuss> OndraSter, I don't really understand how to implement manchester - that's what is wrong with it ;)
[07:03:08] <OndraSter> oh
[07:03:10] <OndraSter> that sucks
[07:04:31] <Malinuss> hehe
[07:05:18] <OndraSter> for you though :D
[07:10:59] <darknite> haha
[07:32:00] <OndraSter> I am glad I studied the middle school I did and I work with AVRs
[07:32:08] <OndraSter> that is what is about the #1 class in 2nd semester :P
[07:46:32] <darknite> Malinuss:
http://www.atmel.com/images/doc9164.pdf maybe it's a bit of a thick read though, but code examples in it as well
[07:47:42] <OndraSter> yay even atmel has got appnote for manchester
[07:47:42] <OndraSter> nice
[07:48:30] <Malinuss> I don't really think I will be needing all that
[07:48:46] <Malinuss> it's clever if you have multiple devices though
[07:53:25] <Malinuss> OndraSter, I'm sorry for you that you actually studied in middle school. I think childhood should be the time for play, not study.. you can study all you want in high school/uni
[07:53:39] <bitd> http://doc.chipfind.ru/irf/irg4bc20fd.htm - I don't seem to be able to turn this on with 5V out of my arduino, am I missing something?
[07:54:53] <OndraSter> Malinuss, high school = middle school here
[07:54:55] <OndraSter> heh
[07:55:18] <OndraSter> basically - the teacher explained it and I remembered it and understood it just fine
[07:55:24] <OndraSter> never learned at home for anything
[07:55:28] <Malinuss> yeah
[07:58:50] <OndraSter> WHAT THE FU.K ARE THEY TEACHING THEM.
http://clip2net.com/s/2MUWc
[07:58:53] <OndraSter> translation of the comments
[07:59:06] <OndraSter> 2) "low" is pseudoinstruction blabla, "high" is pseu..... blabla
[07:59:15] <OndraSter> 3) register Z is made from registers R31:R30
[07:59:20] <OndraSter> why the fu.k should that be in comments?
[07:59:29] <OndraSter> if somebody opens it obviously he knows what he is doing
[08:00:05] <OndraSter> that is in the datasheets
[08:03:29] <specing> Not everyone is that intelligent
[08:03:38] <specing> or patient enaugh to read datasheets
[08:03:47] <specing> (as we have previously establsihed)
[08:03:59] <OndraSter> enough*
[08:04:01] <OndraSter> but still
[08:04:08] <OndraSter> if somebody wants to work with it, he obviously needs to know the basic stuff
[08:04:46] <Malinuss> that looks like my first asm code ;D
[08:04:53] <Malinuss> comments, commenets errywhere
[08:23:00] <theBear> people that don't wanna read datasheets don't wanna be programming micros... least, that's the way it SHOULD be
[08:25:02] <discorpia> i guess that pretty much sums up arduino
[08:25:18] <discorpia> microcontroller programming without datasheets :P
[08:25:36] <seldon> You crack me up.
[08:27:01] <Amadiro> discorpia, I program arduinos occassionally, and I do read datasheets... :P
[08:28:10] <discorpia> in general i assume it's because you want to, not because you need to
[08:28:41] <theBear> and we gonna assume you say arduino to be cheeky, and not cos you don't know the chip that sits on it :)
[08:29:13] <Amadiro> Well, I occassionally teach people to use them, so
[08:29:34] <Amadiro> There's a class at my university that is about user-interaction, and they have the option to use arduinos
[08:30:03] <Amadiro> For my own purposes I usually just put an attiny on a veroboard :)
[08:30:54] <Amadiro> but the arduino was the reason I got first interested in AVRs, so there's that
[08:37:40] <theBear> good boy :)
[08:38:40] <discorpia> i still use arduino as my primary board and then transfer whatever project to separate 328 or tinys on stripboard when things work, i do about 50/50 avr-gcc and arduino "sketches". but i've been using my mkii more and more
[08:39:38] <Amadiro> The only gripe I really have with them is that they are pretty overpriced for what they give you, by modern standards
[08:39:49] <Amadiro> for instance the due is like 50$.
[08:40:51] <Amadiro> for 12$ I can get an ARM chip on a launchpad stellaris with full debugging capabilities, and for ~20$ I can get a 400MHz ARM dev board that runs linux with ethernet, and for 35$ even one with a graphics card
[08:41:58] <specing> and for $50 a cortex-a8 one with hardware SATA
[09:04:59] <Malinuss> Amadiro, you can't really compare a 8-bit to a full-blown computer (what I would consider a arm chip with linux running on it to be)
[09:07:30] <specing> Amadiro: what $20 ARM deb board do you have in mind?
[09:18:36] <Amadiro> Malinuss, the arduino due for 50$ is a 32-bit ARM too, but runs only at like 84MHz or so.
[09:18:45] <Amadiro> specing, there's one called aria or so, one second
[09:19:00] <Amadiro> specing,
http://www.acmesystems.it/aria 24€
[09:19:21] <Malinuss> yeah okay
[09:19:23] <Amadiro> We've used that before in a RC-blimp-project thing
[09:21:01] <specing> Amadiro: Ah, I already have an AT91SAM devboard
[10:16:24] <OndraSter_> wow
[10:16:29] <OndraSter_> atxmega128b1 is fairly cheap
[10:16:39] <OndraSter_> we are talking <$4 ON FARNELL
[10:16:42] <OndraSter_> on digikey they cost more
[10:22:13] <specing> Uhmm...
[10:22:18] <specing> farnell is actually cheaper?
[10:22:21] <specing> CANT BE
[10:22:49] <OndraSter_> yes
[10:22:51] <OndraSter_> it is cheaper by far
[10:23:10] <OndraSter_> plus they are UK based so none of the import papers and stupid US laws crap :P
[10:23:20] <OndraSter_> hmm why are those segment LCDs so.. expensive?
