Back
[00:53:48] <inflex> okay, nice PCB arrived today... hpoe this is a money maker
[00:54:00] <inflex> good indicator is that the stencil cut came out perfect first time
[00:54:23] <inflex> ooh, and other news, I can now cut the individual pads on TQFP32 and even MSOP20's now *mwhahaah*
[03:12:36] * scuzzy pats inflex on the head
[03:32:12] <inflex> grrr
[03:32:14] <inflex> :p
[03:38:43] <scuzzy> lol
[03:40:00] <scuzzy> it's just, I can't help but be patronising when someone sounds like they are trying to take over the world
[03:59:01] <soul-d> doing that every day
[03:59:05] <soul-d> planning for it
[04:04:05] <karlp> if you're not planning to take over the world, what are you doing?
[04:06:35] <scuzzy> taking over the galaxy?
[04:06:44] <scuzzy> Pffft, who isn't?
[04:08:20] <karlp> I'm pretty sure that taking over the galaxy includes taking over the world
[04:08:29] <karlp> I guess it doesn't have to though
[04:21:52] <soul-d> no some plans involve intergalatic highway's
[05:31:19] <inflex> wooo yeah... sexy looking regulator
[05:34:41] <inflex> http://dxp.me/i/5AR.jpg yum yum
[06:20:41] <soul-d> looks nice i realize now that 90degree pins might actualy work better as edge con :P
[06:21:57] <inflex> how so?
[06:22:19] <soul-d> well for stuff on electronic bench
[06:22:25] <soul-d> so wires pull less
[06:23:19] <soul-d> if you had normal straight up pins on such smal board
[06:23:31] <soul-d> it probably want to lift up board
[06:23:56] <soul-d> cause of wire tension
[06:27:30] <inflex> ah yes
[07:09:22] <alabd> good day all , in
http://wenku.baidu.com/view/9acdca0b79563c1ec5da7172.html?from=rec&pos=0&weight=3&lastweight=1&count=5 page 45 there is an example it's telling we should send 0x01 for SOH parametere and 0xBF for code parametere and .... but when we write Micro Controller code , should we write SOH=0x01; Code=0xBF ... in text format and compile it ? mean how to write code to build that 8 bytes packet ?
[07:22:05] <inflex> erm... ?
[07:22:10] <ziph> I don't suppose anyone knows of a quadrature decoder with say two channels and SPI? :)
[07:22:22] <inflex> ziph: oh sure.... lemme pull something out of my behind
[07:22:36] <inflex> ziph: see the sexy beast of a new regulator that I made up today?
[07:22:41] <ziph> inflex: No?
[07:22:57] <inflex> http://dxp.me/i/5AR.jpg
[07:23:18] <ziph> I am actually probably going to have to pull it out of.. a CPLD.
[07:23:44] <ziph> inflex: Cute, how is the heatsink bonded to the regulator?
[07:23:56] <inflex> ziph: they come with an adhesive bonding pad
[07:24:17] <ziph> Adhesive and thermally low R too?
[07:24:22] <inflex> http://static.rcgroups.net/forums/attachments/2/4/1/8/1/a4438831-51-PLD-BAC-ESJ3AR-bot.jpg
[07:25:17] <inflex> Of course not - it's about 50'C/W ;) 5.9'C/W total actually
[07:26:31] <ziph> And I presume you've got a via city under the regulator?
[07:26:33] <Steffanx> All the things you make have a buzzer, or isn't that thing on the same side as the heat sink a buzzer inflex ?
[07:26:52] <inflex> yes, almost all the things I make have buzzers... well, my logger didn't have one!
[07:26:57] <inflex> ziph: yes, a lot of vias
[07:27:19] <Steffanx> Where's that buzzer used for?
[07:27:33] <inflex> low voltage warning
[07:27:37] <ziph> Buzzers are mandatory on all of inflex's boards.
[07:27:49] <inflex> they are, yes... because I've got a freaking wheel of them :(
[07:28:16] <inflex> http
[07:28:18] <inflex> oops
[07:28:21] <inflex> http://dxp.me/i/vias.jpg
[07:28:40] <ziph> If he was making teletype machines for the deaf he'd still put buzzers on them. ;)
[07:28:56] <inflex> ziph: the dead can still feel the vibrations
[07:28:59] <inflex> deaf even
[07:29:09] <ziph> Gold? GP charges heaps for that don't they?
[07:29:34] <inflex> hah, it's not gold, just the lighting in this room ("warm" CFL)
[07:29:45] <ziph> Hah.
[07:30:14] <inflex> have to wait now about 1 week for my 16MBit EEPROMs to arrive
[07:30:22] <inflex> solves the whole compression problem nicely and only cost 10c more
[07:30:40] <ziph> Digikey?
[07:30:40] <inflex> nice thing is, I can quad-bit pump them
[07:30:47] <inflex> ziph: nah, E14, but on backorder
[07:31:51] <inflex> the quad-bit pumping is quite neat though
[07:32:10] <ziph> Does it do buffering?
[07:32:27] <inflex> not sure, don't see how - since it's still SPI
[07:32:55] <inflex> (but you use !HOLD !WR and SO along with SI all at once
[07:32:56] <ziph> The speed doesn't get you anywhere if you have to sit around for 20uS waiting for each byte to write.
[07:33:16] <inflex> it takes 100MHz clocking... and I'll be doing about... oh... 1MHz :p
[07:33:37] <ziph> Yeah, the 80MHz one I used recently though was useless.
[07:33:38] <inflex> what I'll do though is take advantage (I hope!) of the sequential data sending facility
[07:34:25] <inflex> mmm... this is the Atmel AT25DQ161
[07:35:08] <inflex> Anyhow, happy I found it - 16Mbit does solve a lot of problems... and I see there's 32MBit ones too
[07:35:13] <ziph> Yeah, those have RAM.
[07:35:16] <ziph> They're neat.
[07:35:25] <inflex> I saw some FRAM ones as well
[07:36:06] <ziph> The RAM is nice because you can DMA entire pages in at speed and then work on other things for the 10ms the page write takes.
[07:36:15] <inflex> Still this little 256Kbit one will stay in use, makes for nice easy sampling of my projects when I need
[07:37:11] <inflex> come to think of it though, I should just make a volts/amps dumper (via 115200 serial), no need for the EEPROM
[07:37:48] <inflex> Sure, I've got multimeters all around the place but they'll all do just one thing at once, rather than both v and i
[07:38:59] <ziph> Arrggh.
