#avr | Logs for 2011-12-31

Back
[04:26:33] <OndraSter> moarnin btw
[04:27:35] <OndraSter> why do people want to buy new electronics (namely new amp + speakers) cheap s*it from china, when they can buy like 10 year old amp (like I did) and use it with old speakers? The speakers I have here are about 20 years old, yet they kick anything below $1k in the arse... and you could buy them nowadays REALLY cheap.
[04:28:08] <OndraSter> but "nope, it is not 2.1, sorry, me not want!"
[04:28:19] <OndraSter> too bad I have two subs in each tower...
[05:34:05] <18VAAFGHC> Can't wait for new years eve, made an electronic ignition for a "melkbus", http://bram.orkz.net/wordpress/?attachment_id=244 :)
[05:36:48] <OndraSter> I don't think my mind will be anywhere electronics at NYE
[05:37:32] <OndraSter> http://d24w6bsrhbeh9d.cloudfront.net/photo/1377998_460s.jpg
[05:38:02] <18VAAFGHC> nice
[05:44:04] <Steffann> 18VAAFGHC carbitschieten :D
[05:44:50] <18VAAFGHC> Steffan: Yeah, with a remote button :)
[05:44:54] <Steffann> They're doing that since ~10.00AM here :D
[06:19:07] <Kev> hey
[06:19:31] <Kev> i have those blue leds, and I power one of them with a 50% duty cycle signal, and the other one always on, and both look the same (brightness)
[06:19:34] <Kev> any idea ?
[06:20:05] <Kev> voltage / res is the same and correct
[06:20:56] <OndraSter> how big current is going through them?
[06:21:38] <Kev> 20ma give or take
[06:22:47] <OndraSter> hmm
[06:22:53] <OndraSter> what if you lower the current?
[06:23:01] <OndraSter> your eye might not see it
[06:24:32] <Tom_itx> pwm too fast?
[06:26:34] <Kev> does that really make a difference ? because the average voltage is half no matter how fast the pwm is
[06:27:40] <Kev> OndraSter I'll have to try this later because I unfortunately can't do that right now:/
[06:28:40] <Kev> I begin to see a small difference at apx 40% duty cycle
[06:28:52] <Tom_itx> change the CS0 value and test it
[06:29:01] <Kev> at 25% it appears half as bright
[06:29:29] <Tom_itx> our eyes don't see the same as reality
[06:30:56] <Kev> I couldnt use the avr builtin pwm because I have 24 leds connected using shift registers
[06:32:41] <Kev> I feed the shift registers with the approriate values to have a 24 channel pwm fast enough so you don't see the leds blink
[06:43:42] <grummund> Kev: light intensity isn't linear with current, if you are lucky the datasheet will have a graph for that.
[06:50:03] <OndraSter> Kev, I am doing similar thing... but with 3072 LEDs :D
[06:50:05] <OndraSter> http://dl.dropbox.com/u/3936936/parts.html
[06:50:07] <OndraSter> 'tis my partlist
[06:51:15] <Kev> I still have this 640 led display "in progress"
[06:52:51] <Kev> what do you use to control the lines ?
[06:53:11] <Kev> mosfet ?
[06:53:37] <Kev> grummund i'm going to look into that, thanks
[06:56:25] <Kev> 74HC154 is a great idea btw
[06:56:25] <OndraSter> ye, MOSFETs for lines
[06:56:31] <OndraSter> 154 for decoding the lines
[06:56:44] <OndraSter> there will be 12x Allegro A6275 as column LED drivers
[06:56:48] <OndraSter> and 13th shift will be 595
[06:56:53] <OndraSter> that will be outputting to the 154s
[06:57:11] <Kev> yeah i didn't think of that
[06:57:19] <OndraSter> I was even smart enough that one 154 will be controlled by 0001xxx and the other one with 0010xxx :P
[06:57:38] <OndraSter> aka one's G pin connected to QF and the other one to QE
[06:57:51] <Kev> i have 1x 595 per 64 leds
[06:58:36] <OndraSter> I still hope that they will shine enough
[06:58:45] <OndraSter> because each row will be lit up 1/32th of a second
[06:58:55] <OndraSter> so it might seem really dim :/
[06:59:00] <Kev> check the datasheet of your arrays
[06:59:09] <OndraSter> this is china :D
[06:59:11] <Kev> with mines I can go up to 100ma
[06:59:13] <OndraSter> wow
[06:59:18] <OndraSter> http://www.ebay.com/itm/25-LED-Dot-Matrix-24pin-8x8-3mm-Red-Green-Common-Anode-/250953666745?pt=LH_DefaultDomain_0&hash=item3a6e0110b9
[06:59:27] <OndraSter> Max Peak Forward Current(IFP): 100mA
[06:59:35] <OndraSter> but how long does peak last
[06:59:50] <Kev> 100 ma at 15% pwm that is
[07:00:13] <OndraSter> I think that I can drive them lower, the A6275 can do 80 I think
[07:00:14] <OndraSter> mA
[07:00:36] <OndraSter> I always close the datasheets :(
[07:00:45] <OndraSter> yap, upto 90mA
[07:01:04] <Kev> http://www.youtube.com/watch?v=2bDUOPklGcw
[07:01:16] <Kev> damn i started this 5 years ago
[07:01:23] <OndraSter> haha
[07:01:36] <Kev> well my proto with 128 leds
[07:01:38] <OndraSter> I think I will go for something like 70mA
[07:02:21] <OndraSter> okay, something 250 ohm-ish
[07:03:20] <OndraSter> 240R it is
[07:04:15] <OndraSter> w00t those A6275 will be burning hot I bet :P
[07:04:24] <OndraSter> 70mA per port...
[07:04:26] <Kev> I like your 154 idea
[07:04:32] <OndraSter> :P
[07:04:34] <OndraSter> want full schematics?
[07:04:46] <Kev> yeah why not
[07:05:01] <OndraSter> http://dl.dropbox.com/u/3936936/vse.sch
[07:05:28] <OndraSter> all ERC errors should be even fixed... just some around the FTDI, but those pins are disabled in software (or at least they are left floating on Arduino too)
[07:05:38] <OndraSter> and warnings are about Vdd being connected to vcc etc
[07:05:51] <OndraSter> or Source pins of those MOSFETs
[07:06:17] <OndraSter> brb 20 mins
[07:09:36] <Kev> thanks
[07:33:01] <mapee> morning
[07:34:51] <Kev> OndraSter btw where did you find cheap A6275 ?
[07:35:21] <Kev> morning mapee
[07:38:43] <OndraSter> Kev, I got them from my teacher
[07:38:44] <OndraSter> for free
[07:38:45] <OndraSter> :P
[07:39:13] <OndraSter> I am sure there are some nowadays versions of it
[07:39:25] <OndraSter> my teacher has some connections... they were closing some shop or factory and had many of them left
[07:39:35] <OndraSter> so they give him some remaninings when he needs :)
[07:39:52] <Kev> i looked on ebay and it's like $17 for 1pc
[07:40:00] <OndraSter> wow
[07:40:02] <OndraSter> lol
[07:40:30] <OndraSter> these are Allegro
[07:40:42] <OndraSter> I'd say look at farnell or any other shop and grab something similar
[07:40:51] <OndraSter> serial in, serial out, parallel out, constant current pins
[09:19:53] <OndraSter> happy new year dudes, I am off to the parteh. Drink more than much! Bb :)
[09:20:53] <Steffann> b
[09:20:55] <Steffann> b
[09:24:48] <amee2k> we still have just under 8 hours to go here
[09:25:10] <Steffann> Yes yes
[09:26:12] <OndraSter> so do we
[09:26:17] <OndraSter> but party starts at 1800 :D
[09:26:29] <amee2k> http://ompldr.org/vYnp4eg/2011-12-31-161712_1024x768_scrot.png
[09:26:34] <amee2k> ah, i see
[09:26:46] <OndraSter> I forgot stepmania lol
[09:26:57] <OndraSter> I managed to get special compiled version for 4 pads
[09:26:59] <OndraSter> and a lotta music :P
[09:27:02] <Steffann> What is that amee2k ?
[09:27:08] <OndraSter> Steffann, binary clock
[09:27:09] <OndraSter> I presume
[09:27:19] <amee2k> Steffann: countdown clock :)
[09:27:24] <amee2k> well, not countdown but clock
[09:27:37] <OndraSter> once all lights shut down..
[09:27:52] <OndraSter> anyway, now real bb. ENjoy lads!