[10:23:32] <OndraSter_> even regular 3 digit 7segment is $1.5 and up
[10:23:37] <OndraSter_> nearly $2
[10:23:55] <OndraSter_> if I wanted 6digit 12segment... I would have to shove out a lot of money
[10:24:01] <OndraSter_> afk
[10:33:25] <specing> because they are like huuugeee
[10:33:44] <specing> Also buy them from china
[10:33:46] <specing> I guess
[10:37:40] <OndraSter_> huge?
[10:37:41] <OndraSter_> why huge?
[10:49:52] <theBear> one of these days i wanna make a 'fake' fig-8 giant display for a clock, you know, you get some kinda translucent kinda panel, a chunk of foam the same size and maybe a couple inches thick and cut shapes into it, then mount real bright leds in the spaces :)
[10:50:23] <OndraSter_> THE SIMPSONS ALREADY DID IT!
[10:50:26] <OndraSter_> no wait, that was sparkfun
[10:50:34] <OndraSter_> afk
[11:54:59] <Tom_itx> OndraSter what was that smps you found a while back
[11:57:43] <theBear> oh, some dude with a webpage did the giant display years ago, but that doesn't stop it being cool, so long as you make it yourself
[12:05:09] <wisehippy> hey, I was wondering if I could get an opinion about a problem I have. it's on the topic of scanning and doing POV with an LED 5x7 display
[12:05:57] <Horologium> opinion, 42
[12:06:00] <Horologium> next question?
[12:06:18] <wisehippy> Hmm, should have went to #comedians instead.
[12:06:36] <Horologium> they aren't as funny as we are.
[12:06:43] <wisehippy> Ha, totally.
[12:06:56] <wisehippy> to save some typing, here's my post:
http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&p=1034007#1034007
[12:07:25] <rue_bed> theBear, I used my cnc to cut them
[12:07:43] <rue_bed> I was orig gonna use little christmas light bulbs
[12:07:53] <rue_bed> my digiits are about 20cm
[12:08:04] <rue_bed> I'm gonna do it with an avr now
[12:08:22] <rue_bed> wall clock, I have a frosted piece of glass for the front
[12:09:01] <rue_bed> wisehippy, whats the question
[12:09:30] <rue_bed> oh a webpage
[12:10:22] <rue_bed> wisehippy, sketch or real avr C code direct?
[12:11:50] <wisehippy> real avr C code. should have been more direct.
[12:11:56] <wisehippy> my bad.
[12:13:19] <rue_bed> real avr code
[12:14:38] <rue_bed> hmm, what if I say your doing it all non optimally :)
[12:14:58] <wisehippy> wouldn't hurt my feeling one bit. that's how I thought of attacking it.
[12:15:12] <rue_bed> if you set the count on interrupt, your going to lose counts
[12:15:29] <rue_bed> you want to use the divider so that it just overflows at 1 second
[12:17:20] <wisehippy> I thought that's what I did? It does keep time and the logic works.. it's really close to ~1Hz on overflow now. I've clocked it and looked at it with a USB scope
[12:17:54] <rue_bed> your resetting tcnt
[12:18:27] <rue_bed> but right now your problem is flicker, yes?
[12:18:31] <wisehippy> right.
[12:19:03] <rue_bed> I cant make out how the hardware is set up
[12:19:13] <rue_bed> is it an array or a spinning set?
[12:21:23] <wisehippy> the LED matrix?
[12:22:08] <rue_bed> do the leds move or is there an array of then
[12:22:40] <rue_bed> do yo have a schematic posted?
[12:23:07] <wisehippy> I don't but I could draft one up quick.
[12:23:17] <wisehippy> it's an LED array. cathode column, anode row
[12:23:23] <wisehippy> this is the array:
http://www.jameco.com/Jameco/Products/ProdDS/2005497.pdf
[12:23:25] <rue_bed> ah c is column and r is row, ok
[12:23:28] <rue_bed> ok
[12:23:35] <rue_bed> how many 595 are in there?
[12:23:51] <wisehippy> just one. 595 8-bit register controls the row on or off
[12:24:31] <wisehippy> then PD0-4 go to a BC547c to do the inversion to open a gate to pull the corresponding looped column to ground
[12:25:07] <wisehippy> and I write the 8-bit row register (which is in the array datetime[]) one bit at a time via SPI in that loop structure.
[12:25:16] <rue_bed> ok so its portd against the 595
[12:26:19] <rue_bed> do the bits stack horizontally or vertically for the fields?
[12:28:39] <rue_bed> I dont understand why it dosn't smear
[12:28:51] <wisehippy> rue_bed, sorry had to go look at the circuit quick.
[12:29:12] <wisehippy> portd pins 0 through 4 go to the columns
[12:29:26] <rue_bed> may I make a minor design suggestion
[12:30:24] <wisehippy> I'm using SPI master to write out the 8-bit row byte for each column SPDR. it doesn't stack up because I order the datetime array in reverse order
[12:31:03] <wisehippy> so I write out minute column first, it's stacks and gets pushed to the correct column in the end after I write all 5 bytes of date info
[12:31:13] <rue_bed> you want to do the following order (to my experiance)
[12:31:20] <wisehippy> ok
[12:31:41] <rue_bed> a) turn off all the row bits
[12:31:46] <rue_bed> b) set the column
[12:31:56] <rue_bed> c) turn on the row bits
[12:32:04] <rue_bed> d) delay a little bit
[12:32:18] <rue_bed> e) loop
[12:32:30] <rue_bed> your doing something funny tho
[12:32:43] <wisehippy> let me stop you there.
