Back
[00:36:34] <Richard_Cavell> Is this a legitimate reset circuit?
http://i.imgur.com/qwoBj.png
[00:40:55] <RikusW> yep\
[00:41:03] <RikusW> why the 100 Ohm resistor ?
[00:41:24] <RikusW> debouncing a little bit ?
[00:41:51] <RikusW> Richard_Cavell: just be aware that the Reset pin in particular wan't to be pulled really low
[00:42:30] <RikusW> check the ds and do a little Ohm's law calculation to see if it will interfere or not
[00:42:40] <Richard_Cavell> So just get rid of the 100?
[00:42:41] <RikusW> probably nt
[00:42:51] <RikusW> if it works leqve it
[00:43:27] <RikusW> the 100nF will interfere with dW
[00:43:35] <Richard_Cavell> I don't care
[00:43:44] <RikusW> yeah, stk600 can't do that...
[00:44:06] <RikusW> just do it in such a way that you can remove the cap later
[00:44:45] <RikusW> Richard_Cavell: a single 10k pullup works perfectly fine too
[00:44:56] <Richard_Cavell> yep
[00:44:58] <RikusW> and a switch
[00:45:01] <Richard_Cavell> But I need to have a way of grounding it
[00:45:04] <Richard_Cavell> that's the switch
[00:45:17] <RikusW> I used only that on my U2S board
[00:45:23] <RikusW> no problems so far
[00:45:38] <RikusW> (by that I mean more than a year, maybe two)
[00:46:14] <RikusW> actually I used 12k 0603, they were cheaper than 10k ;)
[00:46:49] <Richard_Cavell> By the way I've made up my mind about Jewish peopel
[00:46:51] <Richard_Cavell> I hate them now
[00:46:56] <RikusW> oooh
[00:47:32] <Richard_Cavell> I have nothing against their religion, or their race, it's their culture and the Israeli government
[00:47:33] * RikusW is more or leass neutral, though I didn't have too much to do with actual Jews
[00:48:03] <Richard_Cavell> Here's the rest of the circuit if you want to take a look
http://i.imgur.com/B6rSl.png
[00:48:13] <Richard_Cavell> I want to get this right and I've only been working on it for one day
[00:49:18] <RikusW> Richard_Cavell: put a 100nF cap on either side of the 7805 too
[00:49:25] <RikusW> *both
[00:49:28] <Richard_Cavell> ah yes
[00:50:08] <RikusW> are you using ADC ?
[00:51:22] <Richard_Cavell> no
[00:51:29] <RikusW> if you do you'll have to put at least a 10uH coil between Vcc and AVcc
[00:52:12] <RikusW> I'm not sure what the 100nF cap on reset will do to ISP programming
[00:53:18] <Richard_Cavell> I think all the reset pin does is just stay high
[00:53:33] <Richard_Cavell> If it carried data then I'd be worried about the 10k resistor between Vcc and reset interfering
[00:53:40] <RikusW> and for extra protection of the reset pin put a diode between it and Vcc (1n4148 will be fine) rst ->A>C -> vcc
[00:53:58] <RikusW> 10k don't interfere with data, the cap does
[00:54:10] <RikusW> but ISP only pulls reset low to enter programming mode
[00:54:45] <RikusW> I've tested it to work fine without the cap, maybe leave it off for now
[00:54:55] <RikusW> but do put a place for it on the pcb
[00:56:29] <RikusW> Richard_Cavell: why do you connect data received to MOSI ?
[00:56:39] <RikusW> I'd rather connect it to on of those NC pins
[00:57:18] <RikusW> its safer not to have anything else on the ISP lines
[00:57:22] <Richard_Cavell> Hang on where does the diode go? Just from Vcc to Reset?
[00:57:30] <Richard_Cavell> In reverse bias? What for?
[00:57:38] <RikusW> its a protection diode
[00:57:44] <RikusW> all other pins have two
[00:58:00] <RikusW> but the high one on reset was left off to enable 12V for HVPP
[00:58:38] <RikusW> basically gnd ---->|------PIN------->|---------vcc
[00:58:51] <RikusW> its for ESD protection
[00:59:07] <Richard_Cavell> I don't understand it but I'll take your word for it
[00:59:21] <Richard_Cavell> Now as to why I'm using MOSI, it just happens to be the same pin as PB3
[00:59:27] <RikusW> and if you put a resistor in series with the pins like you do, it will protect against higher voltages too
[01:00:19] <RikusW> what difference will it make if you move it to PORTC ?
[01:00:30] <RikusW> or move B0-3 to C0-3 ?
[01:00:39] <Richard_Cavell> well that would work
[01:00:44] <Richard_Cavell> just more programming I suppose
[01:01:35] <RikusW> if you use 500kHz and 2MHz ISP clocks having other wires connected to the ISP lines will interfere....
[01:01:42] <RikusW> particularly at 2MHz
[01:02:04] <RikusW> should be fine at lower speeds
[01:02:18] <Richard_Cavell> should I have a jumper to remove the PB3 functionality?
[01:02:44] <RikusW> that would be a safer way of doing things if there is trouble later
[01:02:46] <Richard_Cavell> or just replicate PB3 on PC3
[01:02:50] <RikusW> yep
[01:03:00] <RikusW> or PC0 ;)
[01:03:04] <RikusW> or move them all
[01:03:45] <RikusW> I'd put like a 4k7 on the base of the BC548
[01:03:58] <RikusW> or at least 1 or 2k
[01:04:07] <RikusW> you don't need that much current on the base
[01:04:42] <RikusW> say hFE is 100 on the BC548 then you only need 1mA on the base since its max Ic is 100mA
[01:04:57] * RikusW guessed at the hFE, it might be higher
[01:05:24] <Richard_Cavell> I think it's 90
[01:05:51] <Richard_Cavell> ok I moved PB3 and changed the resistor
[01:06:22] <RikusW> if its 90, use say a 2k7
[01:06:35] <RikusW> that should give ~~2mA
[01:07:06] <RikusW> *90 = 180, more than the transistor can handle anyways
[01:07:41] <RikusW> for the rest it looks good to me
[01:20:24] <Richard_Cavell> RikusW: I don't think it can be improved much
[01:20:28] <Richard_Cavell> next step is to redraw it properly
[01:20:33] <Richard_Cavell> then make the master circuit
[01:20:40] <Richard_Cavell> which will allow the attachment of up to 4 slaves
[01:28:40] <RikusW> how will you be doing slave selection ?
[01:29:04] <RikusW> the fun issue of bus arbitration ;)
[01:29:40] <RikusW> btw, why did you decide to use parallel and not I2C or SPI ?
[01:31:15] <Richard_Cavell> I'll have separate lines for each slave
[01:31:36] <Richard_Cavell> And I'm not using the parallel protocol, I'm just using the connector because it's convenient
[01:41:30] <RikusW> but still its parallel right ?
