#avr | Logs for 2012-02-26

Back
[00:36:36] <CapnKernel> rue_house: That's my hope :-)
[00:39:51] <Roklobsta> how's the shenzen food?
[00:42:24] <CapnKernel> I struggle.
[00:42:44] <CapnKernel> I like the cuisine of North-eastern China, and from Sichuan.
[00:43:26] <CapnKernel> Southern chinese cuisine reminds me of the pigslops we used to cook on the farm. An oily boilup of unidentifiable muck, and it tastes as good as it sounds.
[00:43:58] <CapnKernel> On my first tour of duty here (Summer last year) I lost 7kg.
[00:44:25] <CapnKernel> capn.blogspot.com/2012/01/cravings.html
[00:44:29] <CapnKernel> http://capn.blogspot.com/2012/01/cravings.html
[01:16:21] <bwright> If I am just getting started with AVR what documentation is a must read? (Several years programming in C, C++, Haskell but little experience with asm in general)
[01:18:21] <Casper> bwright: the 300-600 pages datasheet
[01:18:24] <Casper> and...
[01:18:38] <Casper> http://www.nongnu.org/avr-libc/user-manual/modules.html
[01:19:40] <Casper> and keep in mind that you have limited ressources avail... so avoid float
[01:20:10] <bwright> Oh this sounds fun ^_^
[01:20:37] <Casper> float is ram and cpu hungry
[01:21:18] <Casper> the avr can multiply in dardware, but not divide... so dividion is cpu intensive
[01:22:41] <bwright> Alright so try avoid floats and division.
[01:28:17] <Casper> also, don't forget to minimise your ram usage.... contrary to pc... a single byte is important...
[02:09:44] <Valen> of your using malloc your doing it wrong ;-P
[02:14:46] <jacekowski> that's not exactly precise
[02:14:59] <jacekowski> division can be replaced with multiplication in some cases
[02:18:14] <Casper> some case, some
[02:18:24] <Casper> but still, to be avoided if possible
[02:18:35] <Casper> or used in moderation
[02:21:27] <Roklobsta> capn: hmmm, so you got worms. nice.
[04:08:48] <RikusW> sleepy sunday time again ;)
[04:20:01] <norbi> guys i have a question, can be spi enabled or disabled on the fly?
[04:20:24] <norbi> like if(sensor){enable spi}else{disable} .. ?
[04:25:37] <RikusW> yes
[04:25:54] <RikusW> just turn on/off the register bits for spi
[04:29:05] <norbi> RikusW: im asking because i want to demonstrate something at school :D
[04:34:55] <RikusW> norbi SPCR - SPE
[04:35:20] <RikusW> have a look at the MSTR bit too
[04:41:18] <norbi> RikusW: i wonder if spi configuration can be saved in eeprom
[04:42:20] <norbi> RikusW: i would like to save configuration of "devices" in eeprom then the bootloader from flash to read them
[04:43:25] <norbi> and then to be able to programm the flash with personalized software (without affecting the eeprom and the flash bootloader)
[04:44:10] <norbi> bootloader will handle device configurations and never need to deal in software to deal with them, just using them
[04:44:33] <norbi> RikusW: is that possible?
[04:50:18] <RikusW> if you write a piece of firmware to do it
[04:50:43] <RikusW> it will need to be included in the bootloader or the actual firmware
[04:51:19] <norbi> the bootloader uses space of internal programm memory right? there is no specific boot sector
[04:51:20] <norbi> ?
[04:52:37] <RikusW> there is on most avr's
[04:52:39] <norbi> RikusW: or bootloader ususally goes in EEPROM?
[04:52:50] <RikusW> if goes in high flash
[04:53:09] <RikusW> an can be protected with lockbits
[04:53:44] <norbi> so bootloader can go or in eeprom or in flash?
[04:53:52] <RikusW> only flash
[04:54:03] <RikusW> eeprom is for data only
[04:54:17] <norbi> after loading the bootloader in flash, you can use lockbits to lock that specific code
[04:54:31] <RikusW> only the bootloader pat yes
[04:54:46] <RikusW> to protect it from being written to
[04:54:53] <RikusW> usually lock = 0xEF
[04:55:16] <norbi> after that you can reprogramm the flash any times, the bootloader remains and the rest is rewrited?
[04:55:40] <RikusW> bootloader rewrites page by page
[04:55:46] <RikusW> no chip erase
[04:56:23] <norbi> bootloader doesnt necesarily mean that you can programm the mcu via it, it can mean anything that needs to be done before running the actual programm
[04:56:53] <norbi> ?
[04:56:58] <RikusW> if you put in code to access IO you can
[04:58:23] <norbi> RikusW: the self programm is just an instruction that will handle the input from USART and put that in the flash?
[04:58:41] <RikusW> read up on the SPM instruction
[04:58:51] <RikusW> and the Bootloader part in the datasheet
[04:58:57] <RikusW> it tells you how
[05:00:35] <RikusW> norbi: http://www.ruemohr.org/code/BLM8.zip
[05:00:43] <RikusW> will work on m64 too
[05:00:47] <RikusW> or it should
[05:01:09] <RikusW> you might need to modify it a little, particulary the USART code
[05:01:20] <RikusW> its in asm
[05:01:28] <RikusW> and use stk500_v2 protocol
[05:01:30] <dehuman> how big?
[05:01:36] <RikusW> 1kb
[05:01:44] <RikusW> afaik
[05:01:46] <dehuman> thats kinda big
[05:02:04] <RikusW> but it works quite well
[05:02:13] <dehuman> there is the one peter fleury wrote too thats like 1k or 512bytes
[05:02:23] <dehuman> optiboot for arduino is 512 bytes
[05:02:33] <dehuman> but it will work with all types of stuff
[05:02:42] <dehuman> dunno how much of it is just arduino
[05:02:44] <RikusW> my USB CDC one is 2kb
[05:03:00] <RikusW> basically the same stk500 code + CDC
[05:04:07] <RikusW> made arduino work with mine
[05:04:19] <RikusW> just changed stk500 to stk500_v2 in the config files
[05:04:27] <norbi> RikusW: im not yet very good in asm, and dunno yet the avr instruction set
[05:04:28] <RikusW> after all arduino use avrdude
[05:04:46] <norbi> is there a way ti set the compiler to reduce size of the code?
[05:04:56] <norbi> to set*
[05:05:07] <RikusW> asm don't get smaller
[05:05:14] <norbi> yea i know
[05:05:17] <RikusW> with gcc use -Os
[05:05:21] <norbi> but i will write it in C
[05:05:41] <norbi> i need some special functions in my bootloader
[05:05:50] * RikusW is fluent in avr asm :)
[05:05:59] <RikusW> like ?
[05:06:08] * norbi is not even fluent in asm
[05:06:41] <norbi> RikusW: like to autoconfigure itself
[05:06:51] <norbi> based on USART instructions
[05:07:09] <RikusW> autoconfigure the IO registers ?
[05:08:06] <RikusW> thats fairly simple, pass it the register adderss and value
[05:08:13] <RikusW> *address
[05:08:58] <RikusW> and use a pointer in the bootloader to write to any memory address
[05:12:00] <ziph> Haha.
[05:12:02] <norbi> ram is automatically used by the avr architecture?
[05:12:04] <ziph> That's a bad sign.
[05:12:17] <ziph> A LED on a completely unpowered board is turning on.
[05:12:22] <norbi> i never adressed data memory directly
[05:14:50] <RikusW> registers is mapped at 0
[05:14:54] <RikusW> and IO at 0x20
[05:15:33] <RikusW> extended IO at 0x60
[05:15:44] <RikusW> ram starts at 0x100 on m64
[05:15:56] <RikusW> and 0x60 is there is no extended io
[05:17:23] <ziph> Wee.
[05:18:08] <ziph> The 4 inch JTAG cable is enough to cause a reset when there's a transmitter sitting on the JTAG adapter.
[05:19:45] <norbi> RikusW: what is the instruction for self programming?
[05:23:35] <norbi> there is an issue in my mind
[05:24:00] <norbi> how do i access bootloader after programed the flash?
[05:24:24] <norbi> i mean how do i comunicate with it?
[05:25:57] <norbi> is there needed a time count down or something?
[05:26:19] <norbi> if you dont connect to it in 5 seconds via usart, then it will jump to main program?
[05:33:27] <RikusW> you gan give either a timeout like arduino, or a command
[05:33:28] <RikusW> have to go
[06:32:55] <OndraSter> anyone building pick&place machine?
[06:32:55] <OndraSter> http://www.ebay.com/itm/5-way-SMT-SMD-Component-Feeder-Tape-and-Reel-Stoarge-/270895753220?pt=LH_DefaultDomain_0&hash=item3f12a52804
[06:32:59] <OndraSter> these are fairly cheap
[06:33:04] <OndraSter> just add some kind of motor :)
[06:35:32] <specing> OndraSter: LOL, we are hobbyists, not PCB manufacturers...
[06:35:37] <OndraSter> :P
[06:35:49] <ziph> There are crazy hobbyists building PnP's.
[06:36:10] <ziph> I'm not sure why, a paste robot or CNC would be more useful for a hobbyist.
[06:36:11] <specing> >_>
[06:36:44] <ziph> But then prototype stencils are so cheap now anyhow.
[06:38:17] <specing> OndraSter: Oh, and 40$ is not cheap
[06:38:25] <OndraSter> considering how much cost regular ones
[06:38:26] <OndraSter> it is
[07:00:27] <OndraSter> EPIC FAIL
[07:00:32] <OndraSter> I put the descriptions wrong on the board
[07:00:36] <OndraSter> and thus I connected all LEDs wrong
[07:00:42] <OndraSter> wrong = red vs green
[07:00:47] <OndraSter> so when red should lit up, green one lits up lol
[07:14:31] <BusError> -fdata-sections \
[07:59:42] <OndraSter> https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!416&parid=40BF7833586103AB!111&authkey=!AF_D9-3nlQUiB0A
[07:59:49] <OndraSter> the mega in bb -- not used
[08:12:06] <OndraSter> lucky me, on the last minute I moved Display Enable signal from some PortG pin to PB7
[08:12:12] <OndraSter> which happens to be also output compare pin
[08:12:31] <OndraSter> I can drive brightness!
[08:13:16] <specing> PORTG?
[08:13:27] <specing> Are you using some massive tqfp-100 monster?
[08:14:03] <abcminiuser> specing, the port names can be a bit wacky
[08:14:09] <abcminiuser> PORTR, for instance, on some XMEGAs
[08:15:12] <specing> Well now that you work at ATMEL, you can ensure that future chips have sane PORT labeling >_>
[08:16:53] <OndraSter> specing, TQFP64 monster :)
[08:17:13] <OndraSter> 45 IO pins
[08:17:16] <specing> close enaugh
[08:17:21] <OndraSter> 20 of them are used just for external memory
[08:17:26] <OndraSter> so it is not _that much_ IOs anymore
[08:17:28] <specing> Niiiiicee
[08:17:42] <specing> How much external memory do you have there?
[08:17:50] <OndraSter> 128kB
[08:17:53] <OndraSter> two banks of 64kB
[08:18:05] <OndraSter> (64kB SRAMs were way more expensive than 128kB one lol)
[08:18:08] <specing> kilo bytes or kilo bits?
[08:18:27] <mrfrenzy_> he did say Bytes, not bits
[08:18:51] <specing> mrfrenzy_: labeling can be misleading
[08:19:16] <mrfrenzy_> I just assume it's correct, and when stuff goes wrong I blame the guy using incorrect abbreviations
[08:20:51] <OndraSter> I say what I mean
[08:20:53] <OndraSter> and I mean kilobytes
[08:20:55] <OndraSter> thus I used kB
[08:21:10] <mrfrenzy_> thanks
[08:21:17] * mrfrenzy_ gives OndraSter a gold star
[08:21:17] <OndraSter> 128Kb wouldn't be too much really
[08:21:26] <OndraSter> 8kB... taking 20 IO lines?
[08:21:35] <OndraSter> actually 19, I counted one signal twice
[08:21:43] <OndraSter> I take it back
[08:21:45] <OndraSter> 20 :P
[08:22:05] <OndraSter> 8 AL + D, 8 AH, 1 ALE, 1 WR, 1RD, 1 Bank switch (or A16 if you want)
[08:22:19] <OndraSter> AL = address low, D = data, AH = address high, ALE = Address Latch Enable
[08:22:23] <dirty_d> wow, this dfll auto calibration works really good
[08:22:50] <specing> OndraSter: what is the latency for accessing that memory?
[08:23:14] <OndraSter> 4 clocks if you have fast memory and don't need any wait states
[08:23:26] <OndraSter> internal memory is 3 clocks "only" I think
[08:23:38] <OndraSter> or was it 2 clocks?
[08:23:50] <OndraSter> simply said, you won't be using xmem for stack
[08:23:58] <OndraSter> stack will be in internal (4kB) memory
[08:29:39] <specing> 2 for internal
[08:32:04] <Kevin`> you can of course use it as stack if there's a need
[08:32:34] <OndraSter> you can
[08:32:38] <OndraSter> but you use stack too often
[08:32:43] <OndraSter> and external cycles would kill perfomance
[08:40:42] <specing> And usually you inline everything so there isn't much use for the stack other than interrupts
[08:41:25] <specing> So you can assume the stack will not grow more than e.g. 10 bytes
[08:41:28] <OndraSter> inline
[08:41:32] <OndraSter> I am doing all in ASM :P
[08:41:56] <specing> OndraSter: I do interrupts in ASM
[09:16:14] <Jan-> hihi
[09:16:15] * Jan- waves
[09:18:50] <OndraSter> hey
[09:19:00] <Jan-> how're you, OndraSter
[09:19:08] <OndraSter> still the same
[09:19:15] <OndraSter> I finally lit up first bunch of LEDs :P
[09:25:51] <Jan-> what's the project?
[09:29:09] * Jan- pokes OndraSter
[09:29:11] <Jan-> what are you trying to make?
[09:29:34] <mrfrenzy_> how goes your hdmi project Jan- ?
[09:30:23] <Jan-> ha ha
[09:30:25] <Jan-> very funny
[09:30:35] <mrfrenzy_> I think it's actually possible
[09:30:42] <Jan-> well it's certainly possible.
[09:30:46] <Jan-> it's just very very very very complicated.
[09:30:46] <mrfrenzy_> found a chip that takes hdmi in and outputs composite
[09:30:57] <mrfrenzy_> take that composite into a micro or fpga and output to your lcd
[09:31:00] <Jan-> Well let me update you
[09:31:16] <Jan-> the latest ideas is: can we remove the TFT display from the back of the camera and put it on a longer cable.
[09:31:40] <Jan-> And actually maybe we can, because it's in one of those tilt-and-swivel mounts, so it must have a reasonably compact flexible cable at some point to go through the hinge.
[09:31:50] <mrfrenzy_> how long cable?
[09:31:59] <mrfrenzy_> it might not work more than half a meter or so
[09:32:00] <Jan-> Six to eight inches.
[09:32:04] <mrfrenzy_> could work
[09:32:07] <Jan-> Oh sure I know, capacitance etc
[09:32:20] <Jan-> look up "panasonic lumix DMC-GH2"
[09:32:25] <Jan-> it's that camera, I've got one right here
[09:33:30] <mrfrenzy_> aah that's a nice camera
[09:33:46] <Jan-> what it is, is a very very cheap way to shoot surprisingly good HD video.
[09:33:49] <mrfrenzy_> I don't like taking such stuff apart though, easy to kill the IP rating
[09:34:06] <OndraSter> Jan-, 32x48 LED display :)
[09:34:07] <OndraSter> bi-color
[09:34:16] <Jan-> Yeah, but on the other hand, the TFT module on this thing really is almost separate.
[09:34:17] <OndraSter> I have pix of one row lit up, but I know that it wouldn't do anything for you
[09:34:30] <Jan-> OndraSter: it's OK, send me a link, my friend Katie is here.
[09:34:40] <OndraSter> ok
[09:34:41] <OndraSter> Jan-, https://skydrive.live.com/redir.aspx?cid=40bf7833586103ab&resid=40BF7833586103AB!416&parid=40BF7833586103AB!111&authkey=!AF_D9-3nlQUiB0A
[09:34:42] <OndraSter> here
[09:34:44] <keenerd> Any pro-level camera should have a monitor out.