[07:39:13] <inflex> Waht I find I need to do those is package up such "gadgets" properly, else I lose them and then end up having to build them all over again
[07:39:19] <inflex> ziph: oh?
[07:39:29] <ziph> inflex: Apple-W again.
[07:41:44] <ziph> inflex: What are your stencils made from?
[07:42:33] <inflex> some plastic...
[07:42:57] <inflex> http://www.preservationaustralia.com.au/products
[07:43:00] <ziph> Do you just lay them on the board and use a blade of some kind?
[07:43:48] <inflex> Yep, just get yourself various width paint-scrapers
[07:43:53] <inflex> flexible is best
[07:44:15] <inflex> and it helps a LOT if you can get the edge as straight/flat as possible.
[07:44:47] <ziph> Sooo.. what do you charge for having some done and thrown in an express post envelope? *g*
[07:45:25] <inflex> depends, the cutting machine has some genuine limits ... eg, no smaller than 0603 typically, no bigger than 50x50mm area quite often, sometimes it's okay beyond that
[07:46:44] <inflex> only just started getting TQFP32's working out.
[07:46:53] <ziph> Ahh, the 50x50 limit would be a problem.
[07:47:03] <inflex> easier perhaps to find yourself a local laser engraver mob and use their cutting machine
[07:47:13] <ziph> I'm currently trying to avoid having to do a 160x160 board.
[07:47:25] <inflex> hooooha, yes, no chance I'd do that
[07:47:29] <ziph> :)
[07:47:43] <ziph> The one I showed you before is 120x100 though.
[07:47:50] <inflex> what happens is that the sheet "drifts" a little or a piece of offcut sometimes gets between the blade/sheet... and it's a disaster then
[07:47:59] <ziph> Laser engraver? What do they normally do?
[07:48:07] <inflex> normal engraving shops
[07:48:17] <inflex> doing plaques, mugs etc
[07:48:36] <inflex> of course, a bit hard to find one that'll actually let you walk in and use their machine :(
[07:48:37] <ziph> They have gear that will cut through shim/the stuff you use?
[07:48:47] <inflex> almost any laser should
[07:49:02] <inflex> 10W and beyond should do the trick... I know someone who's doing it with a BluRay diode
[07:50:26] <ziph> What thickness is the stuff you use?
[07:51:13] <inflex> try get the 50um
[07:51:20] <inflex> 70um~80um is getting too thick
[07:51:49] <ziph> GP's minimum is 0.1
[07:52:04] <inflex> 0.1.... what? mm?
[07:52:09] <ziph> Yeap.
[07:52:23] <inflex> hrmm... that's 100um
[07:52:29] <ziph> Yeap.
[07:52:32] <inflex> okay, maybe my brain is screwed on wrong
[07:52:49] <inflex> yeah, actually that makes sense
[07:52:50] <ziph> They do various sized in 100um to 200um.
[07:53:00] <inflex> go for the lower side, perhaps 120um+
[07:53:42] <inflex> a lot depends on the type of parts you put on
[07:54:01] <inflex> you can always cheat and use a smaller hole with a thicker stencil
[07:54:09] <ziph> Yeap.
[07:54:27] <inflex> but it does make pasting a bit harder if it's too thick, since the paste won't get through
[07:55:03] <inflex> (on the smaller holes)
[07:55:36] <inflex> then ask yourself... do you prefer bridges or dry pads?
[07:55:51] <karlp> who/what is GP?
[07:56:08] <inflex> karlp: Gold Phoenix
[07:56:20] <inflex> supplier of PCBs, assembly and many other things for a lot of us
[07:56:27] <karlp> yeah, I think I've heard of them
[07:56:48] <inflex> Does Sparkfun and BatchPCB still use them?
[07:57:11] <inflex> I know they were griping about the closure every year due to the Chinese New Year.... and were looking for another provider
[07:57:42] <ziph> Pft.
[07:57:52] <ziph> If we complained about Thanksgiving they'd invade us.
[07:58:27] <inflex> well, the Chinese new year does knock production out for a month
[07:58:55] <inflex> yeah, the whole December->March section of the year is a freaking disaster for getting things done, between Western holidays and Eastern ones... it's a deadzone
[07:59:05] <karlp> yeah, and jewish holidays close all online camera stores on different calendars too, you just get used to it
[07:59:16] <karlp> france is effectively shut for all of august
[08:01:25] <inflex> Just August?
[08:01:46] <karlp> well, that's when the maximum impact of summer holidays is
[08:02:03] <karlp> iceland closes for days scattered arbitrarily throughout summer, depending on the weather
[08:04:43] <karlp> http://tweak.net.au/pics2/2008/June/miscstream/pichtml/web_2008_06_20-16_01_30--crw_3225_jfr.html
[08:06:01] <inflex> that's quiet amusing
[08:12:31] <Steffann> lol karlp
[08:13:30] <karlp> ziph: were you looking for quadrature decoder hardware?
[08:13:42] <ziph> karlp: Yeap.
[08:13:45] <karlp> atmel SAM3N has that built in,
[08:14:34] <ziph> Most of the ARM's do but either don't have two or don't have enough 32 bit timers.
[08:15:02] <karlp> use two arms :)
[08:15:20] <karlp> how many 32bit timers do you need/want?
[08:15:33] <ziph> The ST's only have 2.
[08:15:49] <ziph> LPC's have lots but only 1 quadrature decoder.
[08:16:03] <ziph> Which is odd on an IC with 3 motor PWMs. :)
[08:17:42] <ziph> I'm also tempted to put the CLPD on the isolated side and just run SPI back, that would save having many inputs...
[08:56:00] <ys0> oi
[10:08:10] <sh00p> ok, so I have one ATMega32-16PU-0647B, and I have one ATMega32-16PU-1116
[10:08:15] <sh00p> they should work identically right?
[10:21:34] <Tom_itx> sh00p, yes
[10:24:24] <sh00p> nice, thanks
[10:24:33] <sh00p> for some reason the 1116 feels slower
[10:24:38] <sh00p> but it might be shit behind the weels
[10:25:16] <karlp> are you using the internal oscillator? it might be slower :)
[10:25:52] <sh00p> yeah I did that as well
[10:25:59] <sh00p> but then i changed to the external
[10:26:03] <sh00p> but things still wouldnt move
[10:26:05] <sh00p> and when they did
[10:26:12] <sh00p> the mcu didnt respond to switches and stuff
[10:26:17] <sh00p> but for some reason it's working now
[10:26:21] <sh00p> pretty spooky
[10:26:27] <Tom_itx> breadboard?