[12:32:43] <rue_bed> your shifting shifted data into a shift register
[12:33:42] <rue_bed> a,b,c need to happen really quick of course
[12:34:21] <rue_bed> the delay needs to be significantly greater than the off time it had in step b
[12:34:59] * Tom_itx forces rue_bed into the delay loop
[12:36:45] <rue_bed> Tom_itx, I dont think he beleives me
[12:37:24] <rue_bed> all he has to do is send a 1 into the 595, and give it one pulse to get to the next column, then push the time bits to hte port
[12:38:11] <rue_bed> it would be really quick, almost as quick as a 4017, almost as quick as a 74ls138, and almost as quick as running the columns direct
[12:38:46] <Tom_itx> yeah
[12:39:19] <wisehippy> rue_bed, I follow. I guess I was under the impression I needed to send a whole byte to the 595 on write
[12:39:45] <rue_bed> wisehippy, that sequence and delay is what I think is the root of your problem
[12:40:05] <rue_bed> the delay can be up to about 6ms for you
[12:40:29] <rue_bed> a function called loiter() maybe
[12:40:34] <tzanger> lol
[12:42:05] <rue_bed> I got the MAX7219 code working yesterday, this is most cool, muxes up to 64 leds for me
[12:42:18] <rue_bed> now I just need to work out what to do with it
[12:42:34] <rue_bed> I bought 10 more @ 50c ea, still dont know what to use them for
[12:42:50] <rue_bed> but I have green, blue, and red 7 seg displays now
[12:42:54] <rue_bed> :)
[12:43:34] <rue_bed> with intensity control too, I'll take a video of the pulsating count I have it doing in the shop right now
[12:45:00] <wisehippy> rue_bed - so when writing to the 595 ... that for loop for row nested in the column loop didn't exist. I was just writing the whole byte from the array
[12:46:01] <rue_bed> ok, setting port D closer to when the latch is struck on the 595 should help your immediate code
[12:46:43] <wisehippy> ok, I'll move that around. I dont have an EE background, just a hobbyist/dabbler
[12:46:45] <rue_bed> wisehippy, have I given you enough to work on for a while?
[12:47:00] <wisehippy> rue_bed - I think so. at least a few things to consider.
[12:47:07] <rue_bed> yea, I'm a recreational designer/coder
[12:47:25] <wisehippy> let me ask you about 'smearing'... just do I don't mush up what it means to the 'real' hacks here.
[12:47:30] <rue_bed> that is, I do it for fun
[12:47:55] <wisehippy> so when I'd write the whole row array byte out to SPDR.. I'd have what looked to be current bleeding
[12:48:02] <rue_bed> the smearing is from not turning off the leds when you switch columns
[12:48:15] <rue_bed> yup, thats your smear
[12:48:16] <wisehippy> where if I had row 1, 2, 3 and 5 on in column one... LEDs in column two, same row, would be partially light.
[12:48:18] <wisehippy> ok
[12:48:59] <wisehippy> right on. so smearing is just a timing thing then?
[12:49:06] <wisehippy> or in my case, at least.
[12:49:29] <theBear> it can be quite surprising how much different fine timing and gaps make with leds
[12:49:34] <theBear> difference
[12:49:37] <rue_bed> if you turn the leds off during your column switch, you have to make sure to have a delay while they are on, otherwise you lose a lot of brightness
[12:49:56] <wisehippy> ok. good to know.
[12:50:19] <wisehippy> well I think I'll bail for a while and try moving that port D set around latch setting
[12:50:31] <wisehippy> thanks for your help, gents. appreciated.
[12:50:42] <wisehippy> I'll try to get a quick schem hacked out in case I have any more questions.
[14:03:48] <creep> h
[14:03:48] <Horologium> i
[14:07:42] <creep> what's up?
[14:09:42] <Horologium> my temperature.
[14:09:53] <Horologium> been running a fever all day.
[14:10:01] <creep> ;/
[14:10:10] <creep> build air cleaner
[14:10:10] <turutk> hi. i am a beginner to avers. i am trying to learn programming but i need help. i am about to lose my mind
[14:10:35] <Horologium> turutk, must lose one's mind to learn to program.
[14:10:38] <Horologium> it is in the manual.
[14:10:53] <creep> ;>
[14:11:05] <Horologium> but, in the meantime, what is the problem?
[14:11:12] <turutk> some pins at portc go high no matter what i assign them
[14:11:28] <Horologium> ok, what's the chip?
[14:11:40] <turutk> atmega32a
[14:12:09] <creep> if you have pullup resistors and it is tristated then there you have your reason
[14:12:26] <RikusW> turutk: you probably have jtag enabled
[14:12:50] <Horologium> pins 2,3, and 5, eh?
[14:12:52] <turutk> RikusW: how do i disable it?
[14:12:53] <RikusW> let me guess PC4 stays high
[14:12:54] <Horologium> if so, it's jtag.
[14:13:23] <RikusW> the fuses..
[14:13:25] <RikusW> JTAGEN
[14:13:40] <Horologium> 2,3,4, and 5 actually.
[14:13:47] <RikusW> of you can just build a JTAGICE to connect to it
[14:13:54] <turutk> Horologium: now i checked. those pins are correct
[14:14:10] <OndraSter> Tom_itx, found = ??
[14:14:11] <OndraSter> I found many :)
[14:14:12] <turutk> RikusW: today i learned what a counter and a timer is
[14:14:19] <OndraSter> I am using AP5100 from Diodes Inc.
[14:14:21] <Horologium> those pins are connected to the jtag debugging interface by default..you have to turn off jtag with the JTAGEN bit of the fuse.
[14:15:06] <RikusW> turutk: what software are you using to program the avr ?
[14:15:20] <turutk> RikusW: xcode
[14:15:31] <RikusW> and avrdude ?
[14:16:02] <turutk> RikusW: i believe avrdude comes with crosspack-avr
[14:16:18] <turutk> RikusW: so yes
[14:16:29] <Horologium> must be mac-n-stuff
[14:16:49] <RikusW> which programmer ? some ISP one ?