[09:34:59] <Jan-> keenerd: Well, there's a query over "pro level" here :)
[09:35:15] <Jan-> It does have HDMI out, but it's difficult or impossible to get a 3" TFT monitor without paying your first born child for it.
[09:35:52] <keenerd> No, but you can get 6" easily.
[09:36:15] <keenerd> 3" is actually really common to, check any pocket TV.
[09:36:15] <Jan-> Really we want something like a normal broadcast camera's viewfinder.
[09:36:33] <Jan-> look up Panasonic AG-HPX300
[09:36:40] <Jan-> that's what the person who's doing this is used to operating
[09:36:54] <Jan-> the mission is: make the GH2 as much like that as possible.
[09:36:59] <mrfrenzy_> Jan-: "TDA19977 converts HDMI streams into RGB or YCbCr". it's only $15
[09:37:45] <keenerd> Not to be insentive, but there is a huge difference bewteen an optical viewfinder, a digital viewfinder and a normal TTF.
[09:37:47] <Jan-> $15, and a 900-pin Tiny Eency-Meency Little Sub-Micro Quad Flat Package.
[09:37:54] <Tom_itx> so?
[09:38:08] <Jan-> keenerd: Usually broadcast viewfinders are CRTs. Yes, still, seriously.
[09:38:15] <Tom_itx> more pins, more soldering practice
[09:38:20] <OndraSter> hehe
[09:38:25] <OndraSter> BGA with 0.00nothing pitch ftw!
[09:38:27] <OndraSter> at home
[09:38:32] <Jan-> Some extremely high end ones are mono TFTs, which are LED backlit with RGB lights.
[09:38:38] <keenerd> Jan-: They are also black&white, with infinite resolution.
[09:38:43] <Jan-> Sure.
[09:38:50] <Jan-> We have several here.
[09:39:00] * Tom_itx greets abcminiuser in the kindest downhome fashion
[09:39:03] <Jan-> Most of the HD ones want component analog drive, though.
[09:39:11] <Jan-> So we're sort of back to the same problem.
[09:39:14] <abcminiuser> Greetings, professor Falken.
[09:39:26] <Tom_itx> what gives today?
[09:39:31] * abcminiuser is currently up to his arms in code
[09:39:32] <keenerd> Jan-: And if you want to use an LCD in the viewfinder, they use one of those color wheel types so that the RBG overlaps.
[09:39:53] <Jan-> keenerd: sometimes they do, with RGB LEDs like I said, but there's upsides and downsides to those too.
[09:39:53] <abcminiuser> Ripping out some old workarounds in LUFA for the AVR8 arch, so I can move forwards with the XMEGAs
[09:39:55] * Tom_itx thinks abcminiuser should get a LUFA tatoo
[09:39:59] <abcminiuser> Not a quik job :P
[09:40:12] <ziph> abcminiuser: At work
[09:40:13] <ziph> ?
[09:40:17] <abcminiuser> ziph, home
[09:40:20] <abcminiuser> It's Sunday
[09:40:28] <Jan-> Actually the best way we thought of doing it, which is a bit out of price range, is to repurpose one of those pocket DLP projectors into a back projection viewfinder.
[09:40:29] <ziph> Good point. Can you tell I'm self employed? :)
[09:40:34] <keenerd> Jan-: Just saying, it would be easier and get better results with the highest res screen you can get hooked up to the hdmi out.
[09:40:35] <Jan-> It would have DLP performance.
[09:40:59] <Tom_itx> at least dean's on the same day which is kinda out of fashion for an aussie
[09:41:10] <ziph> Pft, LQFP144, that should take 5 minutes with a blunt soldering iron Jan.
[09:41:11] <Jan-> keenerd: Sure, we can already do that, that's trivial, the purpose of this is to create a loupe viewfinder like a broadcast one, which you can use for operating off the shoulder.
[09:41:29] <Jan-> This is really an exercise in ergonomics, not maximum picture quality.
[09:41:59] <Jan-> The current Plan B is to use an HDMI to composite converter and just feed a standard-def CRT viewfinder with that, so that's not going to give us HD pictures :)
[09:42:47] <ziph> That chip (TDA19977B) looks fun.
[09:43:06] <Jan-> The big issue - for anyone - is that nobody makes 1920x1080 TFTs that are only 2" across, because the technology to do that doesn't exist.
[09:43:08] <keenerd> If you are willing to accept low-fi, $200 gets you standard def video glasses.
[09:43:40] <keenerd> (Samsung makes that res in a 6" screen.)
[09:43:43] <Jan-> Well if you go standard def, the display is easy
[09:43:51] <Jan-> getting from HDMI to composite is less easy
[09:44:05] <keenerd> $50 active converter, easy.
[09:44:14] <Jan-> sure, there are cheap converters out there, but HDMI is a bit of a beast, and there can be compatibility issues.
[09:44:30] <keenerd> Won't know till you try :-)
[09:44:38] <Jan-> we're uncomfortably aware of that :/
[09:44:45] <ziph> What does RGB 4:4:4 actually mean?
[09:45:01] <Jan-> Happily, the camera will only produce hdmi displays that are somewhat translatable to standard def modes without having to do any doubling of frames or whatever.
[09:45:06] <Jan-> But even so it's still complex.
[09:45:34] <OndraSter> 4bits per subpixel ziph
[09:45:35] <Jan-> a:b:c notation is ratios of YUV (actually YCrCb) components.
[09:45:48] <OndraSter> oh
[09:45:55] <Jan-> well it's not necessarily about bits and pixels becuase it also strictly speaking applies to analog component video.
[09:45:57] <ziph> 4 bits of dynamic range doesn't make sense though.
[09:46:19] <ziph> So how many bits would a 4:4:4 pixel be, Jan- ?
[09:46:22] <Jan-> In component video, a luminance (Y) signal and two color difference red and blue signals are sent.
[09:46:44] <Jan-> It's done this way because the human eye sees brightness with greater sharpness than color, so the color difference signals can be sent with less bandwidth.
[09:46:52] <Jan-> This is the original design of analog color television.
[09:47:16] <Jan-> When digital came along, they decided to store the Y plane at full resolution, and often the Cr and Cb planes at half that resolution horizontally.
[09:47:23] <Jan-> That's 4:2:2
[09:47:32] <Jan-> for every four luma pixels, there's two Cr pixels and two Cb pixels.
[09:48:01] <Jan-> You can have 4:4:4 YCrCb, in which case the color information is at full res. But for that, people tend to use RGB, which is always 4:4:4
[09:48:16] <Jan-> Make sense?
[09:48:23] <ziph> Why don't they call it 2:1:1 ? :)
[09:48:56] <Jan-> because there are also 4:1:1 modes.
[09:49:05] <Jan-> NTSC-format miniDV works like that
[09:49:27] <ziph> So how many bits is each channel?
[09:49:31] <Jan-> Doesn't matter
[09:49:38] <Jan-> Usually 8, or sometimes 10 for high end work
[09:49:42] <ziph> It can vary in HDMI?
[09:50:03] <Jan-> but it's still four (8 or 10 bit) luma pixels, then two (eight or ten bit) Cr and Cb pixels.
[09:50:21] <Jan-> HDMI can optionally support 8 or 10 bit, RGB or YCrCb, yes.
[09:50:30] <Jan-> All HDMI devices are required to support 8 bit RGB.
[09:50:33] <Jan-> All else is optional.
[09:50:45] <ziph> Hmm, 168 MHz on the video clock.
[09:50:51] <Jan-> yes HDMI is fast.
[09:51:00] <ziph> This is on the decoded side.
[09:51:07] <Jan-> remember DVI will go up to 1920x1080 at 60 frames per second
[09:51:11] <Jan-> that's fast
[09:51:25] <Jan-> (actually it will theoretically do even more)
[09:52:31] <ziph> A 30 bit bus at 165MHz must be a boxful of weasel level fun.
[09:54:11] <Jan-> designing HDMI devices is non trivial
[09:54:14] <Jan-> Tom_itx will of course disagree
[09:54:17] <Jan-> but then, that's Tom_itx :)
[09:54:28] <ziph> I'm talking about the parallel end, not the HDMI side
[09:54:47] <Jan-> well, I guess it'd be sort of parallel to drive a TFT
[09:55:01] <Jan-> hdmi is three pairs of tmds data, twisted pair
[09:55:25] <ziph> TFT's can just take a parallel input with the four sync signals?
[09:55:39] <Jan-> every tft on the market has a different input signal specification
[09:55:43] <Jan-> it's entirely in the lap of the manufacturers
[09:55:50] <Jan-> I'm talking about raw TFT panels here, not finished displays.
[09:56:21] <Jan-> the hdmi signal format itself is not terribly complicated but it is quick.
[09:56:42] <Jan-> 1.6 ish gigabits in 8-in-10 format, differentially signalled
[09:56:56] <Jan-> I guess you could use one of the Xilinx Spartan series FPGAs, which have serial transceivers on them, to implement it.
[09:58:27] <Jan-> but really all this is a pipe dream
[09:58:34] <Jan-> you'd need a fully equipped lab to actually do any of this.
[09:59:49] <ziph> Does HDMI have a delay in it for the back porch?
[09:59:59] <Jan-> I'm not sure.
[10:00:21] <Jan-> I have a feeling a lot of it is based on analog style signal timings, so maybe
[10:00:36] <Jan-> also that may be one of those tricky things that is sometimes done, and sometimes not done.
[10:00:42] <Jan-> hence the compatibility dance.
[10:13:43] <dirty_d> damn, i cant find a pot
[10:13:48] <Jan-> hmm even phone screens have 800x480 displays these days
[10:13:53] <Jan-> barely more than standard definition
[10:19:17] <dirty_d> i think this code is pretty much done!
[10:19:29] <dirty_d> 1200 lines and 7k
[10:19:40] <dirty_d> not bad figured i woulda needed more flash
[10:20:07] <d0ctor_jon> Ah, a whole room for AVR chat, fantastic. :)
[10:20:10] <d0ctor_jon> Hi folks.
[10:20:23] <dirty_d> hi
[10:20:40] <OndraSter> eya
[10:21:47] <d0ctor_jon> I've been looking at (after many years of saying I will) starting an AVR project from scratch. :)
[10:22:00] <dirty_d> go for it, its easier than you think
[10:22:28] <d0ctor_jon> Well, the electronic side of it is quite new to me, so I'm trying to gather some simple project circuits.
[10:22:51] <d0ctor_jon> It's nice to know there are people to chat to as well.
[10:22:56] <dirty_d> whats the project?
[10:23:02] <d0ctor_jon> No idea yet.
[10:23:06] <d0ctor_jon> I'm thinking about that now.
[10:23:33] <dirty_d> d0ctor_jon, there is also #electronics
[10:24:13] <d0ctor_jon> I think, ideally, I'd like to eventually write a simple OS for an AVR based board.
[10:24:15] <Tom_itx> #electronics will get you little
[10:24:20] <keenerd> If I might recommend a product, http://www.pjrc.com/teensy/ is great for a first AVR. The programmer is built in and you don't have to fuss with usb/serial adapters lie Arduino.
[10:24:44] <Tom_itx> d0ctor_jon there are several rtos's out already
[10:24:47] <keenerd> *like
[10:24:55] <dirty_d> d0ctor_jon, avrs are simple enough that an os doesnt really make anything easier
[10:25:02] <d0ctor_jon> That looks good keenerd.
[10:25:08] <keenerd> Tom_itx: Practice is never progress. People learn through reinvention.
[10:25:12] <d0ctor_jon> It's more for my benefit guys, the OS that is.
[10:25:21] <d0ctor_jon> Though, I appreciate that there is little to gain.
[10:25:27] <d0ctor_jon> More...experience.
[10:25:30] <dirty_d> yea
[10:26:01] <keenerd> Still, it bugs me when someone puts down a perfectly good learning experience.
[10:26:24] <d0ctor_jon> That Teensy looks excellent for the money.
[10:27:03] <keenerd> If you are just getting one, it is. More than a few and a regular AVR + programmer is much cheaper.
[10:27:04] <Kevin`> the teensy are nice. fast usb interface and you can stick it into a breadboard
[10:27:36] <d0ctor_jon> Yeah, seems to be plenty of I/O pins too.
[10:27:39] <d0ctor_jon> Enough for me anyway. :)
[10:28:14] <dirty_d> i cant find a single damn piece of junk with a potentiometer in it to salvage
[10:29:46] <d0ctor_jon> Ooh, I'm excited now! I'll be buying one of those Teensy++'s.
[10:31:53] <Tom_itx> keenerd i agree. i would learn how to swim before i jumped off the ship though
[10:32:53] <Tom_itx> once he learns a bit he may realize he didn't really need an os
[10:33:02] <keenerd> So?
[10:33:24] * RikusW was just watching powerlines short in the wind, nice long purple sparks...
[10:33:37] <Tom_itx> RikusW, what whacked you in the head?
[10:33:45] <Tom_itx> bar fight?
[10:33:50] <RikusW> horse
[10:33:52] <keenerd> Tom_itx: A midrange AVR with a few megs of external flash is very close in performance to the PDP11 that unix was written on.
[10:33:52] <Kevin`> RikusW: neat. preferably watching with glasses for less UV.
[10:34:01] <RikusW> was cutting its front hoof
[10:34:08] <Tom_itx> oops
[10:34:08] <keenerd> Tom_itx: There is plenty of room and reason for an OS.
[10:34:18] <Tom_itx> not as a first project
[10:34:19] <RikusW> and it decided to try jumping
[10:34:21] <Tom_itx> but meh
[10:34:30] <Tom_itx> if he want's to i'm all for him
[10:34:32] <RikusW> so the hind hoof got me on the forehead...
[10:34:35] <RikusW> ouch...
[10:35:10] <Kevin`> keenerd: rtos is useful for some projects, but for most of them it would just be a waste of time to integrate
[10:35:13] <Tom_itx> keenerd, i've just seen so many ambitious projects that never got done
[10:35:30] <keenerd> Finish projects is a different skill all on its own ;-)
[10:35:34] <keenerd> *finishing
[10:35:38] <Tom_itx> :)
[10:35:48] * abcminiuser facepalm
[10:35:55] <dirty_d> so true, lol
[10:35:57] <Tom_itx> hehe
[10:36:10] <abcminiuser> GF got a zero-day today, cleaned it up kinda, but told her to be careful for the moment in what she does with it
[10:36:23] <abcminiuser> Not 2 hours later she buys something with her credit card on it
[10:36:34] <abcminiuser> Now she gets another dropper virus warning
[10:36:38] * abcminiuser *sigh*
[10:36:43] <ziph> :)
[10:36:50] <specing> zero-day?
[10:36:56] <Tom_itx> weldome to the better half
[10:37:02] <Tom_itx> welcome*
[10:37:06] <keenerd> abcminiuser: So next time take windows away...
[10:37:10] <Jan-> FIVE HUN... I mean, five hundred and forty-nine dollars, for a TFT display?!
[10:37:13] * Jan- waves her arms
[10:37:17] <ziph> abcminiuser: Buy her a Macbook. ;)
[10:37:55] <abcminiuser> Etch-a-Sketch is next
[10:37:58] <Jan-> so OK
[10:38:02] <abcminiuser> Honestly, I can understand being a little bad at tech
[10:38:11] <Jan-> the prices of lcd displays starts normal, then gets higher as they get bigger
[10:38:11] <abcminiuser> specing, zero-day exploit
[10:38:12] <Kevin`> a 0-dat (as odd as that would be for client-side viruses) wouldn't give you any virus warning
[10:38:15] <Jan-> but it also get shigher as they get smaller
[10:38:20] <Kevin`> 0-day*
[10:38:43] <abcminiuser> Kevin`, the original dropper was zero-day, the payload was known
[10:38:52] <ziph> abcminiuser: How's she finding Norway?
[10:39:02] <Kevin`> abcminiuser: what a lazy virus author :D
[10:39:15] <Kevin`> abcminiuser: although who knows what else the computer has too
[10:39:25] <specing> abcminiuser: wordpress?