[09:27:53] <amee2k> the rings are (from the outside) days, seconds, minutes and hours
[09:28:18] <amee2k> i wrote it as a java exercise last winter, i can upload it if you want
[09:28:56] <amee2k> i have an older version of it too that doesn't have the days ring and an all-green color scheme
[09:29:05] <Steffann> Java :( :P
[09:30:04] <amee2k> Steffann: it is ass slow (8fps on my laptop; 1GHz P3) but last year we ran it on a mac with a huge screen and it worked on the first try
[09:31:13] <amee2k> iirc the ass slow part is mainly due to retarded graphics API that didn't work with incremental updates so i have to re-draw everything for every frame and then refresh the entire screen
[09:32:33] <amee2k> it does antialiasing too for neat straight edges with no pixelation, but then the framerate drops to like 2fps
[09:33:23] <amee2k> (well, still enough to see seconds but if you watch it closely you can see that it lags
[09:33:26] <amee2k> )
[09:34:18] <amee2k> i've been wondering if i could make a stand-alone version of it with a cheap tablet computer
[09:34:34] <amee2k> obviously rewriting it in C for improved performance
[10:39:48] <amee2k> where can i find comprehensive information about debugwire?
[10:40:17] <amee2k> the MCU datasheet only explains the general concept and required circuitry, but has no details beyond that
[10:41:46] <amee2k> is there an appnote or something where the interface is documented?
[10:43:01] <Steffann> You mean.. the protocol details amee2k ?
[10:43:03] <Steffann> No..
[10:43:17] <Steffann> rikusw did some reverse engineering though.
[10:43:24] <Steffann> It's somewhere on rue's website iirc
[10:44:58] <amee2k> yeah, i found that allready
[10:45:12] <amee2k> someone put the link to rue's site on wikipedia
[10:45:31] <amee2k> but except for some hex codes there isn't much there
[10:45:34] <rue_house> if its the communication proto for programming avrs I havn't the slightest clue
[10:45:41] <rue_house> its an obscure datasheet
[10:46:01] <rue_house> they did? cool
[10:46:02] <amee2k> debugwire
[10:46:13] <rue_house> nope havn't a clue
[10:46:17] <amee2k> not the SPI style programming thing
[10:47:21] <amee2k> i vaguely recall that regular ISP can send individual instructions for execution to the MCU, without reflashing for just one instruction
[10:47:48] <amee2k> i was contemplating sort of an "AVR shell" that allows to execute assembly code as it is entered
[10:48:53] <amee2k> would be much more practical for cases where you just need to quickly set up a timer to get a waveform imo
[10:54:46] <rue_house> I think its toms page on the wiki
[11:17:40] <Steffann> amee2k, dW isn't that difficult at all
[11:18:50] <Steffann> I helped rikusw a little with it.. We just used a uart to get the info
[11:19:48] <rue_house> do you remember when?
[11:19:50] <rue_house> we have a log
[11:20:02] <Tom_itx> who posted avrbeginners.net?
[11:20:06] <amee2k> Steffann: yeah, it transmits UART style frames using a TWI style floating signal line with open collector outputs
[11:20:14] <Steffann> Yes
[11:20:34] <amee2k> i've drawn up a concept for a half-active RS232 adapter that i want to try when i'm back home next year
[11:20:37] <rue_house> it'll say me, but I was just restoring whatever it said
[11:20:42] <rue_house> Tom_itx, it was prolly you
[11:20:48] <Tom_itx> no
[11:21:17] <rue_house> we never change our topic, its had to be there for ages
[11:22:00] <amee2k> i know a chan where the topic says last changed mid-2008 :P
[11:22:26] <rue_house> its not that bad yet
[11:23:49] <Steffann> http://www.ruemohr.org/ <= that was in the topic rue_house
[11:24:10] <rue_house> yes, before I gave up on it
[11:26:18] <Tom_itx> when somebody is in a panic for help the last thing they wanna do is search
[11:26:32] <rue_house> ok one more test app to recode and I'm down to 3 layers left
[11:26:36] <Tom_itx> especially a noob who is overwhelmed
[11:27:22] <amee2k> they don't make noobs like they used to anymore ... quality noobs don't panic :P
[11:27:31] <Tom_itx> yeah
[11:27:39] <Tom_itx> kids these days
[11:29:07] <Steffann> poor Tom_itx
[11:32:01] <rue_house> " ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789,./<>?`~!@#$%^&*()_+|-=\'\""
[11:32:04] <Steffann> Yes
[11:32:06] <rue_house> is that all there is to irc?
[11:32:45] <Tom_itx> no
[11:32:57] <Tom_itx> there's ¿ and µ and a few others
[11:34:25] <Steffann> 
[11:35:47] <Tom_itx> °
[11:36:59] <Tom_itx> ¡
[11:37:04] <Steffann> éáç :D
[11:37:41] <keenerd> amee2k: So I was bored and started porting your clock to Pyglet. Need to make it behave when the window changes size and actually update every second.
[11:38:09] <Tom_itx> ohm would be a nice one to have
[11:39:06] <keenerd> Borrow mine. Ω
[11:39:21] <Tom_itx> doesn't display properly
[11:39:52] <keenerd> Get something that does utf8.
[11:40:38] <keenerd> Maybe better fonts.
[11:41:11] <keenerd> amee2k: How do the seconds display, anyway?
[11:42:14] <Tom_itx> would that be 'ansi codes' or 'multibyte char'?
[11:42:51] <rue_house> unicode is multibyte
[11:43:25] <keenerd> amee2k: Nvm, I can't count to three. What is the outer ring though?
[11:43:38] <Steffann> milliseconds?
[12:05:55] <amee2k> keenerd: the segments light up counter-clockwise-ly. rings from the outside are: days, seconds, minutes, hours
[12:06:49] <amee2k> keenerd: i have code in it to use the "center dot" for fractions of seconds (it smoothly lights up like a pie chart with no segments) but due to the slow graphics that looked kinda shit because it was never really smooth
[12:07:48] <amee2k> keenerd: also, mine only does full-screen at the moment since it was intended as a public display, used like a slideshow or presentation
[12:08:02] <keenerd> Cool.
[12:08:13] <keenerd> Dunno if I can get enough FPS out of this for the center dot.
[12:08:41] <amee2k> it should be easy if you can selectively update the display or use some kind of 2D acceleration, like with opengl or something
[12:09:08] <amee2k> but the java graphics API really sucked for this performance-wise
[12:09:31] <amee2k> its polygon handling ability made drawing the segments really nice and easy though when you look at the code
[12:09:48] <keenerd> Pyglet is opengl. I guess I misunderstood the description you gave, I was having it black out rectangles in a CW direction. So it counts down and ends up fully off at midnight.
[12:10:07] <amee2k> ah, i see
[12:10:25] <amee2k> either way should work i suppose :)
[12:10:32] <keenerd> I'll let you fix that later ;-)
[12:10:44] <amee2k> i found the all lit up, then everything goes out effect pretty cool
[12:11:11] <keenerd> Why CCW?
[12:11:24] <amee2k> if you've seen the tv show, the inspiration for it was the 33 minute count-down clock shown at the beginning of the 2004 battlestar galactica show
[12:11:42] <keenerd> Ah.
[12:12:01] <amee2k> in the TV show segments extinguished instead of lit-up because it was a "real" count down, just like yours
[12:12:41] <amee2k> mine was intended as a general purpose clock display at first, but i kept the "lit segments are from the top CCW, unlit ones are from the top CW"
[12:13:27] <amee2k> i also had the same amber color scheme at first before i changed it first to all green, then to blue-and-white
[12:15:15] <amee2k> dividing it into 6 differently colored super-segments makes it easier to read imo because you can count it faster
[12:16:54] <keenerd> Well right now I am doing a full procedural redraw every frame, so my performance is not too hot :-)
[12:17:08] <amee2k> hehe, i know how that goes :)
[12:17:49] <keenerd> Only 75 lines at the moment, so adding a bit for incremental updates won't kill me.
[12:18:07] <amee2k> wow, thats quite short
[12:18:39] <amee2k> i'll be afk soon until next year. if you show off your count down, please do tell how people liked it :)
[12:18:56] <amee2k> also, happy new year and everything when your time comes \o/
[12:46:53] <MrFrakenstine> hello all
[12:47:20] <MrFrakenstine> I'm in a hurry trying to fishing a LED matrix for the final countdown
[12:47:38] <rue_house> hah
[12:47:45] <rue_house> how far along are you?
[12:47:55] <MrFrakenstine> almost done, everything is working except
[12:47:58] <MrFrakenstine> the serial port!!!!
[12:48:02] <rue_house> you should do these things in jan, when you have lots of time :)
[12:48:08] <rue_house> rs232?
[12:48:23] <MrFrakenstine> I cannot get the USART0 to work
[12:48:29] <MrFrakenstine> I'm getting crazy
[12:48:38] <rue_house> what avr is it?
[12:48:44] <MrFrakenstine> atmel168
[12:48:49] <rue_house> what serial baud are you going for
[12:48:54] <MrFrakenstine> lemmi pastebin some code
[12:49:03] <rue_house> ok, what baud?
[12:49:22] <MrFrakenstine> 115200, prescale should be 7
[12:49:32] <keenerd> What is your clock source?