[14:17:00] <turutk> i should have asked before taking all cables out
[14:17:05] <turutk> usbtinyisp
[14:17:44] <RikusW> you'd have to hfuse bit 6 to 1
[14:17:49] <RikusW> *to set
[14:18:27] <RikusW> just do make sure bit 5 stays 0 or you will have disable ISP :-P
[14:18:51] <turutk> i was gonna learn about fuses sooner or later. now seems to be a good time
[14:19:00] <RikusW> have a look in the datasheet in the memory programming section
[14:19:11] <RikusW> I assume you're familiar with hex ?
[14:20:06] <turutk> RikusW: somehow yes. i have been reading and watching tutorials. i know the concept but i don't have much experience
[14:21:01] <Tom_itx> using osx?
[14:21:15] <Tom_itx> you need the crosspak to make it work or it will run very very slow
[14:21:37] <RikusW> turutk: the default hfuse value is 0x99 set it to 0xD9
[14:22:03] <RikusW> that is 99 | 40 => D9
[14:22:05] <turutk> avrdude -c usbtiny -p m32 -U hfuse:w:0xD9:m
[14:22:14] <turutk> this is it right?
[14:22:24] <Tom_itx> you may need -P usb
[14:22:35] <Tom_itx> may not
[14:23:19] <turutk> turns out i dont
[14:23:44] <turutk> finally it works.
[14:23:48] <turutk> RikusW: thank you
[14:24:21] <Tom_itx> got a link to the crosspack you used?
[14:24:30] <Tom_itx> others have asked about osx here
[14:24:45] <RikusW> :)
[14:25:17] <Tom_itx> i finally figured out you need to use the one that comes with crosspack or it will run really slow
[14:26:55] <RikusW> turutk: just be carefull with fuses, you can brick the AVR.... setting the wrong lfuse clock bits will require that you connect a clock source to XT1 otherwise ISP programming will fail
[14:27:29] <RikusW> and don't set hfuse bit 5 = 1, it will turn off spi programming, then you'll need HVPP to regain access
[14:27:33] <turutk> RikusW: i know that. thats why i have been postponing fuse stuff.
[14:27:57] <turutk> at some point i am gonna need to connect a crystal
[14:28:15] <RikusW> some other AVR's got RSTDSBL it will disable ISP as well
[14:28:22] <Tom_itx> http://www.engbedded.com/fusecalc/
[14:28:24] <Tom_itx> let that be your friend
[14:30:20] <RikusW> turutk: a safe value for lfuse is 0xFF, just remember the 22pF caps too
[14:30:31] <turutk> ok
[14:30:36] <RikusW> 0xFF is for an external crystal
[14:30:47] <turutk> what is the default if i mess up
[14:30:55] <RikusW> default is 0xE1 1MHz internal RC osc
[14:31:25] <turutk> if i mess up i will probably end up bricking :D
[14:31:28] <RikusW> if you mess up you'll need a clock source >=4x higher than the ISP clock
[14:31:50] <RikusW> use another avr with code like PORTA++;
[14:32:00] <RikusW> use A0 to XT1
[14:33:00] <RikusW> turutk: you know that both XT1 and XT2 need a 22pF cap to ground right ?
[14:33:21] <turutk> RikusW: yes
[14:33:42] <turutk> i made some boards with atmegas before
[14:33:55] <turutk> this is the first time i am learning programming
[14:35:01] <turutk> that reminds me another question
[14:35:29] <turutk> i have a board with atmega644. i used some open source code which made me burn a boot loader and than program with arduino ide
[14:36:02] <turutk> then i think my friend shorted something in the circuit. it stopped working
[14:36:31] <RikusW> the m644 might be dead
[14:36:40] <turutk> when i connected to usbtiny. avrdude said wrong device signature (0xffffff)
[14:37:15] <turutk> i googled a lot but no solution has worked for me
[14:37:16] <RikusW> that usually happens with nothing connected too...
[14:37:44] <turutk> when it is not connected or when i take out the crystal it says initialization failed
[14:37:58] <turutk> i connect the atmega32 same way. it works
[14:39:53] <turutk> i bought a replacement mcu but still i wanna be sure the same thing won't happen again
[14:40:29] <RikusW> I'd check the board voltages
[14:40:44] <turutk> RikusW: seems ok. even just after the incident
[14:40:45] <RikusW> and other components on it too
[14:41:21] <turutk> it doesn't make sense to me. but i don't know much
[14:41:56] <turutk> my friend connected a fan control circuit to headers
[14:42:11] <turutk> which has 3 pins. vcc gnd and signal
[14:42:12] <RikusW> 12V ?
[14:42:21] <turutk> he shorted gnu and signal
[14:42:24] <turutk> yes 12v
[14:42:40] <RikusW> AVR's don't like 12V....
[14:42:42] <turutk> this things keeps changing gnd to gnu :D
[14:43:00] <turutk> first thing i asked him was if he shorted 12v to anything
[14:43:15] <turutk> he says he is sure it was signal and ground
[14:43:43] <turutk> pd7 was the signal pin if i remember correctly
[14:44:17] <turutk> that was the second atmega that died. we are not off to a good start
[14:44:39] <RikusW> is the signal by any chance 12V ?
[14:45:11] <turutk> no
[14:45:39] <turutk> the circuit was working before. signal is a pwm pin to set fan speed. he made some changes in the circuit but failed to do so
[14:49:35] <rue_house> try a 1k resistor in series with all io pins :)
[14:50:07] <turutk> that is a good idea :D
[14:50:25] <rue_house> turutk, if you set the external crystal fuses, it will not operate without a crystal or external clock source
[14:50:44] <turutk> tomorrow i am going to put in atmega1284p to the same circuit
[14:50:50] <turutk> avr killer
[14:51:24] <turutk> hopefully it will work and as the first job it will mill a new pcb
[14:56:07] <creep> the fan signal is 12V open collector with possible internal pullup
[14:57:26] <creep> PWM pin is driven by an open collector transistor
[15:04:41] <turutk> if a 22pf capacitor is not connected to the crystal
[15:04:50] <turutk> is there any chance it will burn the atmega
[15:05:05] <creep> no
[15:05:06] <Horologium> more like a chance the oscillator won't start oscillating.