[10:39:51] <RikusW> Kevin`: no glasses, but its fairly brief
[10:40:20] <abcminiuser> specing, some random Norwegian site
[10:40:43] <keenerd> abcminiuser: Mint is great for windows refugues. I've had one person think they were still using XP afterwards.
[10:41:01] <abcminiuser> Heh
[10:41:09] <abcminiuser> But she wants to play games
[10:41:14] <abcminiuser> Wait, all her games are DOSBox oldies anyway
[10:41:56] <specing> abcminiuser: Your GF doesen't really have a Ph.D in computer science, right?
[10:41:56] <ziph> Isn't she a bit young for those?
[10:42:01] <Jan-> you shouldn't make a non-computer person use linux.
[10:42:24] <abcminiuser> specing, no, but she has a PHD in breaking computers
[10:42:30] <Tom_itx> teach them young to love linux
[10:42:35] <specing> Ah, that makes a difference
[10:42:35] <Jan-> why
[10:42:38] <Kevin`> Jan-: it works fine, as long as there's someone setting the computer up for them like always
[10:42:39] <keenerd> Jan-: It has worked perfectly every time I've done it.
[10:42:42] <abcminiuser> Jan-, well, she used Ubuntu for a while, she only plays games and surfs the net
[10:42:43] <Jan-> teach them young to love being kicked in the ass
[10:42:56] <Tom_itx> let's not get an os war started today hmm?
[10:42:59] <Tom_itx> it's sunday
[10:43:03] <Tom_itx> we need a day off
[10:43:12] <specing> -The perfect day for OS wars!
[10:43:19] * specing brings in heavy artillery
[10:43:27] * Jan- has cruise missiles
[10:43:30] <dirty_d> i had my parents using archlinux for a couple years, lol
[10:43:33] <ziph> Linux is crap because it has a monolithic kernel.
[10:43:38] <dirty_d> they dont knoe ANYTHING about computers
[10:43:51] <abcminiuser> Hrm, that gives me an idea
[10:43:55] <abcminiuser> Perhaps I'll install Linux, and give her a XP VM
[10:43:55] <specing> ziph: Why is that crap?
[10:44:02] <keenerd> dirty_d: Yeah. The less people know, the less trouble they have with Linux.
[10:44:15] <dirty_d> yea, all they use is the borwser
[10:44:18] <specing> abcminiuser: Yes, and when it breaks, just run a new one :D
[10:44:20] <Kevin`> ziph: everything is crap for various reasons. for example, nobody uses the true microkernel designs, so there's no driver or software support ;p
[10:44:30] <keenerd> dirty_d: There is this weird inverted parabola, peaking at the "windows power user".
[10:44:37] <dirty_d> lol
[10:44:56] <RikusW> keenerd: probably because they arent used to windows yet ;)
[10:44:56] <Jan-> when someone says a TFT display has "1.5 million dots" I guess they're obfuscating and they mean it has 1.5 million including each RGB pixel,
[10:45:06] <Jan-> so the actual number of color pixels is 0.5 million
[10:45:08] <Jan-> right?
[10:45:09] <specing> abcminiuser: Or better yet: startwm() { cp freshwm.img thiswm.img; qemu thiswm.img; rm thiswm.img }
[10:45:20] <ziph> While I could go into a long story here about the relative merits of the two designs, suffice it to say that among the people who actually design operating systems, the debate is essentially over. Microkernels have won. The only real argument for monolithic systems was performance, and there is now enough evidence showing that microkernel systems can be just as fast as monolithic systems.
[10:45:57] <keenerd> ziph: Hybrids have won. No one uses a pure mono or pure micro kernel.
[10:45:58] <dirty_d> how did they solve that problem?
[10:46:29] <specing> ziph: They have?
[10:46:48] <Kevin`> specing: too slow, use snapshots instead (from lvm or from a disk image format)
[10:48:12] <specing> Kevin`: I haven't really went into vserver hosting yet so I don't quite know how it is done
[10:48:15] <dirty_d> or have her use a livecd and when she tells you she has a problem press the reset button and say, "what problem"
[10:48:20] <ziph> LINUX is a monolithic style system. This is a giant step back into the 1970s. That is like taking an existing, working C program and rewriting it in BASIC. To me, writing a monolithic system in 1991 is a truly poor idea.
[10:48:39] <specing> ziph: It works
[10:48:51] <keenerd> ziph: And it is not purely monolithic either.
[10:49:01] <specing> Indeed
[10:49:12] <keenerd> ziph: Also, are you just quoting something? It sounds familiar.
[10:49:32] <ziph> Yeah, you're all being trolled by a decades old Tanenbaum rant.
[10:49:40] <Roklobsta> ziph is channelling Tannenbaum
[10:49:49] <specing> xD
[10:50:27] <Roklobsta> avr is so 1970s. we should be using propellors
[10:50:39] <keenerd> We should be using GA144.
[10:50:45] <specing> ziph: Tell me how much is Tanenbaum's system used in practice, since it is obviously so good?
[10:51:07] <ziph> specing: It was a learning system.
[10:51:16] <Kevin`> specing: that's a dumb argument. look at how many people use windows :/
[10:51:22] <Jan-> windows is awsum
[10:51:29] <Roklobsta> yeah we learned monolithis is good
[10:51:30] <Jan-> I'm kidding
[10:51:30] <ziph> It's obsolete now because Universities don't teach students anything other than Java.
[10:51:32] <Jan-> Windows is ass
[10:51:37] <Jan-> It's just that linux is, er, asser.
[10:51:52] <ziph> And web pages.
[10:51:57] <Roklobsta> windows is the least arse of all arses.
[10:52:24] <ziph> Linux users are just Windows wannabes.
[10:52:30] <keenerd> Installing and updating software on Windows is the worst experience possible. Worse than having your fingernails pulled off.
[10:52:34] <Kevin`> sure sure, keep thinking that
[10:52:44] <ziph> That's why all their software internally is just a poor implementation of Windows concepts.
[10:52:52] <specing> Kevin`: What I meant was, why are Linux people not using Minix? And why did Linus create his own system instead of adding to Minix?
[10:53:01] <Roklobsta> mrs roklobsta will stop using her 5yo winxp when you pry it from her cold, dead hands.
[10:53:24] <Roklobsta> minix cost money and linus was broke
[10:53:31] <Kevin`> specing: linux people are using linux because it has more software. dunno on the second question, ask him! :)
[10:53:40] <Roklobsta> and the winters are long and cold so he needed something to do
[10:53:58] <ziph> People that talk about Linux on IRC channels are also sad and lonely.
[10:54:35] <specing> Kevin`: and then you ask yourself *why* does Linux have more software?
[10:54:47] <Kevin`> specing: because more people use it
[10:54:48] <Roklobsta> i doubt linux has more software
[10:55:09] <ziph> Because Linux users are pirates and don't mind misappropriating things like the GNU system.
[10:55:12] <specing> ziph: Im not sad and lonelines comes from the fact that my handling of people is very *ahem*
[10:56:01] <Roklobsta> yes avrs make more sense than people
[10:56:04] <keenerd> ziph: I want to say that listening to you has been fantastic. I am learning so much, and I hope I'll never have to use any of it :-)
[10:56:28] <ziph> I'm being a retarded troll until someone changes the topic, if you hadn't noticed.
[10:56:39] <keenerd> And you are doing a great job too!
[10:56:44] <ziph> Thanks!
[10:57:01] <ziph> Linux users also tend to have sexual deviances.
[10:57:42] <ziph> Tanenbaum proved it a few years subsequent to his discussion with Linus in a paper that was accepted in Nature.
[10:58:13] <Jan-> OK, now we want to make a DLP display device :)
[10:58:25] * Jan- is nothing if not ambitious
[10:58:36] <Kevin`> quite
[10:58:46] <Kevin`> i'd rather build the lcd from scratch ;p
[10:59:01] <Jan-> Well, TI can make a DLP chip that has the full 1920x1080 resolution and is under an inch across.
[10:59:05] <keenerd> I'd rather buy a real camera with detachalbe eyepeice.
[10:59:12] <Jan-> So it's theoretically possible
[10:59:21] <Jan-> I'm not sure anyone's actually used one of those in a viewfinder though.
[10:59:29] <ziph> Are you taking photographs or video?
[10:59:32] <Jan-> Video.
[11:00:33] <keenerd> Jan-: http://www.bhphotovideo.com/c/product/736853-REG/Zacuto_Z_FIND_EVF_Z_Finder_EVF.html
[11:00:41] <Jan-> Have you seen what those cost :)
[11:00:49] <keenerd> Jan-: There are probably cheaper ones out there.
[11:00:54] <Jan-> Only just.
[11:01:12] <Jan-> There's one company that's threatening to bring one out for $375, but they've been claiming they're about to do that for 18 months.
[11:01:20] <Jan-> At that price, yes, we'd just buy one :)
[11:01:25] <keenerd> Jan-: There are several for $250.
[11:01:33] <Jan-> Linky?
[11:02:22] <Jan-> the trick is that it sort of needs to have hdmi input.
[11:02:30] <keenerd> It is not meant for that camera, but it takes HMDI input http://www.amazon.com/Olympus-VF-2-Electronic-ViewFinder-Replacement/dp/B006TJXQ0C
[11:02:31] <Jan-> if it doesn't, then you have to go through a conversion process.
[11:02:49] <ziph> Can't you just get a composite video display?
[11:03:01] <Jan-> keenerd: that I'm fairly sure is an olympus-specific part, we did find that
[11:03:09] <Jan-> ziph: That's Plan B.
[11:03:28] <Jan-> It's plan B because it means we have to have an HDMI to composite converter, which is compatibility problem, costs money, consumes power, and saps quality.
[11:03:39] <keenerd> Jan-: I think you are correct.
[11:03:40] <ziph> Oh, the Canon's have composite out.
[11:03:54] <ziph> Buy a Canon and throw away whatever junk you have. ;)
[11:03:59] <Jan-> so does the panasonic, but - facepalm - it switches off when you go into record.
[11:04:15] <Jan-> there's several reasons for the panasonic, it has been thought about.
[11:04:49] <Jan-> What he keeps saying is "Just an iphone display with an hdmi input."
[11:04:56] <Jan-> Which I'm sure is possible, just complicated.
[11:05:14] <ziph> Wait, he just wants iPhone size?
[11:05:23] <ziph> What size is the LCD on the camera?
[11:06:53] <keenerd> Jan-: B&H has an 800x480 hdmi viewfinder for $529.
[11:07:17] <Jan-> is that the zacuto one
[11:07:28] <Jan-> ziph the one on the camera is about 3.5 inches diagonally
[11:07:32] <keenerd> Cineroid EVF 4Le
[11:07:41] <Jan-> keenerd: basically there's three
[11:07:45] <ziph> Wouldn't an iPhone only be a marginal increase?
[11:07:47] <Jan-> cineroid, zacuto, and the expensive one
[11:07:53] <keenerd> Heh.
[11:08:09] <Jan-> zacuto are the "coming soon" perpetual non-delivery artists.
[11:08:22] <Jan-> ziph: yes, but it would be on a cable and able to be placed in the right spot.
[11:08:26] <ziph> Ah.
[11:08:35] <Jan-> Google for "panasonic hpx-300" for where the viewfinder should go on a broadcast camera.
[11:08:50] <Jan-> this is about ergonomics not image quality
[11:09:04] <Jan-> the display on the back of the camera is probably only 960x540 which is barely more than standar def anyway.
[11:09:15] <ziph> Which camera was it again?
[11:10:17] <Jan-> our target camera is a panasonic dmc-gh2
[11:10:24] <Jan-> but really, "any HDMI source" is the target
[11:10:37] <Jan-> many many cameras have HDMI or DVI, and it's cheap to get an SDI to DVI converter for those that don't.
[11:11:11] <Jan-> I suspect what we'll end up doing is converting HDMI to composite, but man, what a sucky solution. It's actually more complex technically, it's just that it happens to be off the shelf.
[11:13:12] <ziph> Don't any of those cheap portable TV's have HDMI?
[11:13:26] <Jan-> Not that I've been able to find, but that's certainly one reasonable idea.
[11:13:29] <Jan-> Let me know if you find one :)
[11:13:43] <ziph> Or what about degutting one of those portable DVD players? Surely the new ones are HDMI internally.
[11:14:01] <Jan-> There's 1001 devices that have 2" ish displays on them
[11:14:06] <Jan-> all those flash cameras etc
[11:14:14] <Jan-> but I seriously doubt whether any of them use HDMI inside.
[11:15:07] <Jan-> what makes you think they would?
[11:15:12] <keenerd> Jan-: A quick search on alibiba finds a slew of 9" screens and a handful of 8.4" and 8".
[11:15:35] <keenerd> *alibaba
[11:15:38] <Jan-> keenerd: Yeah, that's easy to do. What we ideally wanted was sort of 2-3 inch range
[11:16:16] <Jan-> also you can almost never actually GET things that are on alibaba, unless you want 200k units in six months.
[11:16:17] <ziph> Jan-: Because the ones that aren't just one big giant ASIC might run HDMI between IC's?
[11:16:44] <Jan-> ziph: bit of a long shot isn't it
[11:16:58] <Jan-> I assume they WOULD all be one giant asic
[11:17:05] <ziph> Jan-: You sounded like you were at that stage. :)
[11:17:07] <keenerd> Jan-: You should ask the BenHeck forums. They would probably know.
[11:17:13] <Jan-> but even if not I don't see why they'd use HDMI internally
[11:17:23] <Jan-> hdmi is encoded for long distance runs
[11:17:27] <Jan-> it would be unnecessary complexity
[11:17:35] <Jan-> Ben Heck?
[11:18:04] <ziph> Jan-: What do they run HDCP over on PCB interconnects?
[11:18:27] <keenerd> Jan-: They take video game consoles and turn them into portables.
[11:18:46] <Jan-> Well, when they're down to the point of driving the TFT, I would assume it's whatever the TFT wants, which is usually some sort of horrible parallel thing.
[11:19:05] <keenerd> Jan-: They know everything about every small LCD on the market.
[11:19:06] <Jan-> keenerd: Ohhh Ben Heckendorn
[11:19:09] <ziph> Wouldn't that mean a screw driver an FPGA is all you need to scrape it then?
[11:19:21] <Jan-> ziph: isn't that inevitably the case
[11:20:12] <Jan-> a screw driver, an FPGA, a fully equipped electronics lab and a PhD in being stylish and wonderful.
[11:20:29] <ziph> Why would you need a lab?
[11:20:39] <ziph> Or a PhD? :)
[11:21:07] <Tom_itx> so you can hang a piece of paper on the wall
[11:21:42] <ziph> I thought it was more so that people would call you Dr. :)
[11:22:00] <Jan-> Tom_itx has a PhD in being stylish and wonderful :)
[11:22:22] <Jan-> if you must know, I did look into it
[11:22:27] <Jan-> TFT panel data sheets are impenetrable
[11:22:46] <Jan-> but mainly they seem to use various sorts of parallel interface, and you just squirt n-bit values at the thing in sequence.
[11:22:57] <Jan-> Just... quickly. Very very quicklu.
[11:23:16] <Jan-> So quickly you're looking at 10 revisions of an 8-layer board, just to get the timing issues sorted out,.
[11:23:22] <Jan-> and that's expensive
[11:24:21] <OndraSter> Adding brightness control for each row on my project? Settings two registers!
[11:24:24] <OndraSter> PWM ftw
[11:24:33] <Jan-> Olympus have made a vf3
[11:24:36] <Jan-> I wonder if it's any better
[11:25:47] <Jan-> ...no it's worse
[11:27:34] <keenerd> Sigh. So many ir receiver modules, all with range estimates. But different powers on their tx source, so everything has to be normalized by hand.
[11:29:14] <dirty_d> i cant find thin enough aluminum square tube for this quad
[11:29:25] <dirty_d> i guess 0.5"x0.125" will have to do
[11:37:04] <Jan-> keenerd: just so you know, the FAQ at benheck.com says "if it doesn't have a composite input already, don't even think about it" :D
[11:37:15] * Jan- thinks those guys are stuck in 1989
[11:38:02] <Jan-> there *are* commercial tft driver boards available
[11:38:08] <Jan-> but none of them seem to target really small displays.