[10:26:32] <sh00p> holecard
[10:49:12] <ziph> How well do AVR's do at being SPI slaves?
[11:14:54] <carp3> Hi, I want to build a simple PID Velocity controller with AVR , whitch AVR do you recommend ? Orginal AVR/XMEGA or AVR32 ? i think faster is not always better.
[11:25:46] <Tom_itx> either one
[11:26:01] <Tom_itx> 8bit would work as well
[12:39:17] <RikusW> hi scuzzy
[12:39:29] <Steffanx> Hi RikusW
[12:39:42] <RikusW> Hi Steffanx
[12:39:59] <Steffanx> How is it in SA nowadays?
[12:40:09] <RikusW> rainy here :)
[12:40:16] <RikusW> was very dry and hot
[12:40:28] <RikusW> cold over there ?
[12:42:33] <RikusW> Been thinking about the flash writing, to make things interesting HVPP and HVSP wants a mode bit set (0x40) when writing the last page.... the stk500 use it to clear the command that was set in the avr
[12:43:13] <RikusW> jtag mkI and stk500 use word addressing for flash, jtag mkII use byte addressing....
[12:43:23] <Steffanx> ~0-9C the last days
[12:43:38] <Steffanx> And VERY foggy the last few days/ week
[12:43:55] <RikusW> snow ?
[12:45:25] <Steffanx> Now
[12:45:27] <Steffanx> *-w
[12:46:35] <RikusW> does it snow every year over there ?
[12:47:27] <Steffanx> Yes i think so, but some years it's not cold enough so it melts very fast when it hits the ground
[12:48:11] <RikusW> the slushy part afterwards is worst ;)
[13:11:56] <RikusW> Now I know what AVRPART/ICE_SETTINGS/JTAGICEmkII/ucAllowFullPageBitstream is for
[13:12:19] <RikusW> its true for older avr's for programming with IR=6, the algorithm changed
[13:12:39] <RikusW> explaining why jtag mkI cannot program newer avr's....
[13:33:29] <Fleck> hey all, what do you use to flash avr using SPI in linux?
[13:35:08] <RikusW> avrdude
[13:35:31] <Fleck> any gui tools? :)
[13:36:20] <Fleck> whats usbprog-gui?
[13:36:28] <RikusW> I'm working on a Qt4 gui
[13:36:33] <RikusW> but its alpha
[13:36:39] <Fleck> wow, nice RikusW :)
[13:36:41] <RikusW> not even beta yet
[13:37:01] <RikusW> what programmer do you use /
[13:37:02] <RikusW> ?
[13:37:13] <Fleck> cheapest on the ebay ;D
[13:37:14] <Fleck> sec
[13:37:30] <RikusW> you already have one ?
[13:37:35] <Fleck> http://www.ebay.com/itm/USB-ISP-Programmer-ATMEL-AVR-51-ATMega-ATTiny-New-/390358378857?pt=LH_DefaultDomain_0&hash=item5ae32c2969
[13:37:39] <Fleck> something like this
[13:37:43] <Fleck> yes i have it! :)
[13:37:57] <Fleck> works great in windows :D
[13:38:14] <Fleck> but i am sick of that OS :P
[13:38:15] <RikusW> you don't know what protocol it use ?
[13:38:27] <RikusW> does it work with AVRStudio ?
[13:38:31] <RikusW> me too ;)
[13:38:47] <Fleck> no clue about avrstudio, but works with bascom
[13:42:44] <RikusW> does bascom support more than one programmer type ? what does bascom call it ?
[13:43:24] <Fleck> RikusW can you name some common? as dont have laptop near me at the moment ;(
[13:43:37] <RikusW> stk500 ?
[13:43:40] <RikusW> usbasp
[13:43:48] <RikusW> usbtiny
[13:44:05] <Fleck> usbasp sounds familiar, any more? like progisp?
[13:44:10] <Fleck> or something like that?
[13:44:17] <Fleck> 1 X USBASP
[13:44:22] <Fleck> in that item descr
[13:44:28] <Fleck> at ebay
[13:45:23] <RikusW> I only started adding usbasp support, its far from finished...
[13:46:29] <RikusW> use avrdude
[13:46:36] <RikusW> it works with usbasp
[13:47:51] <Fleck> ok, what about coding/compiling in linux ?
[13:47:54] <Fleck> any gui tools?
[13:48:28] <RikusW> there is an atmel toolchain for Linux about 70MB
[13:48:37] <Fleck> ohh
[13:48:39] <RikusW> don't know about gui tools
[13:48:51] <RikusW> there is frontends I think
[13:48:54] <RikusW> don't know
[13:48:58] <RikusW> any names
[13:53:43] <carp3> RikusW: most cheap programmers are based on this project:
http://www.fischl.de/usbasp/
[13:55:09] <RikusW> there seems to be quite a few firmware versions of usbasp out there...
[13:58:25] <RikusW> it complicates writing gui sw for it
[14:19:32] <tlvb> why do you feel the need for an usbasp gui? on the other hand, why not have the gui generate a commandline to avrdude that handles the communication with the usbasp?
[14:22:19] <Fleck> soo, i have to create makefile to compile for atmega in linux?
[14:22:30] <RikusW> yes
[14:22:36] <Fleck> ;(
[14:22:43] <RikusW> that will be easiest
[14:23:28] <RikusW> AVRStudio4 generates a makefile anyways...
[14:23:33] <RikusW> you just don't see it
[14:23:45] <RikusW> unless you go looking for it
[14:24:00] <Fleck> would be great to have gui that does that for me in linux :/
[14:24:13] <ben1066> Hey hey
[14:24:33] <ben1066> So im trying to use usart, and avr studio is telling me UBBR is non existent
[14:24:54] <Fleck> UBRR0 ?
[14:25:49] <RikusW> ben1066: on what avr ?
[14:26:00] <ben1066> Hm aparently it doesnt have a single register, only high and low, its an atmega8
[14:27:36] <tlvb> it is a 16 bit register = two 8 bit registers
[14:28:07] <RikusW> first set UBRRH then UBRRL
[14:28:28] <ben1066> Yea, I only noticed via the datasheet
[14:28:41] <RikusW> and when setting UCSRC = (1<<URSEL) | your bits
[14:28:58] <RikusW> it is shared with UBRRH
[14:30:17] <ben1066> yea
[14:30:28] <ben1066> and hurr, avr has decided the standard lib it doesnt like it :p
[14:39:11] <ben1066> any idea why its complaining with "storage class specified for parameter?