[15:05:19] <creep> it is not a resonant mode power supply
[15:05:37] <creep> it only operates in the 100mW range
[15:06:47] <creep> at least 10000 times lower power level
[15:07:52] <Horologium> it's really fun doing it on a solderless breadboard,,,the breadboard provides enough capacitance to make it work.
[15:07:59] <Casper> turutk: a missing capacitor on the crystal will cause at worse a non-oscillating crystal. It might oscillate, but unreliably or at the wrong frequency. Or it might just work. But no damage can happend
[15:08:40] <turutk> so that doesn't explain the mysterious death of a 644p
[15:08:53] <creep> Horologium<< yeah, you can make capacitors on pcb... i believe they have worse tempco than caps
[15:08:55] <Casper> turutk: have you played with the fuses?
[15:09:09] <turutk> it was on the board for a month or so
[15:09:16] <turutk> working without a problem
[15:09:27] <turutk> someday i saw smoke coming out of it
[15:09:32] <creep> but don't know if it is worse than a ceramic cap
[15:09:56] <Casper> turutk: smoke usually indicate a short of an over voltage condition
[15:09:57] <turutk> the only problem i found was that solder of one capacitor was broken
[15:10:18] <turutk> Casper: that is probably it
[15:11:08] <turutk> the board has both 12v and 5v on different places
[15:11:10] <creep> turutk<< if the oscillator fails, microcontroller will not work
[15:11:30] <turutk> maybe i just put it on a screw or something. and shorted somewhere
[15:12:07] <creep> 230VAC on the osc pin
[15:12:48] <Casper> turutk: if you have a 7805 or alike and the ground break, then it will send the full voltage out
[15:13:41] <creep> yeah nice, i have seen that with cold soldered double sided board and to-220 7805 ;)
[15:13:58] <turutk> Casper: that should be the cause
[15:23:54] <creep> metal film resistors can also be made on pcb ;)
[15:26:24] <Horologium> creep..wasn't talking about PCB...was talking about the inherent capacitance of the metal strips in a solderless breadboard.
[15:27:12] <creep> oh the solderless stuff
[15:28:03] <creep> never used thath
[15:28:45] <turutk> it is both easy to use and hard to use at the same time
[15:29:34] <Casper> ah the joy of solderless breadboard :D
[15:29:34] <Horologium> easy to use...hard to find parasitic capacitance issues.
[15:29:49] <Horologium> generally, on digital electronics, they work great...
[15:30:03] <Horologium> on analog stuff, specially high frequency, then it has issues.
[15:30:11] <Casper> capacitance, inductance, resistance, crosstalk and all
[15:30:33] <Casper> bad connection issues
[15:30:35] <creep> what about contact problems ?
[15:30:37] <turutk> i am out of jumper wires :D it is a hell
[15:31:10] <creep> try breadboarding a switched mode power supply
[15:31:28] <creep> it will burn iin case of weak contact for 1us
[15:33:08] <creep> also breadboarding stuff takes more time than making a test smt pcb
[15:33:09] <creep> ;/
[15:34:03] <Horologium> not for me.
[15:34:10] <Horologium> also, it is reusable.
[15:34:33] <Horologium> and cheaper, generally, for quick prototyping,,,at least for me.
[15:34:42] <creep> i doubt you'd want to reuse a $0.5 pcb
[15:34:54] <Horologium> costs more than that for me to make one.
[15:35:00] <Horologium> mostly in time and effort.
[15:35:11] <creep> you don't have a laser printer?
[15:35:17] <Horologium> I do.
[15:35:20] <Horologium> yes, I know,
[15:35:28] <Horologium> I can make boards.
[15:35:32] <Horologium> it's not that difficult.
[15:35:37] <Horologium> soldering them, however, is, for me.
[15:35:43] <creep> ohh
[15:35:58] <turutk> i go to a copy center rather than using a laser printer
[15:36:06] <creep> yeah well reflowing is fast, solder iron is slow
[15:36:22] <turutk> they print the circuit board to a paper called transfer paper
[15:36:26] <Horologium> no reflow gear.
[15:36:33] <Horologium> yes, I know, easy to make, relatively.
[15:36:41] <turutk> then they put the board under the machine that they use to print photos on tshirts
[15:36:41] <creep> actually, i have done boards by using office paper
[15:36:59] <creep> using 0805 stuff and soic
[15:37:02] <Horologium> I have a large supply of press-n-peel blue.
[15:37:06] <Horologium> and a t-shirt press
[15:37:17] <Horologium> so, can make boards relatively easily.
[15:37:24] <turutk> the result is amazing
[15:37:56] <Horologium> have even gotten 2 sided boards to work rather well.
[15:38:30] <Horologium> print top and bottom layers....register sheets(align) and tape together on 3 sides..
[15:38:36] <Horologium> slide the board in between
[15:38:44] <Horologium> put in t-shirt press and press..
[15:42:03] <creep> ever tried a laminator ?:)
[15:44:30] <Horologium> no
[15:44:34] <Horologium> but, don't have one.
[15:44:50] <Horologium> have considered using a fuser from a copier
[15:46:26] <creep> or laser printer
[15:46:58] <creep> that is some lol stuff
[15:47:16] <creep> a 800W halogen light in an aluminium pipe :)
[15:49:32] <Horologium> you should see the big ones.
[15:49:44] <creep> what do those use?
[15:49:55] <Horologium> 1500 watts of heat inside a pair of aluminum tubes covered in half an inch of rubber
[15:50:01] <creep> 2x 800w halogen lights ?:)
[15:50:04] <Horologium> 110 pages per minute.
[15:50:36] <creep> with a 190C temp control
[15:50:41] <Horologium> I work on everything from cheap brother laser printers through 110 page per minute digital press machines.
[15:50:54] <Horologium> 350F for sharps
[15:50:58] <Horologium> 310F for xerox
[15:51:04] <creep> hah, do you have some HUGE motors from them ?