[11:39:43] <keenerd> Find a game console that only has hdmi out. Claim you are doing something with that.
[11:41:47] <Jan-> oled displays seem to need about +4 and -4 volts to operate
[11:42:22] <kdub> can anyone recommend a $20-$40 avr programmer that will work out-of-the box with avrdude/linux?
[11:42:32] <Jan-> kdub: ask Tom_itx
[11:42:51] <dirty_d> kdub, toms
[11:43:37] <rue_house> kdub, I have a parallel port one, if you want usb may I suggest the one tom makes
[11:43:38] <dirty_d> kdub, http://tom-itx.dyndns.org:81/~webpage/commerce/commerce_index.php
[11:43:57] <Tom_itx> kdub, yes
[11:44:02] <kdub> hmm, thanks for the suggestion!
[11:44:12] <kdub> i never could get my avrdragon to work properly
[11:44:14] <Tom_itx> even supported here
[11:44:21] <rue_house> I'm happy to make you a parallel port one
[11:44:23] <Tom_itx> you should be able to
[11:44:24] <rue_house> really
[11:44:26] <rue_house> I can do it
[11:44:29] <OndraSter> AVR Dragon no workie on linux?
[11:44:31] <Jan-> Tom_itx is one of those "disturbing genius" Kinds of people
[11:44:36] <rue_house> lead time is just 2 weeks now
[11:44:36] <OndraSter> then I am happy to use Windoze even moar!
[11:44:44] <rue_house> I can get you any colour you want
[11:44:45] <Tom_itx> kdub, the dragon should work with studio or avrdude
[11:44:48] <ziph> Tom_itx: Is Jan your wife?
[11:44:48] <rue_house> red, black, blue
[11:44:56] <Rikus1> kdub, the newer avrdude supports the dragon
[11:45:02] <Tom_itx> ziph, annoying enough to be huh?
[11:45:03] <Rikus1> like 5.10
[11:45:06] <kdub> hmm, maybe i'll give it a try again
[11:45:09] <ziph> Tom_itx: Haha.
[11:45:15] * Jan- looks glum
[11:45:17] <kdub> it kept complaining about the firmware
[11:45:34] * Jan- isn't annoying
[11:45:39] * Jan- kicks a can, disconsolately
[11:45:51] <Tom_itx> kdub, try at least avdude 5.10 or higher
[11:46:02] <rue_house> Jan-, help me tell everyone that parallel port programmers are the cats meow
[11:46:24] <rue_house> they are awesome, no fiddly usb port and drivers and junk to have to deal with
[11:46:49] <rue_house> its a port on the machine, its operated directly, no hastle required
[11:46:54] <Jan-> errr but USB makes things quick and easy and birds sing and it's sunny and all is great
[11:47:05] * Jan- less-than-three USB
[11:47:29] <ziph> rue_house: Where do I plug one in?
[11:47:32] <kdub> thanks Tom_itx, i'll give it a try sometime today
[11:47:33] <OndraSter> rue_house, tell me where I can find LPT on my board?
[11:47:36] <rue_house> to your parallel port
[11:47:44] <ziph> rue_house: Where is that on my iMac?
[11:47:55] <Tom_itx> kdub, if no go, mine works for sure
[11:47:56] <rue_house> if they forgot to put on onboard, you need to get a pci card
[11:48:33] <rue_house> heh, nobody would seriously try to develop anything on an imac but an iphone app
[11:48:36] <OndraSter> http://www.brainboxes.com/product/items/px-157
[11:48:42] <OndraSter> holy s*it
[11:48:43] <ziph> So I'd get a PCI card to run something that's barely faster than RS232? :)
[11:48:43] <OndraSter> 57 gbp
[11:48:46] <Jan-> Aargh, you can buy iphone screens for $notmuch
[11:48:49] <Jan-> that's so frustrating
[11:49:02] <ziph> rue_house: I do everything on it.
[11:49:10] <Jan-> is there any technical info available for them
[11:50:18] <rue_house> no, apple keeps everything secret, and they obsolete it on a regular basis to keep their market fresh
[11:50:28] <Jan-> nuts
[11:50:50] <dirty_d> GAHH! $20 for my $10 order of aluminum
[11:51:06] <dirty_d> guess i better check out sears hardware first
[11:51:20] <Jan-> Oooooh.
[11:51:27] <ziph> Jan-: Isn't the time billed out at McDonalds rates still going to exceed the price of an off the shelf one?
[11:51:39] <Jan-> Sure, but, y'know :)
[11:51:46] <Jan-> *hack* *bash* *code*
[11:51:56] <Jan-> also then we can add features!
[11:51:58] <Jan-> and things!
[11:52:06] <Jan-> But no, I'm not seriously going to DO this.
[11:52:10] <Jan-> I don't have the ability.
[11:52:47] <Jan-> As I thought would be absolutely crushingly obvious :)
[11:53:32] <Jan-> there are active-matrix oled displays out there which only have about six wires going into them.
[11:53:38] <Jan-> I guess that just means.... it's even faster...
[11:53:57] <OndraSter> what about my LED display... 32x48 pixels!
[11:53:58] <OndraSter> red and green!
[11:54:07] <Jan-> that's not active matrix :)
[11:54:24] <OndraSter> it can be active, if you put active fan into the box! :P
[11:54:34] <Jan-> is there any tech info available on any of these phoen displays?
[11:54:37] <Jan-> I guess they'd all be hard to drive
[11:55:10] <ziph> Jan-: Why are you spending time on it then?
[11:55:22] <Jan-> ziph: well I find it quite mind-expanding, don't you?
[11:55:54] <Jan-> Although on the other hand, I was about to say that it might lead to one day my being able to do it.
[11:56:01] <Jan-> But that's never going to happen, is it
[11:56:04] * Jan- looks depressed
[11:56:09] <ziph> Jan-: It won't take much before you decide electronics sourcing is the least fun thing in the world. :)
[11:56:37] <Jan-> mmmmhmm
[11:57:01] <Jan-> actually I think making printed circuit boards is probably the least fun thing in the world
[11:57:09] <Jan-> you have to make 132 revisions, it costs a fortune, they never work
[11:57:20] <OndraSter> my project works in 1st rev!
[11:57:25] <OndraSter> all boards!
[11:57:28] <OndraSter> at least so far :P
[11:57:35] <ziph> Read more (and check your netlists manually).
[11:57:38] * Jan- doesn't believe OndraSter
[11:57:40] <OndraSter> only one thing is left to test out... switching MSOFETs
[11:57:43] <OndraSter> MOSFETs*
[11:57:56] <ziph> This board I'm working on at the moment has 160 components and worked first revision.
[11:58:16] <OndraSter> wow 160
[11:58:22] <OndraSter> I have about 50 just 100nF caps :D
[11:58:23] * Jan- stamps the word "INHUMAN" on ziph's forehead and resolves to wear an environment suit whenever he's around
[11:59:06] <OndraSter> Tom_L is alter ego of Tom_itx ?
[11:59:57] <Jan-> Oh that just takes the piss :/
[12:00:14] <Jan-> 8gb wrist watch MP3 player with 1.5" OLED screen
[12:00:16] <Jan-> costs $30
[12:00:23] <Jan-> how can something that complex cost that little
[12:01:28] <mrfrenzy_> because they make thousands of it
[12:01:41] <Jan-> don't tell me, because of strange people like Tom_itx and ziph who lie twitching at night muttering "verilog, verilog"
[12:01:50] <shorted_neuron> the chinese have gotten really good at manufacturing
[12:02:49] <Jan-> Seriously, though
[12:02:53] <Jan-> these things all use ASICs
[12:03:06] <Jan-> any homebrew device to drive a TFT display panel or similar would have to use an FPGA
[12:03:12] <Jan-> which would mean its power consumption would be huge.
[12:13:01] <Jan-> who was it who suggested ben heckendorn's forums
[12:13:03] <Jan-> ziph?
[12:13:14] <ziph> Nope.
[12:14:50] <ben1066> fpgas dont have to use a lot of power
[12:14:56] <ben1066> cyclone 4s are pretty good for power
[12:20:18] <Jan-> Phil is typing a huge post up there
[12:20:28] <Jan-> ben1066: many watts
[12:20:35] <Jan-> I presume
[12:22:08] <ziph> Mmm, unprotected JTAG inputs really are bad for susceptibility. This board can take 50W with the antenna an inch away but a 5W transmitter in the same room kills it with JTAG plugged in.
[12:22:15] <ben1066> Not as good as micros sure, but still not aweful
[12:24:12] <ziph> Hmm, Yagi. I wonder if 2000W EIPR would be pushing it.
[12:45:06] * Jan- snickers
[12:45:17] <Jan-> Chinese manufacturers need to not call their company "Feel World"
[12:45:27] <Jan-> ...on the other hand who the hell are we to tell them how to run their companies :(
[12:46:24] <CapnKernel> ziph: Jan-: Sourcing electronic parts and making PCBs == least fun. I guess that makes me mad then :-)
[12:46:43] <Jan-> CapnKernel: I wonder if you can help...
[12:46:50] <CapnKernel> I guess I see my job as "be good at it so others don't have to be"
[12:47:07] <Jan-> I guess your world is full of a billion varieties of things that have tiny TFT displays on them.
[12:47:36] <ziph> CapnKernel: You wouldn't happen to have come across SMA mag mount active GPS antennas with inputs down to 2.85V would you?
[12:48:14] <CapnKernel> ziph: I'm sure there are any number within a km of where I hang out, but I'm afraid I don't know anything about them. Sorry.
[12:48:31] <ziph> CapnKernel: No problem.
[12:48:41] <ziph> The sample I bought recently was DOA.
[12:48:52] <ziph> I took that as a sign and decided not to use those ones.
[12:48:57] <Jan-> hehe
[12:49:00] <Jan-> Fair call
[12:49:09] <CapnKernel> Not a great way to kick things off, eh?
[12:49:56] <Jan-> I thought I was the only person to do things like that
[12:50:05] <CapnKernel> Anyway, 2:40 here, must turn in.
[12:50:09] <Jan-> last week I cancelled a huge order to have our offices re-carpeted because they just lied to us about availability
[12:50:15] <Jan-> so I took the order elsewhere.
[12:50:32] <CapnKernel> I soldered 600 LEDs today, 10m worth of LED strip, to make a light box I can take to the market.
[12:50:45] <Jan-> A light box
[12:50:47] <Jan-> For what purpose
[12:51:58] <CapnKernel> Website photos. http://www.hackvana.com/store/images/mlb/p10-26504sl.jpg
[12:52:13] <Jan-> ok, I believe you
[12:52:26] <CapnKernel> Because it's very depressing to have 7 products on my website, when I really need 70,000.
[12:52:53] <Jan-> GOD I hate blogs
[12:52:59] <Jan-> You find a google result
[12:53:00] <Jan-> then go to the page
[12:53:07] <Jan-> and it's a blog, and that post fell off the bottom ages ago
[12:53:11] <CapnKernel> Jan-'s whining again
[12:53:12] * Jan- runs around in circles, waving her arms
[12:53:19] <Jan-> But it completely breaks the internet!
[12:53:25] <Tom_itx> gawd. here we go
[12:53:32] <CapnKernel> Whiny women. I'm out.
[12:53:57] <Jan-> but is it not stupid?
[12:54:09] <Tom_itx> maybe but not ot
[12:55:00] <Jan-> puh
[12:55:02] <Jan-> you guys suck
[12:55:08] <Steffann> You too
[12:55:12] <Tom_itx> we run a good channel is all
[12:55:27] <Tom_itx> staying mostly on topic is key
[12:55:44] <ziph> Yeah, this is #avr (Australians Vocalizing Remarks), not Americans.
[12:55:48] <Tom_itx> staying mostly off topic is bad
[12:55:54] <Jan-> I'm not an american.
[12:56:01] <Jan-> Not really, anyway.
[12:56:16] <ziph> Do you eat hamburgers?
[12:56:28] * Jan- looks a bit uncomfortable
[12:56:30] <Jan-> Occasionally
[12:56:32] <ziph> QED
[12:56:45] <Tom_itx> only if they're rooburgers?
[12:56:59] <Jan-> I... do actually own a stetson, but I don't own a pickup truck, I don't fire six-shooters into the air, and I never, ever say yeehaw.
[12:57:03] <Jan-> Neither do I put cheese on everything.
[12:57:16] <Jan-> I am not an American.
[12:57:27] <ziph> You know the outback steakhouse? There's on in Australia.
[12:57:34] <ziph> And the people that work there don't do the accent right either.
[12:57:36] <Tom_itx> really?
[12:57:42] <ziph> Yeah, in Sydney.
[12:57:46] <Tom_itx> heh
[12:57:54] <Jan-> oooh I love outback
[12:57:56] <Jan-> that onion thing!
[12:58:00] * Jan- salivates
[12:58:03] <ziph> You go there when you want something completely different from any food you get here.
[12:58:28] <Steffann> Jan- you know you have a unique ability?
[12:59:31] <Jan-> I know, I know. I drag things off topic.
[12:59:33] <Jan-> I'm sorry.
[12:59:49] <Jan-> Did we ever figure out how this serial to USB cable is supposed to work? (That's an AVR question!)(
[13:00:12] <Steffann> You have the 'gift' to turn every conversation into an offtopic one :)
[13:01:20] <Steffann> ziph here!!
[13:01:24] <Steffann> oops
[13:01:33] <Steffann> Wrong one :P
[13:01:37] <ziph> Hmm?
[13:01:41] <Steffann> Nick completion fail
[13:02:32] <ziph> cyanide: That's a killer nick you've got there.
[13:02:45] <Jan-> Well, I seem to have upset CapnKernel
[13:04:05] <Jan-> is anyone aware of any reason that might have happened
[13:04:54] <ziph> Don't know, but if you want a nice IRC friend we could introduce you to flyback.
[13:04:59] <ureif> lol
[13:05:05] <Jan-> oh I've met flyback
[13:05:23] <Jan-> actually I must apologise because flyback is a constant fixture in #electronics where I would be less offtopic.
[13:05:28] <Jan-> But I don't go there because of him.
[13:12:43] <Kevin`> can an avr chip run from a sinewave clock input? I would expect it can since the crystal oscillator input would be similar to that, but I haven't seen it specified
[13:13:30] <RIkusW> Steffann: if Jan- is mrs Offtopic, what am I ? mr StumbleUpon ? :-D
[13:13:44] <keenerd> Jan-: Dragging up something from way back in the buffer, but I tried "fixing" all the problems with blogs a few years ago. No one was interested in it.
[13:14:03] <Jan-> mutter, blogs, grumble
[13:14:09] <keenerd> It lives on as the CMS for my personal site though, kmkeen.com
[13:14:09] <Steffann> Yes RIkusW
[13:17:05] <Jan-> sorry
[13:17:15] <Jan-> but that just makes me think of Commander Keen
[13:17:17] <Jan-> the old PC game :D
[13:17:35] <Jan-> and yes your site is very accessible
[13:17:40] <ziph> Kevin`: I don't recall seeing it defined, but you can get some tiny comparators you could stick in there.
[13:17:50] <OndraSter> http://www.youtube.com/watch?feature=player_embedded&v=cDoRmT0iRic
[13:18:09] <keenerd> Jan-: It is okay, I loved that game and beat all the free ones :-)
[13:19:09] <keenerd> Jan-: The rant about why blogs suck and why mine is better ;-) http://redraftable.com/about/
[13:20:07] <Kevin`> ziph: I don't wanna buy something for it if possible, it's a one-off project
[13:20:40] <Jan-> I'm not sure it's about comment
[13:20:48] <Jan-> it's just about finding a link to foo.blogspot.com which is interesting
[13:20:59] <Jan-> then going there, and finding that Interesting Thing dropped off the bottom of the page 7 months ago.
[13:22:03] <ziph> Kevin`: Oh, for a one off I'd just try it. XTAL1 isn't seeing anything close to a square wave with a crystal anyhow.
[13:22:35] <Kevin`> ziph: I wonder if there's some sort of internal buffer that would be enabled with the crystal modes and disabled for "external clock" though
[13:22:52] <Kevin`> although I could just set it to the crystal mode instead in that case and ignore the output pin
[13:23:23] <Jan-> is your name really kevin?