[01:41:42] <RikusW> you're own homebrew version ;)
[02:04:12] <Richard_Cavell> yes
[03:48:41] <AnalogSound> i have a stupid questin for you guys...i'm using an attiny25 to controll a dc motor with a mosfet with active brake mechanism. the question is: is it normal that after an interrupt is recived and the ISR is excecuted the PC is back from the 0 address?
[03:48:57] <AnalogSound> is like uC is suffering a hardware reset
[03:49:28] <Tom_itx> 2 things. WDT or noise from the motor
[03:50:15] <AnalogSound> this is happening in simulation (don't know if is important i'm using bascom avr)
[03:50:49] <Tom_itx> somtimes isr's don't simulate well
[03:51:00] <AnalogSound> the noise from the motor was a problem but i solve it with a 100n cap. from V+ to GND and that was all
[03:51:53] <AnalogSound> before taht the uC had some problems bc of the parasitic noise ... regarding the fact that both use the samme supply
[03:52:17] <AnalogSound> so ... Tom_itx you say that simulation ... could be a little bit wrong
[04:03:18] <RikusW> yep it can
[04:03:23] <RikusW> using AS4 5 6 ?
[04:07:09] <AnalogSound> RikusW, AS4 5 6 ?
[04:07:16] <RikusW> AVR Studio
[04:07:36] <AnalogSound> aa no i'm using Bascom AVR
[04:07:44] <AnalogSound> and the integrated simulator
[04:07:54] <RikusW> ah yeah, /me got short memory :-P
[04:08:04] <AnalogSound> :P
[04:10:42] <AnalogSound> is a little strange what is happening
[04:11:20] <RikusW> simulators and compilers have been known to be buggy
[04:11:28] <RikusW> in particular avr-gcc
[04:11:49] <RikusW> so sometimes its the hw or fw or compiler.... :S
[04:12:10] <RikusW> or the OS api's or the PC sw....
[04:12:16] <AnalogSound> :))
[04:12:30] <RikusW> in particular linux serial api's
[04:12:34] <RikusW> on older kernels
[04:13:25] <RikusW> afk
[04:28:50] <Richard_Cavell> Hi, guys. I've redrawn my circuit and looking for other people's opinions
http://www.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&p=1002726#1002726
[04:49:15] <RikusW> Richard_Cavell: where is the link in the forum ?
[04:49:47] <Richard_Cavell> you mean the picture? It's here
http://www.avrfreaks.net/modules/PNphpBB2/files/avr_slave_135.png
[04:51:08] <RikusW> D1 is perfect :)
[04:52:39] <RikusW> R4 might cause LED 3 to be somewhat dim
[04:58:17] <Richard_Cavell> it's connected to 9volts not 5
[04:58:24] <Richard_Cavell> should be about 10 mA
[06:12:26] <AnalogSound> another question: if an interrupt pin (like INT0 on tiny25) is controlled by a switch and let's say that every time the sw is pressed the program jumps to ISR and execute it ... the switch allso will bounce that is correct right? and if the sw will bounce ... then that bounce could be interpreted as a new interrupt call.. is that correct ? so i will have to reset de interrupt vector somehow so that only the first contact will be treated as an interrupt
[06:12:26] <AnalogSound> call and not all the bounces that come after... is my logic correct?
[06:13:22] <Richard_Cavell> By the way if anyone's interested, I successfully programmed an AVR using a dapa cable last night. If anyone's interested in discussing it with me, you're welcome to.
[06:23:09] <AnalogSound> Richard_Cavell, even for high speed uC?
[06:23:29] <Richard_Cavell> well I didn't change any fuses so I didn't test it
[06:23:43] <Richard_Cavell> I mentioned it on the avr chat mailing list and they said there's no problem programming high speed chips with it
[06:23:49] <Richard_Cavell> Since ISP has no minimum frequency
[06:24:26] <Horologium> AnalogSound, good practice is to turn off global interrupts as first command of an ISR.
[06:25:00] <Horologium> Richard_Cavell, that was my first avr programmer. just a bunch of wires connected between avr and parallel port.
[06:25:06] <AnalogSound> Horologium, and then enable again in the main program?
[06:25:19] <Horologium> or as the last thing before you exit the ISR.
[06:25:45] <AnalogSound> Horologium, thank you a lot for this tip
[06:26:04] <Horologium> Richard_Cavell, I've programmed everything from attiny26 at 1MHz to atmega1284p at 20MHz with dapa.
[06:26:20] <Horologium> that is my primary programmer. right now I use one based on a 74ls125 buffer.
[06:26:26] <Richard_Cavell> goood
[06:26:34] <Richard_Cavell> well I demonstrated last night that it works
[06:26:36] <Horologium> and you mentioned last night(my time) that you couldn't get power from parallel port....wrong...
[06:26:39] <Richard_Cavell> my only concern is that you don't get voltage out of it
[06:26:47] <Richard_Cavell> enough to keep the AVR alive?
[06:26:57] <Horologium> you CAN...not always recommended but doable...you just tie several output pins together and turn them all on.
[06:27:16] <Horologium> the other way to get good power out is to tap your ps2 keyboard or mouse port.
[06:27:53] <Richard_Cavell> hacky
[06:27:59] <Horologium> if you use avrdude then you can setup a special programmer that specifies several parallel port output pins as power.
[06:28:41] <Horologium> actually, I have some old parallel port interfaced ethernet adapters that power off the keyboard port...have a keyboard vampire tap...little thing that goes in line with keyboard and feeds power over to the adapter.
[06:28:56] <Horologium> I've used those power taps for other things too.
[06:29:39] <Horologium> to get real hacky, grab the 5V and GND from the old 3.5 inch floppy power cable on your power supply in the PC,,,if you are using a desktop and not a notebook/laptop.
[06:30:16] <OndraSterver> <Horologium> AnalogSound, good practice is to turn off global interrupts as first command of an ISR.
[06:30:18] <Horologium> I have an adapter in this box that is used for external SATA that brings a power connector to the back of the computer and just plug into that.
[06:30:19] <OndraSterver> it is done automatically FYI
[06:30:34] <Horologium> OndraSterver, aahh, didn't know that...solves that issue then.
[06:31:02] <AnalogSound> OndraSterver, :) thank you, i saw the answer :)
[06:31:16] <OndraSterver> I just got home
[06:31:22] <OndraSterver> so I am sometimes reading the log :P
[06:31:51] <Horologium> till I came here a while back I never talked AVR with anybody and learned what I might know from datasheets,,and don't always read every page.
[06:32:21] <AnalogSound> OndraSterver :D
[08:13:48] <OndraSterver> well this is cool
[08:13:48] <OndraSterver> https://skydrive.live.com/?cid=182ef9bc3a8bdda4&id=182EF9BC3A8BDDA4%2125057&sff=1#cid=182EF9BC3A8BDDA4&id=182EF9BC3A8BDDA4%211774
[09:49:16] <OndraSterver> http://clip2net.com/s/2paUO
[09:49:17] <OndraSterver> routing hell :)
[09:49:23] <OndraSterver> still need to put somewhere the PDI header
[09:49:31] <OndraSterver> the two vias are NOT that big of a deal on PDI
[09:51:18] <OndraSterver> one on each line
[09:51:21] <OndraSterver> at least it is balanced
[09:53:52] <Essobi> :D
[09:53:57] <creep> anaif you post schematix then maybe ...