[14:39:58] <Fleck> woah, eclipse has AVR plugin! :D
[14:47:30] <ben1066> Hmm, I seem to be doing it wrong :p
[14:47:39] <ben1066> my data doesnt ever appear :S
[14:48:48] <ben1066> UCSRC = (1<<URSEL)|(3<<UCSZ0);
[14:48:53] <ben1066> thats 8 bit, 1 stop, no parity right?
[14:52:34] <RikusW> r16, (1<<URSEL)|(1<<UCSZ1)|(1<<UCSZ0) //8N1
[14:53:15] <RikusW> UCSRB = (1<<RXEN)|(1<<TXEN)
[14:53:53] <ben1066> I only want transmit so I ommited the RXEN
[14:54:13] <ben1066> UBRRH = (unsigned char)(16>>8); UBRRL = (unsigned char)16;
[14:54:25] <ben1066> is how im setting UBRR, its from the documentation :p
[14:55:36] <RikusW> UBRRH = 0;
[14:55:42] <RikusW> 16>>8 == 0
[14:56:07] <RikusW> UBRRL = 16; should be fine without the cast
[14:58:05] <ben1066> Alright, I still dont understand where my data is going :S
[14:58:46] <ben1066> http://pastebin.com/UeQbmE3b That should transmit correctly yes?
[14:59:13] <ben1066> the puts works, ive used it before, the putc is pretty much the avr documentation standard version
[15:00:43] <ben1066> an arduino with Serial.begin( 57600 ) works but this doesnt :S
[15:01:38] <RikusW> while(*text) uart_putc(*text++); ;)
[15:02:01] <RikusW> your code seems fine
[15:02:44] <RikusW> what crystal do you have connected ?
[15:02:59] <RikusW> your baud is probably wrong
[15:03:28] <ben1066> i have a 16mhz crystal
[15:05:51] <tlvb> have you confirmed that you are using the 16MHz crystal for the clocking? not just that it is connected?
[15:05:56] <RikusW> UBRR = 16 should only just work
[15:06:38] <ben1066> Yea I am...
[15:06:49] <ben1066> Are you sure my while loop is right? the simulator trips up on it...
[15:06:55] <ben1066> for the putc
[15:07:40] <tlvb> may be operator precedence, try *(text++) intstead
[15:07:44] <tlvb> but I'm not sure
[15:08:38] <ben1066> Nono
[15:08:46] <ben1066> I mean the loop above, in uart_putc
[15:08:53] <ben1066> it stops on the while loop there in the simulator
[15:08:59] <tlvb> ah
[15:09:17] <RikusW> wait: sbis UCSRA,UDRE rjmp wait
[15:09:21] <RikusW> thats what I used
[15:09:41] <ben1066> that should be the same? no?
[15:09:50] <RikusW> yes
[15:10:23] <RikusW> the simulator won't like that because there is no real serial port....
[15:10:40] <ben1066> the simulator doesnt simulate usart?
[15:10:56] <RikusW> no transmitting actual data
[15:11:05] <RikusW> or receiving...
[15:11:33] <RikusW> a virtual serial port in the OS connected to the simulator would be really nice
[15:11:47] <RikusW> but I don't know of any such thing
[15:12:04] <RikusW> working one anyways
[15:12:09] <mrfrenzy> have a look at proteus vsm
[15:12:48] <mrfrenzy> even invludes examples with avrs running serial ports, ethernet etc
[15:14:31] <carp3> you can use Proteus ISIS and Virtual Serial Bridge but they are not free ( Proteus and Virtual Serial Bridge )
[15:16:36] <ben1066> I dont understand!
[15:16:53] <ben1066> im now using my arduino as a usb to serial
[15:17:01] <ben1066> still no luck, I dont see ANYTHIGN
[15:17:43] <carp3> use Codevision or even Bascom to check Hardware is OK.
[15:17:54] <ben1066> how?
[15:18:04] <ben1066> ive only got avr studio installed :s
[15:18:29] <carp3> download a free trial version of Codevision . it's easy to use
[15:19:14] <ben1066> is there an easier way to test usart/
[15:20:35] <carp3> well first you should confim that your hardware is OK
[15:20:48] <carp3> Codevision :
http://www.hpinfotech.ro/html/cvavr.htm
[15:21:44] <RikusW> he said arduino does work...
[15:21:55] <ben1066> Yea...But still how, a test case in avr studio is going to be the same as test case in codevision no?
[15:22:35] <ben1066> This AVR is brand new also
[15:22:42] <ben1066> So afaik I havent fried it :P
[15:23:13] <ben1066> i can use JUST transmit or JUST recieve right?
[15:23:47] <carp3> no, your can use both at the same time
[15:23:58] <carp3> you*
[15:24:11] <ben1066> nono, but I mean
[15:24:20] <ben1066> I dont need boht enabled if I only need transmit?
[15:24:26] <ben1066> EN_TX or whatever
[15:24:44] <carp3> yes it's ok
[15:24:52] <carp3> what is your avr ?
[15:24:58] <ben1066> atmega8
[15:25:38] <Fleck> avrdude doesnt support Atmega88PA?
[15:25:59] <ben1066> Im not using it...?
[15:26:01] <carp3> what's your clock frequency ? ben1066
[15:26:27] <ben1066> 16mhz crystal, with 22pf caps
[15:26:38] <ben1066> it works since it programs...
[15:26:45] <carp3> did you change the fusebits ?
[15:27:11] <ben1066> ye
[15:27:41] <ben1066> extalhighres_16ck_64ms
[15:28:14] <ben1066> http://pastebin.com/vhQAKqxf
[15:29:10] <carp3> UCSRB=0x18;
[15:29:12] <carp3> UCSRC=0x86;
[15:29:13] <carp3> UBRRL=0x67;
[15:29:41] <carp3> try this
[15:30:17] <carp3> 9600 , 8 Data, 1 Stop, No Parity
[15:31:44] <ben1066> noting
[15:31:47] <ben1066> nothing8
[15:31:49] <ben1066> nothing*
[15:32:26] <carp3> hmmm code is OK , can you upload hex file ? i'll test it.
[15:33:11] <ben1066> http://dl.dropbox.com/u/11197643/uart.hex
[15:33:58] <carp3> 9600 ?