[13:23:30] <ziph> Kevin`: Can you drive it with a 0 to VCC clock anyhow? Or do you have to put a divider on it?
[13:23:53] <keenerd> Jan-: Usually search results link to the exact post, not the front of the blog. Linking to an old, incorrect page is worse imho.
[13:24:26] <Jan-> keenerd: not in my experience
[13:24:59] <Kevin`> ziph: vcc will probably be 0-5v, the clock signal is -0.5v to +0.5v
[13:25:34] <OndraSter> depends on the source of the sinus
[13:25:39] <OndraSter> you can add some DC offset
[13:25:42] <OndraSter> just as well set amplitude
[13:25:58] <OndraSter> I have got here 20Hz - 50MHz RC oscillator back from 70s/80s :P
[13:26:06] <Kevin`> the source is a little atomic clock, i'd rather avoid modifying it :D
[13:26:29] <ziph> Kevin`: If it supports 0 to VCC for oscillator mode I imagine it is less likely to work. Some uC's just require you to capacitively divide your clock input to < 1V +/- though.
[13:26:32] <OndraSter> you can drive mega with regular square wave (probably even sinus?) on XTAL1
[13:26:53] <OndraSter> with 5V amplitude
[13:27:08] <OndraSter> 0-5V
[13:27:17] <OndraSter> aka 2.5V amplitude with 2.5V offset
[13:27:19] <OndraSter> DC
[13:28:23] <Kevin`> making the output centered on a specific voltage wouldn't be hard, but if I have to increase the amplitude i'd just use a comparator or similar circuit instead and make it into a square wave
[13:29:26] <ziph> What's the input, 10MHz?
[13:29:32] <Kevin`> yeah
[13:29:43] <ziph> The rise time of that is close enough to square wave anyhow. ;)
[13:30:05] <OndraSter> wasn't there schematics of xtal input in some datasheet for mega?
[14:31:46] <OndraSter> so, can you drive small piezzo speaker from atmega directly or should I add some driving transistor?
[14:37:24] <OndraSter> I have got here this
[14:37:25] <OndraSter> http://www.ebay.com/itm/EH-Research-Labs-G-710-Pulse-Generator-2-200-Khz-1-50-Mhz-See-My-Store-/260917485112?pt=LH_DefaultDomain_0&hash=item3cbfe4de38
[14:37:25] <OndraSter> :P
[14:39:51] <Kevin`> OndraSter: you can drive it directly
[14:40:06] <OndraSter> cool
[14:40:07] <OndraSter> thanks
[14:44:53] <JanneP> hi, is there some known problems with avr-libc's delay.h function?
[14:45:12] <JanneP> using _delay_ms, im getting delays that are off by 50% or so
[14:45:30] <Tom_itx> have you set F_CPU?
[14:45:30] <Kevin`> JanneP: iirc they don't work corretly if you are using some unusual optimziation settings. try with -Os
[14:45:53] <JanneP> F_CPU is set, and also optimization Os
[14:45:59] <Kevin`> JanneP: and indeed you need to set F_CPU directly. also they expect a CONSTANT as input, don't feed them a variable or they won't work as expected
[14:46:00] <JanneP> the delay is too short
[14:46:55] <Kevin`> JanneP: http://www.nongnu.org/avr-libc/user-manual/group__util__delay.html - everything I said is covered here, and probably some extra stuff
[14:47:21] <JanneP> thanks, i'll have a look and see if i find the problem from there
[14:50:53] <JanneP> nothing that i didn't know already pops out.. odd, i've been using it long with winavr without any trouble..
[14:52:34] <JanneP> i installed the gcc toolchain into my linux box from the build scripts that were on avrfreaks some time ago,, mayby there are some settings inthere that are wrong, i'll have to investigate
[14:52:59] <Kevin`> JanneP: does your F_CPU declaration have the UL postfix?
[14:53:29] <JanneP> i'll have to check that one
[14:56:29] <keenerd> Any chance that your distro already packages the AVR toolchain? That should be more reliable.
[14:57:16] <JanneP> yeah the UL is there
[14:57:19] <ben1066> all debian and fedora based ones will...
[14:57:57] <JanneP> my distro's ubunty 11.04
[14:58:18] <specing> Theres your problem
[15:00:00] <specing> :D
[15:00:12] <Kevin`> indeed ubuntu packages the avr toolchain and related
[15:00:22] <Kevin`> (and it works, use it)
[15:01:12] <Kevin`> apt-get install avr-libc gcc-avr avrdude
[15:04:07] <JanneP> well i guess that's next on the list if this doesn't work out anyway else
[15:05:01] <ben1066> why 11.04
[15:05:11] <ben1066> thats old and not LTS, there is no reason not to go to 11.11
[15:05:14] <ben1066> 11.10*
[15:05:42] <JanneP> ben1066: well i figured there was no reason to update yet
[15:06:00] <ben1066> make sure you update to 12.04 since thats lts
[15:07:45] <chupas> Can anyone recomend a good serial term software?
[15:07:58] <ben1066> chupas: what OS?
[15:08:02] <JanneP> ok just booted into windows, that same program compiles correcly on winavr
[15:08:02] <chupas> win
[15:08:09] <JanneP> definetely something wrong in my toolchain
[15:08:13] <RIkusW> hyperterm will work...
[15:08:20] <ben1066> putty works if windows
[15:08:22] <chupas> looking for something more powerfull
[15:08:24] <OndraSter> yap, putty
[15:08:26] <chupas> Putty maybe
[15:08:31] <chupas> realterm has been bugging me
[15:08:35] <chupas> gona try putty....
[15:08:36] <RIkusW> minicom on linux
[15:08:53] <chupas> Im looking for something to make it easier to pick out sepcfic bytes
[15:09:05] <chupas> in a large 1028 byte stream
[15:09:23] <chupas> by pick out i mean, locate and make sure they are what they should be
[15:09:25] <Kevin`> perl?
[15:09:45] <Kevin`> chupas: I use screen as a serial terminal, fwiw
[15:10:18] <chupas> is that a program? or how do you mean?
[15:10:34] <Kevin`> yes, it is, although it's normally used for other stuff
[15:10:44] <Kevin`> 'screen /dev/ttyUSB0 115200' for example
[15:10:45] <specing> screen is sufficient until you want to use ymodem transfers
[15:11:27] <keenerd> Kevin`: Out of curiosity, what happens when you try to multiplex terminals through that?
[15:11:45] <shorted_neuron> i like "logserial"
[15:12:10] <Kevin`> keenerd: it will open a local terminal if you request a new window. if you request a second terminal connection via 'screen something' again it will open in another window as you'd expect
[15:12:36] <keenerd> That sounds like the most sane thing that could happen, nice :-)
[15:22:59] <iR0b0t1> Do I have to do anything on an AVR to enable interrupts besides set the register?
[15:23:13] <specing> iR0b0t1: sei() :D
[15:23:14] <iR0b0t1> Like, something the arduino environment would do for me automatically, because now my code does not work outside of it
[15:23:58] <Kevin`> indeed, the arduino libraries often use interrupts, so it's probably enabled by default somewhere
[15:24:23] <timemage> probably off during setup and on during loop.
[15:24:37] <iR0b0t1> :(
[15:24:38] <timemage> you should be able to look at the code generated.
[15:24:41] <iR0b0t1> that didn't solve the problem
[15:24:57] <iR0b0t1> I am trying to use timer 1 to decode PPM data
[15:25:13] <iR0b0t1> The interrupt for pin 8, counting or something, doesn't seem to trigger
[15:25:39] <dirty_d> you setup the PCINT regs?
[15:25:52] <dirty_d> or INT if its one of those
[15:26:07] <iR0b0t1> Here one second.
[15:26:08] <Kevin`> iR0b0t1: what type of interrupt are you using, pin change or one of the dedicated interrupts? I dunno what "pin 8" is ;p
[15:26:56] <iR0b0t1> http://www.mediafire.com/?17a1wiczav895iz
[15:27:11] <iR0b0t1> TIMER1_CAPT_vect
[15:27:26] <specing> Haha, and you expect anyone to open that link?
[15:27:34] <dirty_d> lol
[15:27:36] <Kevin`> I did, why wouldn't you open it
[15:28:01] <iR0b0t1> It's... a mediafire link.
[15:28:02] <specing> Because its a flash/javascript/timer/redirect/advertising infested site?
[15:28:08] <OndraSter> because they are linux users
[15:28:12] <OndraSter> scared of viruses!
[15:28:16] <specing> to sum up: internet junk
[15:28:19] <iR0b0t1> Well, that's the best one I could think of.
[15:28:20] <ben1066> mediafire is one of the better file hosters
[15:28:39] <specing> bayfiles.org is the best :D
[15:28:55] <Kevin`> iR0b0t1: did you set up the timer and put it in the correct mode for input capture?
[15:29:34] <iR0b0t1> In ppm_decode.c in ppmd_begin, I basically copied code I saw elsewhere. I set up TCCR1A/B and TIMSK1
[15:29:45] <specing> err, bayfiles.com
[15:30:02] <iR0b0t1> I also set "pin 8", or pin 0 of port B, to input
[15:32:04] <timemage> iR0b0t1, i take in "pin 8" is referring to the arduino pinout.
[15:32:10] <iR0b0t1> Yeah, my bad.
[15:32:44] <JanneP> that _delay_ms problem i mentioned earlier, i copied delay.h and delay_basic.h from my winavr installation to the toolchain i installed in linux. it now works as expected
[15:34:40] <iR0b0t1> I seriously don't know what could be wrong at this point, I've been staring at the code for like 2 days
[15:34:51] <iR0b0t1> The interrupt doesn't seem to be firing
[15:35:57] <OndraSter> interrupts not firing = fun
[15:36:18] <OndraSter> so
[15:36:23] <OndraSter> you setup the pin as input
[15:36:34] <OndraSter> you enable the PCINT/INT interrupt mask
[15:36:37] <OndraSter> you enable interrupts
[15:36:53] <OndraSter> you have procedure that takes care of this interrupt
[15:36:59] <OndraSter> did you check all points?
[15:37:12] <iR0b0t1> If it is a timer interrupt must I enable a PCINT/INT mask?
[15:37:22] <OndraSter> ah timer
[15:37:24] <amee2k> CapnKernel: you still back by chance? :)
[15:37:25] <OndraSter> never tried one of those
[15:37:52] <dctalen> iR0b0t1: if it's a timer issue could your fuses be messed up?
[15:38:13] <dctalen> i had a problem a long time ago and turned out that my clock source fuses were screwed up
[15:38:15] <iR0b0t1> They... shouldn't. I never changed them after I got very similar code working under the arduino environment
[15:38:28] <iR0b0t1> Unless avrdude changes them without you telling it to
[15:38:40] <OndraSter> what are you using to programm it?
[15:38:49] <dctalen> avrdude should be able to dump them out for you to verify
[15:38:53] <iR0b0t1> avrdude -p m328p -c arduino -P /dev/ttyACM1 -e -U flash:w:$(PRG).hex
[15:38:58] <OndraSter> sounds linuxey
[15:39:04] <OndraSter> ergo no AVR Studio available
[15:39:11] <OndraSter> or does gdb now work with Dragon or any other debugger?
[15:39:30] <iR0b0t1> I do not have a programmer, I must use the USB bootloader.
[15:39:36] <OndraSter> oh...
[15:41:02] <dctalen> -U lfuse:r hfuse:r (or similar) should dump them
[15:41:15] <Kevin`> OndraSter: gdb does indeed work
[15:41:39] <Kevin`> OndraSter: I normally just print out stuff to serial though
[15:41:50] <OndraSter> I use my Dragon :)
[15:42:01] <OndraSter> without JTAG or any other way to debug I'd be often lost
[15:42:14] <Kevin`> printf =p
[15:42:26] <OndraSter> stepping, editing registers on-the-go
[15:42:42] <Kevin`> oh sure, it's nice, but it's not normally necessary
[15:42:44] <OndraSter> I can edit registers only from main loop on serial...
[15:42:54] <OndraSter> I came to conclusion that it is... at least for me
[15:43:38] <OndraSter> (remember, I have my o-future bootloader(?) that will sport "debug" functions that enable writing/reading any address in human-readable hex format :P)
[15:43:53] <Kevin`> I have a draon too, but it's normally being used as a programmer except in the strangest faults
[15:44:12] <iR0b0t1> avrdude -p m328p -c arduino -P /dev/ttyACM1 lfuse:r hfuse:r
[15:44:17] <iR0b0t1> what is that supposed to output...?
[15:44:21] <Kevin`> can you break back to the bootloader from the app to fiddle?
[15:45:13] <Kevin`> iR0b0t1: i'm not sure if the arduino bootloader supports reading fuses. since you can't WRITE them, it's not a normally necessary function. it might though.
[15:45:32] <iR0b0t1> OH
[15:45:34] <iR0b0t1> Oh
[15:45:37] <iR0b0t1> then wtf do I do now
[15:45:56] <OndraSter> Kevin`, I want to add that option aswell
[15:45:59] <OndraSter> ...somehow
[15:46:27] <OndraSter> but I am not sure how, the app would need to know about this kind of bootloader
[15:46:29] <Kevin`> iR0b0t1: if your code is running, the fuses are fine enough
[15:46:53] <Kevin`> iR0b0t1: unless you were randomly using the clock output pin or similar fuse-enabled feature
[15:46:57] <iR0b0t1> But the TIMER1_CAPT_vect does not seem to trigger.
[15:47:34] <iR0b0t1> the TIMER1_OVF_vect does trigger though
[15:48:52] <OndraSter> hmm I just realized I need some safety when my project is turned on
[15:48:52] <Kevin`> http://pastebin.com/E27zwakD - how do you re-write the signature stuff or otherwise troubleshoot that? both are new chips from their conductive foam home
[15:48:56] <OndraSter> that the LEDs do NOT light up
[15:49:15] <OndraSter> I will probably use pin to disable the power supply for the LEDs
[15:49:25] <OndraSter> and the first thing in code will actually pull them to disable them
[15:50:04] <Kevin`> hey abcminiuser_
[15:50:13] <OndraSter> or dump all 00s on serial port
[15:50:17] <Kevin`> you know about these avr things, right?
[15:50:55] <iR0b0t1> wait I got it to trigger
[15:51:24] <abcminiuser_> Heyo
[15:51:32] <abcminiuser_> Kevin`, somewhat
[15:52:31] <Kevin`> abcminiuser_: I have an atmega328p that seems to return the signature '0x010305', as seen in the paste a few seconds ago
[15:52:58] <abcminiuser_> Kevin`, what programmer?
[15:53:06] <Kevin`> dragon in isp mode
[15:53:47] <RIkusW> Kevin`: try lowering ISP speed ?
[15:54:13] <abcminiuser_> Could be too high a ISP clock
[15:54:14] <abcminiuser_> RIkusW, snap
[15:54:22] <Kevin`> RIkusW wins. doh
[15:54:30] <Kevin`> I wonder what it was set to before that it wouldn't just work
[15:54:36] <OndraSter> abcminiuser_, during flashing, all IOs go to HiZ I presume?
[15:54:49] * Kevin` goes to calculate what the proper max clock would be in microseconds for a stock device
[15:58:28] <iR0b0t1> Yay!
[15:58:31] <iR0b0t1> The interrupt works.
[15:58:41] <iR0b0t1> Now to figure out why it it isn't decoding the PPM signal
[15:58:52] <dctalen> what fixed it?
[15:58:57] <timemage> exactly
[15:59:22] <iR0b0t1> Dunno.
[15:59:28] <OndraSter> and during RESET...
[15:59:29] <iR0b0t1> Probably never will.
[15:59:31] <OndraSter> ime to check datasheet
[15:59:45] * iR0b0t1 experiences this a lot.
[15:59:50] <OndraSter> haha
[15:59:52] <OndraSter> I know that feeling
[16:00:21] <abcminiuser_> OndraSter, yes in ISP
[16:00:39] <abcminiuser_> ISP requires the /RESET line to be active during programming, so all pins are Hi-Z
[16:00:43] <OndraSter> ah ok
[16:00:51] <OndraSter> and during programming through JTAG?