[09:54:32] <creep> oh exited
[09:55:49] <Richard_Cavell> dumb question - can any AVR put any GPIO pin into a tri-state? ie inactive?
[09:55:55] <OndraSterver> yes
[09:56:00] <OndraSterver> input = HiZ
[10:01:09] <Richard_Cavell> well I've finished my circuit designs
[10:01:28] * Richard_Cavell took me 25 hours
[10:08:47] <Richard_Cavell> http://i.imgur.com/cBaGr.png << Here it is
[10:16:20] <OndraSterver> huh
[10:16:23] <OndraSterver> this 25 hours? :)
[10:16:51] <OndraSterver> and why are they split?
[10:16:52] <OndraSterver> the megas
[10:16:55] <OndraSterver> as in, once for power on the top
[10:16:57] <OndraSterver> once for data on the bottom
[10:17:01] <Richard_Cavell> there's only one mega in the circuit
[10:17:07] <Richard_Cavell> I've just drawn it multiple times
[10:17:10] <OndraSterver> oh
[10:17:18] <OndraSterver> connector slave
[10:17:19] <OndraSterver> I see
[10:17:35] <OndraSterver> why miso and mosi but no sck?
[10:17:59] <creep> i'd put the power led pulled low by a resistor to a port, so i can switch it off and not eat battery so fast
[10:18:01] <OndraSterver> or is it PB0 aka D7
[10:18:18] <Richard_Cavell> OndraSterver: The MISO and MOSI in the lower right corner are my own protocol, nothing to do with the ISP protocol
[10:18:31] <OndraSterver> huh
[10:18:36] <OndraSterver> why reinvent the wheel? :)
[10:18:59] <Richard_Cavell> creep: I originally had it before the regulator and people told me to put it after so that you can confirm the regulator's working and not use so much power
[10:19:18] <Richard_Cavell> OndraSterver: I use MOSI to tell one of the chips that it can use the bus
[10:19:23] <OndraSterver> oh
[10:19:35] <creep> the 220 ohm resistors look unneeded between the 2 up
[10:19:50] <OndraSterver> creep, protection if the slave drives the ouputs
[10:19:50] <creep> but they could be used as jumpers though
[10:20:13] <creep> oh, both will drive ?
[10:20:16] <OndraSterver> no idea
[10:20:20] <OndraSterver> but I presume
[10:20:20] <Richard_Cavell> all of those Connector pins are going to another AVR with identical pins
[10:20:30] <creep> this is what tristate latches for
[10:20:30] <Richard_Cavell> both slave and master can drive the data pins
[10:20:42] <creep> or multiplexers
[10:21:49] <Richard_Cavell> is this going to work?
[10:21:52] <Richard_Cavell> Have I got the right idea?
[10:22:03] <OndraSterver> depends on your programming skills
[10:22:09] <Richard_Cavell> 2 slave AVRs share a data bus with a master, and only one can use the bus at any time by requesting it and being given control by the master
[10:22:27] <creep> Richard_Cavell<< tristate one up's outputs, and set the other to output mode
[10:22:37] <Richard_Cavell> yes
[10:23:12] <creep> the one that wants to set output mode should knoiw if the other is is tristate mode
[10:23:36] <Richard_Cavell> well it has to request permission from the master
[10:23:42] <Richard_Cavell> the master gives control of the bus
[10:23:58] <creep> or just check a port if its high or low on the other one
[10:33:05] <OndraSterver> you can try using the superweak internal pullup when set as input
[10:33:07] <OndraSterver> to try driving it high
[10:33:19] <OndraSterver> if it works, the other side is input
[10:46:39] <OndraSterver> I need a third screen
[10:46:41] <OndraSterver> first: .sch
[10:46:43] <OndraSterver> second: .brd
[10:46:45] <OndraSterver> third: .pdf :P
[10:53:24] <tld> fourth: .mpg?
[10:53:47] <OndraSterver> no
[10:53:51] <OndraSterver> music is enough in the background
[10:56:11] <tld> fourth: .xls?
[10:57:00] <OndraSterver> :P
[10:57:01] <OndraSterver> http://cz.mouser.com/ProductDetail/NXP-Semiconductors/IP4234CZ6125/?qs=sGAEpiMZZMtDT69Hefc2iT2MF1403AcE
[10:57:03] <OndraSterver> this is not bad either
[10:57:07] <OndraSterver> it has not just the ESD suppressors
[10:57:10] <OndraSterver> but also series resistors
[10:57:12] <OndraSterver> but what value are they!
[11:02:43] <r00t|home> i own a set of five identical 19'' TFT screens :P
[11:23:21] <OndraSterver> IMPOSSIBLE HAS JUST BECOME REAL!
[11:23:21] <OndraSterver> http://clip2net.com/s/2pcBL
[11:23:30] <OndraSterver> and there is so much still space left now!
[11:23:36] <OndraSterver> I am THE ROUTING KING
[11:23:51] <OndraSterver> I will maybe need to re-route it a bit to get proper grounds to the 100nFs
[11:27:23] <tld> not bad!
[11:27:25] <tld> eagle?
[11:31:35] <OndraSterver> of course
[11:32:07] <megal0maniac> About bloody time :)
[11:32:11] <megal0maniac> How about Ultra?
[11:32:19] <OndraSterver> that one was the first one I finished
[11:32:21] <OndraSterver> in arduino size
[11:32:47] <megal0maniac> I thought you had unresolved issues
[11:32:55] <OndraSterver> no
[11:32:57] <OndraSterver> I solved it all
[11:33:03] <OndraSterver> but I am thinking about changing ESD suppressors
[11:33:06] <tld> OndraSterver: Two thoughts on xboard btw… Atmel often recommends an inductor near the ADC PSU cap. And I'd have loved to see a fuse (that might just be me though, I love fuses… )
[11:33:08] <OndraSterver> for something "integrated" and cheaper
[11:33:17] <OndraSterver> fuse?
[11:33:19] <OndraSterver> there is polyfuse FYI
[11:33:27] <OndraSterver> and inductor as a filter is on new boards
[11:33:28] <tld> oh, crap, sorry, I must've missed it. :(
[11:33:41] <OndraSterver> I have picked one already... I have it in my xls
[11:33:50] <tld> :)
[11:33:53] <OndraSterver> xboard coco breadboard will have it in revB, the others in revA :P
[11:35:06] <megal0maniac> But coco revA :(
[11:35:14] <OndraSterver> yeah
[11:35:18] <OndraSterver> you aren't missing on much really
[11:35:22] <megal0maniac> I know :)
[11:35:23] <OndraSterver> unless you are going to do some heavy switching
[11:35:29] <megal0maniac> I just like to moan sometimes
[11:35:44] <megal0maniac> I'm installing GTA3 on my phone
[11:35:55] <megal0maniac> Becuase work is just THAT much fun
[11:36:04] <OndraSterver> GTA3 is THE fun
[11:36:06] <OndraSterver> for your work :P
[11:36:25] <megal0maniac> 500mb download...