[15:51:15] <Horologium> 330F to 350F for HP
[15:51:18] <Horologium> I do.
[15:51:31] <creep> i mean BLDC
[15:51:32] <Horologium> big honking steppers basically.
[15:51:39] <Horologium> hybrid type steppers.
[15:51:39] <creep> with 3 halls
[15:53:07] <Horologium> some interesting hybrid stepper type constant speed motors too..
[15:53:28] <Horologium> have a control board connected to it.
[15:53:40] <Horologium> problem is most of those motors are 24V
[15:56:29] <creep> Horologium<< you in us ?
[15:56:36] <Horologium> yes.
[15:57:09] <creep> would be expensive to mail stuff to eu then
[15:57:22] <Horologium> yup.
[15:57:30] <Horologium> find a local copier repair/sales office.
[15:57:37] <creep> this is supercool stuff from printer btw
http://gzkds.en.alibaba.com/product/554343379-0/for_Samsung_Ml_3561n_Laser_Printer_Motor_JC31_00040A.html
[15:57:41] <Horologium> see if they have a bone pile they might let you buy junk from.
[15:57:58] <creep> well i tried in a few, but they don't give a shit
[15:58:09] <creep> they just throw stuff out
[15:58:14] <Horologium> then dumpster dive!
[15:58:20] <creep> even though i'd pay for them
[15:58:26] <creep> haha
[15:59:01] <creep> they throw it out in a hazardous waste facility, where workers share cool stuff...
[15:59:28] <creep> not much thing comes out of there
[15:59:35] <Horologium> we strip parts we can use then have a company haul off what's left as scrap metal.
[15:59:46] <creep> good idea
[16:00:07] <creep> there are stupid people though
[16:00:10] <Horologium> when I'm in the shop around lunchtime I dive into the bone pile and pull small parts and bits and pieces.
[16:03:55] <Horologium> lots of good things get tossed out.
[16:04:32] <Horologium> sensors and gears and belts and stuff.
[16:06:39] <rue_shop2> AAAAA
[16:06:39] <rue_shop2> F B
[16:06:39] <rue_shop2> F B
[16:06:39] <rue_shop2> F B
[16:06:39] <rue_shop2> GGGGG
[16:06:41] <rue_shop2> E C
[16:06:43] <rue_shop2> E C
[16:06:47] <rue_shop2> E C
[16:06:49] <rue_shop2> DDDDD dp
[16:06:51] <rue_shop2> */
[16:06:54] <rue_shop2> it took a second to draw, so that shoud save you all a second
[16:07:19] <tzanger> rue_shop2: nice
[16:07:28] <OndraSter> everything here has to be eco-wasted
[16:07:30] <OndraSter> :(
[16:07:35] <OndraSter> in companies
[16:07:41] <OndraSter> it is hard to dumpster dive here
[16:07:41] <rue_shop2> unsigned char CG[] = {
[16:07:42] <rue_shop2> (1<<A|1<<B|1<<C|1<<D|1<<E|1<<F), /* 0 */
[16:07:42] <rue_shop2> (1<<B|1<<C), /* 1 */
[16:07:42] <rue_shop2> (1<<A|1<<B|1<<D|1<<E|1<<G), /* 2 */
[16:07:42] <rue_shop2> (1<<A|1<<B|1<<C|1<<D|1<<G), /* 3 */
[16:07:42] <rue_shop2> (1<<B|1<<C|1<<F|1<<G), /* 4 */
[16:07:46] <rue_shop2> (1<<A|1<<C|1<<D|1<<F|1<<G), /* 5 */
[16:07:48] <rue_shop2> (1<<A|1<<C|1<<D|1<<E|1<<F|1<<G), /* 6 */
[16:07:50] <rue_shop2> (1<<A|1<<B|1<<C), /* 7 */
[16:07:52] <rue_shop2> (1<<A|1<<B|1<<C|1<<D|1<<E|1<<F|1<<G),/* 8 */
[16:07:55] <rue_shop2> (1<<A|1<<B|1<<C|1<<D|1<<F|1<<G) /* 9 */
[16:08:00] <rue_shop2> };
[16:08:00] <rue_shop2> that has tails on the 6 9 and 7
[16:08:15] <rue_shop2> er no hook on the 7 sorry
[16:08:22] <tzanger> I just pick up stuff I see out on garbage day, or that I find offices I work at throwing out
[16:08:34] <Horologium> you are missing abcdEf
[16:09:13] <rue_shop2> not in that font :)
[16:09:17] <rue_shop2> unsupported codes :)
[16:09:32] <rue_shop2> try accessing them anyhow, see what you get :)
[16:09:50] <rue_shop2> one of the fonts I'v seen uses HELP instead of abcdef
[16:10:41] <rue_shop2> wisehippy,
[16:10:57] <rue_shop2> I'll post this code for you
[16:11:22] <wisehippy> that's cool. as a matter of fact... i *think* I got my flicker issue solved.
[16:11:40] <rue_shop2> http://codepad.org/7ymwAOFj
[16:11:50] <rue_shop2> that is my code for doing digits on a VFD
[16:11:59] <rue_shop2> you can see how I do my scanning
[16:12:51] <wisehippy> rue_shop2: alright, I'll take a look at it.
[16:13:05] <Malinuss> go into this channel... dat spam ;D
[16:13:31] <rue_shop2> Malinuss, say what?
[16:13:46] <Malinuss> hehe ascii art?
[16:13:50] <rue_shop2> hell yea
[16:14:08] <rue_shop2> hahah I guess I should change the title when I make a program from my skel
[16:14:38] <wisehippy> rue_shop2: so you remember my code from earlier... I went back to sending the whole row byte I had defined in the array. I tacked on a few small us delays, but what solved my smearing issue was setting PORTD=0 (where columns are hooked) at the top of loop
[16:14:55] <rue_shop2> yup
[16:15:02] <rue_shop2> see how dI do mine there?