[12:49:37] <MrFrakenstine> UBRR0H = 0, UBRR0L = 7
[12:49:47] <rue_house> internal or external clock
[12:50:00] <MrFrakenstine> internal
[12:52:12] <MrFrakenstine> 16Mhz
[12:52:12] <rue_house> aha
[12:52:12] <MrFrakenstine> its a arduino board, a diecimilia
[12:52:12] <rue_house> you cant do 115200 with the internal clock
[12:52:12] <rue_house> its only stable enough for about 2400
[12:52:12] <MrFrakenstine> damn
[12:52:12] <rue_house> or so has been the case for me
[12:52:12] <MrFrakenstine> sh*t lemmi go through the datasheet to see how to use the external clock
[12:52:12] <rue_house> you need a crystal or a freq gen module
[12:52:12] <rue_house> can you just dial down the baud rate?
[12:52:12] <rue_house> MrFrakenstine, but finish posting the source
[12:53:30] <rue_house> in my experiance, its prolly too high baud for an internal clock, lets see the source anyhow
[12:54:00] <MrFrakenstine> I even implemented some kind of region invalidate
[12:54:10] <MrFrakenstine> that is, only pixels who changed through time are transmitted
[12:54:18] <MrFrakenstine> s/who/that/
[12:54:21] <MrFrakenstine> rue_bed, but wait, the arduino code seems to be using the internal oscillator for high bauds
[12:54:43] <rue_house> it dosn't have an external crystal?
[12:55:10] <MrFrakenstine> https://github.com/arduino/Arduino/blob/f520bb505134893b36182085fc1bfdc301d5bf89/hardware/arduino/cores/arduino/HardwareSerial.cpp
[12:55:14] <Tom_itx> internal is 8Mhz i believe
[12:55:39] <MrFrakenstine> the diecimilia? yes it does, a 16mhz one
[12:55:50] <Tom_itx> not internal is it?
[12:55:52] <rue_house> I dont what they are up to now
[12:56:38] <MrFrakenstine> http://pastebin.com/DBwJsD3J
[12:56:42] <MrFrakenstine> and this is my code
[12:57:15] <MrFrakenstine> Tom_itx, no the crystal is not internal. does that mean I have to enable some other register to correctly clock the USART?
[12:57:17] <rue_house> can you give me the rest of the part number for your specific chip?
[12:57:22] <rue_house> mega168...
[12:57:50] <rue_house> A-AU>
[12:57:51] <rue_house> ?
[12:59:10] <rue_house> according to this datasheet the internal clock only goes to 8Mhz
[12:59:20] <Tom_itx> ats right
[12:59:31] <MrFrakenstine> rue_bed, 20PU
[12:59:38] <Tom_itx> and ckdiv8 is set by default i believe
[12:59:44] <Tom_itx> 20PU shouldn't matter
[13:00:01] <Tom_itx> so fresh outta the box it should be 1Mhz internal osc
[13:00:17] <rue_house> mmm same datasheet
[13:00:21] <Tom_itx> if you want 8, set ckdiv8
[13:00:22] <Tom_itx> fuse
[13:00:27] <rue_house> MrFrakenstine, try recalculating for 8Mhz
[13:00:37] <Tom_itx> AND check that fuse
[13:00:43] <rue_house> yup
[13:00:59] <rue_house> MrFrakenstine, did we lose ya?
[13:01:03] <MrFrakenstine> no
[13:01:44] <rue_house> so are you gonna try assuming its at 8Mhz instead of 16?
[13:01:49] <MrFrakenstine> yes
[13:01:54] <MrFrakenstine> writting down code
[13:02:02] <rue_house> k, so divisor of 17 or so?
[13:02:10] <rue_house> er 15
[13:02:12] <rue_house> ish
[13:02:15] <rue_house> haha
[13:02:26] <rue_house> !assist http://ruemohr.org
[13:02:26] <tobbor> Possibly http://eds.dyndns.org/~ircjunk/http://ruemohr.org
[13:02:38] <rue_house> hehe
[13:02:44] <rue_house> http://ruemohr.org
[13:02:51] <MrFrakenstine> 15? more like 3
[13:03:15] <rue_house> http://www.wormfood.net/avrbaudcalc.php
[13:03:21] <rue_house> oh right I went the wrong way
[13:03:28] <MrFrakenstine> 8000000 / (16 * 115200) - 1
[13:03:35] <Tom_itx> AMD
[13:03:38] <Tom_itx> AND
[13:03:40] <MrFrakenstine> correct? at least is what the datasheet says
[13:03:40] <Tom_itx> http://www.engbedded.com/fusecalc/
[13:03:50] <Tom_itx> sounds right
[13:04:06] <Tom_itx> check wormfood's chart
[13:04:08] <rue_house> 3, with a 7.8% error margin
[13:04:17] <MrFrakenstine> mm.. how the hell do I check the fuses? I'm using the ftdi cable and arduino bootlaoder to progream it
[13:04:20] <MrFrakenstine> (shame on me)
[13:04:30] <Tom_itx> yep shame on you
[13:04:44] <Tom_itx> read their documentation on fuses
[13:04:47] <rue_house> hmm not avrdude?
[13:04:52] <MrFrakenstine> yes avrdude
[13:05:04] <MrFrakenstine> but the bootloader is still arduino
[13:05:13] <rue_house> avrdude -p m168 ....
[13:05:17] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/using_avrdude_index.php
[13:05:22] <Tom_itx> find the one on reading fuses
[13:05:38] <Tom_itx> make the necessary chip and programmer changes
[13:05:40] <MrFrakenstine> but does that matter? if it works with arduino code, it should work with me
[13:05:44] <rue_house> MrFrakenstine, the highest good baud rate to run at with 8Mhz is 38400
[13:05:57] <rue_house> MrFrakenstine, it'll talk to the fuses
[13:06:00] <MrFrakenstine> I mean, using the arduino IDE, serial works fine
[13:06:12] <rue_house> hmm
[13:06:12] <MrFrakenstine> m.m. i see
[13:06:15] <Tom_itx> then read the fuses
[13:06:28] <MrFrakenstine> what talks with the fuses? the bootloader, correct? which should be still doing it
[13:06:49] <Tom_itx> i dunno what their bootloader does
[13:06:54] <rue_house> they might tweek the clock, if you run at 7.3728Mhz your spot on for 115200 at a divisor of 3
[13:06:57] <Tom_itx> i know avrdude will set and read fuses
[13:07:07] <MrFrakenstine> ok, Im going to stop bitching and going to read the fuses
[13:08:02] <MrFrakenstine> but I'm just going to add that the problem is that in the computer, when I write to the serial device, the write function *blocks*
[13:08:19] <rue_house> linux or windows?
[13:08:22] <keenerd> amee2k: On the off chance you un-afk, here is your clock: http://kmkeen.com/tmp/amee-clock.py.txt
[13:08:24] <MrFrakenstine> linux course
[13:08:38] <MrFrakenstine> that is, write(driver->fd, data, 2) <-- blocks
[13:08:38] <rue_house> hmm how did I unblock writes...
[13:09:01] <rue_house> ssls lib "super simple linux serial"
[13:09:35] <MrFrakenstine> if the write operation blocks, it means data was not shifted out
[13:09:47] <MrFrakenstine> so either the computer or MCU is not working
[13:10:01] <MrFrakenstine> but my computer program was working with the arduino IDE code
[13:10:08] <rue_house> writes just hang?
[13:10:11] <MrFrakenstine> yes
[13:10:14] <rue_house> do you have handshaking off?
[13:10:18] <MrFrakenstine> I do
[13:10:52] <rue_house> 1 min
[13:10:52] <MrFrakenstine> or at least I think I do :P
[13:11:02] <rue_house> !assist programming/c/
[13:11:02] <tobbor> Possibly http://eds.dyndns.org/~ircjunk/programming/c/
[13:11:08] <MrFrakenstine> http://pastebin.com/3Pw4pvWu
[13:11:27] <MrFrakenstine> cfmakeraw() <- I think that turns off handshaking
[13:12:14] <rue_house> MrFrakenstine, hold on, I put a lot of time into making code for doing what I wanted with serial ports
[13:12:42] <MrFrakenstine> me too! that code is definitely working, I wrote it years ago and Ive been using it ever since
[13:12:57] <rue_house> http://eds.dyndns.org/~ircjunk/programming/c/ <-- take a look at ssls2.c and ssls2.h
[13:13:07] <MrFrakenstine> so the problem must be in the MCU
[13:13:14] <MrFrakenstine> ok thank having a look
[13:13:27] <rue_house> if you have handshaking off, no
[13:13:36] <rue_house> with all the handshaking off, the computer just pushes data
[13:14:27] <rue_house> I'll be damned, its there too...
[13:14:38] * rue_house goes off on a tangent
[13:15:04] <MrFrakenstine> there too where?
[13:15:11] <MrFrakenstine> didnt get that one
[13:15:34] <rue_house> I just see some code there I was just beating my head against somewhere else
[13:15:41] <MrFrakenstine> mm... that is true LOL, with handshaking off computer just pushes data
[13:15:45] <MrFrakenstine> dman
[13:15:46] <MrFrakenstine> damn
[13:16:34] <rue_house> can you jump pins ... 6 and 7? or is it 7 and 8...