[11:36:29] <megal0maniac> for 320x480
[11:36:35] <OndraSterver> :D
[11:36:41] <OndraSterver> original was 2 CDs
[11:36:49] <OndraSterver> was on*
[11:36:51] <megal0maniac> But it's cool because it's a port, not a remake
[11:37:05] <OndraSterver> aye
[11:37:11] <OndraSterver> I played AoE many years ago on my HTC Kaiser
[11:37:12] <OndraSterver> when I had it
[11:37:22] <OndraSterver> just recompiled by Eidos
[11:37:25] <OndraSterver> or w/e was it back in the day
[11:37:33] <OndraSterver> and ofc fixed to use touchscreen rather mouse :D
[11:38:24] <megal0maniac> I'm going to find that now
[11:38:59] <tld> OndraSterver: can I ask who you used to make the boards, or is that a trade-secret?
[11:39:06] <OndraSterver> my hands
[11:39:11] <OndraSterver> and my hotair rework station
[11:39:38] <tld> for the PCBs themselves?
[11:39:41] <OndraSterver> heh
[11:39:42] <OndraSterver> no
[11:39:48] <tld> (bad phrasing on my part, sorry)
[11:40:02] <OndraSterver> my teacher had some spare space on panel or something, so he offered it for me for price close to zero
[11:40:12] <OndraSterver> I am glad he had only 10 panels or boards or whatever made
[11:40:15] <tld> nice! :)
[11:40:18] <OndraSterver> so I could fix that LDO bug on it :P
[11:40:23] <tld> auch?
[11:40:37] <OndraSterver> I used wrong part for LDO, so on revA it is rotated by 90°
[11:40:43] <OndraSterver> to fit regular 1117 LDO
[11:40:49] <tld> ahh
[11:41:09] <tld> offcourse I've *never* done anything similar… *whistles innocently*
[11:41:09] <OndraSterver> so, I am going to use either this
[11:41:10] <OndraSterver> http://cz.mouser.com/ProductDetail/NXP-Semiconductors/IP4234CZ6125/?qs=sGAEpiMZZMtDT69Hefc2iT2MF1403AcE
[11:41:15] <OndraSterver> or this
[11:41:16] <OndraSterver> http://cz.mouser.com/ProductDetail/NXP-Semiconductors/IP4221CZ6-S115/?qs=sGAEpiMZZMvKM5ialpXrmsjax6qgD1yt
[11:41:17] <OndraSterver> to replace
[11:41:25] <OndraSterver> 5x this
http://cz.mouser.com/ProductDetail/TE-Connectivity/PESD0603-240/?qs=%2fha2pyFadujGuGtIpoz6WFjei900wYi3gpIiSsfe8VA%3d
[11:41:41] <megal0maniac> Too lazy to check, ESD?
[11:41:45] <OndraSterver> (well, two times since I am lazy ass who does not use enough ESD suppressors)
[11:41:46] <OndraSterver> :)
[11:42:14] <OndraSterver> but for the first one I don't know the value of the internal R, they specify only Rs... is it the same? I will have to mail them
[11:42:24] <OndraSterver> for the second one I am not sure any usable layout for USB exists
[11:42:34] <OndraSterver> although I had some rough idea of routing the USB traces AROUND the chip
[11:42:41] <OndraSterver> and putting the chip between D+ and D-
[11:53:35] <tld> any of you ever experienced fuse, ESD-protection or similar saving anything?
[11:53:55] <megal0maniac> Download has -52mb remaining and -3minutes -48seconds :/
[11:53:57] <tld> (I know it can without you noticing, I'm not arguing against them, just curious about stories to tell)
[12:11:14] <OndraSterver> tld, polyfuses - yes
[12:11:20] <OndraSterver> esd - never
[12:11:25] <OndraSterver> clamping diodes - yes
[12:13:38] <tld> makes sense
[12:14:21] <tld> I accidentally forgot to clip a lead once, and it ended touching from power to ground… I suddenly got *very* pleased with my habit of dropping in 1A fuses first thing after power-intake. ;)
[12:18:56] <creep> how you like it? hahaha
http://tinypic.com/view.php?pic=261i5hg&s=6
[12:20:03] <OndraSterver> DAT SYMBOL!
[12:20:06] <creep> pentagram on groundplane?
[12:20:07] <creep> ;>>
[12:20:21] <OndraSterver> black magic circuitry
[12:21:10] <karlp> polyfuses have saved me plenty of times, esd has actually caused problems
[12:21:35] <OndraSterver> hmm?
[12:21:58] <karlp> was replying to tld :)
[12:22:14] <OndraSterver> I ment, what problems
[12:22:16] <karlp> we had esd protection chips that clamped things we weren't expecting to clamp.
[12:22:20] <OndraSterver> oh
[12:22:28] <OndraSterver> fail
[12:23:02] <karlp> well, our fault for doing things the board was't designed for :)
[12:23:21] <karlp> it was designed for inputs, we were pushing poower out instead
[12:23:36] <OndraSterver> eh
[12:25:23] <karlp> what is even more useful is the built in thermal overload on lots of regulator chips :)
[12:25:56] <OndraSterver> eh
[12:28:46] <Amadiro> Somebody here who got a launchpad stellaris and wants to help me out with some debugging? I'm currently writing a gdb driver bridge for it, and the debugging circuit doesn't want to respond to my handshake
[12:29:41] <Amadiro> I'd need somebody to record the traffic that goes back and forth to the ICDI while programming the device and upload me the dump
[12:32:35] <Amadiro> specing, ...?
[12:32:45] <theorifice1> I would, but still haven't received mine ...
[12:33:25] <creep> i sometimes use TVS diodes
[12:33:28] <Landon> Amadiro: doing teescope stuffs?
[12:33:40] <Amadiro> Landon, what's teescope?
[12:33:42] <Landon> er
[12:33:43] <Landon> telescope
[12:33:51] <Landon> oh, I'm thinking of stellarium
[12:34:04] <Landon> thought you were making your own fancy telescope mount
[12:34:06] <Amadiro> nah, it's a microcontroller board by TI that's super-cheap
[12:34:14] <Amadiro> but unfortunately they haven't released working linux drivers
[12:34:30] <creep> ti has much money to burn, they send out free samples of stuff
[12:34:48] <Amadiro> creep, atmel, microchip & co do that too
[12:35:08] <Landon> atmel does?
[12:35:12] <Landon> I thought that was just a myth :P
[12:35:28] <Amadiro> sure, I ordered some xmega samples from them a week ago and got them in the post two days later.
[12:35:37] <Amadiro> faster than actually buying them
[12:35:41] <creep> haha
[12:35:55] <creep> and did you sell your soul to them for a few chips ?
[12:36:20] <Amadiro> creep, nah, they ain't even sending me spam.