[16:15:10] <wisehippy> ya i do
[16:15:26] <wisehippy> I'm just perplexed 'why' you'd have to turn off all the ports on PORTd
[16:15:49] <wisehippy> doesn't my bit shift only keep the one I want to deal with on at any given iteration through the loop?
[16:16:06] <rue_shop2> otherwise there will be a brief appearance of the digit on the next row before its updated
[16:16:19] <wisehippy> ok, so it's totally a timing thing.
[16:16:34] <rue_shop2> see how I turn off the digit when I switch to the next row?
[16:17:01] <rue_shop2> hahah I totally copy/pasted that code
[16:17:19] <rue_shop2> apparently I have 3 first digits
[16:18:28] <wisehippy> rue_shop2: I do see how you do it. So I guess it's something I just need to accept, always implement and apply to future projects ;-)
[16:20:45] <wisehippy> very cool code. thanks for sharing.
[16:33:49] <tandoori> what use is the avr studio simulator?
[16:34:16] <Horologium> to simulate your program
[16:34:20] <Horologium> without loading it onto a chip.
[16:34:28] <Horologium> good for debugging sometimes.\
[16:36:33] <Tom_itx> not for interrupts
[16:36:36] <Tom_itx> or adc
[16:37:44] <tandoori> hum
[16:37:50] <tandoori> im having problems using it
[16:37:50] <tandoori> heh
[16:38:12] <tandoori> because i run it and it doesnt seem to do anything
[16:39:17] <Horologium> I said sometimes.
[16:40:38] <tandoori> heh
[16:41:52] <Horologium> you put breakpoints in the code.
[16:41:58] <Horologium> and look for things working or not working.
[16:43:01] <creep> Horologium<< what is your price?
[16:43:12] <Horologium> my price for what?
[16:43:26] <creep> assuming you work /h
[16:43:35] <tandoori> this is so stupid
[16:43:42] <Horologium> again, price for what?
[16:43:56] <creep> service, assembly coding?
[16:44:01] <tandoori> on the i/o watch list, it doesn't reflect anything thats being shown in code
[16:44:01] <Horologium> oh,
[16:44:19] <Horologium> I don't have time for outside projects...there are others here who do I bet.
[16:44:44] <jacekowski> hapsim!
[16:44:45] <Horologium> I could make time but I have lots of stuff of my own I'm working on.
[16:45:05] <Horologium> I bet Tom_itx or rue_house would be more likely sources.
[16:52:24] <tandoori> LOL! LMAO! Oh lord
[16:52:26] <tandoori> it works
[16:52:37] <tandoori> you will never guess what i was doing
[16:53:07] <Horologium> jacking off?
[16:53:08] <Horologium> err
[16:53:11] <Horologium> wrong channel.
[16:53:15] <tandoori> I was looking at portd instead of portb in the i/o view and was wondering why it wasn't doing anything (i made no use of portd)
[16:53:26] <tandoori> Horologium: already did that this morning. twice
[16:58:02] <creep> i got a request assembling a few lithium packs for electric bikes, 10Ah 36V, what do you think is a fair payment would be for the job for assembling 20 packs ?
[16:58:40] <creep> $800 ?
[16:58:41] <Horologium> no clue..I don't do lithium batteries...
[16:59:03] <creep> sure, there is no place for errors
[16:59:18] <creep> a short makes heavy wiring glow
[16:59:52] <specing> fuses :D
[17:02:10] <Casper> creep: I want a LiFePO4 pack :D
[17:02:34] <Casper> not only I can't find one, but I can't find them inexpensivelly...
[17:03:22] <creep> Casper<< i'm building that
[17:03:23] <creep> :)
[17:04:06] <creep> 12 cell LiFePo4 10Ah
[17:04:21] <Casper> $$$ isn't it?
[17:04:40] <creep> not that much you think
[17:04:43] <Casper> I was considering a like 12V 40AH
[17:04:55] <creep> hm, that is not cheap stuff
[17:05:03] <Casper> yeah
[17:05:05] <creep> original or cheaper chinese?
[17:05:18] <Casper> something that work well for standby application
[17:06:18] <creep> i remember the price of the cheap chinese 20Ah cell is ~$120, the original is ~$250 and you need 4 of them in series
[17:06:45] <Casper> might be cheaper to just use SLA
[17:06:55] <creep> what is that?
[17:07:05] <specing> Casper: might? It surely will be
[17:07:08] <Casper> sealed lead acid
[17:07:11] <creep> oh...
[17:07:16] <specing> though weight
[17:07:19] <creep> well an original A123 wil llast
[17:07:55] <Casper> the prob is that I couln't find any info for them in standby application
[17:08:00] <creep> initial cost is high
[17:08:31] <creep> well, you need 13.8V CCCV power supply for it
[17:08:56] <creep> but those have memory effect, also they don't like to be fully discharged.
[17:09:40] <creep> and the "solar sealed lead acid" type is expensive too.
[17:09:51] <creep> that was ment to be fully discharged
[17:10:55] <Casper> I'm planning to make a DC ups
[17:11:05] <Casper> when I'll have some spare money
[17:11:40] <creep> nah looked at the prices now
[17:11:56] <creep> the above was only from local reseller
[17:12:09] <creep> divide by 2 for larger quantities
[17:12:39] <specing> Casper: UPS howto:
http://www.1wt.eu/articles/alix-ups/ :)
[17:12:42] <Casper> I'll just use SLA...
[17:12:42] <creep> a 20Ah A123 cell is about $60
[17:13:11] <Casper> specing: that's one of the worse one ever
[17:13:36] <creep> lol
[17:13:38] <creep> :)
[17:14:17] <specing> I just need a 9v6 zener
[17:14:24] <specing> they seem to be impossible to find
[17:15:49] <Casper> specing: they do not exists
[17:15:56] <Casper> 9.1V and 10V
[17:15:57] <creep> http://electroschematics.com/374/pb-and-acid-batteries-charger/
[17:16:05] <creep> far better
[17:16:12] <creep> but reverse flow from battery should be prevented.