[13:17:28] <MrFrakenstine> not really
[13:18:05] <rue_house> k
[13:18:09] <MrFrakenstine> mm.. CLOCAL
[13:18:18] <rue_house> I think my code turns off all hanshaking
[13:18:41] <Tom_itx> umm
[13:18:44] <Tom_itx> is it USB?
[13:18:51] <MrFrakenstine> it is
[13:18:55] <MrFrakenstine> FTDI chip
[13:18:58] <Tom_itx> i wonder if the signal is inverted
[13:19:06] <Tom_itx> mmm i think FTDI takes care of that
[13:19:25] <rue_house> I suspect he has all the converters in place
[13:19:33] <MrFrakenstine> thing is my code in the computer was working with my code written in arduino IDE
[13:19:34] <rue_house> MrFrakenstine, want to know how to do a pc loopback test?
[13:20:01] <MrFrakenstine> then I decided to convert the code to pure avr C code
[13:20:14] <MrFrakenstine> and now Im in a rampage trying to fishing this off before new year :(
[13:20:30] <MrFrakenstine> and serial doesnt work
[13:20:33] <Tom_itx> time is running out
[13:20:38] <Tom_itx> you'd better hurry
[13:20:53] <MrFrakenstine> yeah maybe I should just start to get drunk and forget about this
[13:20:53] <MrFrakenstine> :D
[13:21:01] <rue_house> http://www.easysw.com/~mike/serial/serial.html
[13:21:11] <rue_house> by co-inncodince that may be a link our interested in
[13:21:12] <Tom_itx> that _would_ be the easy way out
[13:21:29] <rue_house> MrFrakenstine, can you quickly spin off another pc program?
[13:21:51] <rue_house> connect the rx and tx of your serial converter togethor
[13:21:55] <MrFrakenstine> yeah, I can even attach a router with a serial port to the FTDI pins in the arduino
[13:22:13] <rue_house> forget making a test program just get minicom and see if you can get the characters echo'd back
[13:22:13] <MrFrakenstine> (my serial converter is *in* the arduino9
[13:22:22] <rue_house> MrFrakenstine, wait wait
[13:22:27] <rue_house> can you get minicom?
[13:22:31] <MrFrakenstine> yes
[13:22:39] <rue_house> do that, we will do a quick comm test
[13:22:45] <rue_house> say when
[13:22:49] <MrFrakenstine> now
[13:22:59] <MrFrakenstine> want me to put the arduino beacon in data?
[13:23:09] <MrFrakenstine> the MCU I mean
[13:23:18] <rue_house> ok so , on the converter, jump the RX and TX pins,
[13:23:28] <MrFrakenstine> ah :)
[13:23:37] <rue_house> can you do with at the microcontroller or do you have to do that on the db9?
[13:24:11] <MrFrakenstine> k done
[13:24:28] <rue_house> ok run minicom as root, minicom ttyUSB0 or whatever
[13:24:42] <rue_house> say when ready
[13:24:51] <rue_house> we need to configure it a bit
[13:24:53] <MrFrakenstine> jsut a sec, looks like I dont have the pkg at this computer
[13:25:03] <rue_house> yup, you have to isntall it
[13:26:00] <MrFrakenstine> k, I just remembered I could just connect my router to the FTDI pins and try to get some data into the computer that will tell me if it is working or not
[13:26:04] <MrFrakenstine> but your idea first
[13:26:16] <rue_house> installed??
[13:26:24] <MrFrakenstine> relax, this is gentoo
[13:26:26] <MrFrakenstine> :P
[13:26:30] <rue_house> k
[13:26:52] <MrFrakenstine> looks like minicom takes a while in this athlonxp 2100
[13:27:15] <MrFrakenstine> k done
[13:27:27] <rue_house> ok run minicom as root, minicom ttyUSB0 or whatever
[13:27:37] <rue_house> it'll compalain
[13:27:41] <rue_house> say when its up
[13:28:01] <rue_house> console only program
[13:28:32] <MrFrakenstine> lol you treating me like total dumbass lol
[13:28:38] <MrFrakenstine> K is up
[13:28:52] <rue_house> ctrl-a ctrl-o
[13:28:55] <rue_house> for options
[13:29:01] <rue_house> go to serial port paramiters
[13:29:20] <MrFrakenstine> wait shit
[13:29:22] <rue_house> sory ctrl-A o
[13:29:33] <MrFrakenstine> ctrl+a is interferring with my screen session
[13:29:43] <MrFrakenstine> gotta launch from another term
[13:29:48] * rue_house snickers
[13:29:57] * rue_house waits
[13:30:54] <rue_house> CTRL-A O go to serial port setup, change the serial device, baud rate, and turn all the flow control off
[13:31:56] <MrFrakenstine> done
[13:32:15] <rue_house> hit escape, and down and safe the session
[13:32:24] <MrFrakenstine> k
[13:32:24] <rue_house> "as ttyUBS)" or whatever
[13:32:30] <MrFrakenstine> yes done
[13:32:35] <rue_house> exit with ctrl-A q
[13:32:59] <rue_house> then re-enter the program (it dosn't switch ports or baud with out restart ... *(&&^*)
[13:33:21] <rue_house> check the settings are now right at the bottom
[13:33:42] <rue_house> now, with the serial line looped back, whatever you type should come back to you
[13:33:55] <rue_house> if not, you have a hardware problem to the micro and back
[13:33:57] <MrFrakenstine> meh it didnt reload my config
[13:34:08] <MrFrakenstine> lemmi check if I need to pass the config as param or something
[13:34:11] <rue_house> did you use the same name on the command line?
[13:34:24] <specing> MrFrakenstine: use screen for the terminal
[13:34:31] <specing> minicom is ...crap
[13:34:38] <rue_house> yes, when you run the program do like minicom ttyUSB0
[13:34:49] <specing> screen /dev/tty... 9600,ixoff,cs8
[13:34:55] <MrFrakenstine> ok done
[13:35:03] <rue_house> did the settings stick?
[13:35:13] <MrFrakenstine> specing, lemmi just try rue_house thing first
[13:35:22] <MrFrakenstine> rue_house, yes they did
[13:35:25] <specing> screen works better...
[13:35:32] <rue_house> ok, whatever you type should come back to you
[13:35:42] <specing> since it doesent send those nasty terminal initialisation codes...
[13:35:45] <MrFrakenstine> rue_house but it doesnt
[13:35:57] <rue_house> nothing comes up at all when you type?
[13:36:01] <MrFrakenstine> nope
[13:36:06] <MrFrakenstine> I cant even type
[13:36:11] <rue_house> ok, you have a comm problem
[13:36:22] <rue_house> are you sure you have the right port?
[13:36:22] <specing> MrFrakenstine: have you tried screen?
[13:36:29] <rue_house> specing, NOT NOW!
[13:36:34] <specing> NOW!
[13:36:36] <rue_house> specing, but I want to know more LATER
[13:36:43] <rue_house> specing, he dosn't have time now
[13:36:48] <MrFrakenstine> mmm.. it might be the MCU which is using the pins, let me try to disable the MCU
[13:37:04] <rue_house> MrFrakenstine, can you get to a db9 on it somewhere?
[13:37:04] <MrFrakenstine> how is the AVR reset? putting it on LOW?
[13:37:04] <specing> rue_house: that is because you have filled his head with minicom...
[13:37:35] <rue_house> specing, I know exactly how minicom works, I dont know how much serial port debugging you have done with screen, so not now!
[13:37:54] <rue_house> MrFrakenstine, hmm
[13:38:17] <rue_house> MrFrakenstine, locking the reset low would keep the serial pins as both input meaning you could just short them
[13:38:31] <rue_house> *BUT* if you can get to a db9 connector, lets do that first
[13:39:14] <rue_house> MrFrakenstine, on a db9 you would short pins 2 and 3
[13:40:21] <rue_house> specing, isn't ixoff just software handshaking?
[13:41:10] <specing> nope
[13:41:15] <specing> no handshaking at all
[13:41:34] <rue_house> what would just hardare just software and both be?
[13:41:54] <specing> I don't know
[13:42:19] <specing> and/or care
[13:42:24] <MrFrakenstine> ok so
[13:42:44] <rue_house> MrFrakenstine, a standard pcb jumper works between the db9 pins, or just a peice of wire
[13:44:19] <MrFrakenstine> rue_house, the serial port is a USB device. it is the FTDI chip in the arduino board. I have the MCU reset so effectively disabled which means the arduino 0 & 1 are directly connected to the FTDI chip serial line
[13:44:37] <MrFrakenstine> now in minicom when I type, I can see the RX line in the arduino blinking
[13:44:41] <MrFrakenstine> but there is not reply back
[13:44:53] <rue_house> you have to short the rx and tx lines on the baord
[13:45:33] <MrFrakenstine> they are
[13:45:42] <rue_house> and nothing back?
[13:46:06] <rue_house> the only time I'v seen this is when you have another program running thats reading the serial port
[13:46:14] <MrFrakenstine> ah ok
[13:46:16] <rue_house> lsof |grep ttyS...