[12:36:31] <Amadiro> TI does, but oh well, I've received around 300$ worth of samples from them by now
[12:37:03] <creep> maybe they wait until your million $ inventory till the end of the year, if they don't receive it, they'll track you down
[12:37:08] <Amadiro> linear technology, microchip & co also automatically sign you up for their newsletter, but it's not horribly annoying
[12:46:04] <creep> :)
http://tinypic.com/view.php?pic=w0i16o&s=6
[12:46:28] <creep> the capacitor near the ic is 0805, resistors are 0603
[12:47:16] <megal0maniac> Heh. Phone camera with a 10x eyepiece?
[12:47:26] <creep> no, it is a microscope
[12:47:44] <megal0maniac> Whoops :)
[12:47:51] <megal0maniac> Well, that's how I do mine
[12:48:30] <creep> but accidentally i have noticed phone cameracan work with magnifier too hehe
[12:48:46] <creep> making supermacros
[12:49:53] <tld> karlp: I've ruined voltage dividers by dumping clamping zener in the middle of it, forgetting to account for reverse leakage current (I think that must have been it)
[12:51:38] <creep> do voltage dividers need clamping? ;/
[12:54:11] <creep> megal0maniac<< so, that black box and brown capacitor block is 2mm/80 mil long
[13:35:35] <tld> creep: not normally
[13:35:48] <tld> creep: What can I say? I was young, careful and overprotective.
[13:35:54] <tld> (it was also a couple of days ago)
[14:27:01] <megal0maniac> creep: I know, I've gone as far as 0805 :)
[14:27:05] <megal0maniac> And SOT23
[14:27:28] <megal0maniac> TIME TO GO HOME! Been here 15mins too long and I'm not getting paid for it :/
[14:27:31] <megal0maniac> 'Night all
[14:28:32] <Amadiro> Ok, I can communicate with the device -- now I'll just have to learn the gdb protocol and decipher the proprietary commands used (only two) and I'll be done
[14:37:51] <OndraSterver> http://clip2net.com/s/2pgov
[14:37:55] <OndraSterver> abcminiuser,
http://clip2net.com/s/2pgov
[14:37:58] <OndraSterver> atxmega going arduino :)
[14:38:20] <OndraSterver> third one actually :D
[14:38:38] <OndraSterver> 256a3u, 128a1u and 32a4u
[14:40:32] <abcminiuser> OndraSterver, neat :)
[14:40:48] <abcminiuser> I've been wondering if the Arduino guys will cook their own version up
[14:41:02] <abcminiuser> Since it would offer so much more than the current platform, without costing millions
[14:41:16] <OndraSterver> yep
[14:41:18] <OndraSterver> well
[14:41:20] <OndraSterver> I will beat them
[14:41:23] <OndraSterver> but I need software!
[14:41:42] <abcminiuser> ASF?
[14:41:46] <OndraSterver> yeah
[14:41:53] <OndraSterver> but that is not arduino :)
[14:41:57] <OndraSterver> no pinMode
[14:41:59] <OndraSterver> no analogRead
[14:42:01] <OndraSterver> no digitalWrite
[14:42:10] <OndraSterver> no LCD.Start(16, 2)
[14:42:20] <abcminiuser> Honestly all that stuff would be dead easy to write for the XMEGAs
[14:42:23] <OndraSterver> aye
[14:42:26] <OndraSterver> that's my plan :P
[14:42:27] <OndraSterver> http://clip2net.com/s/2pgtK
[14:42:29] <OndraSterver> this is the 32a4u
[14:42:29] <abcminiuser> Given all the peripherals are actually modular
[14:43:01] <OndraSterver> aand 128a1u
[14:43:01] <OndraSterver> http://clip2net.com/s/2pguf
[14:43:35] <OndraSterver> when will be a1u series available at distributors? Was it nov15 or oct15?
[14:44:51] <ferdna> I can't program this thing:
[14:44:51] <ferdna> https://www.sparkfun.com/products/11055?
[14:45:00] <OndraSterver> I never expected to see anything like arduino leonardo
[14:45:05] <OndraSterver> I expected them to move to xmega instead
[14:45:13] <ferdna> how should i connect my ISCP programmer to it? do i need to power my board or what?
[14:45:58] <ferdna> or should i just purchase the FTDI cable?
[14:46:08] <abcminiuser> ferdna, I'm guessing it has a bootloader but you could just hook up to the ISP lines
[14:46:17] <abcminiuser> OndraSterver, aren't they already available?
[14:46:42] <OndraSterver> arduino xmegas?
[14:46:58] <OndraSterver> I know only akafuino x
[14:47:04] <OndraSterver> or something around that, it is some japanese name
[14:47:08] <OndraSterver> it has got one major fault
[14:47:14] <OndraSterver> he used series resistors on each pin to make it 5V compatible
[14:47:16] <OndraSterver> with zeners
[14:47:26] <OndraSterver> try driving anything with >few uA with it
[14:47:33] <OndraSterver> I don't know how big resistors he used though
[14:47:35] <OndraSterver> but not small ones
[14:47:41] <OndraSterver> and it is expensive
[14:47:47] <OndraSterver> and has only xmega32a4
[14:48:42] <OndraSterver> afk, sister needs help with simple math
[14:49:38] <ferdna> https://dlnmh9ip6v2uc.cloudfront.net/images/products/1/1/0/5/5/11055-03a.jpg
[14:49:55] <ferdna> so if i power GND and 5V the ones on the top left.
[14:50:15] <ferdna> and then attach my SCK, MISO, MOSI and RST to my programmer it has to work right?
[14:50:42] <ferdna> or do i need to power the two lonely pins on the top right? (where it says VIN 6-12V)
[14:51:29] <Paul92> hi. I'm trying to use a stk500 to program an attiny26. Whatever I try, it is not recognized. what I could do, in order to find an fix the problem?
[14:54:01] <abcminiuser> ferdna, it looks like you power the top right pins, and the 5V/GND pins on the left are for your external stuff
[14:54:05] <abcminiuser> But check the manual
[14:54:20] <abcminiuser> Paul92, how is it hooked up?
[14:54:31] <abcminiuser> Try the lowest ISP speed
[14:55:13] <Paul92> it's hooked up by serial cable.
[14:58:04] <Paul92> I've tried on linux with avrdude and on windows with avr studio. Same problem. So, I think it's a hardware problem
[14:58:21] <abcminiuser> I mean, the STK500 to the TINY
[14:58:28] <abcminiuser> What connections, what socket, etc
[15:01:03] <Paul92> oh, well, now it is in SCKT3700A (blue one). I tried putting it in SCKT30003 also (red one). I use ISP programming, so I connect the ISP6PIN header to the corresponding header of the socket
[15:11:24] <Paul92> as I am a begginer, should I build and use a simple programmer (something like few resistors conected to paralel port)?