[16:00:59] <OndraSter> where the RESET line is not being controlled
[16:01:31] <OndraSter> ... doesn't matter, I can not use JTAG anyway... if I'd break the program, the LEDs would die in a horrible pain of 60mA per LED :P
[16:02:21] <iR0b0t1> Now this *really* doesn't make sense.
[16:02:48] <OndraSter> ergo I will write the application as simplest as possible, make sure it works, connect the LEDs afterwards and test only through there
[16:05:32] <OndraSter> I have to put software protection that says "NO" to writing to SREG, TIMSK...
[16:05:39] <OndraSter> into my debug console
[16:05:53] <OndraSter> and probably TCCR0 aswell
[16:06:00] <OndraSter> it would be bad if it burnt, mmkay
[16:08:26] <abcminiuser_> OndraSter, watchdog with short timeout
[16:08:44] <abcminiuser_> Have a init check that aborts and Hi-Zs everything if a reset occurs
[16:08:44] <OndraSter> that won't help in almost any case
[16:08:59] <OndraSter> when JTAG connects and does breakpoint
[16:09:10] <OndraSter> it will be waiting for my input so it can do the timer ovf interrupt
[16:09:17] <OndraSter> that moves to another line... etc
[16:09:52] <abcminiuser_> Anyone know where I can buy an Etch-a-Sketch engraved with the Apple logo?
[16:23:41] <Kevin`> avr seems to work properly with a sine wave clock input centered around vcc/2. I think work properly is good enough for me here
[16:24:04] <Kevin`> and I now have a very consistant software-delay blinky led :)
[16:25:56] <Kevin`> it would be interesting to use an antenna tuned to the desired frequency as a clock input
[16:36:45] <amee2k> Kevin`: hmm may or may not be related... i once read a neat trick about DCF77 and a 5MHz xtal
[16:38:07] <amee2k> the DCF carrier is 77.5kHz and phase locked to the atomic clock used as local reference. now if you use a 5MHz (or multiple thereof with adequate frequency divider) and divide it by 64 you get 78.125kHz
[16:38:27] <amee2k> mixing that with the antenna input yields the DCF carrier at 625Hz where you can demodulate it
[16:39:09] <amee2k> now if you divide the 78.125kHz signal by 125 you get 625Hz too. the replace the loading cap on the xtal with a tuning diode so you can pull it around a bit
[16:39:30] <amee2k> and use a PLL to lock the downmixed DCF carrier and the divided xtal output together
[16:39:45] <amee2k> and you get a 5MHz xtak phase locked to an atomic clock >_>
[16:40:51] <Steffanx> Yay amee2k
[16:40:58] <amee2k> hey Steffanx \o/
[16:41:06] <amee2k> whats buzzin'
[16:41:22] <Steffanx> Not muc
[16:41:22] <Steffanx> h
[16:42:21] <amee2k> that trick sounds pretty cool for implementing an RTC in software. i'd love to try it sometime, but my DCF receiver project a few years back didn't manage to get a lock on the carrier :/
[16:45:18] <Kevin`> amee2k: I just have an atomic clock driving the microcontroller directly, simpler than all that radio stuff :D. but yes, sounds useful
[16:45:45] <amee2k> well, useful in an educational kind of way
[16:45:52] <Steffanx> Kevin` has an atomic clock in his backyard?
[16:46:02] <Kevin`> Steffanx: there's a bunch of them on ebay right now
[16:46:12] <amee2k> i'm not sure if there is a good way to get the circuit to run stable in a situation with intermittent reception only
[16:46:15] <Steffanx> Chinese crap?
[16:46:30] <Kevin`> Steffanx: little ones used in cellular towers to maintain data sync without a master clock present
[16:46:36] <Kevin`> Steffanx: used
[16:47:00] <amee2k> how much do they cist?
[16:47:10] <amee2k> cost* even
[16:47:15] <Kevin`> $45 or so
[16:47:22] <amee2k> neat
[16:47:24] <Kevin`> (used, of course)
[16:47:39] <amee2k> thats pretty cheap compared to the normal rackmount size ones
[16:47:58] <keenerd> amee2k: Aw, looks like DCF77 is german only.
[16:48:17] <Steffanx> No it isn't
[16:48:22] <amee2k> keenerd: its longwave. how good is your antenna? ;)
[16:48:57] <keenerd> Okay, Europe only.
[16:49:05] <Steffanx> :)
[16:49:23] <Kevin`> keenerd: there are other timesource signals in other countries
[16:49:41] <keenerd> "Katastrophenwarnung" germany language is great.
[16:49:47] <Kevin`> (or gps everywhere)
[16:49:50] <keenerd> *german
[16:50:03] <Kevin`> but the simple time signals are a lot easier to decode than gps
[16:50:07] <amee2k> shitty 5$ radio clocks work pretty well within most of europe and if you really want to it is receivable from north america to japan as well
[16:50:56] <amee2k> Kevin`: i haven't heard of a nice PLL trick like that for other time station frequencies
[16:51:19] <amee2k> should be possible to come up with one anyway though
[16:51:48] <amee2k> WWVsomething is 40kHz, right?
[16:52:18] <amee2k> hmm.... 5MHz xtal and common IF of 500Hz should work
[16:54:26] <keenerd> WWVB is 60kHz.
[16:54:59] <keenerd> Good frequency to target, transmitters at the frequency in us, uk, jp.
[16:55:39] <amee2k> interresting
[16:55:55] <amee2k> well, it only makes sense if all these transmitters have better stability than your xtal
[16:56:29] <keenerd> But think of the buzzwords!
[16:57:30] <keenerd> It is something I'd still like to try, just because it sounds fun :-)
[16:58:21] <keenerd> 3MHz works, though 50 is awkward to divide by.
[16:59:19] <amee2k> :)
[16:59:49] <OndraSter> isn't it easier to grab rubidium clock?
[16:59:53] <OndraSter> it is 0,000nothing ppm
[16:59:58] <OndraSter> 10MHz
[17:00:04] <OndraSter> they are about $50
[17:00:50] <amee2k> temperature controlled crystal, then calibrate every so-many hours with a GPS reference
[17:01:10] <OndraSter> I like the rubidium :P
[17:01:23] <OndraSter> I might grab one sometime
[17:01:33] <OndraSter> they are from old cell towers from china I think
[17:01:36] <OndraSter> there is dozen of them on ebay
[17:01:46] <amee2k> neat
[17:02:55] <amee2k> http://www.ebay.de/itm/10MHz-Rubidium-Frequenzstandard-Atom-frequenznormal-/290656400287 << like these?
[17:02:57] <OndraSter> yes
[17:03:09] <OndraSter> except cheaper
[17:03:11] <OndraSter> 80€ is ripoff
[17:03:15] <OndraSter> they are $50 :P
[17:03:16] <amee2k> well, welcome to germany
[17:03:18] <OndraSter> $40 - $60
[17:03:21] <OndraSter> this one is in china
[17:03:26] <OndraSter> Artikelstandort: China, China
[17:03:26] <OndraSter> Versand nach: Weltweit
[17:03:33] <amee2k> the country where everyone else will think you're stink rich
[17:03:45] <amee2k> "product location: china / ships to: worldwide"
[17:03:55] <OndraSter> http://www.ebay.com/itm/10-MHz-sinewave-FE-5680A-Rubidium-Oscillator-Std-50ohm-/180518264100?pt=LH_DefaultDomain_0&hash=item2a07ba1d24
[17:03:57] <OndraSter> $70
[17:04:07] <OndraSter> they used to be cheaper
[17:04:09] <OndraSter> :/
[17:04:10] <OndraSter> damn
[17:04:16] <OndraSter> I should've grabbed it back then
[17:04:40] <amee2k> does the current generation towers use atomic clocks too?
[17:05:07] <OndraSter> humm.... use internet, or if anybody else knows it here.... I am not that kind of guy :P
[17:05:43] <OndraSter> anyway, getting up in a while
[17:05:43] <OndraSter> bb
[17:05:53] <amee2k> if so, when the next generation goes out of comission they should get cheap again
[17:06:23] <amee2k> speaking of time
[17:06:33] <amee2k> does anyone know by chance what timezone CapnKernel is in again?
[17:06:33] <Roklobsta> ooohhh good find, rubidium crock...
[17:06:48] <Roklobsta> yeah, het in line, i am asking about rubidium clocks first
[17:07:21] <Roklobsta> er google timezone china. China only has 1 timezone.
[17:07:51] <Roklobsta> almost 7am
[17:08:06] <amee2k> China Standard Time, UTC+8
[17:08:10] <Roklobsta> you can get gos modules that spit out 10MHz.
[17:08:14] <amee2k> yuck
[17:08:16] <Roklobsta> gps
[17:08:40] <amee2k> thats 6 hours ahead of me. explains why he is never on when i am >_<
[17:08:45] <Roklobsta> they are usually the proper timing modules. your bog standard GPS kind sucks for accurate time.
[17:08:52] <Roklobsta> he'll be here soon.
[17:09:13] <amee2k> it should be like 5 in the morning at his place now
[17:09:15] <Roklobsta> if he's not already at the mega mega mega electronics parts district drumming up business.
[17:09:26] <Roklobsta> it's 7am in china now.
[17:09:33] <amee2k> oh
[17:09:43] <Roklobsta> 7am Monday
[17:09:45] <amee2k> err, right. we have winter time
[17:09:50] <amee2k> so 7 hours ahead of me >_<
[17:10:06] <Roklobsta> 10am monday here
[17:11:16] <Roklobsta> your 50hz/60hz should be very stable
[17:11:39] <Kevin`> Roklobsta: http://www.leapsecond.com/pages/mains/
[17:15:32] <Roklobsta> here's an example of a 10MHz GPS reference. http://www.navsync.com/docs/CW12_TIM.pdf
[17:15:51] <Roklobsta> it seems many timing modules mimic some old Motrola M12+ device.
[17:18:32] <jacekowski> it's still sunday in here
[17:18:53] <jacekowski> not much of it left though
[17:20:00] <iR0b0t1> Hey okay
[17:20:04] <iR0b0t1> I think I know my problem.
[17:20:21] <iR0b0t1> I think avr-gcc is optimizing out an access, even though the variable is declared volatile.
[17:21:20] <iR0b0t1> No, no...
[17:21:24] <iR0b0t1> It keeps getting reset.
[17:21:25] <iR0b0t1> Somehow.
[17:22:35] <iR0b0t1> SOMEHOW
[17:25:16] <iR0b0t1> FUCK.
[17:25:53] <specing> iR0b0t1: PEBKAC
[17:26:22] <iR0b0t1> It... halts.
[17:26:29] <iR0b0t1> It stops responding to interrupts after some condition.
[17:28:35] <iR0b0t1> .
[17:29:55] <specing> Do you have a cap on power supply rails?
[17:30:02] <specing> Do you have a stable power supply?
[17:30:42] <iR0b0t1> Indeed, 'tis stable.
[17:30:45] <iR0b0t1> found my error.
[17:30:48] <iR0b0t1> Well, part of it.
[17:30:53] <iR0b0t1> It loses sync on the ppm channel.
[17:30:59] <iR0b0t1> Now, why it does that, I have no clue.
[17:35:27] <iR0b0t1> I literally copied this code from a working program I had
[17:35:30] <iR0b0t1> how did it randomly stop working
[17:36:33] <Tom_itx> appears random
[17:50:04] <dirty_d> iR0b0t1, what are you building?
[17:58:47] <dirty_d> iR0b0t1, each pulse is at a specific position from the beginning pulse?
[17:59:28] <iR0b0t1> Yeah
[17:59:35] <iR0b0t1> It is older RC equipment
[17:59:51] <iR0b0t1> now they like PCM, which is more digital, but it has longer frames and less resolution
[18:00:00] <dirty_d> http://www.pabr.org/pxarc/1.1/doc/opwm_ppm.gif
[18:00:03] <dirty_d> like that?
[18:00:08] <iR0b0t1> Yup
[18:00:35] <dirty_d> so the length of the actual pulses is fixed?
[18:00:49] <dirty_d> but time between them varies
[18:01:23] <iR0b0t1> Yeah
[18:01:37] <dirty_d> and you just time the rising edges id guess right?
[18:01:41] <iR0b0t1> Mhmm
[18:01:45] <dirty_d> cool
[18:01:50] <iR0b0t1> You can actually make a decoder with just a shift register.
[18:01:56] <iR0b0t1> But god forbid you encounter any interference :p
[18:01:56] <dirty_d> i had to decode regular servo pulses for my project
[18:02:08] <dirty_d> 4 separate channels not one like yours
[18:02:15] <dirty_d> so i suppose its easier in a way
[18:02:35] <dirty_d> in my case it is because im using xmega and it has pulse width capture, so i dont really have to do anything
[18:03:05] <dirty_d> youre losing sync?
[18:03:12] <iR0b0t1> Well, I think
[18:03:14] <iR0b0t1> I am not really sure.
[18:03:19] <iR0b0t1> Something is really funky
[18:03:28] <dirty_d> what happens?
[18:03:49] <iR0b0t1> I would say losing sync I guess. I will get proper readings, but interspersed between them I keep getting exactly 1.5ms times
[18:04:20] <iR0b0t1> I think I keep losing sync and then grabbing the sync pulse.
[18:04:33] <iR0b0t1> But, why do I lose sync? I cannot understand this
[18:05:03] <dirty_d> is your code able to run fast enough to keep up?
[18:05:47] <dirty_d> youre using the INT interrupt not PCINT right?
[18:06:02] <iR0b0t1> I am counting using a timer.
[18:06:06] <dirty_d> not that you couldnt use PCINT, but INT would be easier
[18:06:17] <iR0b0t1> The atmega328 has an external interrupt to trigger a timer capture on pin change
[18:06:43] <iR0b0t1> Now see, I get a hint... My code hangs if I go to a certain position on the controller.
[18:06:44] <dirty_d> oh right
[18:07:04] <dirty_d> it just triggers on rising edge right?
[18:10:53] <iR0b0t1> Well, I start it at zero, it counts in .5us steps until a pulse arrives
[18:10:55] <iR0b0t1> I record that
[18:10:59] <iR0b0t1> Then set it back to 0
[18:11:11] <iR0b0t1> This happens asynchronously in regards to the rest of the program.
[18:12:20] <dirty_d> would it be easier to just set an INT pin interrupt that saves the count reg then resets it to 0
[18:12:25] <dirty_d> with a special case for first pulse
[18:12:39] <dirty_d> so the timer starts from 0 at the rising edge
[18:12:54] <dirty_d> then at the next rising edge you save the time betwwen the first edge and this one
[18:12:57] <dirty_d> then reset to 0
[18:13:00] <dirty_d> for the next one and so on
[18:13:09] <iR0b0t1> Well that is basically what I am doing
[18:13:12] <dirty_d> oh ok
[18:13:22] <iR0b0t1> just using timer-assisted interrupts
[18:15:06] <dirty_d> you should be able to do better than 0.5us though right?
[18:15:06] <dirty_d> you should be able to do better than 0.5us though right?
[18:15:42] <dirty_d> if your clock is 8Mhz you can have prescaler 1
[18:19:08] <iR0b0t1> Fuck.
[18:19:12] <iR0b0t1> I turned a > into a <
[18:19:16] <iR0b0t1> Lets see if this fixes it.
[18:20:25] <iR0b0t1> Well, shit. I still can't *read* the data.
[18:20:30] <iR0b0t1> I think this is an optimization problem..
[18:22:00] <iR0b0t1> dirty_d: I've got some other chips on order
[18:22:04] <iR0b0t1> dirty_d: I think I'll try that
[18:28:03] <jadew> are sine wave generators useful for noobs?
[18:28:14] <jadew> function generators in general
[18:29:02] <Roklobsta> can be
[18:29:19] <Roklobsta> why do you ask, oh n00b?
[18:29:31] <jadew> need a new project
[18:29:35] <Roklobsta> you ar ebest getting an arbitrary wave form generator
[18:29:59] <Roklobsta> for example, chirps.