[13:46:22] <MrFrakenstine> the jumper was not very well connected
[13:46:26] <MrFrakenstine> yes it works
[13:46:29] <rue_house> ah ok
[13:46:31] <rue_house> ok, good
[13:47:01] <rue_house> how, I suggest you make a program for the avr to continious send the letter U
[13:47:08] <MrFrakenstine> doing it
[13:47:09] <rue_house> and make sure you get it comming in ok on the pc
[13:47:25] <rue_house> U is chosen cause its 01010101
[13:47:46] <rue_house> so, if you dont get back U, you can usually tell when baud its really sending at
[13:48:15] <rue_house> specing, hmm I dont have screen on my machine anymore...
[13:49:02] <MrFrakenstine> so which baud do you recomend
[13:49:04] <MrFrakenstine> say a safe one?
[13:49:41] <rue_house> completely safe is 2400
[13:49:42] <MrFrakenstine> we talked about 38400 previously
[13:49:44] <MrFrakenstine> going to yuse that
[13:49:55] <rue_house> 38400 is a bit daring but should work
[13:51:40] <rue_house> screen is one of those simple programs that became a huge mess of every feature ever come up with
[13:51:51] <rue_house> it can do anything, if you can work out how to ask it to
[13:51:57] <rue_house> like emacs
[13:53:17] <MrFrakenstine> mmm..
[13:53:20] <rue_house> I think you will be a few mins, I'll be back in a bit
[13:53:24] <MrFrakenstine> Im receiving data from arduinp
[13:53:28] <MrFrakenstine> from the MCU I mean
[13:53:31] <rue_house> if you dont get U
[13:53:35] <MrFrakenstine> I dont
[13:53:42] <MrFrakenstine> its something else
[13:53:56] <rue_house> go up and down a few x in the baud rate to find what rate its really sending at
[13:54:29] <rue_house> I'm guessing you will find U at 4800
[13:54:53] <rue_house> when your code is set to 38400
[13:55:37] <MrFrakenstine> nope
[13:55:52] <MrFrakenstine> I only receive it at 38400
[13:56:17] <MrFrakenstine> lemmi send a known ascii symbol
[13:56:45] <rue_house> ? did you send "U" ?
[13:56:52] <MrFrakenstine> y
[13:56:54] <rue_house> always use U
[13:57:01] <rue_house> cause its binary 01010101
[13:57:43] <rue_house> as soon as you get what you expect, your baud settings are right
[13:57:48] <rue_house> then you can go back to your program
[13:57:58] <MrFrakenstine> but I dont
[13:58:10] <MrFrakenstine> so the clock prescaler is definitely wrong
[14:00:47] <MrFrakenstine> Tom_itx, checking the fuses
[14:00:49] <rue_house> on the avr try to send U at 38400
[14:00:57] <rue_house> on minicom, see if you GET U at 4800
[14:01:10] <MrFrakenstine> rue_house, I dont, I get somethign else
[14:01:19] <rue_house> find the baud rate you get U
[14:01:22] <MrFrakenstine> at 38400 that is
[14:01:23] <rue_house> maybe 1200
[14:01:33] <MrFrakenstine> k
[14:01:37] <rue_house> avr at 38400 right, assuming a 8Mhz clock?
[14:02:17] <rue_house> with minicom at either 2400, 4800, or 38400 you should get a U
[14:02:52] <rue_house> remember to save, exit, and restart minicom after each baud rate change :)
[14:03:07] <MrFrakenstine> it applies instantaneously
[14:03:16] <rue_house> wonder when they did that
[14:03:30] <rue_house> ok, any results?
[14:04:04] <MrFrakenstine> ahh crap
[14:04:12] <rue_house> when you wrote the code to send U, what cpu clock freq did you assume?
[14:04:15] <MrFrakenstine> guess I need to close the menus to see that
[14:04:17] <MrFrakenstine> data
[14:04:28] <rue_house> what did you get?
[14:04:39] <MrFrakenstine> I assumed 8Mhz
[14:04:52] <rue_house> ok and did you find the baud that gets you U?
[14:05:49] <MrFrakenstine> went all the way down to 300, no data
[14:05:52] <MrFrakenstine> lemmi go up now
[14:06:42] <rue_house> my guesses are 2400 4800 or 38400 :)
[14:06:53] <MrFrakenstine> so far only 38400
[14:07:03] <rue_house> so you get U at 38400?
[14:07:07] <MrFrakenstine> lets try beyond that
[14:07:11] <MrFrakenstine> no U
[14:07:13] <MrFrakenstine> not U
[14:07:15] <MrFrakenstine> some other symbol
[14:07:23] <rue_house> try 76800?
[14:08:19] <MrFrakenstine> dont have that one
[14:08:42] <MrFrakenstine> 57600
[14:08:44] <MrFrakenstine> nope
[14:08:44] <rue_house> k, try the code at soemthing lower, like 4800 and scan for that rate
[14:08:59] <rue_house> all you need to know is the difference between what your telling it and what its doing
[14:09:20] <rue_house> all I can tell is your not running at 1, 8, or 16Mhz
[14:09:49] <rue_house> if you set the avr for 4800 and try 9600 on minicom
[14:09:50] <MrFrakenstine> yeah I see the point, all the rates are multiples of each other
[14:09:56] <rue_house> yes
[14:11:40] <MrFrakenstine> LOL
[14:11:46] <MrFrakenstine> Im receiving the U
[14:12:02] <rue_house> it that time of year, I have to find *ALL* the 2011 reciepts and stuff them in the box...
[14:12:13] <rue_house> MrFrakenstine, with the avr set to and minicom set to?
[14:12:17] <MrFrakenstine> at the MCU its transmitting with FCLK=16Mhz, baud=4800
[14:12:27] <MrFrakenstine> minicom is 4800
[14:12:29] <MrFrakenstine> FAIL
[14:12:32] <MrFrakenstine> lol
[14:12:37] <MrFrakenstine> it has been correct all the long
[14:12:44] <rue_house> oh, cool, but now we know for sure
[14:13:00] <MrFrakenstine> so why the f*ck the computer was not transmitting?
[14:13:09] <rue_house> cause of a handshaking thing
[14:13:26] <rue_house> one thing that specing said was that ixoff is all handshaking off
[14:13:37] <rue_house> there is an ioctl to switch a serial port to that
[14:14:01] <rue_house> I got too many windows open right now, but if you google for linux ioctl ixoff it shoud come right up
[14:14:43] <specing> SEE?
[14:14:54] <specing> screen 4eva
[14:15:11] <rue_house> specing, if your not right about ixoff being all handshaking off, I'll be upset
[14:16:09] * specing jumps on rue_house's head
[14:16:43] <specing> CANUCK !! :D
[14:16:43] <tobbor> Yankie.
[14:17:27] <rue_house> specing, I wasn't saying its not better, I was simply trying not to introduce anyhting new into the mix
[14:17:58] <MrFrakenstine> there
[14:18:22] <rue_house> got it?
[14:18:26] <MrFrakenstine> tcsetattr(fd, IXOFF, &attrs)
[14:18:30] <MrFrakenstine> but SAME
[14:18:36] <MrFrakenstine> it blocks on write to fd
[14:18:56] <rue_house> you need another flag :/
[14:19:16] <rue_house> http://www.easysw.com/~mike/serial/serial.html
[14:19:21] <rue_house> lunch biab
[14:23:45] <rue_house> we talked too long I burned it
[14:28:25] <MrFrakenstine> rue_house, sometimes it tastes better that way
[14:28:29] <MrFrakenstine> *but not always* :D
[14:31:52] <MrFrakenstine> mmm.. in another computer, write returns -1
[14:33:37] <rue_house> ?
[14:34:08] <MrFrakenstine> FAIL
[14:34:12] <MrFrakenstine> HUGE EPIC FAIL
[14:34:12] <rue_house> can you make a quick program for your avr that will turn a led on if it recieves the character U and off for anyhting else?
[14:34:21] <MrFrakenstine> I cannot believe I lost 2 hours in this
[14:34:24] <rue_house> did you find the error?
[14:34:28] <MrFrakenstine> yes I did
[14:34:32] <MrFrakenstine> and I want to kill myslef
[14:34:44] <MrFrakenstine> FAIL!: Bad file descriptor
[14:34:49] <rue_house> you found it, so I'v give yourself a break on that
[14:34:58] <MrFrakenstine> thats the error when I try to write
[14:35:04] <MrFrakenstine> so I checked the open()
[14:35:07] <MrFrakenstine> guess what?
[14:35:29] <MrFrakenstine> "/dev/ttyS0"
[14:35:39] <MrFrakenstine> my device is ttyUSB0
[14:36:11] <MrFrakenstine> my other computer had a ttyS0 device cause it indeed had a hw serial port
[14:36:31] <rue_house> that dosn't help
[14:36:37] <rue_house> how much time you have left?
[14:36:50] <MrFrakenstine> just waiting for friends to call and go
[14:36:56] <MrFrakenstine> maybe 30-1h
[14:38:25] <rue_house> think you can finish?