[15:11:54] <OndraSterver> no
[15:11:55] <prpplague> Paul92: if you are getting started it is best to start with know good items, buy a programmer, they are not expensive
[15:11:57] <OndraSterver> just get something working
[15:12:15] <OndraSterver> the cheapest option is from ebay, $4 for programmer that can do any older mega
[15:12:27] <OndraSterver> after that you can visit Tom_itx 's website for some AVRISP mkII clone
[15:12:28] <specing> cheapes: DASA
[15:12:32] <OndraSterver> that can program virtually anything
[15:12:33] <OndraSterver> ..
[15:12:43] <OndraSterver> unless you have to buy serial card
[15:12:55] <Paul92> as I said earlier, I have a STK500, but I have some trouble with it
[15:16:38] <eric_j> Paul92: the stk500 is a good way to start
[15:17:15] <vectory> if it doesnt have to be avr, check launchpads from ti
[15:17:36] <vectory> very cheap, too
[15:17:56] * vectory reading comprehension [ ]
[15:20:50] <Paul92> I would like to use avr, because I documented about it and I wouldn't like to start from scratch. But, what can I do to check if things like my serial port or my stk500 is working (by the way, I use a 9v supply for stk. Is that right?(
[15:22:36] <eric_j> Paul92: yes, it should work. check your LEDs on the stk500, power should be red, and two green ones on
[15:23:07] <Paul92> yes, it works.
[15:23:24] <Paul92> is there any way to check my serial port and cable?
[15:23:37] <OndraSterver> let's try mission impossibru: SINGLE SIDE LOADED XBOARD ULTRA!
[15:23:47] <eric_j> best way is to try to program an AVR, i think
[15:23:48] <Tom_itx> u da man
[15:24:20] <eric_j> Paul92: do you have an AVR which is known to have its fuses set to internal RC oscillator mode?
[15:25:47] <Paul92> I have attiny26, so I think yes.
[15:26:20] <Paul92> And programming an avr isn't an option, the software isn't recognising my stk
[15:29:12] <creep> Tom_itx<< [190549] <creep> how you like it? hahaha
http://tinypic.com/view.php?pic=261i5hg&s=6
[15:30:17] <Amadiro> creep, pretty nice
[15:30:28] <Amadiro> creep, are you planning to galvanize/paint it for corrosion protection?
[15:31:51] <creep> what's bad in lacquer ?
[15:32:13] <creep> well ok, it could have paint too
[15:32:23] <creep> galvanize? no
[15:33:12] <creep> there are no edge connectors
[15:33:42] <Amadiro> creep, ah, is there already lack on it?
[15:34:19] <creep> no, and that would make photo blurry
[15:34:42] <Tom_itx> did you etch or mill it?
[15:34:45] <creep> lacquer reflects light
[15:34:49] <creep> its etched
[15:34:57] <Tom_itx> what is it?
[15:35:03] <Amadiro> creep, well, the photo is already extremely blurry...
[15:35:16] <Tom_itx> BLDC controller?
[15:35:35] <creep> Amadiro<< unfortunately, i had only a mobile
[15:35:47] <creep> Tom_itx<< yeah, 3 phase motor controller
[15:35:56] <Tom_itx> for what?
[15:36:40] <creep> meant to be used in an aeroplane, but the kind customer canceled it after it was almost done
[15:37:03] <creep> filters to the output, testing
[15:37:26] <Tom_itx> you know inflex made those?
[15:37:41] <creep> inflex?
[15:37:59] <creep> no i dont
[15:38:10] <Tom_itx> for model airplanes
[15:38:19] <creep> this was meant to be in real aeroplane
[15:38:40] <Tom_itx> like Boeing 737?
[15:38:43] <Amadiro> lolwat
[15:38:57] <creep> no, private plane
[15:39:25] <creep> and not main motor, only some gyro spinner stuff
[15:39:40] <creep> that requires 400hz AC
[15:39:53] <Tom_itx> k
[15:40:01] <creep> so, it requires pwm sine
[15:41:54] <creep> currently creating pcb for my hid device with buttons and analog inputs
[15:42:35] <creep> still haven't settled with layout ;/
[15:43:21] <creep> i will use jumpers, but solder wires on board, and if it breaks off somehow anywhere, quickfix would be possible
[15:43:59] <creep> you can always do wire-wrapping without a soldering iron;>
[15:45:02] <creep> this will be useful practice in electric car/bike
[15:45:42] <OndraSterver> http://clip2net.com/s/2phod
[15:45:43] <OndraSterver> so far so goo
[15:45:43] <OndraSterver> d
[15:45:57] <OndraSterver> about halfway done
[15:46:09] <OndraSterver> and I want to murder somebody for 6mil being the smallest I can use :)
[15:46:29] <creep> you have a few wires there
[15:46:39] <OndraSterver> ye
[15:46:44] <OndraSterver> moving the caps from the bottom to the top
[15:46:53] <creep> yeah 6/6 mil is doable
[15:47:01] <OndraSterver> not sure how I will do it for the cap next to the PDI header
[15:47:05] <OndraSterver> black magic!
[15:47:13] <creep> use groundplane
[15:47:15] <OndraSterver> (not black as in black guy, that is racist)
[15:47:17] <OndraSterver> ..?
[15:47:20] <OndraSterver> how will that help?
[15:47:32] <creep> lower the number of crossings?
[15:47:52] <creep> reduce EMI
[15:48:05] <OndraSterver> http://clip2net.com/s/2phpS
[15:48:08] <OndraSterver> that is my problem :P
[15:48:10] <OndraSterver> move that
[15:48:16] <OndraSterver> I know what ground plane is
[15:48:20] <OndraSterver> I am not a n00b
[15:48:30] <creep> what is that?
[15:48:38] <creep> bottom side?
[15:48:41] <OndraSterver> top
[15:48:44] <OndraSterver> the cap is right now on the bottom
[15:48:46] <OndraSterver> which sucks
[15:48:55] <OndraSterver> single loaded boards = 10000 times better
[15:49:59] <creep> only upto a few khz
[15:50:06] <creep> with this long wires
[15:50:08] <OndraSterver> I said LOADED
[15:50:10] <OndraSterver> not sided
[15:50:14] <OndraSterver> aka parts are on one side only
[15:50:44] <creep> o i see
[15:50:59] <creep> yes, if it fits why not
[15:56:52] <specing> OndraSterver: do you need the cap?
[15:58:20] <creep> (i have design i routed for about 2 weeks, >8hrs/day)
[15:58:53] <OndraSterver> specing, decoupling
[15:58:55] <OndraSterver> 100n
[15:59:18] <creep> decoupling is nice at the other side of the board near bga pads
[15:59:33] <OndraSterver> because you know, xmega can do 200mA sink/source per 8 pins
[15:59:42] <OndraSterver> how's that stellaris about sinking/sourcing current
[15:59:53] <OndraSterver> I think that under some circumstanses some (x)megas can go upto 40mA even
[16:00:14] <creep> 100nf can be larger in case you use x7r
[16:00:32] <creep> 1u- 10u ...