[18:30:06] <jadew> thing is, I don't know what I would use them for
[18:30:17] <Roklobsta> well, don't get one until you need one
[18:30:18] <jadew> sure, testing input and output signals
[18:30:42] <Roklobsta> you can use the DAC on some AVRs to make your own
[18:30:58] <jadew> but.. I'm not entirely sure how useful that is, cuz you can check the input signal directly from the thing that generates it
[18:31:09] <Roklobsta> i have made DDS with an 8051 and its DAC before. fine for audio range stuff.
[18:31:54] <jadew> I see, so there really isn't a need for this
[18:32:06] <Roklobsta> not yet
[18:32:13] <Roklobsta> how is the cro?
[18:32:19] <jadew> awesome
[18:32:36] <jadew> can't believe I was fooled into thinking I can do any kind of serious work with out one
[18:32:48] <Roklobsta> yeah
[18:32:54] <Roklobsta> you can't just program your problems away
[18:33:16] <jadew> yeah, there are problems you don't even know are there before you take a look with a scope
[18:33:39] <Roklobsta> right
[18:33:46] <Roklobsta> have you used it to time loops?
[18:33:47] <jadew> and did some things that couldn't have done with out the scope
[18:34:00] <jadew> no
[18:34:05] <jadew> but that sounds interresting
[18:34:18] <jadew> how do you do that?
[18:34:31] <dirty_d> iR0b0t1, wanna pastebin your code?
[18:34:32] <Roklobsta> yeah, set a port pin at the start of a function and set it to 0 at the end and you can see exactly how long the function takes to run
[18:34:35] <dirty_d> its not much right?
[18:34:55] <jadew> ah, those kind of loops
[18:35:09] <Roklobsta> loops functions etc.
[18:35:15] <jadew> yeah, I used it a lot like that to test interrupts
[18:35:23] <jadew> very helpful there
[18:35:42] <Roklobsta> oh yes
[18:36:01] <jadew> particularly important in soft serial and soft pwm generation
[18:36:38] <jadew> wrote some soft usart code and it was really handy there
[18:37:09] <Roklobsta> yep
[18:37:18] <Kevin`> no need to count cycles to make a soft uart, just look at it on the scope and change the number till it's right :D
[18:37:58] <jadew> well, you have to make sure your interrupt won't take longer than the freq/2
[18:37:59] <iR0b0t1> http://www.mediafire.com/?s865u3jgcbvt1bl
[18:38:05] <iR0b0t1> dirty_d: ^
[18:42:29] <iR0b0t1> dirty_d: ppm_decode.c
[18:43:11] <iR0b0t1> It drops into failsafe a whole lot. I don't know why!
[18:43:16] <iR0b0t1> It should be receiving successfully...
[18:43:37] <iR0b0t1> It is because of the check in the TIMER1_CAPT_vect
[18:43:53] <iR0b0t1> Ergo, the signal must not be in range... but why?
[18:43:57] <dirty_d> iR0b0t1, when it overflows?
[18:43:59] <iR0b0t1> I look at it on a scope.
[18:44:05] <iR0b0t1> No, the overflow doesn't trigger.
[18:44:07] <iR0b0t1> Below that.
[18:44:18] <iR0b0t1> The second interrupt routine
[18:45:33] <dirty_d> ok
[18:45:43] <dirty_d> is the timer set up to only cap on rising edge?
[18:46:03] <iR0b0t1> Yes
[18:46:06] <iR0b0t1> You can pick one or the other
[18:46:09] <iR0b0t1> I picked rising edge.
[18:46:09] <dirty_d> ok
[18:46:25] <iR0b0t1> There are three sync checks, the one which fails is in TIMER1_CAPT_vect.
[18:46:45] <dirty_d> if((ICR1 >= MIN_PULSEWIDTH_TICKS) && (ICR1 <= MAX_PULSEWIDTH_TICKS))
[18:46:50] <dirty_d> thats whats failing you think?
[18:47:00] <iR0b0t1> Yeah
[18:47:11] <iR0b0t1> I commented them all out in combinations, etc. That one generates the failsafe state.
[18:47:14] <iR0b0t1> But... why?
[18:47:20] <iR0b0t1> I simply can't figure out why.
[18:47:25] <dirty_d> and you can see on the scope that the pulses are in that range?
[18:47:29] <iR0b0t1> Mhm
[18:48:15] <dirty_d> iR0b0t1, i dont see where your clock is set to 16mhz
[18:48:39] <iR0b0t1> In main.h, F_CPU?
[18:48:44] <iR0b0t1> It is an external clock.
[18:48:51] <iR0b0t1> Well, shit. Do I need to set that up?
[18:48:56] <dirty_d> yea thats your problem
[18:49:05] <iR0b0t1> :3
[18:49:06] <dirty_d> youre running at 2mhz not 16
[18:49:08] <iR0b0t1> How do I enable it?
[18:49:12] <iR0b0t1> WELL FUCK THAT EXPLAINS A LOT.
[18:49:13] <dirty_d> i bet it works after you fix that, lol
[18:49:20] <Tom_itx> woops
[18:49:28] <iR0b0t1> ffffffffffffffff
[18:50:04] <dirty_d> iR0b0t1, look on page 38 of your manual
[18:50:15] <Tom_itx> just diving in here but if that don't work check ckdiv8
[18:51:00] <Tom_itx> those 2 will bite your arse
[18:51:55] <dirty_d> iR0b0t1, actually just do clock_prescale_set(clock_div_1);
[18:52:02] <dirty_d> and you should be set
[18:52:45] <dirty_d> thats in avr/power.h i think
[18:52:47] <iR0b0t1> If it is a fuse I can't change it. But I should be able to use CLKPS, right?
[18:52:54] <dirty_d> yup, that function does it for you
[18:54:50] <dirty_d> CLKPR = 0x80; CLKPR = 0x0;
[18:54:54] <dirty_d> that should do it also
[18:55:19] <iR0b0t1> dirty_d, which header are those in? Google fails me
[18:58:58] <dirty_d> the function?
[18:59:01] <dirty_d> avr/power.h
[19:00:12] <dirty_d> if youre using avrstudio it might be different
[19:00:19] <dirty_d> ive never used it, im using avr-gcc
[19:00:24] <iR0b0t1> Same here.
[19:00:28] <dirty_d> ok
[19:00:36] <iR0b0t1> I just used CLKPR = 0, I couldn't find clock_div_1.
[19:01:06] <dirty_d> you put CLKPR = 0x80; immedietly before right?
[19:01:30] <dirty_d> thats the protection mechanism
[19:01:46] <dirty_d> once you do that you have 4 clock cycles to set the prescaler in the same register
[19:02:43] <iR0b0t1> Okay.
[19:02:52] <iR0b0t1> CLKPR = 0x80, then CLKPR = 0x00.
[19:03:09] <dirty_d> oh, this will have you running on the 8Mhz internal osc
[19:03:11] <dirty_d> yea
[19:03:23] <dirty_d> therse more to do if you wanna use a crystal
[19:03:39] <dirty_d> but you should be fine on 8MHz for now as long as you change your defines in there
[19:03:49] <iR0b0t1> Wait, how do I use an external crystal?
[19:04:09] <Tom_itx> set the fuses to look for it
[19:04:16] <Tom_itx> be careful, you can brick it
[19:04:31] <dirty_d> iR0b0t1, i wouldnt mess with it yet
[19:04:48] <iR0b0t1> No I am going to mess with it.
[19:04:52] <iR0b0t1> Well
[19:04:57] <Tom_itx> better listen to him
[19:05:07] <iR0b0t1> Okay, this is on an arduino board. The fuses should be set to look at the external oscillator.
[19:05:14] <iR0b0t1> Otherwise, I can't change them, I don't have a programmer.
[19:05:25] <Tom_itx> we can fix that too
[19:06:01] <dirty_d> how are you programming it now?
[19:06:06] <iR0b0t1> USB bootloader
[19:06:09] <dirty_d> ive never use an arduinop
[19:06:13] <dirty_d> oh
[19:06:14] <iR0b0t1> well, serial bootloader, going through an atmega32u2
[19:06:27] <iR0b0t1> turns it to USB-serial
[19:07:27] <iR0b0t1> Either way, still keeps losing sync :(
[19:07:43] <dirty_d> cant you switch to an external clock from the interanl in software to not have to mess with fuses?
[19:07:56] <dirty_d> iR0b0t1, well your defines would need to change right
[19:08:04] <dirty_d> theyre set for 16mhz
[19:08:15] <dirty_d> min pulse width etc
[19:08:17] <iR0b0t1> They shouldn't, I think I've been using a 16MHz clock this whole time.
[19:08:32] <iR0b0t1> I mean, these values worked on this exact chip before
[19:08:34] <dirty_d> oh, i dunno maybe with the arduino stuff
[19:08:36] <iR0b0t1> ... under the arduino environment
[19:09:03] <dirty_d> an easy wat to test that is
[19:09:43] <dirty_d> while(1){PORTA = 1; _delay_ms(1000); PORTA = 0; _delay_ms(1000); }
[19:09:49] <dirty_d> with an led or something on the port
[19:10:03] <dirty_d> if its not blinking every second, youre not at the speed you think you are
[19:13:21] <iR0b0t1> It's going faster than one secomd.
[19:14:14] <Roamin> do you have optimizations turned on ?
[19:14:38] <Roamin> and also you can'T really do _delay_ms(1000) because there is a maximum value for delay
[19:14:40] <iR0b0t1> Perhaps.
[19:14:46] <Roamin> depending on the frequency you'Ve selected
[19:14:51] <iR0b0t1> 16MHz
[19:15:23] <Roamin> let me pull up the equation
[19:16:01] <keenerd> Wait really? I've done delays that long.
[19:16:40] <Roamin> The maximal possible delay is 262.14 ms / F_CPU in MHz.
[19:17:54] <iR0b0t1> Using a 500ms delay
[19:17:59] <iR0b0t1> I get three cycles in 50ms
[19:18:02] <iR0b0t1> wut
[19:19:24] <dirty_d> huh? i use a 1000ms delay a lot
[19:19:27] <dirty_d> and it works
[19:19:28] <Roamin> at 16mhz, the longest possible delay using the function would be 16.38375ms
[19:22:04] <iR0b0t1> Also the port doesn't seem to turn all the way on :\
[19:22:12] <dirty_d> whaaa
[19:22:28] <dirty_d> what do you mean
[19:22:38] <Roamin> what port is it? are you changing anything on port C at all?
[19:23:28] <iR0b0t1> PORTB
[19:23:46] <iR0b0t1> doh
[19:23:54] <iR0b0t1> accidentally wrote "PORTB" instead of "DDRB"
[19:26:01] <iR0b0t1> I was reading the pullup resistor :p
[19:26:20] <dirty_d> ahh
[19:27:07] <iR0b0t1> Aaanyway.
[19:27:14] <iR0b0t1> I can't get a delay of 1s
[19:27:19] <dirty_d> what is the time
[19:27:56] <Roamin> you'Ve defined F_CPU and have optimizations on ?
[19:28:05] <dirty_d> iR0b0t1, youre missing a zero #define F_CPU 1600000UL
[19:28:40] <iR0b0t1> Telling it to delay for 500ms and get 20ms
[19:28:56] <iR0b0t1> Roamin: Are you not supposed to do that?
[19:29:10] <dirty_d> you are supposed to do both
[19:29:19] <Roamin> you are supposed to, but you need to do it correctly, as dirty_d pointed
[19:29:31] <iR0b0t1> Well even with the proper define
[19:29:32] <Roamin> seems you are saying you have a 1.6mhz crystal instead
[19:29:37] <iR0b0t1> I still get 20ms delay for 500ms
[19:30:07] <dirty_d> iR0b0t1, did you see that message above about the missing 0
[19:30:19] <iR0b0t1> I did.
[19:30:21] <iR0b0t1> I changed it
[19:30:24] <iR0b0t1> Same output
[19:30:37] <dirty_d> it must not have made it through
[19:30:42] <dirty_d> forgot to save/compile/anything?
[19:31:04] <dirty_d> and changed in every file its defined in?
[19:31:42] <iR0b0t1> oops
[19:31:48] <iR0b0t1> it goes before the include of delay.h
[19:31:49] <iR0b0t1> :3
[19:31:54] <Roamin> is it me or you are defining F_CPU after you have included delay.h ?
[19:32:03] <Roamin> heh
[19:32:05] <Roamin> ;)
[19:32:17] <dirty_d> hehehe
[19:32:51] <Roamin> you should have gotten warnings at compile time
[19:33:44] <iR0b0t1> I did :\. Way up top.
[19:33:57] <iR0b0t1> But... I turn on, delay for 1000 ms
[19:34:03] <iR0b0t1> it stays on for... half a second?
[19:34:31] <Roamin> you said your delay was 500ms no?
[19:34:37] <Roamin> 500ms on, 500ms off
[19:34:39] <iR0b0t1> I changed it.
[19:34:51] <iR0b0t1> the fuck
[19:34:52] <iR0b0t1> hold on
[19:35:01] <iR0b0t1> I delay for 2000ms between state change
[19:35:16] <iR0b0t1> I get a state change inside of a second
[19:35:17] <iR0b0t1> that is wrong
[19:36:03] <dirty_d> hmm?
[19:37:34] <dirty_d> i dont see how it could be going faster than youretelling it
[19:37:42] <dirty_d> that would mean your clock is actually faster than 16MHz
[19:37:48] <dirty_d> which i dont think it is
[19:38:19] <iR0b0t1> 1000ms = 1s
[19:38:19] <iR0b0t1> right
[19:38:37] <Roamin> what board are you using?
[19:38:55] <iR0b0t1> nevermind I'm an idiot somehow
[19:39:01] <iR0b0t1> I must be reading the scope wrong :\
[19:39:11] <iR0b0t1> AH
[19:39:13] <iR0b0t1> indeed I am
[19:39:20] <iR0b0t1> Five divs per box :3
[19:39:25] <iR0b0t1> Not 1box = div
[19:39:35] <dirty_d> ahh lol
[19:39:38] <dirty_d> what scope do you have?
[19:39:39] <dirty_d> i want one
[19:39:57] <Roamin> i have a tektronix 2445 150mhz
[19:40:05] <Roamin> its nice to have
[19:40:10] <dirty_d> sounds expenisve
[19:40:26] <Roamin> think it was under 300$ from ebay
[19:40:32] <dirty_d> oh not that bad
[19:40:57] <dirty_d> all i have is my soundcard which is ok for my ~400Hz stuff im working on now
[19:41:04] <vectory> :)
[19:41:23] <vectory> though, you dont have divs in your example code, or do you?
[19:42:14] <iR0b0t1> dirty_d: Fluke 95 50MHz
[19:42:27] <Roamin> 156$ from ebay i paid, free shipping
[19:42:43] <dirty_d> http://www.ebay.com/itm/Owon-PDS6062S-60-MHz-Portable-Dig-Oscilloscope-USB-250MS-s-Dual-Channel-/320855860167?pt=BI_Oscilloscopes&hash=item4ab47fe3c7#ht_1266wt_1114
[19:42:43] <dirty_d> that looks ok
[19:43:22] <dirty_d> i wnt a digital one, but i guess i dont really need one
[19:48:21] <iR0b0t1> Well, shit. Still fails.
[19:48:31] <iR0b0t1> You guys have been helpful :|
[19:50:11] <Roklobsta> rigol is $329 now and hackable to 100Mhz.
[19:50:17] <dirty_d> hmm, no prob
[19:50:27] <dirty_d> yea was lookin at that one
[19:50:36] <Roklobsta> i got one. jadew got one.
[19:50:47] <dirty_d> they havent fixed it still?
[19:50:53] <Roklobsta> everyone here except you got one
[19:50:57] <dirty_d> lol
[19:53:01] * iR0b0t1 wanted to get one
[20:01:09] <keenerd> I don't have a Rigol...
[20:01:32] <keenerd> But I am happy with the portability of my Hantek ;-)
[20:12:26] <Roklobsta> do i need to download the ASF extenstions for studio 5.1 or are they built into the 5.1 installers?
[20:14:50] <Tom_L> Roklobsta, did you hack yours?