[14:41:25] <MrFrakenstine> not sure
[14:41:37] <MrFrakenstine> anyway thanks all
[14:43:09] <rue_house> MrFrakenstine, come back later and how us
[14:43:11] <rue_house> show
[14:43:30] <Tom_itx> so did we get it fixed?
[14:43:32] <Tom_itx> i was gone
[14:44:45] <rue_house> we worked out his converters worked and that his bausd rate was right (must have an external lcok he dosn't know abotu)
[14:45:08] <rue_house> he found out he was opening the wrong serial port on the pc, he had been changing pc's
[14:45:23] <Tom_itx> we could have pulled the arduino schematic but i think that would be considered going above and beyond
[14:49:03] <MrFrakenstine> yeah I'll put the complete project in the internet when its finished
[14:49:09] <MrFrakenstine> but not today
[14:49:12] <MrFrakenstine> :(
[14:49:21] <MrFrakenstine> I gotta go, I failed the deadline
[14:49:54] <MrFrakenstine> opening the wrong serial device was huge fail
[14:50:02] <MrFrakenstine> thank you guys see you around
[14:58:26] <rue_house> well, couldn't have guessed that one
[15:59:38] <dofidum> if i attach a shift register to a atmega's normal output pins what is the maximum data rate i can achieve? is it 1/4th of the clock rate?
[16:00:57] <dofidum> or 1/2?
[16:01:28] <Tom_itx> depends on how much code is in main and how fast you can flip the pin
[16:05:38] <dofidum> Tom_itx, assuming best-case...
[16:06:35] <dofidum> perhaps i should try and connect it to the SPI interface but it seems all schematics on the web refrain from doing so...
[16:06:37] <Casper> best case would be 1/2 if there is no issue with setting the bit AND the clock at the same time
[16:07:32] <Tom_itx> you want an 8bit port out?
[16:07:57] <Tom_itx> use a latched shift register
[16:08:03] <Tom_itx> par/ser
[16:08:21] <dofidum> Tom_itx, that's what i am doing?
[16:08:24] <Tom_itx> ok
[16:08:37] <dofidum> Tom_itx, just figuring out how to connect it to the atmega8
[16:08:39] <Tom_itx> you mentioned spi so i wondered if you shifting it out serially
[16:09:02] <Tom_itx> did you just buy the mega8?
[16:09:17] <dofidum> Tom_itx, not yet... just trying to design a layout
[16:09:21] <Tom_itx> i wish they'd take that danm thing off the market
[16:09:28] <Tom_itx> use a mega 48
[16:09:30] <dofidum> Tom_itx, why is that?
[16:09:30] <Tom_itx> or 88
[16:09:33] <Tom_itx> newer
[16:09:34] <Tom_itx> cheaper
[16:09:36] <Tom_itx> same pinout
[16:09:39] <Tom_itx> more features
[16:09:55] <Tom_itx> 48 88 168 328
[16:09:56] <pc_magas> Happy 2012
[16:10:00] <Tom_itx> pick your poison
[16:10:06] <dofidum> Tom_itx, ah, :)
[16:11:08] <dofidum> component overload
[16:11:11] <jacekowski> dofidum: i used shift registers to sample @10x avr clock
[16:11:32] <jacekowski> dofidum: but i had external PLL to generate clock signal for shift register
[16:12:07] <dofidum> Tom_itx, but perhaps I misunderstood what you meant with latched shift register, I was thinking serial in, parallel out? like 74hc595
[16:12:25] <dofidum> jacekowski, what kind of shift register?
[16:12:26] <rue_house> wow, the linux kernel keyboard code has hooks for text to speech
[16:12:28] <Tom_itx> i'm not sure what you're trying to do
[16:12:34] <jacekowski> dofidum: serial in parallel out
[16:12:56] <dofidum> Tom_itx, I just want to chain a whole bunch of registers so I can drive a whole bunch of LEDs
[16:13:10] <dofidum> using only a few output pins
[16:13:28] <BrentBXR^> you can control a shift register with one pin
[16:13:43] <Tom_itx> i'd use a port on the avr and clock 8 bits out at once
[16:13:48] <BrentBXR^> or an array of shift registers
[16:13:56] <Tom_itx> load the port, then enable the tristate on the par/ser shift reg
[16:14:20] <Tom_itx> so it reads what's on the port and latches it
[16:14:23] <dofidum> Tom_itx, what do you mean "clock 8 bits out at once"... doesn't sound very serial
[16:14:36] <Tom_itx> load em 8 at a time
[16:14:45] <Tom_itx> on like PORTB
[16:14:45] <Tom_itx> or something
[16:14:55] <jacekowski> dofidum: few registers connected in parallel
[16:15:03] <dofidum> Tom_itx, ah ok... but you manually do one bit at a time in the code?
[16:15:08] <jacekowski> dofidum: and you would use 8 pins as inputs for registers
[16:15:18] <jacekowski> dofidum: outputs to registers
[16:15:27] <Tom_itx> just set the port with the data you want clocked out
[16:15:32] <jacekowski> dofidum: and then another few pins to control which register will latch that data
[16:15:37] <Tom_itx> and spit it out parallel
[16:16:04] <dofidum> Tom_itx, so you mean use some shift register with parallel input?
[16:16:09] <Tom_itx> you could chain a couple and use 2 ports if you felt like it
[16:16:36] <Tom_itx> show us a schematic of what you're doing
[16:16:38] <BrentBXR__> whats wrong with spi?
[16:17:05] <BrentBXR__> better yet what shift register u are trying to use
[16:17:29] <dofidum> Tom_itx, all i have is a bunch of 75hc595 (serial in parallel out) that i am trying to hook up to an atmega of some description
[16:17:37] <Tom_itx> abcminiuser how's it lookin in the new year so far?
[16:17:46] <abcminiuser> Tom_itx, bleak
[16:17:49] <BrentBXR__> dofidum; just use SPI
[16:17:50] <Tom_itx> damn
[16:17:57] <BrentBXR__> chain them together; and your good.
[16:18:00] <abcminiuser> It's all fire and brimstone here
[16:18:10] <BrentBXR__> if you dont have that peripheral; then do it in software
[16:18:14] <BrentBXR__> couldnt get much easier
[16:18:26] <dofidum> BrentBXR__, what kind of resistors should go on the lines to prevent problems during ISP programming?
[16:18:38] <Tom_itx> so you're gonna use the parallel out to the leds
[16:18:45] <BrentBXR__> look at page 2 of the 74 595 dsheet
[16:18:47] <Tom_itx> then spi may be your bet
[16:18:53] <dofidum> Tom_itx, that's right...
[16:18:59] <BrentBXR__> depends
[16:19:03] <BrentBXR__> what programming are you using?
[16:19:19] <Tom_itx> or bitbang it
[16:19:25] <dofidum> BrentBXR__, AVR ISP mkii I think
[16:19:39] <BrentBXR__> just a header direct should work then right?
[16:19:39] <Tom_itx> not what he's asking
[16:19:49] <BrentBXR__> Oh i see
[16:20:32] <BrentBXR__> to be honest im not quite sure how that would interfere
[16:20:51] <Tom_itx> problem here is i'm trying to do 5 things at once
[16:20:52] <BrentBXR__> I would just make your own software based solution to use a differnt pin; bit banging in some data to a shift register
[16:20:55] <BrentBXR__> is pretty simple
[16:21:06] <BrentBXR__> like i said before
[16:21:20] <BrentBXR__> you can run a 74HCT595
[16:21:23] <BrentBXR__> with one pin
[16:21:28] <BrentBXR__> a resistor and a cap
[16:21:44] <BrentBXR__> i think there is examples on it already; let me find one
[16:22:43] <dofidum> BrentBXR__, am I right that I can toggle a output pin every cycle? having a bit of trouble working out whether it wouldn't be too slow
[16:23:13] <BrentBXR__> i dont think so
[16:23:28] <BrentBXR__> if i remember correctly; setting a bit is one cycle; clearing a bit is two
[16:24:12] <BrentBXR__> if your just driving LEDs
[16:24:17] <BrentBXR__> ...
[16:24:29] <BrentBXR__> your not exactly making a logic analyzer or anything :3
[16:24:43] <BrentBXR__> even if you ran at 1/2 MIPS youd be good
[16:25:19] <BrentBXR__> even using the internal clock can run leds, even like 10 shift registers im sure it would look fine
[16:25:36] <dofidum> BrentBXR__, well that depends how many leds but yes
[16:26:03] <dofidum> BrentBXR__, i just have this crazy obsession with trying to understand how it all works
[16:26:26] <BrentBXR__> ah
[16:26:41] <BrentBXR__> i mean if your interested in how many cycles somthing takes
[16:26:47] <BrentBXR__> the datasheet says it; datasheet says all
[16:27:05] <dofidum> BrentBXR__, yeah it's a bit daunting though...