[16:04:05] <OndraSterver> I have got here two reels of 100nF
[16:04:07] <OndraSterver> 0603
[16:04:24] <creep> neat
[16:04:37] <OndraSterver> I have accidentaly bought two :D
[16:04:39] <OndraSterver> each has 4k
[16:04:47] <creep> and for serious decoupling i have 10u x7r
[16:05:02] <creep> and some 4u7 x7r
[16:05:27] <creep> haha that wil last a while
[16:05:54] <creep> but they are cheap on reel...
[16:06:43] <specing> OndraSterver: does it work without it?
[16:06:50] <OndraSterver> it surely does
[16:06:53] <OndraSterver> but..
[16:06:56] <OndraSterver> every pin should be decoupled
[16:06:58] <OndraSterver> properly
[16:06:58] <specing> :D
[16:07:03] <OndraSterver> even 100nF on its own is not enough
[16:07:23] <OndraSterver> creep, 10uF/4u7uF is not considered as proper decoupling
[16:07:25] <creep> that decouples radio signals from vccs...
[16:07:26] <OndraSterver> too big
[16:07:41] <creep> OndraSterver<< x7r is ceramic ...
[16:07:45] <OndraSterver> so?
[16:07:54] <creep> read datasheets
[16:08:06] <creep> so, 10u is not worsethan 100n
[16:09:08] <creep> and 10u ceramic is better than 220u aluminium foil...
[16:09:43] <OndraSterver> creep,
http://www.atmel.com/Images/doc8278.pdf
[16:09:48] <OndraSterver> 10uF and 100nF
[16:09:49] <OndraSterver> suggested
[16:09:58] <OndraSterver> somewhere they go 100n, 10n, 100p or 1n
[16:10:28] <creep> sure for radio frequeny, it may matter, 100pf vs 1nf vs 10nf
[16:11:09] <creep> i used 7 1nf x7r chips on my fm transmitter and it was crystal clear
[16:11:30] <creep> and a 100nf for power supply ( 2x AA )
[16:20:07] <creep> electrolytic is crap because the aluminium foil is not stable, and it is wound up, so it ieven has inductance
[16:20:17] <creep> *is
[16:20:25] <creep> -*is
[16:20:32] <creep> s/ieven/even
[16:21:58] <creep> any pro board makers here?
[16:25:09] <OndraSterver> <-
[16:25:31] <creep> sometimes if i'm in the mood i just use my edding 140s to draw smt board by hand
[16:25:40] <OndraSterver> huh
[16:25:46] <creep> and cut 0.4mm thick pcb with a scissor
[16:26:25] <creep> then toner transfer, etching
[16:26:56] <creep> and photolithography, positive, with UV, then developing and etching
[16:27:32] <creep> led boards are candidate to milling, but i have not completed my cnc mill yet ;< can't wait to start using it
[16:28:25] <creep> it would be pro if the machine calibrates the z axis (board relative position)
[16:29:15] <creep> i have some diamond milling bits 1-3mm diam
[16:29:50] <OndraSterver> specing,
http://clip2net.com/s/2phOf
[16:29:53] <OndraSterver> I'VE DONE IT
[16:30:43] <specing> OndraSterver: make it so it can be plugged into a protoboard
[16:31:04] <OndraSterver> ...
[16:31:19] <OndraSterver> arduino ones have got priority
[16:31:36] <creep> OndraSterver<< you do positive photolithography and etching too?
[16:31:43] <OndraSterver> yes
[16:31:52] <creep> and solder mask?
[16:32:06] <OndraSterver> yes
[16:32:36] <creep> soldermask with photolithography too? then paint?
[16:32:51] <OndraSterver> solder mask with UV paint
[16:32:58] <creep> hmm
[16:33:06] <OndraSterver> stick it on, then negative mask
[16:33:09] <OndraSterver> and UV it again
[16:33:10] <creep> and is never fixed after?
[16:33:13] <OndraSterver> and wash the remaining
[16:33:27] <OndraSterver> I have tried it once so far only :P
[16:33:38] <OndraSterver> but I used too much ink
[16:33:50] <creep> but it will be damaged by uv
[16:33:52] <creep> ;/
[16:34:01] <creep> otherwise sounds good
[16:34:07] <OndraSterver> damaged?
[16:34:10] <OndraSterver> why should it be damaged?
[16:34:21] <OndraSterver> it already *has* been under UV
[16:34:25] <OndraSterver> had*
[16:34:37] <creep> yoyou meant UV curing paint?
[16:34:48] <OndraSterver> of course
[16:34:49] <OndraSterver> what else
[16:34:54] <creep> oh thats cool
[16:35:04] <OndraSterver> it is ridiculously cheap
[16:35:11] <creep> if you can get some...
[16:35:15] <OndraSterver> 10ml for $3.5 or $2.5, exists for green only though
[16:35:17] <creep> currently i do not know of any
[16:35:25] <OndraSterver> otherwise 100ml of green/blue/red/white for $9
[16:35:28] <OndraSterver> I know of many :)
[16:35:38] <creep> but you are not in wurope
[16:35:41] <creep> europe
[16:35:42] <OndraSterver> I am
[16:35:55] <creep> country?
[16:36:08] <OndraSterver> you can check my DNS
[16:36:41] <OndraSterver> http://www.ebay.com/itm/UV-Curable-Solder-Mask-PCB-Repairing-Paint-Green-New-/170846431660?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item27c73da9ac
[16:36:42] <OndraSterver> cheapest
[16:36:46] <OndraSterver> other colours are in 100ml packs only
[16:37:14] <OndraSterver> I have bought now this
[16:37:15] <OndraSterver> http://www.ebay.com/itm/PCB-UV-Curable-Solder-Mask-Repairing-Paint-Red-100g-/180917181592?pt=LH_DefaultDomain_0&hash=item2a1f811c98
[16:37:22] <OndraSterver> and donated the green one to my previous school
[16:37:27] <OndraSterver> since they do PCBs there manually
[16:38:11] <creep> nice
[16:38:29] <creep> in hungary its trade secret how they bastardize pcbs
[16:39:00] <creep> they use some powder, or ink smudging with stencil...
[16:39:18] <creep> to be the cheapest
[16:39:55] <creep> your dns does not tell much about your location btw
[16:40:14] <OndraSterver> it does
[16:40:36] <creep> traceroute gets ****-ed too
[16:40:54] <OndraSterver> gw.southprague.net
[16:40:57] <OndraSterver> South Prague
[16:40:58] <OndraSterver> ..?
[16:41:30] <creep> ah, i started net tools on it
[16:42:26] <creep> double sided photo board? ;)
[16:42:57] <OndraSterver> at home?
[16:43:00] <creep> sure
[16:43:00] <ferdna> there was someone in here who was into RC stuff...
[16:43:10] <ferdna> just dont remember who he was...
[16:43:14] <OndraSterver> I have got PCB for it, but not sure if I will ever do it - without through hole and via plating...
[16:43:16] <ferdna> was it tomitx?
[16:43:30] <creep> OndraSterver<< do you plate vias at home?