[20:15:03] <Roklobsta> tom: not yet
[20:15:05] <Roklobsta> no need
[20:15:12] <Tom_L> i heard it's hard on the adc
[20:15:13] <Roklobsta> i don't need over 50MHz ./... <yet>
[20:15:38] <Roklobsta> yeah but the 100mhz version has the same guts, diff firmware.
[20:16:00] <Tom_L> i got a saleae analizer instead
[20:16:05] <Roklobsta> the adcs work at 1GS regardless if it's 100MHz or 50Mhz
[20:16:11] <Tom_L> but i have an analog scope already
[20:16:12] <keenerd> Tom_L: 8 or 16 channel?
[20:16:17] <Tom_L> 8
[20:16:21] <Roklobsta> analogue? ahahahahohohheheheee
[20:16:25] <Tom_L> they didn't have the 16 at the time
[20:16:31] <Tom_L> Roklobsta, it's old
[20:16:32] <Roklobsta> that's so 1924.
[20:16:46] <Tom_L> it gets it done
[20:16:57] <iR0b0t1> Whoah.
[20:17:01] <Tom_L> no need to buy a new one for just a hobby
[20:17:02] <Roklobsta> 1MHz?
[20:17:07] <iR0b0t1> dirty_d: You still here?
[20:17:15] <iR0b0t1> My values... seem to be 2x the value I was expecting.
[20:18:02] <iR0b0t1> nopenevermind
[20:19:33] <iR0b0t1> They're... within range.
[20:32:50] <iR0b0t1> MY GOD.
[20:32:52] <iR0b0t1> YOU ARE FUCKING ME.
[20:32:53] <iR0b0t1> FUCK.
[20:32:57] <iR0b0t1> FUCKFUCKFUCKFUCKFUCK
[20:32:57] <iR0b0t1> FUCK
[20:32:59] <iR0b0t1> .
[20:33:06] <iR0b0t1> Like all of my major problems, it was a typo.
[20:33:13] <iR0b0t1> :|
[20:34:06] <Tom_L> iR0b0t1 watch your language in this channel
[20:35:21] <iR0b0t1> Tom_L, I plead dire stress.
[20:36:56] <Tom_L> ok but you have been warned
[21:58:07] <iR0b0t1> On an ATmega328, which PWM pins do Timer0 and Timer2 effect? Does Timer0 have PWM pins hooked to it?
[21:58:22] <iR0b0t1> I know this should be in the manual but how I am reading it I can't find it
[21:59:36] <Tom_L> look under the register summary in the back of the pdf
[21:59:47] <Tom_L> it will show
[22:00:35] <Tom_L> they will be labelled something like OCR0x OCR2x etc
[22:01:26] <Tom_L> also a table in the timer section will tell you how to set up pwm for them
[22:02:10] <iR0b0t1> Oh right now I see it.
[22:02:10] <iR0b0t1> Thanks
[22:02:21] * iR0b0t1 wonders how he missed it ...
[22:47:33] <CapnKernel> amee2k: Sorry, I was off emitting Zeds :-)
[22:48:06] <CapnKernel> I was a Zed Emitting Dude.
[22:51:59] <timemage> another recursive acronym.
[22:54:11] * CapnKernel giggles
[23:18:36] <Roklobsta> capnk:
[23:18:42] <Roklobsta> i have to axe you a question
[23:19:26] <Roklobsta> a) Have you seen the Aurora Borealis yet?
[23:20:28] <Roklobsta> b) Do I need to get the seperate ASF 2.11.0 Extension package or is it already in the 5.1 installer?
[23:20:59] <Roklobsta> c) Does your broad band have blimin' caps like here in .au?
[23:21:07] <CapnKernel> axe away
[23:21:30] <CapnKernel> RokLobsta: In China, you're lucky to see any stars at all. I'm not kidding.
[23:21:37] <CapnKernel> In the cities at least.
[23:21:44] <CapnKernel> City of 10m+ people here
[23:21:58] <Roklobsta> oh fsck i mistook you for Dean.
[23:22:01] <Roklobsta> sorry
[23:22:08] <CapnKernel> I only wish :-)
[23:22:15] <Roklobsta> doh.
[23:22:20] * Roklobsta wipes eggs off keyboard
[23:22:35] <Roklobsta> anyway, can you get rubidium atomic clocks cheap?
[23:22:51] <CapnKernel> Shenzhen is only 20N, so it's unlikely the AB would be seen so far south (although it's been seen in Honolulu)
[23:23:03] <CapnKernel> If you're referring to the ones on eBay, best buy them from eBay.
[23:23:07] <Roklobsta> yeah i was wondering if dean had gone to see the norhten lights
[23:23:11] <Roklobsta> ok
[23:23:19] <CapnKernel> I can help with new components that are going into products being made now.
[23:23:27] <Roklobsta> how about modules like gps modules?
[23:23:42] <CapnKernel> I've seen the southern lights in Australia many times
[23:23:56] <Roklobsta> i havent' where did you see them?
[23:24:04] <Roklobsta> melb?
[23:24:16] <CapnKernel> Port Campbell, Wilson's Prom. Esp latter.
[23:24:33] <CapnKernel> Port Campbell not so far from you :-)
[23:24:36] <Roklobsta> i live close by i should keep my eyes out on a moonless night
[23:26:19] <CapnKernel> Re GPS modules, not at this stage.
[23:26:39] <Roklobsta> ok just components
[23:26:45] <Roklobsta> and pcbs
[23:27:03] <CapnKernel> Yes, everyday components, and PCBs.
[23:27:11] <Roklobsta> what about paste silkscreen foils?
[23:27:26] <CapnKernel> I'd like to, but not at this stage.
[23:27:42] <CapnKernel> Easier and cheaper to get me to make your PCBs.
[23:27:58] <rue_house> iR0b0t1, uh keep the language clean or you could get unexpectadly banned eh?
[23:28:00] <Roklobsta> ok. do you havea general rice guide?
[23:28:05] <Roklobsta> haha i mean price
[23:28:12] <CapnKernel> I was going to say :-)
[23:28:31] <CapnKernel> I'm working on my online ordering system now, and you'll be able to get instant quotes. But not at the moment
[23:28:40] <Roklobsta> the chinese people here say aussie chinese food is nothing like "home". i think I am glad.
[23:28:43] <CapnKernel> I've quoted for a couple of people, and they have been pleasantly surprised.
[23:29:23] <Roklobsta> max layers?
[23:29:25] <Roklobsta> 8
[23:29:28] <Roklobsta> 10
[23:29:30] <Roklobsta> 20
[23:29:35] <CapnKernel> 1 2 4 6
[23:29:47] <CapnKernel> I'm aiming at hobbyists, and most hobbyists will do 4 tops, I think.
[23:29:53] <Roklobsta> can you use the vialess pcb process?
[23:30:15] <CapnKernel> Ya, I luff to listen on ze vialess :-)
[23:30:24] <CapnKernel> (I'm not sure what you mean)
[23:30:36] <Landon> CapnKernel: that pun was amazing.
[23:30:42] * Landon goes back to lurking
[23:30:44] <Roklobsta> nono, no vias, a pad and a small bump to interconnect layers
[23:31:01] <Roklobsta> i had a 6 layer board made using it. looks weird with no vias.
[23:31:13] <CapnKernel> Roklobsta: They are not specialist PCBs, they are good quality, garden variety PCBs.
[23:31:36] <iR0b0t1> >quality
[23:31:37] <iR0b0t1> >china
[23:31:43] <iR0b0t1> How much did you say this costs? :p
[23:31:51] <Roklobsta> we get lenty of good quality crap from china
[23:31:56] <CapnKernel> Let's talk about it in PM
[23:32:19] <CapnKernel> Seriously, these are nice boards. If they weren't, I wouldn't sell them.
[23:32:42] <iR0b0t1> Why PM?
[23:32:57] <Roklobsta> you sound like a used car salesman.
[23:33:07] <Roklobsta> your personal guaratee!
[23:33:07] <iR0b0t1> ^
[23:33:12] <CapnKernel> I'd rather not spam #avr with stuff not everyone's interested in. I'll leave that to Jan-
[23:33:27] <iR0b0t1> No one is talking anyway.
[23:33:30] <iR0b0t1> You might as well.
[23:33:30] <Roklobsta> do you have a hair oil combover and a cheap tweed jacket?
[23:33:37] <CapnKernel> I'm a hacker just like you guys.
[23:33:46] <iR0b0t1> NO
[23:33:51] <iR0b0t1> THOU ART A CAPITALIST
[23:33:57] <CapnKernel> Come on guys, give me a break.
[23:34:05] <iR0b0t1> We're not attacking you.
[23:34:13] <Roklobsta> they used to kill capitalists in china.
[23:34:22] <iR0b0t1> It's just, people selling things... Are usually interested in money.
[23:34:47] <CapnKernel> Usually
[23:35:04] <iR0b0t1> Okay, that was a fluke of my local colloquial dialect.
[23:35:07] <Roklobsta> only the raspberry pi guys arre a charity
[23:35:09] <iR0b0t1> Always is the proper word.
[23:35:20] <CapnKernel> I'm into permaculture, hackerspaces, beer making, and I love to see people making things.
[23:35:32] <CapnKernel> People making things gives me a buzz.
[23:35:34] <iR0b0t1> Oh god I hate most of those hackerspace guys the most.
[23:35:46] <iR0b0t1> Someone even made a company type deal for those :|
[23:35:58] <iR0b0t1> Well I guess I mainly hate THOSE guys, but not a lot are much better.
[23:36:03] <Landon> iR0b0t1: all managers are bad too?
[23:36:20] <CapnKernel> I hope I can support my family with this, but I'm not doing it to make my fortune.
[23:36:25] <Roklobsta> is 4seconds for a 128kb write/verify to an atmega fast or average?
[23:36:25] <iR0b0t1> Landon: No, I've had a few good managers. Then they started giving salary incentives...
[23:36:28] <CapnKernel> Money doesn't inererest me
[23:36:57] <iR0b0t1> CapnKernel: We'll see what you say in a few months, or years, depending on your timeframe...
[23:37:18] <CapnKernel> Meanwhile, PCBws.
[23:37:21] <CapnKernel> PCBs even
[23:37:36] <Roklobsta> do you just want gerbers or will you take pcb files?
[23:38:18] <CapnKernel> Gerbers plus a drill file are fine.
[23:38:22] <Roklobsta> cool
[23:38:40] <Roklobsta> any colour?
[23:38:44] <CapnKernel> Any size board, any shape. Min qty 5
[23:38:44] <Roklobsta> i like blue
[23:41:29] <CapnKernel> (Why the iR0b0t1 won't my VPN stay up??? :-( )
[23:43:01] <Roklobsta> it's the GFC playing with your packets.
[23:43:10] <Roklobsta> are you using openvpn?
[23:43:19] <CapnKernel> Yes
[23:43:40] <CapnKernel> I was terminating at my house in Australia, but I recently switched to my Linode instance. It doesn't seem to work as well.
[23:43:58] <CapnKernel> Roklobsta: I presume you mean GFW not GFC :-)
[23:45:17] <Roklobsta> yer
[23:46:08] <CapnKernel> Roklobsta; Board colours: Green Red Blue Yellow Black White, Green is cheaper
[23:46:15] <Roklobsta> k
[23:46:18] <CapnKernel> Silkscreen: Black or white
[23:46:20] <Roklobsta> and boring
[23:46:47] <CapnKernel> If you're on a budget, go for green. If you want folks to drool, go for blue.
[23:46:51] <CapnKernel> With gold finish :-
[23:46:53] <CapnKernel> :-)
[23:47:28] <iR0b0t1> You said you were Australian, right? Do you have people on the street want to take pictures with you?
[23:47:32] <iR0b0t1> I've heard that's kinda... a thing.
[23:47:52] <CapnKernel> Shenzhen's ok. Other places, yes. It's annoying
[23:48:52] <CapnKernel> I went to the aviation museum in Beijing (you never saw so many MiG-15s and MiG-21s), and a group of schoolgirls asked me to be in their photo. Bizarre.
[23:49:21] <CapnKernel> I'm not quite sure why a group of schoolgirls were at an aviation museum though...
[23:49:56] <CapnKernel> iR0b0t1: I was born in (and my family live in) Box Hill.
[23:50:01] * Kening wonders why schoolgirl is in his hilight list.
[23:50:35] <iR0b0t1> o.o
[23:51:07] <CapnKernel> Oh sorry, that was aimed at RokLobsta. "Box Hill" isn't going to mean anything to you :-)
[23:51:30] <Roklobsta> it's a very rough part of melbourne
[23:51:47] <Roklobsta> locally known as "stab city"
[23:51:51] <CapnKernel> It is? First I heard.
[23:52:22] <Roklobsta> ok I had to talk it up. Box Hill is a very boring, white bread suburb of Melbourne.
[23:52:35] <CapnKernel> Indeed.
[23:52:57] <Roklobsta> yes I have been accosted by japanese schoolgirls for photos.
[23:53:05] <CapnKernel> Only claims to fame are almost reasonable public transport, and a *huge* asian presence in the central area
[23:53:16] <CapnKernel> Go there on the weekend, and you can play "spot the westerner"
[23:53:34] <Roklobsta> yeah Canberra Uni got like that. We played spot the aussie too.
[23:54:05] <CapnKernel> So, two Japanese guys are talking. One says "how was your overseas holiday?". The other replies, "I don't know, I haven't seen the photos yet"
[23:54:08] <Roklobsta> can you get cheap board loading?
[23:54:29] <CapnKernel> I'm looking into it, but not something I can do today.
[23:54:32] <CapnKernel> First things first.
[23:54:37] <Roklobsta> ok 1 thing at a time
[23:54:46] <Roklobsta> i just want cheap cheap cheap 10uf caps.
[23:54:49] <Roklobsta> lots of them.
[23:54:52] <CapnKernel> Can do.
[23:54:55] <Roklobsta> all sizes and voltages
[23:55:01] <CapnKernel> Whatever you want.
[23:55:02] <Roklobsta> high esr, low esr.
[23:55:04] <CapnKernel> And good quality too
[23:55:28] <CapnKernel> There are hundreds of shops that just sell caps. There's one shop I know that JUST sells one value of cap.
[23:55:32] <Roklobsta> i need lots of low esr caps for my rail gun.
[23:55:41] <CapnKernel> Let's talk PM then.
[23:56:04] <Roklobsta> i estimate about 2million 10uf caps.
[23:56:04] <CapnKernel> Also, getting boards loaded sounds like making products. Which is different to hobbyist projects.
[23:56:09] <iR0b0t1> CapnKernel, I don't get that joke?
[23:56:33] <iR0b0t1> Roklobsta: Why would you buy 2 million 10uF caps.
[23:56:35] <CapnKernel> It isn't a joke. Therre's a guy here who makes his living from selling one value of capacitor.
[23:56:50] <Roklobsta> the recent us navy rail gun tests inspired me.
[23:56:53] <CapnKernel> I want to help hobbyists with their projects.
[23:56:55] <iR0b0t1> Nah, the one about the pictures. I'll believe the "one-value-cap" guy.
[23:57:02] <Roklobsta> i need to better it. so i can get on hackaday
[23:57:08] <iR0b0t1> Roklobsta: At least buy 1000uF
[23:57:16] <CapnKernel> Roklobsta: How you going to connect them all? How many seconds for each cap connection? Multiply that by 2m.
[23:57:48] <Roklobsta> look, i need 2mill for each firing. I epexct that all the caps and wires will explode and catch fie upon firing.
[23:57:54] <iR0b0t1> I always thought the Chinese economy was slightly strange. You don't sell a product, you sell a part to a part to a product :p
[23:58:05] <Roklobsta> oh here he is
[23:58:29] <CapnKernel> Japanese seem to define the quality of their holiday by the number of photos they took. Or alternatively, it can be said that they spend their whole time behind the lens of the camera, and not getting out there and experiencing it.
[23:58:38] <CapnKernel> A gross generalisation of course.
[23:58:42] <iR0b0t1> Hah.
[23:58:43] <iR0b0t1> I see.
[23:59:13] <CapnKernel> Roklobsta: Suddenly my "I don't think you're serious about the caps" detector kicked in./
[23:59:24] <Roklobsta> yeah, it needs some tuning.