[15:34:09] <ben1066> What ever you gave me
[15:34:57] <carp3> works
[15:35:34] <carp3> check your hardware
[15:35:40] <ben1066> well the heck am I doing wrong then
[15:36:06] <ben1066> avcc should be ok unconnected right?
[15:36:20] <RikusW> ben1066: you're using a ftdi chip on the arduino ?
[15:36:33] <ben1066> Ye
[15:36:40] <ben1066> When I loopback the arduino it works
[15:36:50] <ben1066> so I figured connected the avr to the arduino would also work
[15:37:07] <carp3> http://1.aseo.hostei.com/images/360a502598a4.jpg
[15:37:09] <RikusW> and you set the PC side to 9600 bps too ?
[15:37:25] <ben1066> i set putty to 9600
[15:37:35] <RikusW> 8n1 ?
[15:37:57] <RikusW> how about hw flow control ? did you turn that off ?
[15:38:03] <carp3> putty's default is 8n1
[15:38:08] <ben1066> yesyes and yes
[15:38:13] <ben1066> and the arduino data showed
[15:38:17] <ben1066> so idk why mine isnt
[15:40:21] <ben1066> And I dont get ANY data
[15:40:29] <ben1066> so the avr doesnt seem to be sending anything
[15:41:17] <ben1066> any other ideas?
[15:41:54] <carp3> check voltage of TX pin
[15:42:30] <ben1066> 5v
[15:43:02] <carp3> well it should be something like 2.5V
[15:43:33] <ben1066> well its not :p
[15:45:20] <ben1066> Oh, the arduino was pulling it high
[15:49:42] <ben1066> now its holding itself high
[15:49:43] <ben1066> :S
[15:49:47] <ben1066> Im comfused
[15:49:51] <ben1066> I thought this would be easy
[15:51:04] <ben1066> my arduino also pulls the pin high
[15:52:04] <ben1066> help :(
[15:53:27] <ben1066> i really dont understand
[15:54:49] <carp3> :p do you set com port name properly in putty ?
[15:55:07] <ben1066> yes
[15:55:23] <ben1066> I also have a telly mate thingy, that outputs serial to a tv, that isnt seeing it either
[15:56:10] <carp3> TV ?
[15:56:17] <ben1066> composite
[15:56:28] <ben1066> again
[15:56:37] <ben1066> my arduino outputs valid serial
[15:56:39] <ben1066> the atmega doesnt
[15:57:15] <carp3> so maybe your atmega is dead or clock is not set properly.
[15:57:32] <ben1066> how can it die, by me taking it from the tube :s
[15:57:50] <carp3> try to toggle LED every 5sec and check it's really 5sec.
[16:04:30] <ben1066> the led is toggling
[16:04:35] <ben1066> it seems to be at right speed
[16:05:03] <carp3> so everything is ok
[16:06:02] <ben1066> So I understand even less :S
[16:06:20] <carp3> Program-->OK avr-->OK clock-->OK Circuit-->OK FT232-->OK USBCable-->OK Putty-->OK
[16:06:31] <carp3> so what's wrong ? :D
[16:08:30] <ben1066> I dont know!
[16:11:15] <ben1066> carp3: that doesnt help me :S
[16:13:44] <carp3> i don't know, maybe you should ask someone that done this using arduino . maybe there is a trick for arduino
[16:13:47] <carp3> i don't know.
[16:19:03] <ben1066> NO
[16:19:45] <ben1066> But the other serial out i have isnt working either
[17:02:51] <Fleck> whats the best replacement for atmega8 ?
[17:03:12] <Casper> depend on your needs
[17:03:16] <Casper> and why a replacement?
[17:03:21] <Casper> it's still in production
[17:03:57] <Fleck> is it? thought its deprecated :D
[17:04:21] <Steffanx> A mega88 is a good replacement though.
[17:04:29] <Steffanx> It's pin compatible, and has more features
[17:04:46] <Fleck> is 88 the best/fastest/feature rich replacement?
[17:05:03] <Casper> depend what you want
[17:05:25] <Casper> best would be like the 328 but it's big so not pin compatible
[17:05:42] <Fleck> i like PWM :D
[17:06:04] <Steffanx> The 328 is not pin compatible?
[17:06:23] <Casper> it's a pdip40
[17:06:33] <Steffanx> No, not really
[17:06:47] <Casper> the mega88 is pdip28
[17:06:56] <Steffanx> 328 too
[17:07:13] <Casper> yeah? hmmm maybe it's the wrong number then
[17:07:23] <Casper> I was thinking the big fat 40 pins one
[17:07:33] <Steffanx> 324/644 ?
[17:07:41] <Casper> I think so
[17:08:29] <Fleck> i need 32-TQFP
[17:08:51] <Casper> Fleck: check the parametric table at atmel
[17:08:55] <Steffanx> Go to atmel.com and use their "Find it" tool
[17:09:19] <Fleck> uhh didnt know about findit tool :D
[17:09:22] <Fleck> let me try it! :)
[17:13:07] <Fleck> Steffanx where is it :D ? i cant find such tool
[17:13:28] <Steffanx> Here it's one the homepage .. on the left side
[17:24:13] -verne.freenode.net:#avr- [freenode-info] if you're at a conference and other people are having trouble connecting, please mention it to staff:
http://freenode.net/faq.shtml#gettinghelp
[17:41:24] <Fleck> well yes, looks like 88 is the best replacement
[17:41:27] <Fleck> or 168
[17:42:28] <Fleck> strange, ATmega88 datasheet says 20Mhz, FindIt tool says max freq = 16 ;D
[17:44:21] <Fleck> ohh Automotive is different, what does that mean?
[17:59:22] <tlvb> I think atmega168 is tqfp32
[18:00:07] <tlvb> yep, that's the one I'm using, and at 20MHz
[18:01:04] <Tom_itx> it is both
[18:01:07] <Fleck> what does automotive to chips?
[18:01:10] <Tom_itx> dip and tqfp
[18:01:31] <tlvb> yes, of course there is dip too, but Fleck needed tqfp32
[18:01:42] <Tom_itx> either specialized chips or wider temperature range
[18:01:44] <Fleck> atmega88 can run @ 20MHz, atmega88 automotive only 16MHz
[18:02:44] <Fleck> ATMEGA88PA-AU << this is automotive?
[18:03:39] <Fleck> and whats the difference between 88, 88A, 88P and 88PA ?