[16:27:11] <BrentBXR__> you will never get 1 cycle to run multiple LEDs
[16:27:25] <BrentBXR__> because you will have if statments; and clears and sets, and so on and so on
[16:27:31] <BrentBXR__> each take 1 or more cycles
[16:27:36] <dofidum> BrentBXR__, well no, just trying to work out how quickly in I could refresh all leds
[16:27:43] <BrentBXR__> most in one but when you add it all upl your talking like 8 (rough)
[16:27:44] <Tom_itx> use an fpga
[16:27:52] <Tom_itx> then you can make it damn quick
[16:27:54] <BrentBXR__> lol bit overkill
[16:28:14] <BrentBXR__> dofidum; the best way to learn that; is to try it and find out
[16:28:16] <dofidum> BrentBXR__, it doesn't hurt to work out whether the refresh rate would be roughly in the right ballpark
[16:28:18] <Tom_itx> or ram
[16:28:43] <BrentBXR__> of course not i agree; what you should do is finish the source (in c i assume) and get the ASM output
[16:28:51] <BrentBXR__> then you can actually look at each instruction
[16:28:55] <BrentBXR__> map it out
[16:29:05] <BrentBXR__> 2 cycles, 1 cycles, 5 cycles, ...
[16:29:12] <BrentBXR__> then match that with your clock speed
[16:29:18] <BrentBXR__> and you got yourself an OK guestimate
[16:29:39] <BrentBXR__> factor in everything else i guess; external parts like the shift registers
[16:29:52] <BrentBXR__> i dont know off the top of my head how fast they react
[16:30:31] <Tom_itx> it depends on the series of 74xx parts
[16:30:39] <Tom_itx> and that's all given in data sheets
[16:30:39] <BrentBXR__> 74HC595
[16:30:48] <BrentBXR__> yep
[16:31:15] <BrentBXR__> how many LEDs are you planning on driving?
[16:31:29] <dofidum> BrentBXR__, only 40 or so...
[16:31:55] <BrentBXR__> oh man imo i wouldnt even bother doing all that; i thought we were running 200x500 or somthing :3
[16:32:02] <BrentBXR__> in some crazy
[16:32:07] <BrentBXR__> :D
[16:32:15] <BrentBXR__> how many shift registers?
[16:32:45] <Tom_itx> then there's charlieplexing
[16:32:48] <dofidum> BrentBXR__, nahh... that is a crazy size (even motorway signs are smaller than that)
[16:32:58] <dofidum> i was thinking 5 shift registers
[16:33:14] <BrentBXR__> will you be using an external clock source
[16:33:40] <dofidum> BrentBXR__, not sure yet
[16:34:16] <BrentBXR__> should be fine either way
[16:34:35] <BrentBXR__> disable clock divide
[16:34:40] <BrentBXR__> run that baby full powa :D
[16:35:07] <Tom_itx> use the xtal out to drive a buffer that drives them
[16:35:22] <dofidum> BrentBXR__, yeah just working it out helps me understand... will be driving a 192x128 + 192x92 later this year ;-)
[16:35:22] <Tom_itx> i think you can do that
[16:35:46] <BrentBXR__> nice; i never got interested in the LED stuff
[16:35:54] <Tom_itx> me either
[16:35:56] <BrentBXR__> im more into digital communications and stuff
[16:36:10] <BrentBXR__> where timing is somtimes critical
[16:36:19] <BrentBXR__> so iv played that game quite a bit myself :3
[16:36:26] <dofidum> BrentBXR__, its not really by choice... but the application is pretty cool, there is quite a few of them on UK motorways
[16:36:40] <BrentBXR__> infact i just bought an FPGA board to simulate trace issues (to long, to short) and how it affects the signals
[16:37:09] <BrentBXR__> when your running at likw 210MIPS i think a MM is a huge problem :D
[16:37:19] <dofidum> BrentBXR__, the electronics of the big signs won't be much to do with me but just trying to improve my electronics skills :)
[16:37:32] <BrentBXR__> nothing wrong with that :D
[16:38:10] <BrentBXR__> if you ever need any help let me know; im forsure not the best but i know a few things :3 ill help if i can
[16:38:35] <dofidum> BrentBXR__, thanks :)
[16:38:37] <BrentBXR__> far from it
[16:38:38] <BrentBXR__> :D
[16:38:40] <BrentBXR__> np
[16:58:37] * dofidum is swapping his atmega8 for an atmega88
[17:07:32] <BrentBXR__> swap to an attiny
[17:29:36] <BrentBXR__> anyone here interfaced a serial FRAM chip to an AVR
[17:30:01] <BrentBXR__> more specificly FM25V10
[17:30:10] <BrentBXR__> ramtron 1Mb F-RAM
[18:12:16] <inflex> I2C or SPI chip isn't it?
[18:12:18] <inflex> (just woke up)
[18:14:06] <BrentBXR_> spi
[18:14:34] <tech2077> ah, wonderful spi
[18:14:55] <BrentBXR_> :3
[18:15:00] <inflex> well, it'll be easy then
[18:15:06] <tech2077> then you have a cRIO based platform which is nearly impossible to do spi on
[18:15:14] <tech2077> yet it cost ~$1000
[18:15:14] <inflex> most use the same standard control sequences
[18:15:38] <BrentBXR_> can it be used to extend the AVRs normal RAM?
[18:15:43] <BrentBXR_> or no
[18:15:49] <tech2077> not usually
[18:15:51] <BrentBXR_> or can it only be used for somthing else
[18:15:54] <BrentBXR_> lame
[18:16:02] <tech2077> you can do some things which can do that
[18:16:06] <BrentBXR_> there are megas tat allow that right?
[18:16:09] <tech2077> and there are some avr's that allow that
[18:16:12] <BrentBXR_> but i think parallel only or somthing
[18:16:13] <tech2077> natively
[18:16:40] <tech2077> you can write low level code and basically do that
[18:19:02] <BrentBXR> guys wanna play a game
[18:19:19] <Tom_itx> thermalnuclearwar?
[18:19:35] <tech2077> fun!
[18:26:22] <BrentBXR> I have a little board
[18:26:26] <BrentBXR> from a past project
[18:26:43] <BrentBXR> its usless little thing with a attiny45.
[18:26:47] <BrentBXR> anyone want it
[18:26:52] <tech2077> yep
[18:28:08] <BrentBXR> http://dangerousprototypes.com/forum/download/file.php?id=5690
[18:28:09] <BrentBXR> can u see that
[18:28:37] <tech2077> yeah
[18:29:01] <BrentBXR> http://imgur.com/Sg5uc
[18:29:04] <BrentBXR> thats the schematic
[18:29:34] <BrentBXR> any ideas of what you could do with it
[18:29:36] <BrentBXR> ?
[18:30:19] <BrentBXR> to program if; you need SMD clips or desoldier the chip
[18:30:29] <BrentBXR> iv programed it a couple times so i know its unlocked
[18:30:47] <tech2077> i have smd clips
[18:31:19] <BrentBXR> where do u live
[18:31:31] <tech2077> texas coast
[18:31:36] <BrentBXR> can you etch PCBs
[18:32:40] <tech2077> i have
[18:32:50] <tech2077> but don't have the materials right now
[18:34:49] <BrentBXR> ah
[18:35:26] <BrentBXR> you really want it? I dont really need or want it. it will just end up sitting in my part closet forever
[18:35:53] <BrentBXR> i was thinking about stripping the AVR off, but its not even worth my time
[18:36:04] <BrentBXR> i have tons of AVRs i never even programmed
[18:36:18] <BrentBXR> so stripping an ttiny is just waste of time
[18:36:19] <tech2077> looking at the schem, not really
[18:36:27] <BrentBXR> lol yeah its nothing special
[18:36:41] <BrentBXR> i mean if you dont etch PCBS its nice
[18:36:49] <BrentBXR> use it for somthing or other i dont know
[18:37:00] <BrentBXR> just a simple little breakout board :3
[18:38:36] <tech2077> yeah, but it's contorted a bit :P
[18:45:46] <keenerd> Throw them on ebay for $2 each or something.
[18:49:18] <tech2077> yep
[18:50:36] <Tom_itx> how about a new version of it?
[18:50:38] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/tiny/tinyTPI1.jpg
[19:00:38] <tech2077> thats one of the 6 pin tiny's
[19:00:47] <tech2077> how are they
[19:01:29] <BrentBXR> i only have two :P i didnt make them
[19:01:38] <BrentBXR> 6pin?
[19:01:39] <Tom_itx> 16bit timer
[19:01:40] <Tom_itx> adc
[19:01:43] <BrentBXR> lol
[19:01:43] <Tom_itx> pwm
[19:01:49] <BrentBXR> u mean usable pin?
[19:01:53] <Tom_itx> 4
[19:01:56] <BrentBXR> its just an attiny45
[19:01:57] <Tom_itx> useable
[19:02:15] <BrentBXR> how do you figure?
[19:02:39] <BrentBXR> i count six
[19:02:39] <BrentBXR> :P
[19:02:51] <BrentBXR> 2 = pw/gnd
[19:03:01] <BrentBXR> 6=general io and other
[19:03:06] <Tom_itx> and?