[16:43:36] <OndraSterver> I don't plate anything
[16:43:43] <OndraSterver> that's why I am not sure I will bother with dual sided stuff at all
[16:43:51] <creep> but i believe i saw many vias on your board ;/
[16:43:58] <OndraSterver> this PCB will not be home made LOL
[16:44:03] <creep> haha ok
[16:44:04] <OndraSterver> I only do it to test stuff which I am not sure about
[16:44:10] <OndraSterver> (SMPS and such)
[16:45:53] <creep> i only dislike bga currently
[16:46:09] <creep> i'm fine with any dfn, tqfp package
[16:46:33] <OndraSterver> everybody dislikes BGA
[16:46:42] <OndraSterver> it is hard to check if it has been properly soldered
[16:46:46] <OndraSterver> you have to have xray for that
[16:47:06] <creep> also it will become damaged after some time
[16:47:17] <creep> especially if its flexed
[16:47:39] <creep> for example heavy heatsink on vga card
[16:48:04] <creep> and due to heating... different materials
[16:48:29] <creep> best is military pjcc
[16:48:36] <creep> the ceramic pack
[16:48:52] <OndraSterver> best is regular wood
[16:48:59] <OndraSterver> wood machines mmmm
[16:49:58] <creep> wooden gold plated silicon, germanium chips with butter and ketchup ?
[16:50:40] <OndraSterver> no
[16:50:45] <OndraSterver> pieces of wood
[16:51:04] <creep> bga can be breadboarded if you want
[16:51:19] <OndraSterver> I don't really want any BGA near my PCBs
[16:51:37] <creep> just solder 268 wires to the balls, and the wires to your board
[16:51:46] <OndraSterver> hmm let me cound the size of my breadboard
[16:55:24] <Amadiro> creep, if you're doing that on a breadboard, you're probably realistically only going to wire up like 20 wires anyway for testing purposes, so it shouldn't be as bad...
[16:58:21] <creep> hahaha check out this:
http://www.agilitymfg.com/images/assets/nonwarrantyrepair.jpg
[17:01:56] <OndraSterver> nice
[17:04:02] <Amadiro> wow
[17:04:03] <creep> if you are not concerned about crosstalk ;/
[17:04:15] <creep> this won't work for a ga chip
[17:04:18] <creep> *vga
[17:08:29] <specing> Oh btw, Im studying CS+maths now and the CS part is totaly pathetic.
[17:09:43] <creep> OndraSterver<< anyway good thing my things usually don't need service, so i can just spraypaint the whole board if i want to
[17:26:38] <OndraSterver> spray with tan?
[17:34:02] <Tom_itx> evening
[17:35:23] <OndraSterver> evenin
[17:37:15] <Tom_itx> http://www2.electronicproducts.com/Carbon_nanotube_pencil_allows_designers_to_draw_sensors_on_paper-article-fajb_carbon_nanotube_pencil_oct2012-html.aspx
[17:40:17] <OndraSterver> mm
[17:40:18] <OndraSterver> DAT SENSOR!
[17:40:20] <OndraSterver> lol
[18:40:23] <w|zzy_> jadew: any more updates to OLS?
[21:46:37] <skorket> When accessing TCNT1H and TCNT1L, how are you guaranteed an atomic read? The examples show reading TCNT1L first then TCNT1H next, and mention something about TCNT1H having a temporary variable it writes into. The chip knows when you do a read of TCNT1L followed by a read of TCNT1H to ensure atomic reads?
[21:46:55] <skorket> I'm looking at the datasheet for the atmega328 and family
[22:04:38] <creep> skorket<< H is high part, that will be 256 times slower than the low part
[22:05:40] <skorket> creep, but if you read the low part and there's an increment in between and the low part happened to be 255, then reading the high part will be off
[22:05:49] <skorket> that's the point about it being atomic
[22:06:21] <skorket> the datasheet says it's atomic, I'm just wondering what the mechanism is by which it achieves that
[22:06:55] <creep> it may be possible for the architecture to prefetch a low then a high read request and execute in parallel for you
[22:07:21] <creep> lik on x86 many instructions can be paired
[22:12:55] <JViz> !seen abcminiuser
[22:12:56] <tobbor> abcminiuser was last seen in #avr on Oct 17 13:24 2012
[22:15:20] <JViz> why does the HID keyboard class use key press status instead of key up and key down events?
[22:16:50] <JViz> doesn't that require continuous report generation instead of just events for up and down?
[22:17:49] <JViz> man, i have really no idea where to ask these questions, but it's in the LUFA library, so...
[22:18:12] <Tom_itx> maybe you should wait for abcminiuser
[22:18:45] <JViz> i seem to be really bad at catching him
[22:19:05] <Richard_Cavell> Tom_itx: Mate it looks like OS X requires a hand-recompilation of avrdude. I'm not going to bother with it because I'm quite happy programming from Windows and Linux.
[22:19:13] <Richard_Cavell> Otherwise very happy with your programmer
[22:20:18] <Tom_itx> sry i can't help much with osx since i don't own a mac
[22:20:40] <Richard_Cavell> Sure, well just for future reference you need to hand-compile avrdude
[22:20:44] <Richard_Cavell> with some modifications
[22:21:18] <Tom_itx> so write up a how to for it
[22:21:23] <skorket> JViz, why don't you try it and see?
[22:23:00] <JViz> try it?
[22:23:24] <JViz> you mean wait?
[22:24:15] <skorket> press a key and see if it generates multiple events
[22:24:43] <JViz> it's the way the class is set up
[22:24:57] <JViz> it's basically polling
[22:25:30] <JViz> if it's not going to generate multiple reports, there should be a way to skip the polling if i don't need it
[22:32:43] <JViz> well
[22:32:53] <JViz> i think i answered my question
[22:33:14] <Xark> skorket: I believe reads are atomic because it latches the high byte into a buffer when the low is read, this is why the Atmel manual says -> "For a 16-bit read, the low byte must be read before the high byte." It does something similar for writes ("To do a 16-bit write, the high byte must be written before the low byte.").
[22:34:44] <skorket> Xark, so there's basically some internal circuitry that copies over the high byte to the temporary buffer when the low byte is read?
[22:35:32] <skorket> and writes the high byte in the buffer to the high byte when the low byte is written to as well?
[22:35:42] <skorket> I think this was essentially what creep was trying to say...
[22:35:45] <Xark> skorket: Yes, from 16.3 in 328 manual "Each 16-bit timer has a single 8-bit register for temporary storing of the high byte of the 16-bit access. The same temporary register is shared between all 16-bit registers within each 16-bit timer."
[22:36:46] <skorket> so if you did a read of the low byte, then another read of the low byte, it would write twice into the temporary high buffer?
[22:37:38] <skorket> so basically you are triggering a 16 byte read when you read the low byte
[22:37:42] <Xark> skorket: Something like that. However, I think that case is harmless but reading the high first you may read a stale value.
[22:38:56] <Xark> skorket: This bit me when I was polling the value of a timer (for cycle accurate stuff).
[22:39:45] <skorket> yeah, I'm doing some timing critical stuff so I made sure to pay attention to that detail