[18:04:08] <Tom_itx> there is a parametric search table on atmel
[18:04:52] <Fleck> yeah well - did not find any differences
[18:05:02] <karlp> the more Ps and As the better
[18:05:15] <karlp> lower power, wide range of speed vs voltage,
[18:05:36] <karlp> automotive normally only rates them for higher voltages, and higher temps
[18:05:37] <Tom_itx> lower power, lower speed
[18:05:54] <Tom_itx> run it fast and loose power
[18:06:01] <Tom_itx> higher v
[18:07:47] <Fleck> ok thx
[18:08:00] <Fleck> ill get ATMEGA88PA-AU then
[18:08:47] <Tom_itx> 328
[18:08:55] <Tom_itx> not much more
[18:09:27] <Tom_itx> same features
[18:10:18] <Fleck> :D 9$ more
[18:10:44] <Tom_itx> mmm
[18:10:49] <Tom_itx> not here that i know of
[18:12:00] <Fleck> 88PA-AU is 3$ :D 168 is 5$ and 328 ir 12$ USD
[18:13:42] <Fleck> strange... cause arduino nano with 328 is 12.80$ :D
[18:13:59] <Tom_itx> they buy in quantity
[18:14:20] <Fleck> yeah, i guess so
[18:14:58] <Fleck> ok /me sleep! bye all and thx for info/help
[18:14:58] <Tom_itx> they're always cheaper by the dozen:
http://tom-itx.dyndns.org:81/~webpage/boards/new_batch/USB_chips.jpg
[18:15:19] <Fleck> send one Tom_itx :D
[18:15:26] <Fleck> to me i mean :D
[18:15:30] <Tom_itx> those are usb
[18:15:49] <Tom_itx> at90usb162 and atmega32U2
[18:16:44] <Fleck> ohh
[19:02:50] <amee2k> if a power LED package has a "heatsink pad" and there are no specifications about it in the datasheet, is it acceptable to electrically connect it to one of the power pins?
[19:04:02] <amee2k> that would make my life a lot easier since i can just drop 2/3rd of it on a ground plane. as opposed to having to get the heat out of a 2x0.5mm area in the middle of the footprint
[19:04:56] <Tom_itx> amee2k, more than likely it's connected to gnd
[19:05:10] <Tom_itx> or not at all
[19:05:20] <amee2k> ground being... cathode?
[19:05:29] <Tom_itx> put a meter on it
[19:05:45] <amee2k> i'm selecting parts to order, i don't have them yet
[19:06:13] <Tom_itx> cathode is negative
[19:06:15] <amee2k> the datasheet makes no mention of how to handle that pad electrically that i can see
[19:07:45] <amee2k> hence why i'm wondering, is it usually acceptable to connect it to either anode or cathode potential when nothing is specified?
[19:08:51] <Tom_itx> i would see what the meter says but i would tend to ground it
[19:09:17] <amee2k> i don't want to spend a lot of money on parts only to find out that they're a bi*ch to use
[19:09:49] <Tom_itx> that's why we make prototypes
[19:10:33] <amee2k> hehe
[19:40:16] <inflex> going to be a busy w/e here
[19:40:22] <inflex> 5 more PCB designs just arrived
[19:48:06] <ziph> amee2k: If there's nothing in the datasheet I'd try E-mailing the manufacturer asking for recommendations.
[19:53:51] <geckosenator> anyone know where I can get avrs cheaper than digikey in low quantities?
[19:54:19] <ziph> Tried Future Electronics?
[19:54:43] <ziph> Oh, low quantities.
[19:55:04] <ziph> http://futurlec.com.au/ ? :)
[19:56:28] <geckosenator> cool let me see
[19:57:27] <geckosenator> i need 64k flash min
[19:58:17] <ziph> They're fairly slow to deliver, but for low quantities I can't think of anyone cheaper.
[19:58:41] <ziph> Are you in Australia?
[19:59:37] <geckosenator> new zealand
[20:01:03] <ziph> On a boat?
[20:01:07] <inflex> Just go with digikey... their prices are acceptable... you could try ES but they're flooded still.
[20:01:17] <ziph> ES?
[20:02:01] <inflex> http://www.es.co.th/
[20:02:01] <geckosenator> yeah ok fine
[20:02:11] <geckosenator> ziph:: yes I am on a boat
[20:02:21] <inflex> ziph: they're drowning atm of course because of the Thailand flooding
[20:02:29] <geckosenator> heh
[20:02:33] <ziph> Yeap.
[20:02:58] <inflex> which is a pita because they have good prices in bulk for a lot of the stuff I need
[20:03:03] <inflex> ...and I'm running low
[20:03:21] <geckosenator> and digikey is a rip
[20:03:30] <geckosenator> it may seem cheap that it costs only $3
[20:03:33] <inflex> pffft... you should try RS or E14
[20:03:36] <geckosenator> but in reality it should cost $1
[20:03:48] <inflex> you're paying for the convenience
[20:04:09] <ziph> This looks cute:
http://www.es.co.th/detail.asp?Prod=99928206
[20:05:06] <ziph> Dontronics is sometimes cheap, but it depends on where the exchange rate was when he last bought the particular tube/tray. :)
[20:06:26] <inflex> http://dxp.me/i/pcbs.jpg --- too many to choose from!
[20:06:38] <inflex> ziph: oh yes, I've often considered getting that ARM dev kit
[20:06:47] <ziph> inflex: For what?
[20:06:54] <inflex> *shrug*
[20:06:54] <geckosenator> did you know 24bit types are supported in avr-gcc now?
[20:07:04] <inflex> geckosenator: oh really - what, as uint24_t ?
[20:07:06] <geckosenator> I figured just for pointers, but
[20:07:16] <ziph> inflex: Something fancy with an LCD, or something towards the embedded end?
[20:07:21] <geckosenator> they have stuff for negation too in the compiler so it must do integers too
[20:07:23] <inflex> that's cute, could be useful too in tight math situations where 16 is too small and 32 is obscene
[20:07:39] <inflex> ziph: no idea --- it just appealed to me in terms of tinkering :p
[20:07:42] <geckosenator> yeah and it actually only performs 3 instructions not 4 to add and subtract
[20:08:03] <geckosenator> its mainly to use pointers to address space larger than 128k
[20:08:04] <inflex> it'd be useful for 16x8 multiply's
[20:08:06] <ziph> inflex: Have a play with the M3's.
[20:08:13] <inflex> ziph: BMW M3?