[19:03:13] <BrentBXR> =8 pin attiny
[19:04:23] <tech2077> perfect for a proof of concept i want to do
[19:04:23] <BrentBXR> oops
[19:04:23] <BrentBXR> i think we are on two differnt thoughts :3 my bad
[19:04:23] <BrentBXR> i thought tech was talkin about my thingy didnt notice your pic :3
[19:04:23] <BrentBXR> my bad
[19:04:45] <Tom_itx> what are you getting at?
[19:05:35] <BrentBXR> me? nothing it was a misunderstanding.
[19:05:45] <tech2077> use one pin to measure light and change the brightness of a led
[19:06:16] <BrentBXR> that just a nice little breakout board tom?
[19:06:24] <BrentBXR> or is there plans for that guy
[19:07:21] <tech2077> while using a coil for power
[19:07:41] <BrentBXR> On the little board i showed you; the top right 3 pins are for a IRd reciver originally
[19:07:56] <BrentBXR> i think its on an interrupt pin
[19:07:59] <Tom_itx> the tiny10 was kinda fun to mess with
[19:08:25] <tech2077> how is it programmed
[19:11:09] <tech2077> standard spi way or someway else
[19:13:52] <Tom_itx> TPI
[19:15:02] <Tom_itx> i think inflex is using some in his boards
[19:16:02] <tech2077> i wish i had one of your programmers :P
[19:16:26] <tech2077> may buy one and some attiny10's
[19:16:37] <Tom_itx> i need to make some more up soon
[19:17:06] <tech2077> how much are they again
[19:17:35] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/commerce/commerce_index.php
[19:17:53] <Tom_itx> the blue one is all i have stocked right now
[19:18:19] <tech2077> ah
[19:18:32] <tech2077> i don't see a price for a not boxed one
[19:18:41] <Tom_itx> i'm out right now
[19:18:51] <tech2077> how much when you have them
[19:18:53] <Tom_itx> soon as the chinese new year is past i'll order more
[19:18:57] <tech2077> ah
[19:19:11] <Tom_itx> boards that is
[19:20:06] <Tom_itx> it takes quite a bit more time to make the blue ones
[19:20:23] <Tom_itx> since i cut the enclosures etc myself
[19:21:15] <tech2077> i think i'll wait until you get the one with the enclosures
[19:21:29] <Tom_itx> with or without?
[19:21:35] <tech2077> without*
[19:21:38] <Tom_itx> k
[19:21:56] <tech2077> i'm stocking up one a few with things and hoping to get started with FPGAs
[19:21:58] <inflex> Tom_itx: yueah, I'm moving over to a lot more T10 stuff
[19:22:13] <inflex> Tom_itx: my primary product line is likely to move to it... soon as I can get some PCBs done :(
[19:22:14] <Tom_itx> i made a board rev to them so the new batch will reflect that
[19:22:32] <Tom_itx> yeah i missed their cutoff i think
[19:22:41] <Tom_itx> i wonder if they would sit on it if i sent it in now
[19:24:35] <inflex> dunno... I'm tring to decide if I should give iTead a shot with their $45 option... but by the time you add courier delivery it's not far off the GP 100in.sq option
[19:24:59] <Tom_itx> how's the quality compare?
[19:25:18] <Tom_itx> you have a link to them again?
[19:25:33] <inflex> I'm a little dubious of iTead, sure they're fine for proto's... but I am worried about this strange conformal layer they have on the boards which makes soldering-after-reflow near impossible
[19:25:50] <Tom_itx> oh
[19:25:53] <inflex> http://iteadstudio.com/store/index.php?main_page=index&cPath=19
[19:25:54] <Tom_itx> i forgot about that
[19:27:19] <tech2077> just put in a sample request for to atmel, likely not going to get anything :P
[19:28:17] <tech2077> has anyone had luck with them
[19:28:33] <Tom_itx> inflex, where's the one you were gonna use? all i see is 10 pc
[19:28:42] <Tom_itx> tech2077 i did once
[19:28:49] <Tom_itx> after that not so good
[19:29:06] <tech2077> ah
[19:29:26] <tech2077> i usually by my avrs, so not that big of a deal
[19:29:40] <inflex> Tom_itx: http://iteadstudio.com/store/index.php?main_page=product_info&products_id=509
[19:30:43] <Tom_itx> will they cut shapes?
[19:31:12] <inflex> don't know - but they do have 2 sides
[19:31:20] <inflex> erm, silk-screen that is
[19:31:48] <inflex> Anyhow, I just use them for the proto's (hard to argue with $9.90)... but i'm getting annoyed by this solderability issue
[19:32:20] <keenerd> I never had that problem with iTead boards, but I'm just doing hot air on them.
[19:32:34] <inflex> keenerd: it's not on all of them, and only jus the recent ones I got
[19:33:15] <inflex> out of the reflow oven they attain a dull grey appearance on any non-reflowed HAL exposed surfaces
[19:33:42] <inflex> and it's near impossible to solderthrough, you need 420'C+ tip and a lot of flux
[19:33:51] <keenerd> Weird.
[19:34:10] <inflex> damned right :(
[19:34:32] <keenerd> Have you asked them about that? I know they are cheap, but they probably would not want to make their service less attractive if it could be helped.
[19:34:40] <inflex> Not yet - but will
[19:34:52] <inflex> I doubt many people reflow their PCBs
[19:36:20] <Tom_itx> they charge more for colored mask
[19:36:59] <Tom_itx> inflex, how's their shipping?
[19:41:10] <inflex> $4.40 I think for standard
[19:41:56] <Tom_itx> mmm
[19:41:59] <Tom_itx> 4.9 x 3
[19:42:02] <Tom_itx> so it'll fit ok
[19:46:42] <Tom_itx> i think they are cheaper
[19:47:53] <Tom_itx> inflex, do they panelize em or cut em for you?
[19:48:53] <inflex> you can have them panalized if you ask when you place the order
[19:48:57] <inflex> the protos all come precut
[19:49:09] <Tom_itx> i don't care either way really
[19:49:33] <Tom_itx> it's easy to pick n place if they're on a strip together
[19:49:41] <inflex> I'd prefer the protos in panels... but they'd make it a lot more $$$ to ship
[19:49:50] <Tom_itx> i may give em a try
[19:49:58] <Tom_itx> i wonder what the lead time is
[19:50:17] <inflex> about 3 days from submission till they typically ship
[19:50:20] <inflex> sometimes 4~5
[19:50:23] <Tom_itx> are their gerbers the same as GP?
[19:50:28] <inflex> shipping to me are about 10 days
[19:50:44] <inflex> no, they have their own CAM file to use - and put them all in a .zip
[19:50:51] <inflex> they also have their own DRU
[19:50:54] <Tom_itx> crap
[19:50:56] <Tom_itx> ok
[19:51:06] <tech2077> wow, that looks like a great deal
[19:51:08] * inflex just does zip foo_pcb.zip foo.[GT]*
[19:51:46] <Tom_itx> i'll see if mine pass as is with their dru
[19:51:55] <inflex> Tom_itx: their DRU isn't any worse than GP... it'll bitch about silk's mostly :)
[19:53:08] <keenerd> All of mine were accepted on the first try. They seem to like Kicad gerber files.
[19:53:09] <Tom_itx> have you used their cam?
[19:53:32] <Tom_itx> i like to preview them before i send em
[19:53:41] <Tom_itx> i use gcpreview for that
[19:54:16] <inflex> I use gerbv
[19:54:32] <inflex> ag, that's right you're on Windows though aren't you?
[19:54:42] * inflex wonders if they've ported gerbv to Win
[19:54:51] <keenerd> 2.6.0 of gerbv released a few days ago :-)
[19:54:59] <inflex> http://gerbv.sourceforge.net/
[19:55:16] <inflex> the biggest annoyance I have with gerbv is that it doesn't seem to remember layer colour configurations :(
[19:55:17] <Tom_itx> mine should work ok
[19:55:20] <Tom_itx> i'm gonna try it
[19:55:39] <inflex> ok
[20:00:25] <Tom_itx> inflex, will they reject dru or just ignore them?
[20:03:24] <inflex> ignore afaik
[20:03:32] <inflex> the DRU is really more about protecting -you- :D
[20:03:49] <Tom_itx> do you rename the gerbers?
[20:04:11] <Tom_itx> mine came out with %N.GBL etc
[20:05:57] <inflex> I don't rename them, I just package them up in the zip and send it
[20:06:18] <inflex> (as per their comprehensive but slightly ambigious explanation on their page)
[20:06:26] <Tom_itx> you send all they generate or just the ones in their website list?
[20:26:54] <Tom_itx> inflex do you have an example of their silk i could see?
[20:27:16] <Tom_itx> is it pretty clean compared to GP?
[20:30:55] <Tom_itx> GCpreview worked ok. the drill registry was off a bit but that's pretty common with some of em
[20:37:41] <Tom_itx> inflex, do you know which file has the outline (layer20)?
[20:38:40] <Tom_itx> GML
[20:38:58] <Tom_itx> but their site doesn't request that one