[17:16:20] <rue_shop2> yes they do
[17:16:50] <Casper> creep: yeah 317 work nice as a charger
[17:17:03] <creep> until you burn them
[17:17:11] <Casper> and a simple diode from out to in work nice to protect it
[17:17:29] <creep> if you burn it, then lithium will burn too :)
[17:17:48] <creep> in case used for lithium charger
[17:17:57] <Casper> don'T use for lithium :D
[17:18:13] <rue_shop2> closest I can find in my stuff is 9.5
[17:18:30] <rue_shop2> but if your stuck, use a 9V with a forward biased normal diode
[17:18:48] <Casper> or use a tl431
[17:20:21] <student> Hello everyone! I have a question for about CPU-load in uC/OS. It's fairy long, should i post it on pastebin first and paste it here or can i post it directly?
[17:23:56] <student> Here is the question:
http://pastebin.com/gtA07WHx
[17:24:20] <Casper> that's not a long queston
[17:24:26] <student> i have been stuck for more than 2 weeks, can't figure this out...
[17:24:44] <specing> whats the freq?
[17:24:58] <student> hmmm It felt long when i wrote it
[17:25:06] <student> 8Mhz
[17:25:19] <specing> 1200 baud is 120 chars/s
[17:25:21] <Casper> student: general rules of thumb: if it take less than 1-3 lines, paste directly. more than that and you use the pastebin...
[17:25:22] <tandoori> Casper: o/
[17:25:25] <Casper> that's a 1 liner
[17:25:29] <specing> that is hardly any load at all
[17:25:42] * Casper triggers his camera flash in tandoori's face
[17:25:56] <tandoori> i wonder what is the most advanced AVR mcu available
[17:25:59] <Casper> tandoori: the camera flash I have is a big one :D
[17:26:00] <student> ok!
[17:26:01] <tandoori> aww
[17:26:09] * tandoori is wearing cool shades
[17:26:33] <tandoori> love my paparazzi
[17:27:28] <Casper> tandoori: my camera can in theory take 5 pics a second
[17:27:38] <Casper> my sd card allow 3 a second
[17:27:55] <Casper> the flash allow about 3 in a row, then 2 per seconds
[17:28:06] <student_> sorry, my client dissconnected
[17:28:23] <student_> the last thing i saw was something writing 1200char/s
[17:28:27] <student_> someone*
[17:28:45] <student_> if my tick rate is 10ms so it's 1,2char/tick
[17:28:57] <Casper> <Casper> student: general rules of thumb: if it take less than 1-3 lines, paste directly. more than that and you use the pastebin...
[17:29:19] <student_> got that! will remember it to the next time
[17:31:02] <student_> or do you mean i should paste it in here anyway now?
[17:31:26] <Casper> dunnot if people will go read it there or not
[17:31:30] <Casper> might just paste here
[17:31:51] <student_> okok, here it is people:
[17:31:51] <student_> The question is how do i know/calculate the approximate theoretical CPU-load when i have three tasks sending one character each with usart? Some variables i think is useful, baud rate: 1200Baud, Usart buffer (1 Frame, unsure about this, running on atmega644pa), 5tick delay after each character sent. The theoretical value should be approximate 16% and i have measured the practical value to 17% if it's helping.
[17:32:43] <Casper> really, it depend if you do interrupt based or pooling
[17:33:20] <student> i don't know why but the webchat client is dissconnecting me
[17:33:49] <tandoori> thats pretty cool
[17:34:32] <hrr> now this should be more reliable
[17:35:05] <Casper> <Casper> really, it depend if you do interrupt based or pooling
[17:35:46] <Xark> student: Probably best to actually measure it. You can try something simple (e.g.) see how many increments of a counter you get in one second, and see how this varies with no interrupts/traffic to determine overhead.
[17:36:57] <hrr> It's actually a question in my assignment paper asking why the theoretical cpu-load should be 16%
[17:38:20] <hrr> <Casper>if pooling means that all three tasks is running at the same time so it's pooling i'm using
[17:38:44] <hrr> not at the "same" time, but i hope you know what i tried to say
[17:40:02] <hrr> Our professor gave us two leads, how many character USART can buffer and how long time it takes to send a character with 1200baud
[20:29:58] <creep> this is an interesting graph: MTBF versus Metastability Resolution Time
[20:30:04] <creep> http://klabs.org/richcontent/MAPLDCon00/Presentations/Session_A/A5_Erickson_S.PDF
[20:30:12] <creep> http://klabs.org/richcontent/MAPLDCon00/Papers/Session_A/A5_Erickson_P.pdf
[21:17:00] <abetusk> Can you use internal pullups on the attiny's, specifically that attiny85, to drive an i2c bus?
[21:18:14] <Tom_itx> i wouldn't recomend it
[21:18:23] <Tom_itx> i don't think they're strong enough
[21:19:01] <abetusk> Tom_itx, is there any way to quantify that a bit more? And any links on the issue? Would the datasheet say something about it?
[21:24:54] <Tom_itx> P. 214 of the atmega328 data sheet
[21:25:02] <Tom_itx> says use external pullups
[21:26:18] <Tom_itx> sry i was on the wrong data sheet but i bet they're the same
[21:26:45] <Casper> abetusk: i2c ask for 5k pullup.... the internal pullup is closer to 50-100k, so unsuited
[21:27:26] <abetusk> Casper, wouldn't the pullup depend on voltage?
[21:28:37] <timemage> abetusk, have a look at:
http://dsscircuits.com/articles/effects-of-varying-i2c-pull-up-resistors.html
[21:28:43] <Tom_itx> internal pullups aren't near strong enough
[21:29:00] <Casper> abetusk: no
[21:34:05] <abetusk> timemage, thanks