[20:08:20] <ziph> inflex: Yes, those too.
[20:08:22] <ziph> inflex: :)
[20:08:23] <geckosenator> heh, I didnt see 16x8 in there
[20:08:36] <ziph> inflex: The Cortex M3's. They are as easy to wire as an AVR.
[20:08:38] <inflex> geckosenator: bah, I need that often
[20:08:45] <geckosenator> I mean it has special code for 16x8 -> 32
[20:08:46] <inflex> ziph: one day... one day
[20:08:51] <geckosenator> so it probably jsut truncates that
[20:09:10] <inflex> geckosenator: well, damn, already using that... though I was hurting bad when I was doing 32x32
[20:09:15] <geckosenator> look in gcc/config/avr/avr.md
[20:09:29] <geckosenator> its PSI mode for partial single integer (single being 32 heh)
[20:09:43] <geckosenator> no it already knows how to do high side stuff
[20:09:53] <geckosenator> so if you shift then multiply it can emit less muls
[20:09:57] <geckosenator> or whatever
[20:10:04] * inflex has always loved integer math... but I suppose that happens when you start on Z80's and 8088's
[20:10:07] <inflex> (with no copro)
[20:10:25] <geckosenator> it doesnt cover all cases, but they improved recently a lot
[20:10:28] <inflex> I remember when Fractint first came out and it was such a brilliant speedup from the FP algorithms
[20:10:34] <ziph> inflex: I like 16 bit integer math because you can exhaustively test a lot of things.
[20:10:38] <geckosenator> I'm still trying to merge my fixed point support
[20:10:53] <inflex> geckosenator: that's useful too (fixedpoint)
[20:11:04] <inflex> okay,I need to generate some stencils now... oooh, which to do first... which to do
[20:11:29] * inflex has two products of utmost in his mind... the 4 channel programmable LED strobe... and the 4 channel independent power board for servos
[20:11:45] <geckosenator> there are only 8 16 and 32bit nattive types
[20:11:56] <geckosenator> the 64 and 128 are still emulated in c heh
[20:12:16] <geckosenator> I was thinking about making long long native
[20:13:09] <geckosenator> for 64bit integers rather than having to call a c routine
[20:15:26] <inflex> yeah, for the most part, I only end up with a 64-bit result as an intemmediate which I scale back down to 32 quickly, but it does slow things up a lot
[20:15:42] <geckosenator> they also have native support for delays heh
[20:19:19] <geckosenator> in fact there are so many changes in svn-gcc I am almost afraid to use it heh
[20:19:43] <geckosenator> probably get internal compiler errors trying to build something complex
[20:22:45] <inflex> well, I did notice that my _delay_ms() calls were broken recently
[20:22:54] <inflex> comes up saying that it needs to be an integer constant
[20:23:10] <geckosenator> you used floating point values?
[20:23:32] <geckosenator> or you passed a variable?
[20:23:57] <geckosenator> because the code to make delay native requires a const
[20:24:52] <geckosenator> I could probably fix it to make it invoke a routine if you pass a variable..
[20:25:13] <inflex> I was using fixed values... eg, _delay_ms(100);
[20:25:16] <geckosenator> it would break on delay_us though if you passed it a really small number..
[20:25:30] <geckosenator> how did that break?
[20:25:43] <inflex> erm... lemme go show the code
[20:26:17] <inflex> /usr/lib/gcc/avr/4.5.3/../../../avr/include/util/delay.h:153:28: error: __builtin_avr_delay_cycles expects an integer constant.
[20:27:02] <inflex> started happening recently (3~5 days ago) on code that I've never touched for months that used to compile just fine
[20:27:47] <geckosenator> see, told you it is built in :-P
[20:28:09] <geckosenator> but you called with 100?
[20:28:13] <geckosenator> that _is_ and integer constant
[20:28:25] <inflex> yes
[20:28:51] <inflex> not sure what's going on - it doesn't fail on the simpler test (eg, C code with only a while loop with _delay_ms()
[20:29:10] <geckosenator> oh, what is the code then?
[20:29:33] <geckosenator> because it probably trying to be smart but in fact it is dumb
[20:30:20] <inflex> the code it's coming from (where the error comes up) is quite a bit more complex - but the actual _delay_ms() call is still just _delay_ms(100); (or some other number)
[20:30:49] <geckosenator> I realize that
[20:31:18] <geckosenator> you need to take the complex code, and try to simplify it until you get the simplest test case that still gives the error
[20:31:29] <geckosenator> hopefully will only be 10 lines or less
[20:31:30] <inflex> well, seems like it might be in the flags then... or a mix of things
[20:31:36] <inflex> avr-gcc -Wall -mmcu=atmega168 -ffreestanding -fno-split-wide-types -funsigned-char -fno-inline-small-functions -ffunction-sections -fshort-enums -Wl,--relax,--gc-sections -Os simple.c -c
[20:31:39] <inflex> causes the fault
[20:32:57] <inflex> -ffreestanding is the fault maker... but I removed that from my other code and the error persists... some something is interacting somewhere
[20:32:59] * inflex tries again
[20:33:07] <geckosenator> ok
[20:33:12] <geckosenator> a mix of things heh
[20:33:26] <geckosenator> gcc is a pain in the ass
[20:33:40] <geckosenator> it takes me an hour to recompile after changing 1 line
[20:33:55] <inflex> mmm... okay, -ffreestanding purge fixed it now on the older code
[20:34:01] <inflex> didn't work the other night :(
[20:34:34] <inflex> (the other night I dropped right back to just -Os ... and nothing else :(
[20:35:06] <inflex> Still, no one cares, so it's back to stencil making :D
[20:36:47] <geckosenator> what is ffreestanding?
[20:36:59] <geckosenator> takes out inits and stuff?
[20:37:33] <geckosenator> I wonder why that made any difference, it shouldnt
[20:44:22] <geckosenator> I got internal compiler errors trying to build one of my sources
[22:47:04] <geckosenator> they changed stuff so you have to declare variables in pgmspace as const
[22:47:10] <geckosenator> which breaks avr-libc
[22:47:18] <geckosenator> but I think I fixed it heh
[22:50:42] <earth1000> I'm a complete newbie with micro-controllers. I'm hoping to find a place where I can chat about them and which kind of microcontroller is ideal for which kinds of applications.
[22:53:37] <inflex> well, two more products done...
http://nqrc.com/images/PLD-PGMS8.jpg and
http://nqrc.com/images/PLD-POWERBUS-A4.jpg .... aaah, sometimes I love this life
[22:54:28] <Casper> which kind... I'll tell you something, you're in an AVR channel, so we're all going to say that avr is THE thing. BUT you should take a look at PIC, the series 18 seems quite good, so might want to consider it
[22:54:48] <Casper> as to which one do the job, it depend on too many factor, like how good at programmation you are
[22:55:02] * inflex votes to oust Casper from the channel ... heathen!
[22:55:11] <inflex> at least suggest the MSP430, for we mostly respect that !
[22:55:38] <Casper> I don't know that one
[22:55:57] <Casper> inflex: are you good with smps debugging? I made one, I hit 30% efficiency :/
[22:56:25] <inflex> Casper: awesome... but no, I am suckful with SMPS, I only did my first "real" one this week
[22:56:32] <inflex> but it is getting 70%+
[22:57:20] <Casper> I'm suspecting that the problem is the transistors
[22:57:55] <inflex> what's your schem?
[22:58:13] <inflex> oh, the MSP430 is a fantastic chip btw... worth getting yourself one of their $6 dev kits
[22:58:26] <Landon> good for christmas presents too :>
[22:58:37] <Landon> (to appropriate recipients)
[22:58:44] <Casper> inflex: I don't have one, but it's basically a push pull based on SG3524
[22:59:00] <inflex> http://au.element14.com/texas-instruments/msp-exp430g2/kit-dev-msp430-launchpad/dp/185379303
[22:59:01] <Casper> almost straight from the datasheet
[22:59:52] <Casper> gonna try to put another 1k on the input and see if the current drop...
[23:00:16] <inflex> ok
[23:00:24] <inflex> just trying to pull up the datasheet now
[23:01:07] <inflex> anything on the board getting hot?
[23:02:47] <inflex> Casper: is that the Fig15 one?
[23:03:55] <inflex> Casper: wonder why they don't use MOSFETs these days?
[23:09:30] <Casper> push pull transformer coupled circuit
[23:09:33] <Casper> texas instrument
[23:10:14] <Casper> transformer is a toroidal core I had, removed the winding and put a new one in, 20-20:20-20 turns
[23:10:37] <Casper> I removed the output inductor, somehow it was making the current shoot up AND make the output drop alot in power
[23:10:55] <Casper> but the rest is very simmilar
[23:11:33] <Casper> output diode is from a powersupply
[23:11:53] <Casper> (but didn't tested if one was short...)
[23:22:48] <Casper> inflex: that is an old part
[23:23:04] <Casper> the datasheet is from 1977, that's why
[23:23:22] <Casper> but I'm debating to change the transistor for some nchan fet
[23:24:21] <Casper> if I remove the 100 ohms and put some 2k... should be an easy swap...
[23:24:29] <Casper> atleast to try...
[23:24:50] <geckosenator> if I access memory fro 0x00 to 0x20 is that the actual registers?
[23:25:33] <Casper> yes, afaik
[23:27:42] <geckosenator> yes to mine?
[23:41:02] <Casper> yes to your
[23:41:14] <Casper> check the memory map section or the register
[23:41:19] <Casper> it tell you where stuff is
[23:42:00] <geckosenator> it doesnt list below 0x20 so i was wondering
[23:42:08] <Casper> hmm
[23:42:10] <geckosenator> just wondering what's down there
[23:42:34] <geckosenator> the code in svn gcc outputs values outside the range for sbi and cbi
[23:42:41] <Casper> ah there
[23:42:53] <geckosenator> where?
[23:42:55] <Casper> 0-1F is the 32 registers
[23:43:00] <geckosenator> ah ok
[23:43:14] <Casper> in the atmega88 it's page 18
[23:43:19] <Casper> 7.3 SRAM Data Memory
[23:43:30] <Casper> Data Memory
[23:43:31] <Casper> 32 Registers 0x0000 - 0x001F
[23:43:31] <Casper> 64 I/O Registers 0x0020 - 0x005F
[23:43:31] <Casper> 160 Ext I/O Reg. 0x0060 - 0x00FF
[23:43:41] <Casper> then it's the internal sram
[23:44:21] <geckosenator> apparently I still can't read datasheets :-/
[23:44:27] <Casper> apparently yeah :D
[23:44:28] <geckosenator> anyway that is good to know
[23:44:47] <geckosenator> I was at the very end of the datasheet looking at the table
[23:47:11] * inflex returns
[23:47:24] <inflex> geckosenator: any crazy idea why that compiler issue has cropped up?
[23:47:37] <geckosenator> yes
[23:47:43] <inflex> oh okay?
[23:47:47] <geckosenator> because they just added a whole ton of optimizations to avr-gcc
[23:47:48] <Casper> inflex: still trying to understand my smps issues
[23:47:58] <geckosenator> and one of them is to emit sbi and cbi instructions
[23:48:08] <geckosenator> which before it could just do lds or whatever
[23:48:22] <geckosenator> this is new code heh
[23:48:25] <inflex> Casper: could well be the transformer
[23:48:33] <inflex> Casper: esp if nothing else is heating up
[23:48:36] <geckosenator> it also emits in and out instructions
[23:48:44] <geckosenator> which are also brokeen
[23:48:45] <Casper> inflex: what heat up is the 2 switch transistor
[23:48:46] <inflex> geckosenator: ugh, okay, explains why it works on the old system and not this "up to date" one
[23:48:58] <inflex> Casper: mmm... not being turned on hard enough?
[23:49:00] <geckosenator> you have problems too?
[23:49:20] <geckosenator> I am using gcc 4.7.0 I think 4.4 and earlier is ok
[23:49:32] <inflex> geckosenator: well, basically it's breaking all the code I have which use that -f flag, so I'll just drop the flag at least
[23:49:48] <inflex> 4.5.3 is the one that's screaming at me
[23:49:53] <geckosenator> but does it generate sbi instructions with optimizations on?
[23:50:20] <geckosenator> like DDRB |= 1
[23:50:33] <geckosenator> if that is the only line in your main, and you compile with -S to generate assembly
[23:50:34] <inflex> dunno, haven't checked the asm
[23:51:26] <Casper> inflex: doubling the 1k resistor (so 500 ohms) lead to zero change
[23:51:32] <geckosenator> compile a file with that in it, then put -c file.c -S -o - -O
[23:51:57] <geckosenator> and see if you have sbi 36, 1 (clearly invalid)
[23:54:21] <geckosenator> come on heh
